019736 S196 M5 c4591011 protocol

Pfizer Documents (PHMPT/FDA)

Pfizer Bla Submission

Pfizer 16 Plus Documents

171

Document text

BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 1of 170NON-INTERVENTIONAL ( NI) STUDY CONCEPT PR OTOCOL
Title Active Safet y Surveillance of the Pfizer -
BioNTech COVID -19 Vaccine in the U nited 
StatesDepartment of Defense Population 
Following Emergency  Use Authorization
Protocol number C4591011
Protocol version identifier Final Version 1.0
Date of last version of protocol 29January 2021
EUPost Authori zation Study (PAS) 
register numberTo be registered before the start of data 
collection
Active substance COVID-19 mRNA Vaccine is single -stranded, 
5’-capped messenger RNA (mRNA) produced 
using a cell -free in vitro transcription from the 
corresponding DNA templates, encoding the 
viral spike (S) protein of SARS -CoV-2.
Medicinal product Pfizer-BioNTech COVID -19 Vaccine
(BNT162b2)
Research question and objectives Research question: what are the incidence rates 
of safetyeventsof interest (based on adverse 
events of special interest [AESI])among 
individuals vaccinated with the Pfizer-
BioNTech COVID-19 vaccine within the 
United StatesDepartment of Defense (DoD)
Military Health Sy stem (MHS) overall and in
sub-cohortsof interest , as compared to
expected rates of those events?
Primary study objectives:
To assess whether individuals in the 
DoDMHSexperience increased risk of 
safetyevents of interest following 
receipt of the Pfizer- BioNTech 
COVID-19 vaccine;
To assess whether sub- cohorts of 
interest (i .e.,pregnant women, 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073039
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 2of 170immunocompromised, e lderly, 
individuals with specific comorbi dities,
individuals receiving only one dose of 
the Pfizer -BioNTech COVID-19 
vaccine, and individuals with prior 
SARS-CoV-2 infection) in the DoD
MHS experience increased risk of 
safety events of interest following 
receipt of the Pfizer -BioNTech 
COVID-19 vaccine.
Secondary study objective :
To characterize utilization patterns of 
the Pfizer -BioNTech COVID-19 
vaccine among individuals within the 
DoDMHS,including estimating the 
proportion of individuals receiving 
vaccine, 2- dosevaccinecompletion
rate, and distribution of time gaps 
between the first and second dose , 
demographics and health histories of 
recipients, overall and among the sub-
cohorts of interest . 
Authors Renu Garg, PhD, MPH
Safety Surveillance Research Scientist
Pfizer, Inc.
New York, NY
Mei Sheng Duh, ScD, MPH 
Managing Principal and Chief Epidemiologist 
Analysis Group, Inc.
Boston, MA
This document contains confidential information belonging to Pfizer. Except as otherw ise agreed to in writing, 
by accepting or reviewing this document, you agree to hold this information in confidence and not copy or 
disclose it to others (except where required by applicable law) or use it for unauthorized purposes. In the event 
of any actual or suspected breach of this obligation, Pfizer must be promptly notified.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073040
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 3of 1701.TABLE OF CONTENTS
1.TABLE OF CONTENTS.......................................................................................................3
2. LIST OF ABBREVIAT IONS................................ ................................ ................................ 5
3. RESPONSIBLE PARTI ES................................ ................................ ................................ ....8
4.ABSTRACT ................................ ................................ ................................ ........................... 9
5.AMENDMENTS AND UP DATES................................ ................................ ..................... 22
6.MILES TONES................................ ................................ ................................ ..................... 23
7. RATIONALE AND BAC KGROUND ................................ ................................ ................ 24
8. RESEARCH QUESTION AND OBJECTI VES................................ ................................ .25
9. RESEARCH METHODS................................ ................................ ................................ ....26
9.1. Study  Design................................ ................................ ................................ ...........26
9.1.1. Self -Controlled Risk I nterval (SCRI) Design ................................ .............26
9.1.2. Active Comparator Design ................................ ................................ .........29
9.1.3. Contemporary  Unvaccinated Control Design ................................ .............30
9.1.4. Study  Period................................ ................................ ................................ 31
9.2. Setting ................................ ................................ ................................ ...................... 31
9.2.1. Inclusion Criteria ................................ ................................ ........................ 31
9.2.2. Exclusion criteria ................................ ................................ ........................ 31
9.2.3. Subgroups ................................ ................................ ................................ ...31
9.3. Variables ................................ ................................ ................................ .................. 32
9.3.1. Exposure of I nterest................................ ................................ .................... 32
9.3.1.1. Pfizer -BioNTech COVID- 19 Vaccine Groups of Interest ........33
9.3.2. Baseline Characteristics ................................ ................................ ..............33
9.3.3. Outcomes ................................ ................................ ................................ ....35
9.4. Data Source................................ ................................ ................................ .............42
9.5. Study  Size................................ ................................ ................................ ................ 43
9.5.1. Power ................................ ................................ ................................ ..........43
9.6. Data Management ................................ ................................ ................................ ...45
9.6.1. Case report forms (CRFs)/Electronic data record......................................45
9.6.2. Record retention ................................ ................................ .......................... 45
9.7. Data Anal ysis................................ ................................ ................................ ..........46
9.7.1. Identification of Contemporary  Unvaccinated Controls ............................. 46
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073041
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 4of 1709.7.2. Baseline Characteristics ................................ ................................ ..............47
9.7.3. Vaccine Utilization Patterns ................................ ................................ .......47
9.7.4. Safet y Signal Analyses ................................ ................................ ...............47
9.7.4.1. Signal Detection ................................ ................................ ........48
9.7.4.2.Signal Evaluation ................................ ................................ ......52
9.7.4.3. Signal Verification ................................ ................................ ....54
9.7.5. Seasonality -Adjusted Cases -Centered Method ................................ ...........54
9.7.6. End-of- Season and End -of-Surveillance Analy ses................................ .....55
9.7.7. Subgroup Analy sis................................ ................................ ...................... 56
9.7.8. Incidence Rates and Time to Safety  Event of Interest Anal ysis................. 56
9.8. Quality  Control................................ ................................ ................................ ........57
9.9. Strengths and Limitations of the Research Methods ................................ ...............57
9.10. Other Aspects ................................ ................................ ................................ ........58
10. PROTECTI ON OF HU MAN SUBJECTS ................................ ................................ ........58
10.1. Patient I nformation ................................ ................................ ................................ 58
10.2. Patient Consent ................................ ................................ ................................ ......59
10.3. Institutional Review board (I RB)/Independent Ethics Committee (I EC).............59
10.4. Ethical Conduct of the Study ................................ ................................ ................ 59
11. MANAGEMENT AND R EPORTING OF ADVERSE EVENTS/ADVERSE 
REACTIONS................................ ................................ ................................ ...................... 59
12. PLANS FOR DI SSEMINATING AND COMMUNI CATING STUDY RESUL TS........61
13. REFERENCES ................................ ................................ ................................ .................. 62
14. LIST OF TABLES ................................ ................................ ................................ .............66
15. LIST OF FIGURES ................................ ................................ ................................ ...........66
16. ANNEX 1. LIST OF STAND-ALONE DOCUMEN TS................................ ................... 66
17. ANNEX 2. ENCEPP CHECKLIST FOR STUDY PROTOCOL S................................ ...66
18. ANNEX 3. ADDITIO NAL INFORMATION ................................ ................................ ...66
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073042
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 5of 1702. LIST OF ABBREVIATIONS
Abbreviation Definition
ACIP Advisory Committee on Immunization Practices
ADEM Acute disseminated encephalomy elitis
AE Adverse event
AEM Adverse event monitoring
AESI Adverse events of special interest
AIDS Acquired immunodeficiency  syndrome
AMI Acute myocardial infarction 
BMI Body mass index
CAD Coronary  artery disease
CI Confidence Interval
CCI Charlson comorbidity  index
CDC Centers for Disease Control and Prevention
CIDP Chronic inflammatory  demyelinating pol yneuropathy
CMA Conditional Marketing Authorization
COPD Chronic obstructive pulmonary  disease
COVID-19 Coronavirus Disease 2019
CPT Current Procedural Terminology
CRFs Case report forms
DIC Disseminated intravascular coagulation
DoD Department of Defense
DVT Deep vein thrombosis
TDap Diphtheria, tetanusand (acellular)pertussis
Td Diphtheria and tetanus
ED Emergency  department
EMA European Medicines Agency
EMR Electronic medical records
EU European Union
EUA Emergency  Use Authorization
EU PAS European Union Post -Authorization Safety
FDA Food and Drug Administration
GBS Guillain-Barré syndrome
GEP Good Epidemiological Practice
GPP Good Pharmacoepidemiology  Practices
H0 Null hypothesis
Ha Alternative h ypothesis
HBV Hepatitis B virus 
HCPCS Healthcare Common Procedure Coding S ystem
HCV Hepatitis C virus
HIV Human immunodeficiency virus
HPV Human papillomavirus
HRTx HealthResearchT x
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073043
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 6of 170Abbreviation Definition
ICD-10-CM International Classification of D iseases, Tenth Revision, Clinical 
Modification
ICD-10-PCS International Classification of Diseases, Tenth Revision, Procedure 
Coding Sy stem
IDN Integrated delivery  network
IEA International Epidemiological Association
IEC Independent Ethics Committee
IQR Interquartile range
IRB Institutional Review Board
ITP Immune thrombocy topenia
KD Kawasaki disease
LLR Log-likelihood ratio
MaxSPRT Maximized sequential probability  ratio test
MenACWY Meningococcal conjugate
MenB Serogroup B meningococcal 
MDR MHS Data Repository
MHS Military Health Sy stem
MIS-A Multisystem inflammatory  syndromein adults
mRNA Messenger RiboNucleic Acid
MS Multiple sclerosis
NDC National Drug Code
NIS Non-interventional study
ON Optic neuritis
PASS Post-Authorization Safety  Study
PB Privacy board
PDTS Pharmacy  data transaction sy stem
PRISM Post-Licensure Rapid Immunization Safety  Monitoring
RCA Rapid cycle analysis
RR Relative risk
SAP Statistical analy sis plan
SARS-CoV-2 Severe acute respiratory  syndrome coronavirus 2
SAS SAS Institute
SCRI Self-controlled risk interval
SD Standard deviation
SPEAC Safety Platform for Emergency  vACcines
TM Transverse m yelitis
TRICARE US Department of Defense purchased care
UK United Kingdom
US United States
VAED Vaccine-associated enhanced disease
VAERS Vaccine Adverse Event Reporting S ystem
VSD Vaccine Safet yDatalink
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073044
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 7of 170Abbreviation Definition
VTE Venous thromboembolism
WHO World Health Organization
YRR Your Reporting Responsibilities
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073045
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 8of 1703.RESPONSIBLE PARTIES
Principal Investigators of the Protocol
Name, degree(s) Job Title Affiliation Address
Renu Garg,
PhD, MPHSafety Surveillance Research 
ScientistPfizer, Inc. 235 East 42nd Street, 
New York, NY 10017
Mei Sheng Duh, 
ScD, MPHManaging Principal and Chief 
Epidemiologist
Visiting Scientist, Department of 
BiostatisticsAnalysis Group, Inc.
Harvard T. H. Chan 
School of Public 
Health111 Huntington Ave
14thFloor
Boston, MA 02199
677 Huntington Ave
Boston, MA 02115
Maral DerSarkissian, 
PhD Vice President and Senior 
Epidemiologist
Adjunct Assistant ProfessorAnalysis Group, Inc.
Fielding School of 
Public Health, 
University of 
California, Los 
Angeles333 South Hope Street
27th Floor
Los Angeles, CA
90071
650 Charles E Young 
DriveSouth
Los Angeles, CA 
90095
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073046
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 9of 1704.ABSTRACT
Title:Active Safet y Surveillance of the Pfizer -BioNTech COVID- 19 Vaccine in the U nited 
StatesDepartment of Defense Population Following Emergency Use Authorization
Protocol Version: 1.0; Dateof Protocol : 29January 2021
Authors: RenuGarg, PhD, MPH, Pfizer, I nc.; Mei Sheng Duh, ScD, MPH , Analysis Group, 
Inc.
Rationale and background :
In March 2020, the World Health Organization (WHO) declared a global pandemic for the 
coronavirus disease 2019 (COVID -19) due to the severe acute respiratory  syndrome 
coronavirus 2 (SARS -CoV-2), which was first identified by  public health officials in China 
in December 2019.1The COVID -19 pandemic presents an unprecedented public h ealth 
crisis. As of January  7, 2021, over 21.4million COVID -19 cases and 364,000 deaths have 
been reported in the United States (US) alone.2
Pfizer and BioNTech have partnered to develop a novel messenger Ri boNucleic Acid 
(mRNA) vaccine against SARS -CoV-2 for the prevention of COVID -19 (Candidate 
BNT162b2). Pfizer is conducting a Phase 1/2/3, randomized, placebo -controlled, observer -
blind, dose -finding, vaccine candidate- selection, and efficacy  study among he althy 
individuals (NCT04368728). The Food and Drug Administration (FDA) reviewed the 
available safet y data from 37,586 participants 16 years of age and older and did not identify 
any specific safet y concerns.  In addition, the analysis of available efficac y data from 36,523 
participants 12 years of age and older without evidence of prior SARS- CoV-2 infection at 
least 7 day s after receiving the second dose demonstrated 95% efficacy  of the vaccine in the 
prevention of COVID -19 (as confirmed by 8 vs. 162 COVID -19 cases in the vaccine and 
placebo groups, respectively ).3,4Based on these safety  and efficacy  data, as well as a review 
of manufacturing information regarding product quality  and consistency , the FDA 
determined that th e known and potential benefits of the vaccine outweighed the known and 
potential risks for the prevention of COVID -19 in individuals 16 y ears of age and older .4
Therefore on December 11, 2020, the Pfizer -BioNTech COVID- 19 vaccine was granted an 
Emergency  Use Authorization (EUA) by the FDA to prevent COVID -19in individuals 16 
years of age and older.5
With respect to geographic regions other than the US, on December 2, 2020, the United 
Kingdom (UK) wasthe first country  in the world to grant temporary  authorization for 
emergency  use of the Pfizer -BioNTech COVID -19 vaccine.6On December 21, 2020, the 
European Medicines Agency  (EMA) granted the Pfizer -BioNTech COVID- 19 vaccine a 
conditional marketing authorization (CMA) for use among individuals 16 years of age and 
older throughout all of the European Union’s (EU) 27 member states.7
As required b y the EUA, post-authorization observational studies using real -world data are 
needed in order to assessthe association between Pfizer -BioNTech COVID- 19 vaccine and 
pre-determined safet yevents of interest (including deaths , hospitalizations, and severe 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073047
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 10of 170COVID-19) among individuals administered the vaccine in both the population at large and 
in populations of interest (e .g., pregnant women, immunocompromised individuals, elderly , 
and those with specific comorbidities ).4Pfizer,in collaboration with Health ResearchTx
(HRTx)and Analy sis Group , herein proposes post-EUA active surveillance of safetyevents
of interest in the Department of Defense (DoD) population based on the Priority  List of 
Adverse Events of Special I nterest from the Brighton Collaboration’s Safety  Platform for 
Emergency  vACcines (SPEAC) Project, and theFDA and the Centers for Disease Control 
and Prevention’s (CDC) Advisory  Committee on Immunization Practices (ACI P) enhanced 
safety monitoring recommendation . As part of phased allocation of COVID -19 vaccinations, 
all healthcare providers, emergency  services, and public safety personnel within the DoD 
population will qualify  to receive the COVID -19 vaccine.8This safet y surveillance study  will
identify and evaluate rapid ,near real-time potential safet y signals associated with the Pfizer -
BioNTech COVID -19 vaccine in thelarge-scale DoDMilitary Health Sy stem (MHS) 
healthcare database, which includes both administrative claims data and clinical data from
electronic medical record s (EMR). The observed safetyevent of interest rates will be 
compared to expected rates derived from self -controls,active comparators receiving seasonal 
influenza vaccination , and contemporary  unvaccinated controls. Part of the me thodologies 
used in this study  are constructed based on approaches previously used by the Post-Licensure 
Rapid Immunization Safety  Monitoring (PRISM) program for the H1N1 vaccine .9This non-
interventional study  is designated as a Post -Authorization Safety Study (PASS) commitment 
to the US FDA and is a Category  3 commitment in the EU Risk Management Plan .
Research question and o bjectives: 
Research question: what aretheincidence rates of safety eventsof interest (basedonadverse 
events of special interest AESIamong individuals vaccinated with the Pfizer-BioNTech 
COVID-19 vaccine within the US DoDMHSoverall and in sub- cohortsof interest ,as 
comparedto expected rates of those events?
Primary study objectives:
To assess whether individuals in the DoDMHSexperience increased risk of safety 
eventsof interest following receipt of the Pfizer -BioNTech COVID -19 vaccine;
To assess whether sub -cohorts of interest (i.e., pregnant women,
immunocompromised, elderly , individuals with specific comorbidities,individuals
receiving onl y one dose of the Pfizer -BioNTech COVID- 19 vaccine, and individuals 
with prior SARS -CoV-2 infection) in the DoD MHS experience increased risk of 
safety eventsof interest following receipt of the Pfizer -BioNTech COVID -19 
vaccine.
Secondary study objective :
To characterize utilization patterns of the Pfizer -BioNTech COVID -19 vaccine 
among individuals within the DoDMHS,includingestimating the proportion of 
individuals receiving vaccine, 2-dosevaccinecompletion rate,anddistribution of 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073048
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 11of 170time gaps between the first and second dose , demographics and health histories of 
recipients, overall and among the sub-cohorts of interest .
Study design: This post-EUAactive safet y surveillance program will employ  a rapid-cycle, 
longitudinal, observational cohort study  designto provide early  real-world safet y 
information . 
The self-controlled risk interval (SCRI) design will be used to sequentially  monitor 
occurrence of safety eventsof interest while controlling for time- invariant
confounders .The SCRI  design uses data from cases (i .e., individuals who 
experience safet y eventsof interest following vaccination) to compare the risk 
interval following vaccination t o pre-or post-vaccination non -risk intervals (“pre -
vaccination control interval” and “post -vaccination control interval”) in the same 
individual.
Safety eventsof interest associated with Pfizer -BioNTech COVID -19 vaccinations 
will also be sequentially  monitored and compared to two comparator populations:
(a) Recipients of influenza vaccinein the DoDMHSduring2014/2015
through2018/2019flu seasons , as an active comparator . Data in peri -COVID 
time periods from January  2020to present are excluded because of pandemic -
associated underutilization of health resources and underreporting of medical 
events.
(b) Asample of contemporary  unvaccinated matched controls in the DoD 
MHS, as a general population comparator group, who will be identified during 
the same time period as individuals receiving the Pfizer-BioNTech COVID-19 
vaccineto reflect the background rate of current safet y events of interest .The
contemporary  unvaccinated controls will be randomly  sampled to matchthe 
baseline demographic and clinical characteristics of individuals who receive 
the Pfizer -BioNTech COVID- 19 vaccine (via both exact and propensit y score 
matching, using a ratio of 1:N, but no more than 1:4 due to diminishing gains 
in efficiency ) in order to ensure that the cohortsare comparable.
Population : The exposed population will be kept as broad as possible in order to capture 
safety eventsof interest that occu r among all individuals receiving the Pfizer -BioNTech 
COVID-19 vaccine in the period from December 11, 2020 to present . Individuals will be 
included if they have a record of at least one dose of Pfizer-BioNTech COVID-19 vaccine in 
the period. Individuals who receive at least one dose of CO VID-19 vaccine from a 
manufacturer other than Pfizer -BioNTech will be identified and reported, but they  will be 
excluded from further analy sis.
The influenza vaccine comparator cohort will be identified based on a record of at least one 
dose of seasonal influenza vaccine during prior flu seasons, from 2 014/2015 through 
2018/2019.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073049
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 12of 170The contemporary  unvaccinated control cohort will be randomly sampled from individuals in 
DoD MHS who did not receive any COVID-19 vaccine on or after December 11, 2020. They 
will be matched to individuals who received the Pfizer -BioNTech COVID- 19 vaccine on
baseline demographic and clinical characteristics via both exact and propensity  score 
matching in the time period of December 11, 2020 onwards.
All individuals will be required to have at least 1 year of continuous enrollment (i.e., baseline 
period) prior to vaccination date (or matched index date for unvaccinated controls).
Depending on the attrition rate, the length of the baseline period may  be modified to 6 
months.
Variables :
Exposure s: Administration of Pfizer -BioNTech COVID- 19 vaccine post -EUA 
approvalwill be identified based on the following:
oCurrent Procedural Terminology  (CPT) code 91300 (Severe acute respiratory  
syndrome coronavirus 2 (SARS -CoV-2) (coronavirus disease [COVID- 19]) 
vaccine, mRNALNP, spike protein, preservative free, 30 mcg/0.3mL dosage, 
diluent reconstituted, for intramuscular use) and associated vaccine 
administration HCPCScodes corresponding to the first dose: 0001A (ADM 
SARS-CoV-230 mcg/0.3mL  1st),and the second dose: 0002A (ADM SARS -
CoV-2 30 mcg/0.3mL 2nd);10,11OR
o10 and 11-digit National Drug Codes (NDCs) 59267-1000-1 (corresponds to 
first dose) , 59267-1000-01(corresponds to second dose) ;10OR
oImmunization records that contain da ta on vaccine code descriptor, vaccine 
manufacturer ( i.e., Pfizer), lot number, injection site, and date(s) of 
immunization ;10
Relevant codes will be continuously  reviewed and amended if new codes are added. 
Administration of the seasonal influenza vaccine during 2014/2015 through 
2018/2019 flu seasons will be identified based on the following:
oCPT codes 
90654 (Influenza virus vaccine, trivalent (IIV3), split virus, 
preservative -free, for intradermal use) ;OR
90656 (Influe nza virus vaccine, trivalent (IIV3), split virus, 
preservative free, 0.5 mL dosage, for intramuscular use); OR
90658 (Influenza virus vaccine, trivalent (IIV3), split virus, 0.5 mL  
dosage, for intramuscular use) ;OR
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073050
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 13of 170o10 and 11- digit NDCs ;OR
oImmunization re cords that contain data on vaccine code descriptor, vaccine 
manufacturer, lot number, injection site, and date(s) of immunization. 
Outcomes : Safety eventsof interest for active surveillance (see Table 1and 
Appendix Table 2)are based on the Priority  List of Adverse Events of Special 
Interest from the Brigh ton Collaboration’s Safet y Platform for Emergency  vACcines 
(SPEAC) Project, the FDA, 
and the Centers for Disease Control and Prevention’s 
(CDC) Advisory  Committee on Immunization Practices (ACIP) enhanced safet y 
monitoring recommendations . 
The list of safety eventsof interest may be revised over the course of the study , and if 
unanticipated potential safety eventsof interest are identified during the course of 
surveillance, they  will be added to the list and included in the anal ysesof interest . 
The risk and control intervals for each safety eventof interest are based on biological 
plausibility  and precedents in the literature (see Table 1). Outpatient(including 
emergency  department) and/or inpatient setting s will be used to identify  safety events
of interest , depending on the type of event. The specific encounter setting to be 
considered for each safety event of interest is summarized in Table 1and can be 
assigned to 1) the risk interval following Pfizer -BioNTech COVID -19 vaccination, 2) 
the pre-vaccination self-control interval, 3) the post -vaccination self-control interval, 
or 4) risk interval for the active comparators receiving seasonal influenza vaccine,
and 5) risk interval for the contemporary  unvaccinated controls. Events outside the 
intervals will not be counted.
Only the individual’s first instance of asafety eventof interest following a specified 
clean window (i .e., the occurrence -free baseline periodused to define incide nt 
outcomes during which individuals enter the study  cohort only  if the safety event of 
interest did not occur during this period) will be captured; this means that if a safety 
eventof interest is identified but diagnosis codes corresponding to the safety event of 
interest are also observed during the clean window, it will not be counted. The 
duration of the pre-specified clean window will differ by  safety eventof interest (see 
Appendix Table 2) in order to rule out pre -existing events.
Key Covariates
:Baseline demographic ( i.e., age, sex, state) and clinical 
characteristics (i.e., smoking, body mass index [ BMI],history of anaphylaxis/allergic 
reactions, previous anaphy laxistovaccine component, history  of hospitalizations, 
pregnancy , Charlson Comorbidity  Index [CCI ], selectedcomorbidities, and 
concurrent immunizations)12will be assessed based on available data ( i.e., during1-
yearbaseline) prior to the date of vaccination with Pfizer -BioNTech COVID -19 
vaccine,date of seasonal in fluenza vaccination foractive comparator s, or assigned 
index date for contemporary  unvaccinated controls.
Subgroups: Pregnant women, immunocompromised individuals, elderl y, individuals 
with specific comorbidities, those receiving onl y one dose of Pfizer-BioNTech 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073051
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 14of 170COVID-19 vaccine , and individuals with prior SARS -CoV-2 infection will be 
identified .
Data source :The MHS is a single pay er system that provides medical coverage and 
pharmacy  benefits for active duty  and retired military  members , civilian DoD personnel, and 
their families (beneficiaries). There are 9.6 million beneficiaries included in the MHS, of 
whom 1.4 million (14.6%) are active dut y, 1.7million (17.7%) are active duty  family 
members, 392,000 (4.1%) are national guard and reservemembers, 609,000 (6.3%) are 
family members of national guard and reservemembers, and 5.47 (57.0%) million are 
retirees and their family  members.13,14The DoD includes 64 hospitals, hundreds of clinics, 
25,000 uniformed ph ysicians, and 400,000 community  network providers. The population 
within the MHSis demographicall y representative of the US overall, with slight over -
representation of persons >65 y ears of ag e(20.1% in DoD MHS vs. 12.9 % in the general US 
population) .15The gender distribution is approximately 49% female and 51% male.
The DoD prioritized vaccine distribution to healthcare workers and emergency  services 
personnel, personnel performing activities associated with critical national capabilities, select 
deploying individuals, other critical and essential support, individuals at the highest risk for 
developing severe illness from COVID -19, and adults age 7 5 and older.16
Study size:The sample size achieved will depend on the number of recipients of Pfizer -
BioNTech COVID-19 vaccine within the DoD MHS during the stud y period, which will 
increase over time with subsequent anal yses.Preliminary  estimates for the number of 
individuals in the DoDMHSwho received thePfizer -BioNTech COVID- 19 vaccine will be 
reported in the statistical anal ysis plan (SAP).
Data analy sis:A stepwise process, illustrated below,will be performed for signal detection, 
evaluation , and verification. 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073052
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 15of 170
Notes: 
[1] List of safety events of interest and corresponding definitions may be refined as the study progresses based on additional 
available information.
[2] The risk and control intervals selected for the SCRI analysis for each safety event of interest are based on biological 
plausibility and pre cedents in the literature. Only the individual ’s first instance during the specified clean window (i .e., the 
interval used to define incident outcomes) will be included . Note that only the first inpatient or outpatient occurrence of a 
safety event of interest following the clean window will be used to identify incident events (e.g., if an inpatient safety event
of interest occurs in the clean window, a repeat occurrence will not be counted in the risk interval). However, event 
worsening will be counted as a safety event of interest . For example, if an outpatient safety event of interest occurs in the 
clean window and an inpatient occurrence for the same type ofsafety event of interest occurs in the risk interval, the 
inpatient occurrence will be counted as a safety event of interest .
1) Signal detection : The goal is to provide rapid- cycle, near real -time safety  surveillance. In 
the signal detection phase, the SCRI  analysis will only  include pre -vaccination control 
intervals as the post -vaccination control intervals will require a longer time to accumulate 
and will be used in the signal evaluation phase . To account for multiple testing and repeated 
review of the data , e.g., monthly  (to be stipulated in the SAP) , the maximized sequential 
probability  ratio test (MaxSPRT )using a binomial probability  model will be applied. For 
comparison with individuals who received seasonal influenza vaccination, the Poisson -based 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073053
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 16of 170MaxSPRT will be applied. The comparison for the contemporary  unvaccinated controlswill 
be conduct edusing the binomial-based MaxSPRT method. Over time, however, the number 
of eligible contemporary  unvaccinated controls to be matched to vaccinated individuals is 
expected to decrease, which will result in uncertainty  in the expected number of safety events
of interest .  As a result, conditional Poisson MaxSPRT (CMaxSPRT) may  be considered to 
account for error in the estimated expected number of events.
Sequential anal yses for each safet y event of interest will commence once at least 3 events 
occur. This approach is consistent with the FDA’s COVID -19 Vaccine Safety  Surveillance 
Projectto avoid spurious signals from a few early  events.17Signals will be detected if the 
critical values are reached via the SCRI or active comparator anal ysis. Criticalvalues will be 
determined for each safety event of interest based on historical incidence rate, expected upper 
limit of the number of events under the null h ypothesis, pre -specified significance level ,and 
power. Incidence rates will also be calculated ,and Kaplan-Meier methods willbe used to 
analyze time to safety eventsof interest .
2) Signal evaluation : If signals a re detected for safety eventsof interest based on the anal ysis 
described above, further evaluation will be conducted to refine and confirm such detections.
This will include comprehensive quality assurance (for example, check for possible 
duplications of claims or medical records, checking for unusual clustering in claim or 
medical record acc rual by service date for potential coding issues, check for geographical 
distribution of cases that may  be related to lot numbers or diagnostic practice) and 
multivariate adjustment using Poisson regression to account for baselinedifferences between 
Pfizer-BioNTech COVID- 19 vaccinated and active comparator cohorts. SCRI analyses using 
the post-vaccination control intervals will be conducted as an additional inferential analy sis 
once enough post -vaccination time has accumulated. Lastly, theassessment of temporal 
clustering will also be conducted.
3) Signal verification : diagnostic validation of the detected safety eventsof interest via 
adjudication of medical records by DoDMHSclinicians for outcomeverification will be 
conducted in a representative sample of cases. For rare events, potentiall y all cases may  be 
adjudicated.
End-of-season anal yses (over the course of the 30 -month period) and an end-of- surveillance 
analysis (i.e., at 30 months) will be conducted. Various subgroup analyses will also be 
conducted , examining different age groups, immunocompromised individuals, pregnancy , 
individuals with specific comorbidities patients, those who only  received one dose of the 
Pfizer-BioNTech COVID- 19 vaccine, and those with prior SARS -CoV-2infection based on 
medical history  or pre-vaccination serology.
Milestones:
Registration in the EU PAS register : To be registered before the start of data 
collection ;
DoD Institutional Review Board ( IRB)approval (estimated) : 15 March 2021;
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073054
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 17of 170Start of data collection (estimated planned date for starting data extraction for 
analysis): 01 May2021;
Interim reports: 30June 2021; 31 December 2021; 30June 2022, 31 December 2022;
End of data collection (estimated planned date for final data cut) : 10June2023;
Final study  report: 31 December 2023.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073055
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 18of 170SUMMAR Y:
Objective s Primary 1 Primary 2 Secondary
To assess whether individuals in the DoD 
MHS experience increased risk of safety 
eventsof interest following receipt of the 
Pfizer-BioNTech COVID -19 vaccine .To assess whether sub-cohorts of interest (i.e. , 
pregnant women, immunocompromised, 
elderly, individuals with specific comorbidities,
individuals receiving only one dose of the 
Pfizer-BioNTech COVID -19 vaccine, and 
individuals with prior SARS -CoV-2 infection ) 
in the DoD MHS experience increased risk of 
safety events of interest following receipt of the 
Pfizer-BioNTech COVID -19 vaccine .To characterize utilization patterns of the Pfizer-BioNTech 
COVID-19 vaccine among individuals within the DoD MHS, 
including estimating the proportion of individuals re ceiving 
vaccine, 2 -dose vaccine completion rate, and distribution of 
time gaps between the first and second dose, demographics 
and health histories of recipients, overall and among the sub-
cohorts of interest .
Study 
designThis post-EUA active safety surveillance program will employ a rapid -cycle, longitudinal, observational cohort study design to provide early real -world safety 
information. 
The self-controlled risk interval (SCRI) design will be used to sequentially monitor occurrence of safety events of interest while controlling for time -
invariant confounders. The SCRI design uses data from cases (i .e., individuals who experience safety events of interest following vaccination) to 
compare the risk interval following vaccination to pre -or post-vaccination non -risk intervals (“pre -vaccination control interval” and “post -vaccination 
control interval”) in the same individual. 
Safety events of interest associated with Pfizer-BioNTech COVID -19 vaccinations will also be sequentially monitored and compared to two comparator 
populations:
(a) Recipients of influenza vaccine in the DoD MHS during 2014/2015 through 2018/2019 flu seasons, as an active comparator. D ata in peri -
COVID time periods from January 2020 to present are excluded because of pandemic -associated underutilization of health resources and 
underreporting of medical events. 
(b) A sample of contemporary unvaccinated controls from the general population as reflected in the DoD MHS database, who will be identified 
contemporaneously with patients receiving Pfizer -BioNTech COVID -19 vaccine to reflect the background rate of current safety events of interest 
during the same time period. The contemporary unvaccinated controls will be randomly sampled to match the baseline demographi c and clinical 
characteristics of individuals who receive the Pfizer -BioNTech COVID -19 vaccine (via both exact and propensity score matching, using a ratio of 
1:N, but no more than 1:4 due to diminishing gains in efficiency) in order to ensure that the cohorts are comparable. 
Study 
populationThe study will be kept as broad as possible in order to capture safety events of interest that occur among vaccinated individuals.
Inclusion criteria:
Record of at least one dose of Pfizer-BioNTech COVID -19 vaccine ;or
Record of at least one dose of seasonal influenza vaccine during prior flu seasons, from 2014/2015 t hrough2018/2019 ( as anactive comparator ); or
No record of anyCOVID-19 vaccine (i.e. , unvaccinated controls) .
Exclusion criteria:
Individuals who receive at least one dose of Pfizer-BioNTech COVID -19 vaccine in addition to a COVID -19 vaccine from a manufacturer other than 
Pfizer-BioNTech will be identified and reported, but they will be excluded from further analysis.
Study 
PeriodThe studywill be conducted for a period of 30 months, starting on December 11, 2020 onward, with data collection conclu ding onJune 10, 2023.
Exposure Administration of Pfizer -BioNTech COVID -19 vaccine post -EUA approval will be identified based on records of the following:
Current Procedural Terminology (CPT) code 91300 (Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (coronavirus disease [COVID -
19]) vaccine, mRNALNP, spike protein, preservative free, 30 mcg/0.3mL dosage, diluent reconstituted, for intramuscular use) a nd associated vaccine 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073056
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 19of 170administration HCPCS codes corresponding to the first dose: 0001A (A DM SARS -CoV-2 30 mcg/0.3mL 1st), and the second dose: 0002A (ADM 
SARS-CoV-2 30 mcg/0.3mL 2nd); OR
10 and 11 -digit National Drug Codes (NDCs) 59267 -1000-1 (corresponds to first dose), 59267-1000 -01 (corresponds to second dose); OR
Immunization records that contain data on vaccine code descriptor, vaccine manufacturer (i .e., Pfizer), lot number, injection site, and date(s) of 
immunization; 
Administration of the seasonal influenza vaccine during 2014/2015 through 2018/2019 flu seasons will be identified based on records of the following:
CPT codes 
o90654 (Influenza virus vaccine, trivalent (IIV3), split virus, preservative -free, for intradermal use); OR
o90656 (Influenza virus vaccine, trivalent (IIV3), split virus, preservative free, 0.5 mL dosage, for intramuscular use); OR
o90658 (Influenza virus vaccine, trivalent (IIV3), split virus, 0.5 mL dosage, for intramuscular use); OR
10 and 11 -digit NDCs; OR
Immunization records that contain data on vaccine code descriptor, vaccine manufacturer, lot number, in jection site, and date(s) of immunization .
Safety 
events of 
interestSafety events of interest for active surveillance were identified based on the Priority List of Adverse Events of Special Interest from the Brighton Collaboration’s 
Safety Platform for E mergency vACcines (SPEAC) Project, the FDA, and the Centers for Disease Control and Prevention’s (CDC) Advisory Committee on 
Immunization Practices (ACIP) enhanced safety monitoring recommendations. The list of safety events of interest may be revised over the course of the study, 
and if unanticipated potential safety events of interest are identified during the course of surveillance, they will be added to the list and included in the analyses. 
The risk and control intervals for each safety event of interestare based on biological plausibility and precedents in the literature. Outpatient (including 
emergency department) and/or inpatient settings will be used to identify safety events of interest , depending on the type of event. The specific encounter setti ng 
to be considered for each safety events of interest maybe assigned to 1) the risk interval following Pfizer -BioNTech COVID -19 vaccination, 2) the pre -
vaccination self -control interval, 3) the post -vaccination self -control interval, or 4) risk interval for the active comparators receiving seasonal influenza vaccine 
and contemporary unvaccinated controls. Only the individual’s first instance of a safety event of interest following a specified clean window (i .e., the occurrence -
free baseline period used to define incident outcomes during which individuals enter the study cohort only if the safety event of interest did not occur during this 
period) will be included; this means that if a safety event of interest is identified but diagnosis codes corresponding to the safety event of interest are also 
observed during the clean window, it will not be counted. The duration of the pre -specified clean window will differ by safety event of interest (see 
Appendix Table 2) in order to rule out pre -existing events.
Neurologic:
Generalized convulsions/seizures
Guillain-Barré syndrome (GBS)
Aseptic meningitis
Encephalitis/encephalomyelitis
Other acute demyelinating diseases
Transverse myelitis (TM)
Multiple sclerosis (MS)
Optic neuritis (ON)
Bell’s palsy
Immunologic:
Anaphylaxis
Vasculitides
Arthritis and arthralgia/joint pain
Multisystem inflammatory syndrome in adults (MIS-A)
Kawasaki disease (KD)
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073057
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 20of 170Fibromyalgia
Autoimmune thyroiditis
Cardiac:
Myocarditis
Pericarditis
Acute myocardial infarction (AMI)
Hematologic:
Thrombocytopenia
Disseminated intravascular coagulation (DIC)
COVID-19 (for all COVID-19 -related safety events of interest listed below, a diagnosis of COVID- 19 will be required in addition to diagnosis codes or 
laboratory values specified in Appendix Table 2; in addition, CO VID-19 related safety events of interest will only be evaluated using data from 2020 onward 
using the SCRI design and unvaccinated controls only):
Severe COVID -19 disease
Microangiopathy
Heart failure and cardiogenic shock
Stress cardiomyopathy
Coronary artery disease (CAD)
Arrhythmia
Deep vein thrombosis (DVT)
Pulmonary embolus
Cerebrovascular hemorrhagic stroke
Cerebrovascular non -hemorrhagic stroke
Limb ischemia
Hemorrhagic disease
Acute kidney injury
Liver injury
Chilblain-like lesions
Single organ cutaneous vasculitis
Erythema multiforme
Other:
Pregnancy outcomes (note that outcomes related to delivery will only be assessed using active comparators and unvaccinated co ntrols rather than SCRI 
since delivery will only occur at a single time point)
Death
Narcolepsy/cataplexy
Non-anaphylactic allergic reactions
Appendicitis
Data source The DoDMHS Data Repository (MDR) will be used.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073058
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 21of 170Data 
analysisA stepwise process, illustrated below, will be performed for signal detection, evaluation, and verification.  
1) Signal detection: The goal is to provide rapid -cycle, near real -time safety surveillance. In the signal detection phase, the SCRI analysis wil l only include pre -
vaccination control intervals as the post -vaccination control intervals will require a longer time to accumulate and will be used in the signal evaluation phase. To 
account for multiple testing and repeated review of the data , e.g., monthly(to be stipulated in the SAP) , the maximized sequential probability ratio test 
(MaxSPRT) using a binomial probability model will be applied. For comparison with individuals who received seasonal influenza vaccination, the Poisson -based 
MaxSPRT will be applied. The comparison for the contemporary unvaccinated controls will be conducted using the binomial -based MaxSPRT method. Over 
time, however, the number of eligible contemporary unvaccinated controls to be matched to vaccinated individuals is expected to decrease, which will result in 
uncertainty in the expected number of safety events of interest . As a result, conditional Poisson MaxSPRT (CMaxSPRT) may be considered to account for error 
in the estimated expected number of events.
Sequential analyses fo r each safety event of interest will commence once at least 3 events occur. This approach is consistent with the FDA’s COVID -19 Vaccine 
Safety Surveillance Project to avoid spurious signals from a few early events. Signals will be detected if the critical values are reached via the SCRI or active 
comparator analysis. Critical values will be determined for each safety event of interest based on historical incidence rate, expected upper limit of the number of 
events under the null hypothesis, pre -specified significance level, and power. Incidence rates will also be calculated ,and Kaplan -Meier methods will be used to 
analyze time to safety events of interest .
2) Signal evaluation: If signals are detected for safety events of interest based on the analysis de scribed above, further evaluation will be conducted to refine and 
confirm such signals. This will include comprehensive quality assurance (for example, check for possible duplications of clai ms or medical records, checking for 
unusual clustering in claim or medical record accrual by service date for potential coding issues, check for geographical distribution of cases that may be related 
to lot numbers or diagnostic practice) and multivariate adjustment using Poisson regression to account for baseline diffe rences between Pfizer -BioNTech 
COVID-19 vaccinated and active comparator cohorts. SCRI analyses using the post -vaccination control intervals will be conducted as an additional inferential 
analysis once enough post -vaccination time has accumulated. Lastly, the assessment of temporal clustering will also be conducted. 
3) Signal verification: diagnostic validation of the detected safety events of interest via adjudication of medical records by DoD MHS clinicians for outcome 
verification will be conducted in a representative sample of cases. For rare events, potentially all cases may be adjudicated.
End-of-season analyses (over the course of the 30 -month period) and an end-of -surveillance analysis (i.e. , at 30 months) will be conducted. Various subgroup 
analyses will also be conducted, examining different age groups, immunocompromised individuals, pregnancy, individuals with specific comorbidities patients, 
those who only received one dose of the Pfizer -BioNTech COVID -19 vaccine, and those with prior SARS-CoV -2 infection based on medical history or pre -
vaccination serology.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073059
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 22of 1705.AMENDMENTS AND UPDAT ES
None
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073060
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 23of 1706.MILESTONES
Milestone Planned date
Registration in the EU PAS register To be registered before the start of 
data collection
DoD IRB approval(estimated) March 2021
Start of data collection (estimated) May2021[1]
Interim reports 30June 2021
31December 2021 
30June 2022 
31 December 2022 
End of data collection (estimated) 10 June2023[2]
Final study  report 31December 2023 
Abbreviations : DoD, Department of Defense; IRB, Institutional Review Board.
Notes:
[1] Start of data collection is the planned date for starting data extraction for the purposes of the study 
analysis.  The initial data analysis will include the Pfizer -BioNTech COVID- 19 vaccine exposure since 
December 11, 2020, the EUA approval date by the US FDA.
[2] End of data collection is the planned date on which the Pfizer -BioNTech COVID -19 vaccine exposure 
reached 30 months post EUA approval.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073061
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 24of 1707. RATIONALE AND BACKGROUND
In March 2020, the World Health Organization (WHO) declared a global pandemic for the 
coronavirus disease 2019 (COVID -19) due to the severe acute respiratory  syndrome 
coronavirus 2 (SARS -CoV-2), which was first identified by public health officials in China 
in December 2019.1The COVID -19 pandemic presents an unprecedented public health 
crisis. As of January7, 2021, over 21.4million COVID -19 cases and 364,000 deaths have 
been reported in the United States (US) alone.2To date, the incidence of COVID- 19 has 
continued to rise, largel y affecting the elderl y and middle-aged individuals, with worsening 
clinical sequelae linked to increasing age and comorbid conditions (e .g., cardiovascular 
disease, active cancer, obesity , diabetes and chronic lung disease).18,19SARS-CoV-2 is a 
well-adapted highly infectious human pathogen with a case fatality rate that ranges between 
0.5% and 20%, based on the individual’s age, gender, race , and comorbidites .20
Pfizer and BioNTech have partnered to develop a novel messenger RiboNucleic Acid 
(mRNA)vaccine againstSARS-CoV-2for the prevention of COVID -19 (Candidate 
BNT162b2). To this end, Pfizer is conducting a Phase 1/2/3, randomized, placebo -controlled, 
observer-blind, dose -finding, vaccine candidate- selection, and efficacy  study among healthy  
individuals (NCT04368728). In their Phase 1 trial evaluating safety  and immunogenicit y of 
two mRNA vaccine candidates (i.e., BNT162b1, BNT162b2) at various dose levels, 
candidate BNT162b2 was selected for advancement to a pivotal Phase 2/3 safet y and efficacy 
evaluation due to its milder systemic reactogenicity  profile, especially  in older adults.21The 
study was initiated in July  2020 with a target enrollment of 43,998 individuals.22
The US Food and Drug Administration (FDA) announced that regulatory  emergency  use 
authorization (EUA) as well as full approval of any  COVID-19 vaccine will require 
demonstrating prevention of the disease or decrease in its severity in at least 50% of the 
individuals who receive it . In addition, data from Phase 3 studies are required to include a 
median follow -up duration of at least 2 months after completion of the full vaccination 
regimen to assess the vaccine’s benefit -risk profile, especiall y adverse events and cases of 
severe COVID -19 in vaccinated study subjects.23,24The FDA reviewed the available safet y 
data of the Phase 1/2/3 trial from 37,586 participants 16 y ears of age and older and did not 
identify any specific safety  concerns. In addition, the analysis of available efficacy  data from 
36,523 participants 12 years of age and older without evidence of prior SARS -CoV-2 
infection at least 7 day s after receiving the second dose demonstrated 95% efficacy  of the 
vaccine in the prevention of COVID -19 (as confirmed by 8 vs. 162 COVID -19 cases in the 
vaccine and placebo groups, respectivel y).3,4Based on these safety  and efficacy  data, as well 
as a review of manufacturing information regarding product quality  and consistency , the 
FDA determined that the known and potential benefits of the vaccine outweighed the known 
and potential risks for the p revention of COVID -19 in individuals 16 y ears of age and older.4
Therefore on December 11, 2020, the Pfizer -BioNTech COVID- 19 vaccine was granted an 
Emergency Use Authorization (EUA) by the FDA to prevent COVID -19 in individuals 16 
years of age and older.5
With respect to geographic regions other than the US, on December 2, 2020, the United 
Kingdom (UK) was the first country  in the world to grant temporary  authorization for 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073062
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 25of 170emergency  use of the Pfizer -BioNTech COVID -19 vaccine.6On December 21, 2020, the 
European Medicines Agency  (EMA) granted the Pfizer -BioNTech COVID- 19 vaccine a 
conditional marketing authorization (CMA) for use among individuals 16 years of age and 
older throughout all of the European Union’s (EU) 27 member states.7
As required b y the EUA , post-authorization observational studies using real -world data are 
needed in order to assess the association between Pfizer -BioNTech COVID- 19 vaccine and 
pre-determined safet y events of interest (including deaths , hospitalizations, and severe 
COVID-19) among individuals administered the v accine in both the population at large and 
in populations of interest (e .g., pregnant women, immunocompromised individuals, elderly , 
and those with specific comorbidities ).4Post-authorization safety evaluations are important 
for identify ing rare, serious safet yevents of interest in larger populations that may not have 
been detected during clinical trials (either due to sample size or selected study  populations), 
and ensure a favorable benefit- risk ratio post -trial.Pfizer,in collaboration with Health 
ResearchT x(HRTx)and Analy sis Group, herein proposes post-EUAactive safety 
surveillance of safet y eventsof interest in the Department of Defense (DoD) population 
based on the Priority  List of Adverse Events of Special Interest from the Brighton 
Collaboration’s Safet y Platform for Emergency vACcines (SPEAC) Project, the FDA and the 
Centers for Disease Control and Prevention’s (CDC) Advisory  Committee on I mmunization 
Practices (AC IP) enhanced safet y monitoring recommendation . As part of phased allocation 
of COVID -19 vaccinations, all healthcare providers, emergency  services, and public safet y 
personnel within t he DoD popul ation will qualify to receive the COVID -19 vaccine.8This 
safety surveillance stud y willidentify and evaluate rapid, near real- time potential safet y 
signals as sociated with the Pfizer -BioNTech COVID -19 vaccine in the large-scale DoD
Military Health Sy stem (MHS) healthcare database , which includes both administrative 
claims data and clinical data from electronic medicalrecords (EMR). The observed rates of 
safetyevents of interest will be compared to expected rates derived from self -controls,active 
comparators , and contemporary  unvaccinated controls. Part of the methodologies used in this 
study are constructed based on approaches previously  used by the Post -Licensure Rapid 
Immunization Safet y Monitoring (PRI SM) program for the H1N1 vaccine .9
This non- interventional study  is designated as a Post -Authorization Safety Study (PASS) and 
is a commitment to the US FDA and is a Category 3 commitment in the EU Risk 
Management Plan. 
8.RESEARCH QUESTION AND OBJECTIVES
Research question: what aretheincidence rates of safet y eventsof interest (based on adverse 
events of special interest [AESI]) among individuals vaccinated with the Pfizer-BioNTech 
COVID-19 vaccine within the US DoDMHS overall a nd in sub- cohorts of interest, as 
compared to expected rates of those events?
Primary study objectives:
To assess whether individuals in the DoDMHSexperience increased risk of safet y 
events of interest following receipt of the Pfizer -BioNTech COVID -19 vaccine;
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073063
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 26of 170To assess whether sub -cohorts of interest (i.e., pregnant women,
immunocompromised, elderly , individuals with specific comorbidities,individuals
receiving onl y one dose of the Pfizer -BioNTech COVID- 19 vaccine, and individuals 
with prior SARS -CoV-2 infection) in the DoD MHS experience increased risk of 
safety events of interest following receipt of the Pfizer -BioNTech COVID -19 
vaccine.
Secondary study objective :
To characterize utilization patterns of the Pfizer -BioNTech COVID -19 vaccine 
among individuals within the DoD MHS , includingestimating the proportion of 
individuals receiving vaccine, 2-dose vaccine completion rate,anddistribution of 
time gaps between the first and second dose , demographics and health histories of 
recipients, overall and am ong the sub-cohorts of interest .
9.RESEARCH METHODS
9.1.Study Design 
This post-EUAactive safety  surveillance program will employ  arapid-cycle, longitudinal, 
observational cohort study design to provide earl y real-world safet y information . 
The self-controlled risk interval (SCRI) design will be used to sequentially  monitor 
occurrence of safety eventsof interest while controllingfor time-invariant
confounders (such as sex, race, chronic illness, and state). 
Safety events of interest associated with Pfizer -BioNTech COVID -19 vaccinations 
will be sequentially  monitored and compared to two comparator populations : 
(a) Recipients of influenza vaccine in the DoD MHS during 2014/2015 through 
2018/2019 flu seasons, as an active comparator. T his will be particularly helpful to 
assess rarer safety eventsof interest occurring with Pfizer-BioNTech COVID-19 
vaccinations and compared to recipients of influenza vaccine in the DoD MHS 
between2014/2015 to 2018/2019 .9,25
(b) A sample of contemporary  unvaccinated matched controls inthe DoD MHS, as a 
general population comparator group, who will be identified during the same time 
period as individuals receiving Pfizer -BioNTech COVID- 19 vaccine to reflect the 
background rate of current safet y events of interest . This analy sis will be conducted in 
order to evaluate risk as compared to a comparable general population of individuals 
who do not receive anyCOVID-19vaccinein the DoDMHS and provide context for 
interpretation of excess risk identified.
9.1.1.Self-Controlled Risk Interval (SCRI ) Design 
The SCRIdesign uses data from cases(i.e., individuals who experience safety eventsof 
interestfollowing vaccination) to compare the risk interval following vaccination to pre -or
post-vaccination non -risk intervals (“pre-vaccination control interval” and “post -vaccination 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073064
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 27of 170control int erval”) in the same individual.26Whether a pre -or post-vaccination control 
interval is used will depend on the clinical nature, seasonality , and frequency  of the safety 
eventof interest , as described in greater detail below. A length of 42 dayshas been used to 
define therisk interval in SCRIdesign studies for signal detection to ascertain the safet y 
profile of the H1N1 vaccine.9,25The same length of risk interval is proposed here, subject to 
further modification based on clinical input, clinical trial data, biologic plausibility ,and 
published literature . The day of vaccination will only  be included in the risk period for those 
safety eventsof interest for which a same -day occurrence is biologically  plausible 
(e.g.,anaphylaxis).
As some individuals may choose to decline or delay Pfizer-BioNTech COVID-19 
vaccination soon after an illness (known as the “healthy  vaccineeeffect”),27the 
pre-vaccination control interval will exclude the 14 -day period before vaccination.28While 
using a pre -vaccination control period allows for timely  analysis, especially  pertinent for 
rarer safety eventsof interest , a post-vaccination control interval would be more appropriate 
andwill be used for certain safet y eventsof interest for the following reasons (1)a recent 
prior safet y eventof interest might preclude va ccination (i .e.,anaphylaxis), (2) individuals 
might have an underl ying condition that is also a contraindication for vaccination (i .e., 
seizure disorder), or (3)safety eventsof interest and vaccination may be seasonal in nature.29
The time between the risk and control intervals will be determined based on the biological 
mechanism of action for each safety eventof interest assessed , and may  be subject to change 
based on further clinical input. Examplesof the SCRI  design with a pre -vaccination control 
interval and a post -vaccination control interval (in anindividual who only receives the first 
dose of vaccine) arepresented in Figure 1below.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073065
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 28of 170Figure 1. Example of SCRI Design for Assessment of a Safety Eventof Interest with a 
42-day Risk Intervalin an Individual who R eceives O nly One Vaccine Dose, Showing 
Both Pre-and Post-vaccination Control Intervals
*The risk intervalmay include day 0, date of Pfizer -BioNTech COVID -19 vaccination, for some of the safety 
eventsof interest assessed (e.g. , anaphylaxis) . The length of the risk intervalwill vary across each safety event
of interest and may be subject to change based on clinical input .Note that some individuals may not receive the 
complete course of vaccination, and thus may only receive the first dose of vaccine. This is represented in 
Figure1while Figure 2represents an example where the complete course with 2 doses are received.
Two doses of the Pfizer -BioNTech COVID -19 vaccine are recommended 3 weeks apa rt.  
This studyprogram will monitor safety  eventsof interest that occur after dose 1 and before 
dose 2 (i.e., during risk interval 1), after dose 2 (i .e., during risk interval 2), and aggregate for 
doses1 and 2 (i .e., risk interval 1 + risk interval 2), respectively , for individuals receiving 
both doses. 
For individuals whoreceive two doses of the vaccine, two separate control intervals will be 
defined to correspond to the risk interval associated with each dose (re gardless of whether 
pre-or post-vaccination control intervals are used). See Figure 2below for a n example in an 
individual who receives two doses of Pf izer-BioNTech COVID -19 vaccine, with the second 
dose received 21days after the first. S afety eventsof interest that occur during the 
overlapping period of risk interval 1 and risk interval 2 (shown in gray  shading in Figure 2) 
may be flagged for separate anal yses to discern the additive effect of Pfizer -BioNTech 
COVID-19 vaccine dose 1 and dose 2.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073066
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 29of 170Figure 2.Example of SCRI D esign with Overlapping Risk Intervals when Two Doses 
of Pfizer-BioNTech COVID -19 Vaccine are A dministered, Showing a 
Pre-and Post-vaccination Control Interval
9.1.2.Active Comparator Design
In the active comparator design, the frequency  of safety eventsof interest among individuals 
who received Pfizer -BioNTech COVID-19 vaccine from December 11, 2020 onward will be 
compared with the event frequency  among individuals who received the seasonal influenza 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073067
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 30of 170vaccination in five prior seasons, between2014/2015 through2018/2019. Data in peri -
COVID time periods from January  2020 to present areexcluded because of pandemic-
associated underutilization of health resources and underreporting of medical events. The 
same risk interval length (e.g.,42 days)will be used to evaluate safety eventsof interest 
following vaccination with Pfizer-BioNTech COVID-19 vaccine and to assess safet y events
of interest occurring after vaccination for seasonal influenza in prior seasons. The observed 
number of safet y eventsof interest for Pfizer-BioNTech COVID-19 vaccine will be 
compared to the expected number calculated for the influenza vaccine in past seasons.9
9.1.3.Contemporary Unvaccinated Control Design
A contemporary  matched comparator cohort of individuals who are not vaccinated with any
COVID-19vaccine will be identified and serve as contemporary  unvaccinated controls
during the period December 11, 2020 onward. These individuals will be randomly  sampled 
from the general population to match the baseline demographic and clinical characteristics of
individuals who receive the Pfizer-BioNTech COVID- 19 vaccine in order to ensure that the 
populations are comparable. Specificall y, issues of non- comparability  between vaccinated 
and contemporary  unvaccinated controls will be addressed via exact matching (1:N, but no 
greater than 1:4 due to diminishing gains in efficiency  with higher ratios)30on age, sex, state, 
and key conditions known to increase the risk of severe COVID-19 (e.g., cancer, obesity , 
pregnancy , smoking, t ype 2 diabetes, etc.) .12In addition, patients will be matched on whether 
or not they  received a seasonal influenza vaccine and the timing of vaccination with seasonal 
influenza vaccine in relation to vaccination with Pfizer -BioNTech COVID -19 vaccine , as 
receipt of influenza vaccine close to COVID- 19 vaccination may  impact occurrence of safety 
eventsof interest . Additional covariates of clinical significance will be adjusted for via 
propensity  score (PS) matching. M atched sample s allow one to estimate the treatment effect 
by directly comparing the outcome(s) of interest between the vaccinated and unvaccinated
matched sample.31This approach parallels that of a randomized control trial , where the 
distribution of covariates is similar between treatment arms.31
The index date for the contemporary  unvaccinated controls will be selected based on the 
distribution of index dates in the vaccinated cohort. If vaccination is associated with a regular 
healthcare encounter (i.e. , an evaluation and management code or similar), the contemporary  
unvaccinated control will be required to have an encounter within 30 day s of the assigned 
index date, and the date of encounter will be set as the index date to ensure comparability  of 
covariate measurement.
Additionally , there is a theoretical risk that vaccination could result in vaccine -associated 
enhanced disease (VAED), i .e., exacerbation of viral infection resulting in more severe 
illness or specific clinica l manifestations upon exposure to SARS -CoV-2 as compared with 
what would have been experienced without vaccination. To evaluate this potential risk and 
identify such a signal, patterns of serious COVID- 19 illness will be evaluated between 
vaccinated and contemporary  unvaccinated controls. The self -controlled design would not be 
appropriate for evaluating VAED as the timing of exposure to the wild -type virus would 
impact this outcome. I ncidence rates will be calculated ,compared, and stratified by  
categories of age and presence or absence of risk factors for severe disease. An apparent 
excess of serious COVID -19 illness in the reference populations, such as young individuals 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073068
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 31of 170and/or individuals without risk factors (i.e., individuals at low risk for severe disease) may  be 
indicative of VAED, and would warrant further evaluation (e.g. , chart review).
9.1.4.Study Period 
The studywill be conducted for a period of 30 months , starting on December 11, 2020 
onward, with data collection concluding on J une 10, 2023. 
9.2. Setting 
The exposed population will be kept as broad as possible in order to capture safety eventsof 
interestthat occur among all individuals receiving Pfizer-BioNTech COVID-19 vaccine.
9.2.1. Inclusion Criteria
Record of at least one dose of Pfizer-BioNTech COVID-19 vaccine in the period of 
December 11, 2020 to present; or
Record of at least one dose of seasonal influenza vaccine during prior flu seasons, 
from 2014/2015 to 2018/2019 (applies to active comparator s only); or
No recordof anyCOVID-19 vaccine( applies to the contemporary  unvaccinated 
controlsonly); and
At least 1 y ear of continuous enrollment (i .e., the baseline period) prior to dateof 
Pfizer-BioNTech COVID-19 vaccination, seasonal influenza vaccination, or matched 
index date for unvaccinated controls.
9.2.2.Exclusion criteria
Individuals who receive at least one dose of COVID -19 vaccine from a manufacturer 
other than Pfizer- BioNTech will be identified and reported, but they  will be excluded 
from further anal ysis.
9.2.3.Subgroups
Safety surveillance may  be conducted for subgroups of interest, including, but not limited to:
Pregnant women;
Immunocompromised individuals;
Different age groups, with a focus on the elderl y(e.g., <35, 35 to<45, 45 to <55, 55 
to<65, 65 to <75, >75);
Individuals with specific comorbidities;
Individuals receiving only  one dose of Pfizer-BioNTech COVID-19 vaccine ;
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073069
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 32of 170Individuals with prior SARS -CoV-2 infection based on medical history  or 
pre-vaccination serology .
Additional subgroups of interest will be assessed as additional inform ation becomes 
available from ongoing clinical trials, Vaccine Adverse Event Reporting Sy stem (VAERS), 
and other sources that will inform the Pfizer-BioNTech COVID -19 vaccine safet y profile.
9.3.Variables
9.3.1.Exposure of Interest
Administration of Pfizer-BioNTech COVID-19 vaccine post-EUA approval will be identified 
basedon the following:
Current Pr ocedural Terminology  (CPT) code 91300 (Severe acute respiratory  
syndrome coronavirus 2 (SARS -CoV-2) (coronavirus disease [COVID-19]) vaccine, 
mRNALNP, spike protein, preservative free, 30 mcg/0.3mL dosage, diluent 
reconstituted, for intramuscular use ) and associated vaccine administration HCPCS
codescorresponding to the first dose: 0001A (ADM SARS -CoV-2 30 mcg/0.3mL  
1st),and the second dose: 0002A (A DM SARS -CoV-2 30 mcg/0.3mL  2nd);10,11OR
10and 11-digit National Drug Codes (NDCs) 59267-1000-1(corresponds to first 
dose), 59267-1000-01(corresponds to second dose) ;OR
Immunization records that contain data on vaccine code descriptor, v accine 
manufacturer (i.e.,Pfizer), lot number, injection site ,and date(s)of immunization .10
Relevant codes will be continuously  reviewed and amended if new codes are added.
Person-time at-risk exposure to the first dose onl y, overlapping first and second doses, and 
second dose onl y will be anal yzed separatel y.
Administration of the seasonal influenza vaccine during 2014/2015 through 2018/2019 flu 
seasons will be identified based on the following:
CPT codes 
o90654 (Influenza virus vaccine, trivalent (IIV3), split virus, preservative -free, 
for intradermal use) ;OR
o90656 (Influenza virus vaccine, trivalent (IIV3), split virus, preservative free, 
0.5 mL dosage, for intramuscular use) ;OR
o90658 (Influenza virus vacc ine, trivalent (IIV3), split virus, 0.5 mL  dosage, 
for intramuscular use );OR
10 and 11-di git NDCs ;OR
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073070
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 33of 170Immunization records that contain data on vaccine code descriptor, vaccine 
manufacturer, lot number, injection site, and date(s) of immunization. 
9.3.1.1.Pfizer-BioNTech COVID -19 Vaccine Groups of Interest
While the primary  vaccination group of interest is all individuals receiving Pfizer-BioNTech 
COVID-19 vaccine (irrespective of receipt of seasonal influenza vaccination), additional 
subsets of the study population will be studied, similar to the PRI SM safety surveillance 
programof H1N1 vaccine safety:9
Cohort A: Individuals vaccinated with Pfizer-BioNTech COVID- 19 vaccine who did not 
receive the influenza v accineduring the flu seasonin which COVID -19 vaccination 
occurred;
Cohort B: Individuals vaccinated with Pfizer-BioNTech COVID-19 vaccine who received 
the seasonal influenza vaccine at least 42 day s prior to COVID -19 vaccination during the 
same flu season in which COVID -19 vaccination occurred;
Cohort C: Individuals vaccinated with Pfizer-BioNTech COVID-19 vaccine who received 
the seasonal influenza vaccine within 42 day s before or any  time after COVID -19 
vaccination during the same flu season in which CO VID-19 vaccination occurred;
Cohort D: Individuals vaccinated with both Pfizer-BioNTech COVID -19 vaccine and the 
seasonal influenza vaccine on the same day .
The following sub -cohorts will be assessed for each of the Cohorts A -D:
Individuals vaccinated with only 1 dose (i.e., incomplete course) of Pfizer-BioNTech 
COVID-19 vaccine;
Individuals vaccinated with 2 doses (i .e., complete course) of Pfizer-BioNTech 
COVID-19 vaccine.
9.3.2.Baseline Characteristics
The following data elements regarding baseline demographic and clinical characteristics will 
be assessed based on a1-yearbaseline period priortothe date of vaccination with Pfizer-
BioNTech COVID -19 vaccine, date of seasonal influenza vaccination for active comparators, 
and assigned index date for contemporary  unvaccinated controls. Depending on the attrition 
rate, the length of the baseline period may  be modified to 6 months. All diagnoses, 
procedures, and medications will be identified by  the International Classification of Diseases, 
Tenth Revision, Clinical Modification ( ICD-10-CM)diagnosis codes, ICD-10-PCS 
(procedure coding s ystem) codes, ICD-10-CM Current Procedural Terminology  (CPT), and 
Logical Observation Identifiers Names and Codes (LO INC) laboratory  results,orHealthcare 
Common Procedure Coding S ystem (HCPCS) procedure codes, and generic drug names, as 
appropriate (see Appendix Table 1). The follow ing demographic and clinical characteristics 
will be assessed:
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073071
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 34of 170Demographic s:
Age
Sex
State
Sponsor service (e .g., Air Force, Army, Coast Guard, M arine Corps, Navy)
Category of beneficiary (e.g., active dut y, retiree, active guard/reserve, dependent )
Clinical characteristics:
Smoking status
Body mass index (BMI)
History of anaphylaxis/allergic reactions
Previous anaph ylaxisof vaccine component 
History of hospitalizations
Pregnancy
Charlson c omorbidity  index (CCI)
Selectedcomorbidities
oAutoimmune disease
oAsthma
oBleeding diathesis or condition associated with prolonged bleeding
oCancer
oCardiovascular conditions
oChronic kidney  disease/dialy sis
oChronic obstructive pulmonary  disease (COPD)/interstitial lung disease
oDiabetes mellitus
oDown syndrome
oSicklecell disease
oHepatitis B virus ( HBV)
oHepatitis C virus ( HCV)
oHuman immunodeficiency virus (HIV)
oHyperlipidemia
oHypertension
oLiver disease
oNeurological disease
oOther immune deficiencies
oSolid organ transplant
oVenous thromboembolism (VTE)
Concurrent immuniz ations
oSeasonal influenza vaccine
oTetanus diphtheria and pertussis (Tdap or Td )
oChickenpox ( varicella)
oShingles (herpes zoster recombinant and/or live )
oHuman papillomavirus ( HPV)
oPneumococcal conjugate
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073072
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 35of 170oPneumococcal pol ysaccharide
oHepatitis A
oHepatitis B
oMeningococcal conjugate (MenACWY) and serogroup B meningococcal 
(MenB)
oHaemophilus influenza ty pe b
9.3.3.Outcomes
The safety eventsof interest for active surveillance were identified based on thePriority List 
of Adverse Events of Special Interest from the Brighton Collaboration’s Safety  Platform for 
Emergency  vACcines (SPEAC) Project and the FDA and Centers for Disease Control and 
Prevention (CDC)enhanced safet y monitoring recommendations.32,33Endpoints of special 
interest in signal detection, as noted b y CDC’s Advisory Committee on Immunization 
Practices (AC IP) are denoted in i talics.33The list of safet y eventsof interest may be revised 
over the course of the study , and if unanticipated potential safety eventsof interest are 
identified during the course of surveillance , they will be added to the list and included in the 
analyses.See Appendix Table 2for the operational definitions of the outcome variables
based on ICD -10-CM diagnosis codes and LOINC laboratory  codes, which may be refined as 
the study progresses based on additional availab le information and the published literature
(e.g., frequency  of ICD-10 codes) . Outpatient (including emergency  department )and/or
inpatient setting s will be used to identify  safety eventsof interest , depending on the t ype of 
event. The specific encounter setting considered for each safet y eventof interest is 
summarized in Table 1.Any record of death will be captured, regardless of whether the
individual died in a healthcare or non -healthcare setting. The following safety  eventsof 
interestwill be assessed:
Neurologic :
G
eneralized convulsions/seizures
Guillain-Barré syndrome (GBS)
Aseptic meningitis
Encephalitis/encephalomyelitis
Other acute dem yelinating diseases
Transverse m yelitis (TM)
Multiple sclerosis (MS)
Optic neuritis (ON)
Bell’s pals y
Immunologic :
Anaphylaxis
Vasculitides
Arthritis and arthralgia/joint pain
Multisystem inflammatory syndrome in adults (MIS-A)
Kawasaki disease (KD)
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073073
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 36of 170Fibromyalgia
Autoimmune thy roiditis
Cardiac:
Myocarditis
Pericarditis
Acute myocardial infarction (AMI)
Hematologic :
Thrombocytopenia
Disseminated intravascular coagulation (DIC)
COVID-19(for all COVID-19-related safet y eventsof interest listed below, a n inpatient
diagnosis of COVID -19 will be required in combination with the codes or laboratory  values 
specified in Appendix Table 2; in addition, COVID -19 related safety eventsof interest will 
only be evaluated using data from 2020 onward using the SCRI  design and contemporary  
unvaccinated controls ):
Severe COVID -19 disease
Microangiopath y
Heart failure and cardiogenic shock
Stress cardiom yopathy
Coronary  artery disease (CAD)
Arrhythmia
Deep vein thrombosis (DVT)
Pulmonary  embolus
Cerebrovascular hemorrhagic stroke
Cerebrovascular non- hemorrhagic stroke
Limb ischemia
Hemorrhagic dis ease
Acute kidney  injury
Liver injury
Chilblain-like lesions
Single organ cutaneous vasculitis
Erythema multiforme
Other:
Pregnancy  outcomes (note that outcomes related to delivery  will only  be assessed 
using active comparators and contemporary  unvaccinated controls rather than SCRI  
since delivery will only occur at a single ti me point)
Death
Narcolepsy/cataplexy
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073074
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 37of 170Non-anaphylactic allergic reactions
Appendicitis
The risk and control intervals selected for the SCRI anal ysis for each safety eventof interest 
are based on biological plausibility  and precedents in the published literature ( Table 1). A 
safety eventof interest will be co unted if it can be assigned to 1) the risk interval following 
Pfizer-BioNTech COVID-19 vaccination , 2) the pre -vaccination control interval, 3) the post -
vaccination control interval, or 4) the risk interval for the active comparators receiving 
seasonal influenza vaccine ,and 5) risk interval for the contemporary  unvaccinated controls. 
Events outside the intervals will not be counted. Only the individual’s first instance of a 
safety eventof interest following a specified clean window (i.e., the occurrence -free baseline 
periodused to define incident outcomes during which individuals enter the study  cohort onl y 
if the safet y eventof interest did not occur during this period)will be included ; this means 
that if a safety eventof interest is identified but diagnosis codes (or laboratory  values in the 
case of select safety eventsof interest )corresponding to the safet y eventof interest are also 
observed dur ing the clean window, it will not be counted. The duration of the pre -specified 
window will differ b y safety eventsof interest in order to rule out pre- existing events. This 
approach is consistent with the FDA’s COVID -19 Vaccine Safet y Surveillance Projec t.17By 
way ofexample, safety event sof interest for the SCRI  design can be considered in the 
following way s:
If asafety eventof interest occurs in the individual ’s pre-vaccination control 
interval and th ere are no other diagnosis codes for the same safety  eventof 
interestin the clean window (e .g., 1-year prior to that date) , the safet y eventof 
interestshould be assigned to the pre -vaccination control interval.
oIf a safetyevent of interest occurs in the pre-vaccination control interval
but another diagnosis code for the same safety event of interest is 
identified during the risk interval, the n thesafety eventof interest will not 
be assignedtothe risk interval and will only be assigned to the pre -
vaccination control interval as it will have occurred in the required clean 
window preceding the risk interval . However, if an outpatient safety event
of interest occurs in the clean window and an inpatient occurrence for the 
same type of safety eventof interest occurs in the risk interval, the 
inpatient occurrence will be counted in order to capture event 
exacerbation.
If a safety eventof interest occurs in the risk interval and there are no other 
diagnoses for the same safety eventof interest in the cl ean window (e .g., 1-year
prior to this date), which also includes the pre- vaccination control interval, the n 
thesafety eventof interest willbe assigned to the risk interval .
The same approach will be applied for the post -vaccination control intervals .
The risk intervals for outcome evaluation for the active comparators (i.e., individuals who 
received seasonal influenza vaccination )and contemporary  unvaccinated controls (i .e., 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073075
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 38of 170individuals who did not receive the Pfizer -BioNTech COVID- 19 vaccine) will bethe same as 
for the individuals who received Pfizer -BioNTech COVID-19 vaccine. 
However, it is possible that some safety eventsof interest do not have a precise time interval 
from which to evaluate risk, for example if biological plausibility is unknown or the 
diagnostic time window is more delay ed than anticipated. In these cases, misspecification of 
the risk (and control) intervals could result in misclassification and introduce bias, often 
toward the null. For instance, the assumption of a longer risk interval than is true may  result 
in “washing out” the signal, and an erroneousl y short risk interval may  similarly  result in 
underestimation of effect when using post -vaccination time intervals for self -control. To 
address this, sensitivity  analyses may be conducted with vary ing risk intervals (longer as well 
as shorter) in order to increase the likelihood that the safet y risk is detected accuratel y. 
Additionally , if further refinement and evaluation is necessary , temporal scan statistics may  
beused to empiricall y identify the at-risk time interval b y evaluating clusters of safety events
of interest . This will be further described in the statistical anal ysis plan (SAP).
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073076
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 39of 170Table 1.Outcome algorithms for SCRI analysis , with risk and control intervals
Safety event of Interest* Setting
(Inpatient [IP], 
Outpatient [OP])Clean window Pre-vaccination 
control interval 
(days)Risk interval 
(days)Post-vaccination 
control interval 
(days)
Neurologic
Generalized convulsion/seizures9IP or OP96 months N/A 0-14 15-29
GBS9,25IP, primary  position 
only171 year N/A 1-42 43-84
Aseptic meningitis34IP only171 year N/A 1-42 43-84
Encephalitis/ encephalomyelitis9IP only171 year -56 through -15 1-42 N/A
Other acute dem yelinating diseases9IP or OP91 year -98 through -15 1-42 N/A
TMaIP only171 year -98 through -15 1-42 N/A
MS9,25IP or OP91 year -98 through -15 1-42 N/A
ON9,25IP or OP91 year -98 through -15 1-42 N/A
Bell’s pals y9,25IP or OP171 year -56 through -15 1-42 N/A
Immunologic
Anaphylaxis9,25IP or OP176 months N/A 0-2 7-9
VasculitideseIP only 1 year N/A 1-28 29-56
Arthritisand arthralgia /joint paincIP or OP 1 year N/A 1-42 43-84
MIS-AbIP only171 year N/A 1-42 43-84
KD35IP only351 year N/A 1-28 29-56
FibromyalgiacIP or OP 1 year N/A 1-42 43-84
Autoimmune thy roiditiscIP or OP 1 year N/A 1-42 43-84
Cardiac
Myocarditis9,25IP or OP171 year -56 through -15 1-42 N/A
Pericarditi s9,25IP or OP171 year -56 through -15 1-42 N/A
AMIdIPonly171 year -56 through -15 1-42 N/A
Hematologic
Thrombocy topenia34IP or OP171 year N/A 1-42 43-84
DICeIPonly171 year N/A 1-42 43-84
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073077
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 40of 170Table 1.Outcome algorithms for SCRI analysis , with risk and control intervals
Safety event of Interest* Setting
(Inpatient [IP], 
Outpatient [OP])Clean window Pre-vaccination 
control interval 
(days)Risk interval 
(days)Post-vaccination 
control interval 
(days)
COVID-19(for all COVID -19-related safet y events of interest listed below, an inpatient diagnosis of COVID -19 will be required in 
combination with the codes or laboratory  values specified in Appendix Table 2; in addition, COVI D-19 related safety events of interest will 
only be evaluated using data from 2020 onward using the SCRI  design and contemporary  unvaccinated controls )
Severe COVID -19 diseasebIP only 1 year N/A 1-42 43-84
Microangiopath yeIP only 1 year N/A 1-42 43-84
Heart failure and cardiogenic shockdIP only 1 year -56 through -15 1-42 N/A
Stress cardiom yopathydIP only 1 year -56 through -15 1-42 N/A
CADdIP only 1 year -56 through -15 1-42 N/A
ArrhythmiadIP only 1 year -56 through -15 1-42 N/A
DVTeIP or OP171 year N/A 1-42 43-84
Pulmonary  emboluseIP or OP171 year N/A 1-42 43-84
Cerebrovascular hemorrhagic stroke9IP only171 year N/A 1-42 43-84
Cerebrovascular non -hemorrhagic 
stroke9IP only171 year N/A 1-42 43-84
Limb ischemiaeIP only 1 year N/A 1-42 43-84
Hemorrhagic diseaseeIP only 1 year N/A 1-42 43-84
Acute kidney  injurygIP only 6 months N/A 1-42 43-84
Liver injurygIP or OP 1 year N/A 1-42 43-84
Chillblain -like lesionseIP or OP 1 year N/A 1-42 43-84
Single organ cutaneous vasculitiseIP only 1 year N/A 1-42 43-84
Erythema multiformefIP only 6 months N/A 1-2 8-9
Other
Pregnancy  outcomes36,hIP or OP 1 year N/A 1-42 43-84
Narcoleps y and cataplexyaIP or OP171 year -98 through -15 1-42 N/A
Non-anaphylactic allergic reactions9,25IP or OP96 months N/A 1-2 8-9
Appendicitis37IPonly176 months N/A 0-42 43-84
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073078
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 41of 170Table 1.Outcome algorithms for SCRI analysis , with risk and control intervals
Safety event of Interest* Setting
(Inpatient [IP], 
Outpatient [OP])Clean window Pre-vaccination 
control interval 
(days)Risk interval 
(days)Post-vaccination 
control interval 
(days)
*Safety events of interestare based on the Priority List of Adverse Events of Special Interest from the Brighton Collaboration’s Safety Platform for Em ergency 
vACcines (SPEAC) Project, the FDA and the Centers for Disease Control and Prevention’s (CDC) Advisory Committe e on Immunization Practices (ACIP) enhanced 
safety monitoring recommendations.
Notes: 
aPublished risk and control intervals for demyelinating diseases and cranial disorders w ere applied to TM and narcolepsy/catap lexy.
bAs severe COVID -19 ranges from sev ere pneumonia, acute respiratory distress syndrome, and multisystem organ failure/MIS -A, a 1-42 day risk interval was applied 
in order to capture the 14 -day incubation period of the disease and 4 -5 day period from exposure to symptom onset.
c Published ris k and control intervals for autoimmune disorders w ere applied to similar autoimmune rheumatic conditions (i .e., fibromyalgia and autoimmune 
thyroiditis).
dPublished risk and control intervals for myocarditis and pericarditis w ere applied to other cardiova scular conditions (i .e., heart failure and cardiogenic shock, stress 
cardiomyopathy, CAD, arrhythmia, AMI). 
eSimilar risk and control intervals were applied to all cardiovascular and hematological disorders characterized by damage to the blood vessels an d/or arteries and 
clotting (i .e., microangiopathy, DVT, pulmonary embolus, limb ischemia, hemorrhagic disease, DIC, chilblain -like lesions). The published risk and control intervals for 
KD were applied to vasculitides given that KD is a type of medium and small-vessel vasculitis. 
fPublished risk and control intervals for non -anaphylactic allergic reactions were applied to hypersensitivity disorders (i .e., erythema multiform e).
gRisk intervals of 42 days were applied for acute kidney injury and liver injury to be consistent with other COVID -19 related safety events of interest.
hPregnancy outcome of eclampsia/pre -eclampsia only will be assessed using SCRI. Other pregnancy outcom es that are related to delivery (i.e., post -partum hemorrhage, 
premature rupture of membranes, chlorioambionitis, placental abruption, and cesarean section ) will be evaluated using the active comparator and contemporary 
unvaccinated control designs.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073079
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 42of 1709.4.Data Source 
This study  will be conducted in the DoD Military Health Sy stem (MHS)database. The MHS 
is a single pay er system that provides medical coverage and pharmacy  benefits for active 
duty and retired military  members, civilian DoD personnel, and their families (beneficiaries). 
Veterans who receive medical coverage through the Veterans Health Administration are not 
included. There are 9.6 million beneficiaries included in the MHS, of whom 1.4 million
(14.6%)are active dut y, 1.7million (17.7%) are active duty family members, 392,000 (4.1%)
are national guard and reservemembers, 609,000 (6.3%) are family  members of national 
guard and reserve members, and 5.5 (57.3%) million are retirees and their family  
members.13,14The DoD also includes 64 hospitals, hundreds of clinics, 25,000 uniformed 
physicians, and 400,000 community  network providers. The population within the MHS is 
demographicall y representative of the US overall, with slight over -representation of persons 
>65 years of age (20.1% in DoD MHS vs. 12.9% in the general US population).15The gender 
distribution is approximately  49% female and 51% male.
The MHS provides care in two way s: direct and purchased care. Direct care is provided to 
beneficiaries within a global network of military  hospitals and clinics. MHS uses an EMR 
that captures administrative and encounter information, as well as a detailed clinical record. 
Purchased care ( through TRICARE, the DoD health insurance ) is provided to beneficiaries 
by civilian providers who are paid via fee -for-service reimbursements or managed care 
contracts. MHS collects and verifies encou nter and claims records for each service.38
All healthcare encounters, whether received through direct or purchased care, are archived, 
validated, and normalized within a central MHS Data Repository (MDR). For those receiving 
direct care, all medical services are captured, as well as clinical details, diagnostic and 
laboratory  test ordered, and test results. I nformation is collected at the point of care and 
available almost immediately . Direct care accounts for approximately  40% of care within the 
MHS, though this proportion may  change over time .39The MHS purchased care data include 
records of ph ysician services, hospital care (in patient and outpatient), emergency  room visits, 
home health, hospice, and other services. Claims for laboratory and diagnostic testing are 
collected; however, unlike direct care, the results of these tests are not captured.  
Prescription data from both di rect and purchased care are captured within MHS’s electronic 
medication ordering s ystem called the Pharmacy  Data Transaction System (PDTS). 
Dispensing details and physician-administered medication events are coded electronically  
and include the prescribed drug name and NDCs, dose, therapeutic class, quantity, refills,and 
fill location. Vaccination records include data on the vaccine code descriptor, vaccine 
manufacturer, lot number, injection site, and date(s) of immunization.
Each individual is assigned a unique identification number to allow for longitudinal 
follow-up to provide comprehensive information about the individual and his/her medical 
encounters. The MHS is an appropriate data source to evaluate the safet y of the Pfizer-
BioNTech COVID-19vaccine,as the vaccine will bedistributed through government 
facilities (including MHS facilities) as part of initial distribution ,andanalysis of DoD data 
will provide earl y data on the safety of the vaccine .The DoD prioritized vaccine distribu tion 
to healthcare workers and emergency  services personnel, personnel performing activities 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073080
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 43of 170associated with critical national capabilities, select deploy ing individuals, other critical and 
essential support, individuals at the highest risk for developing severe illness from COVID -
19, and adults age 75 and older.16Specifically , as part of Phase 1a, all healthcare providers, 
emergency  services, and public safet y personnel within the DoD population will qualify  for 
the vaccine.8Phase 1b.1 and 1b.2 will include those considered critical for national 
capabilities and individuals preparing to deplo y outside of the US, respectively.8
9.5.Study Size
The sample size achieved will depend on the number of individuals administered the Pfizer -
BioNTech COVID-19 vaccine within the DoDMHSduring the stud y period, which will 
increase over time with subsequent anal yses. Specifically, the data will be refreshed on a 
repeated basis , e.g., monthly  (to be stipulated in the SAP), and a continuous sequential test 
procedure will be used to reevaluate data according to this schedule. Preliminary  estimates 
for the number of individuals in the DoD MHS who received the Pfizer -BioNTech COVID -
19 vaccine will be reported in the SAP.
As a result of the ability  to perform near -real-time analysis, the risk interval (and post-
vaccination control interval, for applicable safety eventsof interest ) may have only partially 
elapsed in some cases. To account for this, we will use methods adopted in previous 
studies,9,28,40whereby risk intervals will be scaled (or truncated) in order to ensure an 
equivalent length (or a fixed ratio) of time is assessed between the control and risk intervals.  
The same approach will also be applied for contemporary  unvaccinated controls.
9.5.1.Power
Power calculations for the rapid cycle analysis (RCA)approaches proposed for safety event 
of interest signal detection will be conducted according to the methods of Kulldorff et al .41,42
Table2illustrates the estimated power for the RCA approach using the Poisson -based
maximized sequential probability  ratio test ( MaxSPRT), and provides an overview of the 
power required to detect vary ingrelative risk (RR)estimates with an alpha level of0.0 1. T 
denotes the expected number of safety eventsof interest to occur during the risk intervalof 
interest(Table2and Table 3). Power of ≥ 80% is typically desirable in drug safet y research. 
Usually the FDA views a RR of >3 as meaningful, so this has been to for power calculations 
here.43As anexample, a s shown in Table2, the surveillance s ystem would h ave sufficient 
power (80.0%) to detect an increased risk of safety eventsof interest associated with the 
Pfizer-BioNTech COVID- 19 vaccine b y 3fold when the expected number of safety eventsof 
interest reaches 6 events. 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073081
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 44of 170Table2.Estimated S tatistical Power for the Poisson -based MaxSPRT41
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073082
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 45of 1709.6.Data Management
Data for this stud y will be stored and extracted from the DoDMHSdatabase (previousl y
described in Section 9.4) that contain information about demographics, vaccinations,
procedures, diagnoses, and death.  
9.6.1.Case report forms (CRFs)/ Electronic data record
As used in this protocol, the term CRF should be understood to refer to either a paper form or 
an electronic data record or both, depending on the data collection method used in this study .
A CRFisrequired and should be completed for each included patientin the signal 
verification phase that requires EMR and chart review (see Section 9.7.4.3). The completed 
original CRFs should not be made available in any form to third parties, except for 
authorized representatives of Pfizer or appropriate regulatory  authorities, without written 
permission from Pfizer. Analy sis Group shall ensure that the CRFs are securely  stored on 
DoDservers in an e ncrypted electronic and/or paper] form and will be password protec ted or 
secured in a locked room to prevent access by unauthorized third parties.
Data abstractors haveultimate responsibility  for the collection and reporting of all clinical, 
safety, and laboratory data entered on the CRFs and any other data collection forms (source 
documents) and ensuring that they are accurate, authentic/original, attributable, complete, 
consistent, legible, t imely (contemporaneous), enduring, and available when required.  The 
CRFs must be signed b y the responsible part y abstracting medical records and/or 
adjudicating the endpoints to attest that the data contained on the formsare trueand accurate 
based on their review of the data . Any corrections to entries made in the CRFs or source 
documents must be dated, initialed, and explained (if necessary) and should not obscure the 
original entry .
The source documents are the hosp ital or the ph ysician's chart. In these cases, data collected 
on the CRFs must match those charts. 
9.6.2.Record retention
To enable evaluations and/or inspections/audits from regulatory  authorities or Pfizer, 
Analysis Group agrees to keep all study -related records, which includes study  documents a nd 
deliverables such as the protocol, SAP, aggregated results tables, SAS programming files, 
and study  report. The records should be retained by Analysis Group according to local 
regulations or as specified in the vendor contract, whichever is longer. Analysis Group must
ensure that the records continue to be stored securely  for so long as they  are retained.
If Analysis Groupbecomes unable for any  reason to continue to retain study  records for the 
required period, Pfizer sho uld be prospectivel y notified. The study  records must be 
transferred to a designee acceptable to Pfizer.
Study records must be kept for a minimum of 15 years after completion or discontinuation of 
the study, unless Analysis Group and Pfizer have expressly  agreed to a different period of 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073083
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 46of 170retention via a separate written agreement. Record must be retained for longer than 15 years 
if required b y applicable local regulations.  
Analysis Group must obtain Pfizer's written permission before disposing of any  records, even 
if retention requir ements have been met.
9.7.Data Analysis
Detailed methodology  for summary  and statistical anal yses of data analyzedin this study  will 
be documented in a SAP, which will be dated, filed, and maintained b y the sponsor. The SAP 
may modify the plans outlined in th e protocol; any major modifications of primary  endpoint 
definitions or their anal yses would be ref lected in a protocol amendment. The SAP will also 
provide additional detail regarding the evaluation of a threshold of excess risk for each of the 
safety events of interest. Consistent with the approach of Kulldorff et al., t his will be 
determined based on background incidences for each event (e .g., based on historical 
influenza vaccinated active comparator cohort data to be evaluated during the study ), in 
addition to pre -specified significance level (e .g., alpha=0.01 or 0.05) and power.41This 
information, in conjunction with a clinically  meaningful RR (e .g., 2 or 3) and the expected 
upper limit of events under the null hypothesis will allow for the calculation of critical values 
of each safet y event of interest using the MaxSPRT method. Greater power (e .g., 80%) is 
also a natural criterion to use when selecting the upper limit on the length of surveillance, 
and in turn, the expected number of events to occur, although there is ultimately  a tradeoff 
between that power and the time allowed t o identify  the expected number of events to occur.
Data analy ses will be conducted using SAS Enterprise Guide version 7.1 (SAS I nstitute Inc., 
Cary, NC) or R Version 3.5.3 or its latest version (R Core Team, Vienna, Austria). In 
addition, SaTScan will als o be used to conduct specific temporal anal yses.
9.7.1.Identification of Contemporary Unvaccinated Controls
Exact and PS matching will be used to identify  contemporary  unvaccinated controls with 
similar baseline characteristics to individuals who receive Pfizer -BioNTech COVID-19 
vaccine. The PS will be defined as the probability  of receiving the Pfizer -BioNTech COVID-
19 vaccine versus not receiving the vaccine conditional on observed baseline characteristics . 
The PS model will be estimated using logistic regression , by regressing receipt of vaccine on 
baseline covariates. In this way , both exact and PS matching will be used to balance the 
distribution of observed baseline covariates between vaccinated and unvaccinated 
individuals, with exact matching used specific ally for the most important prognostic factors.
A matching ratio of 1:N, but no greater than 1:4 will be used due to diminishing gains in 
efficiency  with higher ratios .24The exact matching ratio will be determined pending the 
available sample size of contemporary  unvaccinated controls. A balanced nearest neighbor 
matching approach (with a caliper) will be used in order to require that controls alternate 
between having PS great than and less than the matched vaccinated individual in order to 
avoid contempor ary unvaccinated controls being consistently  clustered to one side of the 
matched vaccinated individual .44This approach has been shown to result in lower bias than 
the more commonly  used greedy  matching approach. If a sufficient number of controls 
cannot be obtained b y matching on these covariates, the n a variable matching ratio will be 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073084
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 47of 170considered, the most important confounders will be chosen for matching ,and the number of 
matching covariates maybe reduced in order to ensure sufficient sample size can be obtained 
for the study.
9.7.2.Baseline Characteristics
Baseline demographics and clinical characteristics for individuals receiving Pfizer -BioNTech 
COVID-19 vaccine, individuals who received seasonal influenza vaccination , and 
contemporary  unvaccinated controls will be summa rized using descriptive statistics, 
consisting of the mean and standard deviation (SD) and median (interquartile range [IQR]) 
values for continuous variables and frequency  distributions for categorical variables.
Incidence rates (i.e., per-patient per -month) for prior hospitalizations may  be calculated as 
the number of events divided by  person-time of observation since the length of the baseline 
period may  vary between individuals. Standardized differences will be calculated between 
individuals who received the Pfizer BioNTech COVID -19 vaccine and with active 
comparator s who received seasonal influenza vaccination. In addition, s tandardized 
differences will be calculated between contemporary  unvaccinated controls and individuals 
receiving the Pfizer -BioNTech COVID-19 vaccine to ensure that the matched cohorts are 
similar with respect to the distribution of baseline characteristics . Standardized differences 
<10% will indicate that matching has appropriately  balanced the characteristics between 
vaccinated and unvaccinated cohorts.
9.7.3.Vaccine Utilization Patterns
Descriptive statistics will also be used to summarize vaccine utilization patterns, including 
proportion of individuals receiving vaccine, 2-dose completion rate , distribution of time gaps 
between the first and second dose , and care setting where immunization was received (e.g., 
outpatient clinic, pharmacy , inpatient ward) .Counts of individuals who received a COVID -
19 vaccine from a different manufacturer in addition to the Pfizer -BioNTech COVID-19 
vaccine will be reported.
9.7.4.Safety Signal Analyses
Several analy ses corresponding to the designs discussed previousl y will be conducted to 
detect safety signals associated with Pfizer-BioNTech COVID -19 vaccine. Analy ses will be 
conducted among all individuals receiving the vaccine, individuals who received 
Pfizer-BioNTech COVID- 19 vaccine without seasonal flu vaccine (Cohort A will be used for 
SCRI; Cohort B +Cwill be used for active comparator analyses), and individuals receiving 
Pfizer-BioNTech COVID- 19 vaccine and seasonal flu vaccine on the same day  (Cohort D) , 
along with sub- cohorts receiving onl y one dose vs. two doses.
A stepwise process, illustrated below, will be performed for signal de tection, evaluation , and 
verification ( Figure 3). This approach has been adapted from the Active Monitoring Protocol 
of the FDA’s COVID -19 Vaccine Safet y Surveillance Project.17The statistical approach 
described below may  be modified further based on data availability , additional clinical input, 
and for consistency  or to complement similar st udies of Pfizer -BioNTech COVID-19 
vaccine.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073085
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 48of 170Figure 3.Steps in Signal Detection, Evaluation , and Verification
Notes: 
[1] List of safety events of interest and corresponding definitions may be refined as the study progresses based on addit ional 
available information.
[2] The risk and control intervals selected for the SCRI analysis for each safety event of interest are based on biological 
plausibility and precedents in the literature. Only the individual’s first instance during the specifie d clean window (i .e., the 
interval used to define incident outcomes) will be included . Note that only the first inpatient or outpatient occurrence of a
safety event of interest following the clean window will be used to identify incident events (e.g., if an inpatient safety event
of interest occurs in the clean window, a repeat occurrence will not be counted in the risk interval ). However, event 
worsening will be counted as a safety event of interest . For example, if an outpatient safety event of interestoccurs in the 
clean window and an inpatient occurrence for the same type of safety event of interest occurs in the risk interval, the 
inpatient occurrence will be counted as a safety eventof interest .  
9.7.4.1.Signal Detection
9.7.4.1.1. Sequential Testing -SCRI Desi gnusing the Binomial -based MaxSPRT for 
Comparison to Pre -vaccination Control Intervals
The goal is to provide rapid- cycle, near real -time safet y surveillance. In the signal detection 
phase, the SCRI  analysis will only  include pre -vaccination control intervalsasthe post-
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073086
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 49of 170vaccination control intervals will require a longer time to accumulate and thus will not allow 
for timely  analysis. The post -vaccination control period will be assessedduringthe signal 
evaluation phase (see Section 9.7.4.2), to allow for additional observation time to accrue as 
well as to more deeply  investigate potential signals. This will allow for timely  RCA without 
the need to wait for data to accumulate for safety eventsof interest with post -vaccination 
control intervals. 
To account for multiple testing and repeated review of the data ,e.g., monthly  (to be 
stipulated in the SAP), the MaxSPRT using a binomial prob ability model will be applied. 
The null hy pothesis(H0) assumes that the risk of a safety eventof interest during the risk 
interval is equivalent to the risk of the same safety eventof interest developing during the 
control interval , accounting for differences in interval durationas needed (e .g., for safety 
eventsof interest such as demyelinating disease) , meaning a RR of 1 is specified under H 0.25
Theone-sided composite alternative h ypothesis (Ha) assumes that the risk of a safety eventof 
interest during the risk interval is greater than the risk of the same safety eventof interest
developing during the control interval, accounting for differences in interval duration(i.e., 
RR>1, Hais applicable across a range of RRs ).41
Specifically, for the Pfizer -BioNTech COVID- 19 vaccine, let xrepresent the total count of 
safety eventsof interest in the control interval ( Figure 4), let y represent the total count of 
safety eventsof interest in the risk interval, and let rrepresent the ratio of yto x under the 
null hypothesis. Thus, when the total control interval duration and total risk interval duration 
are equal, r will be 1. The RR is estimated by  ..
.28The RR and corresponding 99% 
confidence intervals (CIs) will be calculated. 
Figure 4. Example of SCRI Design for a S afety Eventof Interest with a 42 -day Risk 
Interval and a Pre-vaccination Control Interval
For the binomial model, the log-likelihood ratio (LLR) is calculated as the log probability of 
observing this distribution of y under Ha, divided by  the probability  of this occurring under 
H0.41This ratio is calculated whenever new data arereceivedto account for the continuous 
data stream until the full 42 -day risk period is complete.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073087
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 50of 170=ln(|)
(|)
Once the LLR test statistic reaches a pre-specified critical value , a signal is detected.  
Specificall y, the null hypothesis will be rejected if the LLR exceeds the critical value. The 
null hypothesis will not be rejected if the LLR does not reach or exceed the critical value, if 
the total number of safety eventsof interest reache sa pre-specified upper limit, or if 
surveillance ends wit hout reaching this upper limit.28
For each safety eventof interest (and specific to each age group, if age-stratified anal yses are 
conducted ), the critical value of the LLR will be determined based on the safety eventof 
interest-specificupper limit of expected safety eventsof interest and alpha level.28Upper 
limits will be determined based on the expected number of safety eventsof interest under the 
null hypothesis, assuming the risk after Pfizer-BioNTech COVID -19 vaccination is no 
greater than the risk of safety eventsof interest after s easonal influenza vaccination. 
Therefore, upper limits will be chosen such that they would not usually  be reached. 
9.7.4.1.2. S equential Testing -Poisson-based MaxSPRT for Comparison to Active 
Comparators who Received Seasonal Influenza Vaccination
For comparison with active comparator s who received seasonal influenza vaccination, the 
Poisson-basedMaxSPRT will be applied , following the same statistical approach as 
described above, but using a Poisson probabilit y distribution. In the Poisson MaxSPRT 
approach, the event frequency  of safety eventsof interest in the risk interval after Pfizer-
BioNTech COVID-19 vaccination will be compared to a background rate of safety eventsof 
interest in the risk interval after seasonal influenza vaccination in five prior seasons, ranging 
from 2014/15 through2018/19. This approach is particularl y important for extremely rare 
safety eventsof interest (i.e., less than 50 anticipated based on historical influenza vaccine 
safety eventsof interest rates).25Poisson MaxSPRT is used to monitor very rare safety events
of interest as binomial MaxSPRT may  not detect a signal, despite a clinically  meaningful 
RR.28This will also allow for more timel y analysis using historical data, as well as improved 
power and sample size.
GBS is of particular i nterest relative to the safety profile of Pfizer -BioNTech
COVID-19 vaccine. As GBS isan extremely  rare safety eventof interest , the primary RCA 
proposed will focus on Poisson M axSPRT and apply an alpha of 0.05. The Poisson 
MaxSPRT has increased power to det ect a signal with fewer occurrences of the safety event
of interest . However, this method cannot fully  control for confounding by  indication .
9.7.4.1.3. Sequential Testing -Binomial -based MaxSPRT or conditional Poisson 
MaxSPRT for Comparison to Contemporary Unvaccinated Controls
For comparison with the contemporary  unvaccinated controls, the binomial-based MaxSPRT 
will be applied, following the same statistical approach as described above. Data during the 
risk intervals of vaccinated individuals and contempo raryunvaccinated controls will 
accumulate at the same time, and thus follow- up for both cohorts are expected to be 
comparable.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073088
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 51of 170Over time, however, the number of eligible contemporary  unvaccinated controls to be 
matched to vaccinated individuals is expecte d to decrease, which will result in uncertaint y in 
the expected number of safet y eventsof interest .45As a result, conditional Poisson MaxSPRT 
(CMaxSPRT) may  be considered to account for error in the estimated expected number of 
safety eventsof interest and to consider historical data on contemporary  unvaccinated 
controls, as it will not require a baseline risk function and is more appropriate when the 
expected number of cases in the comparative data are small in order to reduce bias toward 
signaling.46The feasibility of continued matched c ontemporary  unvaccinated control anal ysis 
will be reassessed if sufficient matches cannot be identified in the data.
9.7.4.1.4. Critical Values and Alpha Spending
Critical values for the LLR test statistic are shown below in Table 3based on calculations 
conducted b y Kulldorff et al . 2011.41For example, a ssuming T= 6(number of expected 
events under the null) and RR=3, which corresponds to a power of 8 0.0% (See 
Section 9.5.1 ), the critical value would be 5. 14using alpha of 0.01 for the Poisson -based 
MaxSPRT. As noted previously , each safety  event of interest will be evaluated separatel y to 
determine a critical v alue based on background incidence, alpha, power, and clinically  
meaningful RR. These details will be addressed in the SAP.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073089
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 52of 170Table 3. Critical Values for Poisson- based MaxSPRT
Multiple ty pes of alpha spending functions can be emplo yed to calculate the cumulative rate 
at which Ty pe 1 error (alpha) probabilit y is spent during sequential testing.47To achieve 
optimal expected time -to-signal, especiall y when historical Poisson data are used with 
surveillance data, a power-typeconvex alpha spending shape will be used based on published 
literature.47Additionally , ρ =1.5 is referenced as a “rule of thumb” as it is suggested to be 
appropriate in most applications.
9.7.4.2.Signal Evaluation
Signals are detected when the event frequency ofa safety eventof interest during the risk 
intervalfollowing vaccination with Pfizer-BioNTech COVID -19 vaccine is significantl y 
increased compared to the event frequency  of the same safety eventof interest in the control 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073090
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 53of 170comparator (i.e.,the critical value is achieved and surpassed) . If signals are indeed detected 
for safety eventsof interest based on the anal ysis described above, further evaluation is 
warranted to refine and confirm such signals.This will include the following additional 
analyses to assess the robustness of the findings. 
9.7.4.2.1. Post-Signal Quality Assurance
Quality assurance will first be conducted in order to assess the quality  of the data and 
analysis that produced the signal. While quality control measures will be conducte d during 
the signal detection phase (see Section 9.8), post-signal quality assurance will also be 
performed. This will include a comprehensive quality  assurance (for example, check for 
possible duplications of claims or medical records, checking for unusual clustering in claim 
or medical record accrual by  service date for potential coding issues, check for geographical 
distribution of cases that may  be related to lot numbers or diagnostic practice) . In addition, 
for signals detected via active comparison, additional anal yses comparing to pre-vaccination
control intervals may be formed to check for consistency . Signals will also be confirmed 
across all of the safety studies planned to be performed (i.e., C4591008, C4591011, 
C4591012) toconfirm that specific data sources are not biased. 
9.7.4.2.2. Multivariate Adjustment using Poisson Regression
If signals are detected and persist after conducting quality  assurance , further evaluation via 
statistical measures are warranted. Specifically , to investigate whether potential signals 
identified via Poisson MaxSPRT for the compa rison to active comparator s with seasonal 
influenza vaccination are not confounded (i.e., to take into account baseline differences 
between the Pfizer BioNTech COVID -19 vaccinated and active comparator populations) , a 
multivariate Poisson regression analysis will be conducted to compare the incidence rates of 
the safety eventsof interest occurring within the risk intervals. T he predictor would be
whether the individual had received the Pfizer -BioNTech COVID -19 vaccine or had received 
the influenza vaccine during historical seasons. Analy ses will be adjusted for relevant 
baseline and/or clinical characteristics (e .g., age, sex, race, CCI and/or specific comorbidities 
of interest, state, etc.).9
If the signal remains, based on an IRR >3 with a p -value <0.01 from the adjusted Poisson 
regression, further evaluation may be considered via signal verification. 
9.7.4.2.3. Assessment of Temporal Clusters
Vaccine safety  surveillance must allow for sufficient type Ierrorprobability for rapid signal
detection, and statistically  significant signals must be studied further to ensure that a true 
association is present.48Therefore, t he presence of temporal clusters will be assessed using 
the software SaTScan to calculate temporal scan statistic in order to further refine safety 
signals detectedfrom the signal detection analyses.25A temporal scan stati stic accounts for 
multiple testing present during overlapping risk intervals. The null h ypothesis assumes that 
there is no association between the safety eventof interest and immunization, and safety 
eventsof interest are assumed to be distributed indepe ndently and uniformly during a period 
of time subsequent to Pfizer-BioNTech COVID -19 vaccination.25A temporal scan statistic 
will begenerated b y moving a time interval of fixed length across the risk interval, 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073091
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 54of 170comparing the number of observed versus expected safety eventsof interest within the time 
interval under the null h ypothesis.49
9.7.4.2.4. Sequential Testing -SCRI Design using the Binomial MaxSPRT for
Comparison with Post-Vaccination Control Intervals
Similar to the SCRI  design using the binomial -based MaxSPRT method for pre- vaccination 
control intervals, sequential testing anal yses will be conducted using the post -vaccination 
control intervals as appropriate for specific safet y eventsof interest . Thiswill be conducted 
during the signal evaluation phase in order to allow time to accumulate during the post -
vaccination control period. The same statistical methodology  as described for the pre -
vaccination control intervals will be applied.
9.7.4.3.Signal Verificat ion
If a signal persists after conducting signal evaluation , signal verification through medical 
records review may  be conducted .
9.7.4.3.1. Medical Records Review
As part of the signal evaluation process, diagnostic validation of the detected safet y eventsof 
interest(i.e.,cases) via a djudication of medical records by DoD MHS clinicians for outcome 
verification will be conducted ina representative sample of cases . The total number of charts 
to be reviewed wi ll depend on the number of safety eventsof interestdetected, such that all 
cases may  be reviewed for safety eventsof interest where a small number of events result in 
signal detection and a representative sub -sample may  be reviewed for safety eventsof 
interest where a larger number of events results in signal detection .50For rare events, 
potentially  all cases may  be adjudicated. An adjudication charter will be developed to govern 
signal evaluation and medical records review. Specificall y, validation of detected safet y 
events of interest will be performed through patient medical chart review in collaboration 
withan adjudication committee comprised ofthe treating or trained healthcare 
professionals.50
9.7.5.Seasonality- Adjusted Cases -Centered Method
A case-centered anal ysis for specific safety  eventsof interest for which signals were detected
may also be conducted in order to account for bias caused b y seasonality of safety eventsof 
interest and vaccination .26This method will use data on all safety eventof interest cases that 
occur after vaccination with Pfizer-BioNTech COVID -19 vaccine. Logistic regression will 
be used to compare the number of safety eventof interest cases that were vaccinated inside 
versusoutside a pre -specified risk interval , as of the date of the safety eventsof interest ,
where the total number of vaccinations given inside versus outside the risk interval (in the 
population of all vaccine es)is used as the offset term.28Specifically , the association of 
vaccination with risk of safety eventsof interest will be estimated from a logistic regression 
model that includes summarized data with one record per risk set. The key independent 
variable will be the proportion of the risk set who were in the risk interval on the date of the 
safety eventof interest occurrence. In this way , risk sets are anchored to calendar dates, and 
confounding b y seasonality of the safety eventsof interest and vaccination is addr essed.51  
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073092
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 55of 170Note that other confounders may  also be adjusted for by  restricting risk sets to vaccine es 
similar with respect to select characteristics (i. e., through stratification).
9.7.6.End-of-Season and E nd-of-Surveillance Analyses
For any safety eventof interest with signals detected, end -of-season anal yses (over the course 
of the 30-month period) and an end -of-surveillance analysis (i.e., at 30 months)will be 
conducted. S imilar methodology will beapplied for the end -of-surveillance anal ysis and end -
of-seasonanalysis conducted for seasonal influenza vaccine and contemporary  unvaccinated 
controls, respectively ,in order to adjust for the seasonality  of both disease and vaccine 
administration .9This approach will be able to define the true risk intervals after each dose 
and estimate the risk for potential safety eventsof interest after both dose 1 and 2 of the 
Pfizer-BioNTech COVID-19 vaccine, as well as the ability  to discern whether or not one or 
two doses of seasonal influenza vaccine were administered during the same period.  
Thenumber of events in the sum of three distinct risk intervals will be compared to the 
control interval, adjusting for potential differences in interval length, to estimate the RRof 
Pfizer-BioNTech COVID-19 vaccine compared to the influenza vaccine . In order to monitor 
the safety after the first and full course of thevaccine, the number of potential safety events
of interest occurring in three separate risk intervals(P1, P2, P3) will be estimated (Figure 5). 
P1represents the risk intervalafter the first dose only , excluding any  overlap in risk intervals 
with the second dose. P2represents the overlapping risk intervals for first and sec ond dose of 
thevaccine. P3represents the risk intervalof the second dose of the vaccine , excluding the 
overlapping risk intervalalready captured in P 2. This design will allow for the assessment of 
risk during the appropriate periods ,regardless of the time i nterval between vaccine doses. As 
multiple endpoints will be assessed, 99% Ciswill be calculated around the RR in order to 
ascertain whether the Pfizer-BioNTech COVID -19 vaccine isassociated with safety events
of interest . 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073093
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 56of 170Figure 5.Example of R isk (P1, P2, P3) and Pre-vaccination Control Intervals for the 
SCRI End-of-surveillance Analyses of 1 or 2 Doses ofPfizer-BioNTech COVID -19 
Vaccine
In Figure 5A , P1+ P2+ P3represent the risk intervals where a safet y eventof interest may  
occur. In Figure5B, there is no overlapping risk intervalso that P 1+ P3represent the risk
intervals where a safet y eventof interest may  occur.The timing of the risk and control 
intervals maybe adjusted for in order to control for the effect of s easonality across the 
intervals assessed.
9.7.7.Subgroup Analysis
Separate ana lyses of baseline characteristics, vaccine utilization patterns, signal detection, 
signal evaluation, and signal verification in subgroups of interest may  be conducted based on 
feasibility , sample size, and data available. 
9.7.8.Incidence Rates and Time to Safety Eventof Interest Analysis
Incidence rates (and corresponding C is) will be calculated from safet y eventof interest signal 
detection anal yses.Kaplan-Meier methods will be used to anal yze time-to-event (i.e., time to 
safety eventof interest ). If individuals do not experience the safet y eventof interest , they will 
be censored at the end of the risk interval. Median time to safety eventof interest and 
corresponding CI swill be reported.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073094
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 57of 1709.8.Quality Control
DoD data will be accessed through a secured server using encry pted login and passwords. 
Access to de -identified DoD data will be b y HealthResearchT x(HRTx) through HRT x’s 
secure private cloud computing environment. 
Each data content area will be subject to high level variable name/t ype checks, to detailed 
trending comparisons. Asan example, the diagnostic data is subject to the following checks:
Referenced table exists
Diagnosis t ype is correctly assigned by codes defining the diagnosis
Percentages, rates, are as expected (check ranges and for missing)
Both inpatient and outpatient diagnosis codes are captured. Referenced variables exist 
and areof appropriate length and type
Data retrieval will be coordinated b y an experienced programmer/anal yst. The analyst will 
write programming for retrieval of each data element from the electronic databases .Double 
programming will be performed for the first iteration of the anal yses; results/datasets will be 
compared, and if an y discrepancies are identified, both programmers will determine a 
resolution, bringing in a third programmer if needed. Subsequent iterations of analyses (i.e., 
re-runs of the anal yses) will be audited by  a senior programmer. All tables will be reviewed 
by the project manager and the principal investigator to evaluate for internal consistency  of 
counts and totals. All calculated variables will be checked against the component variables 
(cross tabs) to ensure acc uracy. For example, categorical age would be compared with 
continuous age to confirm that each category of age contained onl y individuals of the 
expected age ranges within that category .
9.9.Strengths and Limitations of the Research Methods
To identify  individualswho experienced safety eventsof interest associated with Pfizer-
BioNTech COVID-19 vaccine, the SC RImethod of signal detection offers some key  
advantages. The SCRIapproach inherentl y adjusts for within-individual confounders, such as 
age, sex,andconfounding b y indication or contraindication . Additionally , the inclusion of a 
post-vaccination control period and comparison to unvaccinated controls will account for 
increased detection bias from stimulated safet y eventof interest reporting due to hei ghtened 
vigilance on COVID -19 vaccines.52Specificall y, safety events of interest may be more likel y 
to be reported or sought care for after vaccination with Pfizer-BioNTech COVID -19 vaccine 
than before (i .e., during the pre -vaccination control interval) which may  result in bias against 
the Pfizer -BioNTech COVID-19 vaccine. Exact and PS matching will also be implemented 
for contemporary  unvaccinated controls in order to ensure that baseline characteristics 
between Pfizer -BioNTech COVID-19 vaccinee s and contemporary  unvaccinated controls are 
comparable. Lastly, SCRI allows for near real -time monitoring of safet y risks associated with 
the Pfizer -BioNTech COVID- 19 vaccine. 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073095
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 58of 170The DoD operates the largest cradle- to-grave healthca re database in the US . This data is both 
geographically  and demographicall y representative of the US general population .15The DoD 
also provides comprehensive access to its covered beneficiaries, which allows prescr iptions 
to be obtained at either no or low cost. In addition, as all DoD beneficiaries are eligible for 
healthcare coverage across locations of care, past studies have shown loss to follow -up to be 
minimized.53TheDoD MHS database provides a range of additional benefits, including its 
comprehensive structure, large number of enrollees, and electronic accessibility . The DoD 
MHSdatabasealso comprises of EMR data and allows for the possibility  of chart review.
Importantly , the DoDMHS databaseretainselectronic immunization records that include 
manufacturer name and lot numbers, facilitatingthe identification of brand -specific vaccines, 
such asthe Pfizer -BioNTech COVID -19 vaccine. Moreover, the DoD data are updated
frequently, which will allow for rapid monitoring of potential safet y signals on a repeated 
basis, e.g., monthl y (to be stipula ted in the SAP) .
However, there are several limitations when rely ing on secondary  data sources such as the 
DoDMHS database that should be noted. First, EMR data (such as laboratory  and diagnostic 
test results) will not be available for all individuals in the DoD MHS (i .e., purchased care 
only enrollees). As such, outcomes among these individuals will be identified via 
administrative claims data, which may  be subject to the misspecificatio n of billing codes or 
lack of documentation that may result in potential misclassification. Second, contemporary  
unvaccinated controls may  be systematicall y different from individuals receiving Pfizer -
BioNTech COVID -19 vaccine. While a matched based approach will be performed to 
increase the comparability between cohorts, caution should be exercised when interpreting 
the results of real-world studies due to the potential bias from unmeasured or residual 
confounding. Third, patients who may  have dual coverage through TRI CARE or DoD and 
through Medicare may  not be captur ed in the MHS healthcare database since their 
vaccinations may  be covered through Medicare.  For instance, for patients with Medicare 
Part B, TRICARE may  serve as a second payerto Medicare. Therefore, if there is no cost 
share for a service (such as vaccination) for a Medicare beneficiary  provided outside of the 
DoD MHS, then there will be no evidence of that healthcare encounter within the MHS. This 
may particularly be an issue for misclassification of receipt of seasonal influenza vaccine and 
similarly, for the COVID-19 vaccine , since Medicare covers flu shots at 100% without any  
further cost share, and there will be no TRICARE claim. As such, some patients who appear 
not to have received seasonal influenza vaccine may  have indeed received the influenza 
vaccine. This limitation will be addressed b y conducting stratified anal yses within specific 
age groups that exclude Medicare beneficiaries.
9.10.Other Aspects
Not applicable .
10. PROTECTION OF HUMAN SUBJECTS
10.1.Patient Information
All parties will comply  with all appl icable laws, including laws regarding the implementation 
of organizational and technical measures to ensure protection of patient personal data. Such 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073096
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 59of 170measures will include omitting patient names or other directl y identifiable data in an y 
reports, publicati ons, or other disclosures, except where required b y applicable laws.
To protect the rights and freedoms of natural individuals with regard to the processing of 
personal data, when study  data are compiled for transfer to Pfizer and other authorized 
parties,any patient names will be removed and will be replaced b y a single, specific, 
numerical code. All other identifiable data transferred to Pfizer or other authorized parties 
will be identified by  this single, patient -specific code. In case of data transfer, Pfizer will 
maintain high standards of confidentialit y and protection of individuals’ personal data 
consistent with the vendor contract, and applicable privacy  laws.
10.2.Patient C onsent
As this study  does not involve data subject to privacy  laws according to applicable legal 
requirements, obtaining informed consent from individuals by Pfizer is not required.  
10.3.Institutional R eview board (IRB)/Independent Ethics Committee (IEC)
There must be prospective approval of the study  protocol, protocol amendments, and their 
relevant documents from the relevant IRBs/IECs. All correspondence with the I RB/IEC must 
be retained.Copies of IRB/IEC approvals must be forwarded to Pfizer. The study protocol 
will be reviewed by theUS DoDIRBand affiliated privacy  board (PB) . 
10.4.Ethical Conduct of the Study
The study  will be conducted in accordance with legal and regulatory  requirements, as well as 
withscientific purpose, value and rigor and follow generall y accepted research practices
described in Guidelines for Good Pharmacoepidemiology  Practices (GPP) issued by  the 
International Societ y for Pharmacoepidemiology ,54the FDA Guidance for Industry and FDA 
Staff: Best Practices for Conducting and Reporting, Pharmacoepidemiologic Safet y Studies 
Using Electronic Healthcare Data55and Good Epidemiological Practice (GEP) guidelines 
issued by the International Epidemiological Association (IEA).56
11.MANAGEMENT AND REPORTING OF ADVERSE EVENTS/ADVERSE 
REACTIONS 
Signal Detection and Evaluation
This study  involves data that exist as structured data by  the time of study  start or a 
combination of existing structured data and unstructured data, which will be converted to 
structured form during the implementation of the protocol solely  by a computer using 
automated/algorithmic methods, such as natural langu age processing. In these data sources, 
individual patient data are not retrieved or validated, and it is not possible to link (i.e., 
identify a potential association between) a particular product and medical event for any  
individual. Thus, the minimum crit eria for reporting an adverse event (AE) (i.e., identifiable 
patient, identifiable reporter, a suspect product, and event) cannot be met.
Signal Verification
This study  protocol requires human review of patient- level unstructured data; unstructured 
data refer to verbatim medical data, including text -based descriptions and visual depictions 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073097
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 60of 170of medical information, such as medical records, images of ph ysician notes, neurological 
scans, X-rays, or narrative fields in a database. The reviewer is obligated to r eport adverse 
events (AEs) with explicit attribution to any  Pfizer drug that appear in the reviewed 
information (defined per the patient population and study  period specified in the protocol).  
Explicit attribution is not inferred b y a temporal relationshi p between drug administration 
and an AE, but must be based on a definite statement of causality  by a healthcare provider 
linking drug administration to the AE.
The requirements for reporting safet y events of interest on the non- interventional study  
(NIS) adverse event monitoring (AEM) Report Form to Pfizer Safety  are as follows:
All serious and non- serious AEs with explicit attribution to any  Pfizer drug that 
appear in the reviewed information must be recorded on the data collection tool (e.g., 
chart abstra ction form) and reported, within 24 hours of awareness, to Pfizer Safet y 
using the NIS AEM Report Form.
Scenarios involving drug exposure, including exposure during pregnancy , exposure 
during breast feeding, medication error, overdose, misuse, extravasatio n, lack of 
efficacy, and occupational exposure associated with the use of a Pfizer product must 
be reported, within 24 hours of awareness, to Pfizer Safet y using the NI S AEM 
Report Form.
For these AEs with an explicit attribution or scenarios involving exp osure to a Pfizer 
product, the safety  information identified in the unstructured data reviewed is captured in the 
Event Narrative section of the report form, and constitutes all clinical informat ion known 
regarding these AEs. No follow -up on related AEs wi ll be conducted.
All the demographic fields on the NI S AEM Report Form may  not necessarily  be completed, 
as the form designates, since not all elements will be available due to privacy  concerns with 
the use of secondary  data sources. While not all demograp hic fields will be completed, at the 
very least, at least one patient identifier (e.g., gender, age as captured in the narrative field of 
the form) will be reported on the NI S AEM Report Form, thus allowing the report to be 
considered a valid one in accord ance with pharmacovigilance legislation. All identifiers will 
be limited to generalities, such as the statement “A 35- year-old female...” or “An elderl y 
male...”  Other identifiers will have been removed.      
Additionally , the onset/start dates and stop dates for “Illness”, “Study  Drug”, and “Drug 
Name” may  be documented  in month/y ear (mm/yyyy ) format rather than identify ing the 
actual date of occurrence within the month /y ear of occurrence in the day /month/year 
(DD/MMM/YYYY) format.
All research staff m embers must complete the following Pfizer training requirements:   
Your Reporting Responsibilities ( YRR)Training for Vendors Working on Pfizer 
Studies.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073098
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 61of 170These trainings must be completed by  research staff members prior to the start of data 
collection. Al l trainings include a “Confirmation of Training Certificate” (for signature by  
the trainee) as a record of completion of the training, which must be kept in a retrievable 
format. Copies of all signed training certificates must be provided to Pfizer. 
Re-training must be completed on an annual basis using the most current Your Reporting 
Responsibilities training materials . 
12.PLANS FOR DISSEMINAT ING AND COMMUNICATING STUDY RESULTS
This protocol will be posted on publicly available registers following its finalization. The 
final study  results will be made publicly  available via the European Union Post Authorisation 
Safety (EU PAS) Register and may besubmitted for publication in a peer reviewed medical 
journal.
In the event of any prohibition or restriction imposed (e.g., clinical hold) by an applicable 
competent authorit y in any area of th e world, or if the investigatoris aware of an y new 
information which might influence the evaluation of the benefits and risks of a Pfizer 
product, Pfizer should be inform ed immediately .  
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073099
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 62of 17013.REFERENCES
1. World Health Organization (WHO). WHO Director- General’s opening remarks at the 
media briefing on COVID -19 -11 March 2020.  
https://www.who.int/dg/speeches/detail/who- director-general-s-opening-remarks-at-
the-media-briefing-on-covid-19---11-march-2020. Accessed November 11, 2020.
2. Johns Hopkins University. Coronavirus Resource Center.  
https://coronavirus.jhu.edu/ . Accessed November 10, 2020.
3.U.S. Food and Drug Administration (FDA). Vaccines and Related Biological 
Products Advisory  Committee Meeting: FDA Briefing Document Pfizer -BioNTech 
COVID-19 Vaccine. In:December 10, 2020.
4. U.S. Food & Drug Administration (FDA). Pfizer COVID-19 Vaccine EUA Letter of 
Authorization reissued 12-23 -20. In:December 23, 2020.
5.U.S. Food and Drug Administration (FDA). FDA Takes Key  Action in Fight Against 
COVID-19 By Issuing Emergency  Use Authorization for First COVID -19 Vaccine.  
https://www.fda.gov/news -events/press -announcements/fda -takes-key-action-fight-
against-covid-19- issuing-emergency -use-authorization -first-covid-19.
6.Pfizer Inc. Pfizer and BioNTech Achieve First Authorization in the World for a 
Vaccine to Combat COVID -19.  https://www.pfizer.com/news/press -release/press-
release-detail/pfizer -and-biontech-achieve-first-authorization -world. Accessed 
January 4, 2021.
7.Pfizer Inc. Pfizer and BioNTech Receive Authorization in the European Union for 
COVID-19 Vaccine.  https://www.pfizer.com/news/press -release/press- release-
detail/pfizer -and-biontech-receive-authorization-european- union. Accessed December 
28, 2020.
8. U.S. Department of Defense. DoD Population Schema. 2020.
9.Yih WK, L ee GM, Lieu TA, et al. Surveillance for adverse events following receipt 
of pandemic 2009 H1N1 vaccine in the Post- Licensure Rapid Imm unization Safety  
Monitoring (PRI SM) System, 2009 -2010. Am J Epidemiol. 2012;175(11):1120 -1128.
10.American Medical Association (AMA). Appendix Q: Severe Acute Respiratory 
Syndrome Coronavirus 2 (SARS -CoV-2) (coronavirus disease [COVID- 19]) 
Vaccines.  http s://www.ama -assn.org/s ystem/files/2020-11/covid-19- immunizations -
appendix-q-table.pdf. Accessed January  12, 2021.
11.Centers for Medicare & Medicaid Services. COVID -19 Vaccines and Monoclonal 
Antibodies. 2020; https://www.cms.gov/medicare/medicare -part-b-drug-average-
sales-price/covid -19-vaccines- and-monoclonal -antibodies. Accessed January  14, 
2021.
12. Center for Disease Control (CDC). People with Certain Medical Conditions. 
December 29, 2020; https://www.cdc.gov/coronavirus/2019 -ncov/need -extra-
precautions/people -with-medical-conditions.html . Accessed January  4, 2021.
13.Military Health Sy stem. About the Military  Health Sy stem. 2021; 
https://www.health.mil/About- MHS. Accessed Jan 15, 2021, 2021.
14.U.S. Department of Defense Military  Health Sy stem. Pati ents by Beneficiary  
Category.  https://www.health.mil/I -Am-A/Media/Media -Center/Patient- Population -
Statistics/Patients- by-Beneficiary -Category. Accessed January  18, 2021.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073100
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 63of 17015. Dorrance KA, Ramchandani S, Neil N, Fisher H. Leveraging the military  health 
system as a laboratory  for health care reform. Mil Med. 2013;178(2):142 -145.
16.U.S. Department of Defense Military  Health Sy stem. Getting the COVID -19 Vaccine. 
2020; https://www.health.mil/Military -Health-Topics/Combat -Support/Public -
Health/Coronavirus/COVID -19-Vaccine-Efforts/Getting -the-COVID-19-Vaccine. 
Accessed January  17, 2021.
17. U.S. Food and Drug Administration (FDA). COVID- 19 Vaccine Safet y Surveillance 
Project: Active Monitoring Protocol. December 16, 2020.
18.Center for Disease Control (CDC) COVI D-19 Response Team. Preliminary  Estimates 
of the Prevalence of Selected Underl ying Health Conditions Among Patients with 
Coronavirus Disease 2019 —United States, February  12–March 28, 2020. MMWR 
Morbidity and Mortality Weekly Report 2020;69(13):382 -386.
19.Gupta S, Hay ek SS, Wang W, et al. Factors Associated With Death in Critically  Ill 
Patients With Coronavirus Disease 2019 in the US. JAMA Intern Med. 
2020;180(11):1 -12.
20.Williamson EJ, Walker AJ, Bhaskaran K, et al. Factors associated with COVID -19-
related death using OpenSAFELY. Nature. 2020;584(7821):430-436.
21.Walsh EE, Frenck RW, Jr., Falsey  AR, et al. Safety  and Immunogenicity of Two 
RNA-Based Covid-19 Vaccine Candidates. N Engl J Med. 2020.
22.U.S. National L ibrary of Medicine. Study  to Descri be the Safet y, Tolerability, 
Immunogenicit y, and Efficacy of RNA Vaccine Candidates Against COVID -19 in 
Healthy Individuals. NCT04368728.
23. U.S. Food and Drug Administration (FDA) Center for Biologics Evaluation and 
Research (CBER). Emergency  Use Authorization for Vaccines to Prevent COVID -
19: Guidance for Industry .  https://www.fda.gov/media/142749/download Accessed 
November 10, 2020.
24.U.S. Food and Drug Administration (FDA) Center for Biologics Evaluation and 
Research (CBER). Development and L icensure of Vaccines to Prevent COVID -19; 
Guidance for Industry , June 2020.  https://www.fda.gov/media/142749/download. 
Accessed November 10, 2020.
25.Lee GM, Greene SK, Weintraub ES, et al. H1N1 and seasonal influenza vaccine 
safety in the vaccine safety datalink project. Am J Prev Med. 2011;41(2):121 -128.
26.Fireman B, Lee J, L ewis N, Bembom O, van der Laan M, Baxter R. I nfluenza 
vaccination and mo rtality: differentiating vaccine effects from bias. Am J Epidemiol. 
2009;170(5):650 -656.
27.Virtanen M, Peltola H, Paunio M. Day -to-day reactogenicity  and the healthy  vaccinee 
effect of measles -mumps-rubella vaccination. Pediatrics. 2000;106(5):E62.
28. Greene SK, Kulldorff M, Yin R, et al. Near real -time vaccine safety surveillance with 
partially accrued data. Pharmacoepidemiol Drug Saf. 2011;20(6):583 -590.
29.Keren R, Zaoutis TE, Bridges CB, et al. Neurological and Neuromuscular Disease as 
a Risk Factor for Respiratory  Failure in Children Hospitalized With I nfluenza 
Infection. JAMA. 2005;294(17):2188-2194.
30.Jewell NP. Statistics for Epidemiology. Boca Raton: Chapman & Hall/CRC. Print; 
2004.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073101
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 64of 17031.Austin PC. The performance of different propensity -score methods for estimating 
differences in proportions (risk differences or absolute risk reductions) in 
observational studies. Stat Med. 2010;29(20):2137 -2148.
32.Law B, Sturkenboom M. D2.3 Priority Lis t of Adverse Events of Special Interest: 
COVID-19. 2020.
33.Shimabukuro T. CDC post -authorization/post -licensure safety  monitoring of COVID -
19 vaccines. National Center for Immunization &  Respiratory Diseases. 2020.
34.Liu CH, Yeh YC, Huang WT, Chie WC, Chan KA. Assessment of pre -specified 
adverse events following varicella vaccine: A population -based self -controlled risk 
interval study . Vaccine. 2020;38(11):2495-2502.
35.Baker MA, Baer B, Kulldorff M, et al. Kawasaki disease and 13 -valent pneumococcal 
conjugate vaccination among young children: A self -controlled risk interval and 
cohort study  with null results. PLoS Med. 2019;16(7):e1002844.
36.Layton JB, Butler AM, Li D, et al. Prenatal Tdap immunization and risk of maternal 
and newborn adverse events. Vaccine. 2017;35(33):4072-4078.
37.Gee J, Naleway A, Shui I, et al. Monitoring the safety  of quadrivalent human 
papillomavirus vaccine: findings from the Vaccine Safety  Datalink. Vaccine. 
2011;29(46):8279 -8284.
38.Wendy Funk, Kennell and Associates. Mil itary Health Sy stem Basics.  
https://www.hsrd.research.va.gov/for_researchers/cy ber_seminars/archives/2422 -
notes.pdf.
39.Defense Health Agency . TRICARE Transformation. August 7, 2020; 
https://health.mil/Reference- Center/Presentations/2020/08/07/TRI CARE-Brief. 
Accessed January  17, 2021.
40.Li R, Stewart B, McNeil MM, et al. Post licensure surveillance of influenza vaccines 
in the Vaccine Safet y Datalink in the 2013 -2014 and 2014 -2015 seasons. 
Pharmacoepidemiol Drug Saf. 2016;25(8):928 -934.
41.Kulldorff M, Davis RL, Kolczak† M, Lewis E, L ieu T, Platt R. A Maximized 
Sequential Probability  Ratio Test for Drug and Vaccine Safet y Surveillance. 
Sequential Analysis. 2011;30(1):58 -78.
42.Kulldorff M, Silva I R. Continuous Post -Market Sequential Safet y Surveillance with 
Minimum Events to Signal. arXiv preprint arXiv. 2015(1503.01978).
43.When Are Risks Real? I n. Clinical Nutrition Insight. Vol 332007:6.
44.Rassen JA, Shelat AA, My ers J, Gly nn RJ, Rothman KJ, Schneeweiss S. One -to-
many propensity  score matching in cohort studies. Pharmacoepidemiol Drug Saf. 
2012;21 Suppl 2:69 -80.
45.Li L, Kulldorff M. A conditional maximized sequential probability  ratio test for 
pharmacovigilance. Stat Med. 2010;29(2):284 -295.
46.Yih WK, Kulldorff M, Fireman BH, et al. Active surveillance for adverse events: the 
experience of the Vaccine Safet y Datalink project. Pediatrics. 2011;127 Suppl 1:S54 -
64.
47.Silva IR, Lopes WM, Dias P, Yih WK. Alpha spending for historical versus 
surveillance Poisson data with CMaxSPRT. Stat Med. 2019;38(12):2126-2138.
48.Silva IR. Type I error probability  spending for post -market drug and vaccine safet y 
surveillance with binomial data. Stat Med. 2018;37(1):107-118.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073102
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 65of 17049. Kulldorff M. A spatial scan statistic. Communications in Statistics - Theory and 
Methods. 2007;26(6):1481 -1496.
50. U.S. Food and Drug Administration (FDA) Center for Biologics Evaluation and 
Research (CBER). Guidance for Industry: Good Pharmacovigilance Practices and 
Pharmacoepidemiologic Assessment. 2005. Accessed November 12, 2020.
51.Baker MA, Lieu TA, L i L, et al. A vaccine study  design selection framework for the 
postlicensure rapid immunization safety  monitoring program. Am J Epidemiol. 
2015;181(8):608 -618.
52. Hoffman KB, Demakas AR, Dimbil M, Tatonetti NP, Erdman CB. Stimulated 
reporting: the impact of US food and drug administration -issued alerts on the adverse 
event reporting s ystem (FAERS). Drug Saf. 2014;37(11):971 -980.
53.M Fox K, v J, Alzola C, Evans A, Tod Morris T. Incidence and Predictors of 
Pulmonary  Events among Patients with Hodgkin Ly mphoma Treated with Bleom ycin 
in the US Department of Defense Healthcare Sy stem. Chemotherapy: Open Access. 
2018;07(02).
54.International Societ y for Pharmacoepidemiology  (ISPE). Guidelines for good 
pharmacoepidemiology  practices (GPP). Pharmacoepidemiol Drug Saf. 
2008;17(2):200 -208.
55.U.S. Food and Drug Administration (FDA) Center for Drug Evaluation and Research 
(CDER). Best Practices for Conducting and Reporting Pharmacoepidemiologic 
Safety Studies Using Electronic Healthcare Data. 20 13; 
https://www.fda.gov/media/79922/download . Accessed January  19, 2021.
56. International Epidemiological Association (IEA). Good Epidemiological Practice 
(GEP), IEA Guidelines for Proper Conduct of Epidemiological Research. 2007; 
https://ieaweb.org/IEAWe b/Content/I EA_Publications.aspx . Accessed January  25, 
2021.
57.Centers for Medicare & Medicaid Services. 2021 ICD -10-CM Official Guidelines for 
Coding and Reporting.  https://www.cms.gov/files/document/2021 -coding-
guidelines -updated-12162020.pdf . Accessed January 17, 2021.
58.Baxter R, Eaton A, Hansen J, Aukes L , Caspard H, Ambrose CS. Safet y of 
quadrivalent live attenuated influenza vaccine in subjects aged 2 -49years. Vaccine. 
2017;35(9):1254 -1258.
59.Johns Hopkins Vasculitis Center. Ty pes of Vasculitis.  
https://www.hopkinsvasculitis.org/ty pes-
vasculitis/#:~:text=%E2%80%9CAngiitis%E2%80%9D%20and%20%E2%80%9CAr
teritis%E2%80%9D,lit'%20i%20deez%E2%80%9D . Accessed January  19, 2021.
60.OptumInsight Inc. Gui de to Clinical Validation, Documentation and Coding: Acute 
Kidney Injury.  
https://www.optum360coding.com/upload/pdf/ECDCG14/CDCG14_v2.pdf . 
Accessed January  17, 2021.
61. U.S. Department of Health and Human Services. Common Terminology Criteria for 
AdverseEvents (CTCAE). 2017; 
https://ctep.cancer.gov/protocoldevelopment/electronic_applications/docs/CTCAE_v
5_Quick_Reference_5x7.pdf . Accessed January  17, 2021.
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073103
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 66of 17062.Forns J, Cainzos -Achirica M, Hellfritzsch M, et al. Validity  of ICD-9 and ICD -10 
codes used to identify  acute liver injury : A study in three European data sources. 
Pharmacoepidemiol Drug Saf. 2019;28(7):965 -975.
14.LIST OF TABLES
Table 1. Outcome algorithms for SCRI  analysis, with risk and control 
intervals................................ ................................ ................................ .....39
Table2. Estimated Statistical Power for the Poisson -based MaxSPRT41..............44
Table 3. Critical Values for Poisson -based MaxSPRT ................................ ...........52
Appendix Table 1. Demographic and Clinical Characteristics Definitions ............................ 67
Appendix Table 2. Operationa l Definitions of Safet y Events of Interest.............................. 103
15.LIST OF FIGURES
Figure 1. Example of SCRI  Design for Assessment of a Safety  Event of 
Interest with a 42- day Risk Interval in an Individual who Receives 
Only One Vaccine Dose, Showing Both Pre -and Post-vaccination 
Control Intervals................................ ................................ ....................... 28
Figure 2. Example of SCRI  Design with Overlapping Risk Intervals when 
Two Doses of Pfizer -BioNTech COVID- 19 Vaccine are 
Administered, Showing a Pre -and Post-vaccination Control 
Interval................................ ................................ ................................ ......29
Figure 3. Steps in Signal Detection, Evaluation, and Verification .......................... 48
Figure 4. Example of SCRI  Design for a Safety  Event of Interest with a 42 -
day Risk Interval and a Pre -vaccination Control Interval ........................ 49
Figure 5. Example of Risk (P1, P2, P3) and Pre- vaccination Control I ntervals 
for the SCRI  End-of-surveillance Analy ses of 1 or 2 Doses of 
Pfizer-BioNTech COVID- 19 Vaccine ................................ ...................... 56
16.ANNEX 1. LIST OF STAND- ALONE DOCUMENTS
None. 
17.ANNEX 2. ENCEPP CHEC KLIST FOR STUDY PROT OCOLS
N/A
18.ANNEX 3. ADDITIONAL INFORMATION
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073104
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 67of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
Demographic Characteristics
Age Continuous variable;
Dichotomous 
variable: 18 -64
>65;
Categorical variable:
<35
35 -<45
45 -<55
55 -<65
65 -<75
≥75Age as of the date prior to Pfizer -
BioNTech COVID -19 vaccination 
(and/or date prior to seasonal influenza 
vaccination for active comparators, 
matched index date for contemporary  
unvaccinated controls)
Sex Categorical variable: 
Male
Female
Unknown
State Geographic regions in the 
USState of residence
Sponsor service Categorical variable:
Army
Air Force
Coast Guard
Marine Corps
Navy
Navy Afloat
Other
Unknown
Beneficiary  
categoryCategorical variable:
Active Dut y 
Retirees
Active 
Guard/Reserve
Dependents of Active 
Duty
Dependents of 
Retiree
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073105
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 68of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
Dependent Survivor
Dependent of Active 
Guard/Reserve
Inactive 
Guard/Reserve
Family Member of 
Inactive 
Guard/Reserve 
Other
Unknown
Clinical Characteristics
Smoking Dichotomous variable Defined b y the “tobacco” variable. ‘Y’ 
indicates the person is a tobacco user
ICD-9-CM codes:
305.1, Tobacco use disorder
V15.82, History  of tobacco use
ICD-10-CM codes:
F17.200, Nicotine dependence, 
unspecified, uncomplicated
Z7.20, Tobacco use
Z87.891, Personal history of 
nicotine dependence
Body mass index 
(BMI)Continuous variable; 
Categorical variable:
Underweight (<18.5) 
Normal weight (18.5 -
24.9) 
Overweight (25 -29.9)
Obese (≥30 -<40)
Severe obesit y (>40)Calculated from height and weight data 
(kg/m2) 
ICD-9-CM codes:
V85.0, Bod y Mass Index less 
than 19, adult
V85.1, Bod y Mass Index 
between 19- 24, adult
V85.2, Bod y mass index 
between 25- 29, adult
V85.3, Bod y mass index 
between 30- 39, adult
V85.4, Bod y mass index 40 and 
over, adult
ICD-10-CM codes:
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073106
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 69of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
Z68.1, Bod y Mass Index 19.9 or 
less, adult
Z68.2, Body mass index 20-29, 
adult
Z68.3, Bod y mass index 
between 30- 39, adult
Z68.4, Bod y mass index 40 and 
over, adult
History of 
anaphylaxis/allergic 
reactionsDichotomous variable ICD-9-CM code:
V13.81, Personal history  of 
anaphylaxis
V14.0 -V14.6, V14.8, V14.9, 
Personal history  of allergy  to 
drugs, medications and 
biological substances, excluding 
serum and vaccine
V15.0x, Other allergy
525.66, Allergy  to existing 
dental restorative material
995.0, Other anaph ylactic 
shock, not elsewhere classified
995.1, Angioneurotic edema, 
not elsewhere classified
995.21, Arthus phenomenon
999.27, Other drug allergy
995.3, Allergy , unspecified, not 
elsewhere classified
995.6x, Anaphy lactic shock due 
to food
999.41, Anaph ylactic reaction 
due to administration of blood 
and blood products
999.49, Anaph ylactic reaction 
due to other serum 
ICD-10-CM code:
Z87.892  Personal history of 
anaphylaxis
Z88.0 -Z88.6, Z88.8, Z88.9, 
Allergy status to drugs, 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073107
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 70of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
medications and biological 
substances, excluding serum and 
vaccine
T78.00xx -T78.09xx, 
Anaphylactic reaction due to 
food, initial encounter, 
subsequent encounter and 
sequela
T78.2xxx, Anaphy lactic shock, 
initial encounter, subsequent 
encounter and se quela
T78.3xxx, Angioneurotic 
edema, initial encounter, 
subsequent encounter and 
sequela
T78.41xx, Arthus phenomenon
T80.51xx, Anaphy lactic 
reaction due to administration of 
blood and blood products, initial 
encounter, subsequent encounter 
and sequela
T80.59xx, Anaphy lactic 
reaction due to other serum,  
initial encounter, subsequent 
encounter and sequela
T88.6xxx, Anaphy lactic 
reaction due to adverse effect of 
correct drug or medicament 
properly administered, initial 
encounter, subsequent encounter 
and sequela
Previous 
anaphylaxis of 
vaccine componentDichotomous variable ICD-9-CM code:
999.42, Anaph ylactic reaction 
due to vaccination
V14.7, Personal history  of 
allergy to serum or vaccine
ICD-10-CM codes:
T80.52xx, Anaphy lactic 
reaction due to vaccination, 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073108
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 71of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
initial encounter, subsequent 
encounter and sequela
Z28.04, Immunization not 
carried out because of patient 
allergy to vaccine or component
Z88.7, Allergy  status to serum 
and vaccine
History of 
hospitalizations Dichotomous variable;
Continuous variableDefined b y having any  hospitalizations
(dichotomous) and number of 
hospitalizations (continuous)
Pregnancy Dichotomous variable LOINC code:
82810-3, Pregnancy  status
11449-6, Pregnancy  status -
Reported
ICD-9-CM codes:
V22.x, Normalpregnancy
V23.x, V23.xx, Supervision of 
high-risk pregnancy
ICD-10-CM codes:
Z33.1, Pregnant state, incidental
Z33.3, Pregnant state, 
gestational carrier
Z34, Supervision of normal 
pregnancy
O09, Supervision of high risk 
pregnancy
Charlson 
Comorbidity  Index 
(CCI)Continuous variable ICD-9-CM codes:
410.x, 412.x, My ocardial 
infarction
398.91, 402.01, 402.11, 402.91, 
404.01, 404.03, 404.11, 404.13, 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073109
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 72of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
404.91, 404.93, 425.4 -425.9, 
428.x, Congestive heart failure
093.0, 437.3, 440.x, 441.x, 
443.1 -443.9, 447.1, 557.1,
557.9, V43.4, Peripheral 
vascular disease
362.34, 430.x -438.x, 
Cerebrovascular disease
290.x, 294.1, 331.2, Dementia
416.8, 416.9, 490.x -505.x, 
506.4, 508.1, 508.8, Chronic 
pulmonary  disease
446.5, 710.0 -710.4, 714.0 -
714.2, 714.8, 725.x, Rheumatic
disease
531.x -534.x, Peptic ulcer 
disease
070.22, 070.23, 070.32, 070.33, 
070.44, 070.54, 070.6, 070.9, 
570.x, 571.x, 573.3, 573.4, 
573.8, 573.9, V42.7, Mild liver 
disease
250.0 -250.3, 250.8, 250.9, 
Diabetes without chronic 
complication
250.4 -250.7,Diabetes with 
chronic complication
334.1, 342.x, 343.x, 344.0 -
344.6, 344.9, Hemiplegia or 
paraplegia
403.01, 403.11, 403.91, 404.02, 
404.03, 404.12, 404.13, 404.92, 
404.93, 582.x, 583.0 - 583.7, 
585.x, 586.x, 588.0, V42.0, 
V45.1, V56.x, Renal disease
140.x -172.x, 174.x -195.8, 
200.x -208.x, 238.6, Any 
malignancy , including 
lymphoma and leukemia, except 
malignant neoplasm of skin
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073110
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 73of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
456.0 -456.2, 572.2 -572.8, 
Moderate or severe liver disease
196.x -199.x, Metastatic solid 
tumor
042.x -044.x, Acquir ed 
immunodeficiency  syndrome 
(AIDS)/Human 
immunodeficiency  virus (HIV)
ICD-10-CM codes:
I21.x, I21.xx, I 22.x, I25.2, 
Myocardial infarction
I09.9, I11.0, I13.0, I13.2, I25.5, 
I42.0, 
I42.5 -I42.9, I43, I43.x, I50.x, 
I50.xx, Congestive heart failure
I70.x, I71.x, I73.1, I73.8, I 73.9, 
I77.1, I79.0, I79.2, K55.1, 
K55.8, K55.9, Z95.8, Z95.9, 
Peripheral vascular disease
G45, G45.x, G46.x, H34.0, 
I60.x -I63.x, I60.xx -I63.xx, 
I60.xxx - I63.xxx, I65.x -I69.x, 
I65.xx -I69.xx, I65.xxx -
I69.xxx, Cerebrovascular 
disease
F00.x -F03.x, F00.xx - F03.xx, 
F05, F05.1, G30.x, G31.1, 
Dementia
I27.8, I27.9, J40.x - J47.x, 
J40.xx -J47.xx, J40.xxx -
J47.xxx, J60.x -J67.x, J68.4, 
J70.1, J70.3, Chronic pulmonary  
disease
M05, M05.x, M05.xx, M05.xxx, 
M06, M06.x, M06.xx, M06.x xx, 
M31.5, M32.x - M34.x, M32.xx 
-M34.xx,  M35.1, M35.3, 
M36.0, Rheumatic disease
K25.x -K28.x, Peptic ulcer 
disease
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073111
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 74of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
B18.x, K70.0 - K70.3, K70.9, 
K71.3 -K71.5, K71.7, K73.x, 
K74.x, K74.xx, K76.0, K76.2-
K76.4, K76.8, K76.9, Z94.4, 
Mild liver disease
E10.0, E10.1x, E10.6x, 
E10.6xx, E10.8, E10.9, E11.0x, 
E11.1x, E11.6x, E11.6xx, 
E11.8, E11.9, E12.0, E12.1, 
E12.6, E12.8, E12.9, E13.0x, 
E13.1x, E13.6x, E13.6xx, 
E13.8, E13.9, E14.0, E14.1, 
E14.6, E14.8, E14.9, Diabetes 
without chronic complication
E10.2x -E10.5x, E10.2xx -
E10.5xx, E10.7, E11.2x -
E11.5x, E11.2xx - E11.5xx, 
E11.7, E12.2 -E12.5, E12.7, 
E13.2 -E13.5x, E13.7, E14.2 -
E14.5, E14.7, Diabetes with 
chronic complication
G04.1, G11.4, G80.1, G80.2, 
G81.x, G81.xx, G82.x, G82.xx, 
G83.0, G83.1- G83.3, G8 3.1x-
G83.3x, G83.4, G83.9, 
Hemiplegia or paraplegia
I12.0, I13.1x, N03.2 - N03.7, 
N05.2 -N05.7, N18.x, N19, 
N25.0, Z49.0x -Z49.3x, Z94.0, 
Z99.2, Renal disease
C00-C75, C00.x -C75.x, C00.xx -
C75.xx (excluding C44, C44.x 
and C44.xx), C7A., C7A.x, 
C7A.xx, C7B ., C7B.x, C7B.xx, 
C76-C80, C76.x -C80.x, C76.xx -
C80.xx, C81- C96, C81.x- C96.x, 
C81.xx-C96.xx, Any  
malignancy , including 
lymphoma and leukemia, except 
malignant neoplasm of skin
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073112
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 75of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
I85.0, I85.9, I86.4, I98.2, 
K70.4x, K71.1x, K72.1x, 
K72.9x, K76.5, K76.6, K76.7, 
Moderate or severe liver disease
C77.x -C80.x, C77.xx - C80.xx, 
Metastatic solid tumor
B20, B97.35, AIDS/HIV
Comorbidities Categorical variable:
Autoimmune 
disease 
Asthma
Bleeding diathesis 
or condition 
associated with 
prolonged 
bleeding
Cancer
Cardiovascular 
conditions (e. g., 
heart failure, 
CAD, 
cardiomyopathies)
Chronic kidney  
disease/dial ysis
COPD/interstitial 
lung disease
Diabetes mellitus 
(ie, Type 2 
diabetes)
Down syndrome
Sickle cell disease
HBV
HCV
HIV
Hyperlipidemia
Hypertension
Liver disease
Neurological 
diseaseAutoimmune disease 
(immunocompromised state [weakened 
immune sy stem] from solid organ 
transplant):
ICD-9-CM  codes:
245.2, Chronic ly mphocytic 
thyroiditis
340, Multiple scler osis
357, Acute infective 
polyneuritis
357.4, Poly neuropathy  in 
other diseases classified 
elsewhere
696.1, Other psoriasis
694.3, Impetigo 
herpetiformis
696.1, Other psoriasis
696, Psoriatic arthropathy
695.4, Lupus erythematosus
714, 714.x, 714.xx, 
Rheumatoid arthritis and 
other inflammatory  
polyarthropathies
359.6, Sy mptomatic 
inflammatory  myopathy in 
diseases classified elsewhere
357.1, Poly neuropathy  in 
collagen vascular disease
714.89, Other specified 
inflammatory  
polyarthropathies
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073113
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 76of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
Other immune 
deficiencies
Solid organ 
transplant
VTE714.9, Unspeci fied 
inflammatory  
polyarthropathy
446.5, Giant cell arteritis
710.2, Sicca s yndrome
ICD-10-CM codes:
D69.3, Immune 
thrombocy topenic purpura
E06.3, Autoimmune 
thyroiditis
G35, MS
G61.0 and G65.0, GBS and 
sequelae of GBS
L40.x, L40.5x, Psoriasis
L93.x, Lupuserythematosus
M05.x, M05.xx, M05.xxx, 
Rheumatoid arthritis with 
rheumatoid factor
M06.x, M06.xx, M06.xxx, 
Other rheumatoid arthritis
M31.5, M31.6, Giant cell 
arteritis
M35.0x, Sicca (Sjogren’s) 
syndrome 
E10, E10.x, E10.xx, Ty pe 1 
diabetes mellitus
N05.9,Glomerulonephritis
D84.9, Immunodeficiency , 
unspecified
Asthma: 
ICD-9-CM codes:
o493.xx, Asthma
ICD-10-CM codes:
oJ45.2x -J45.3x, Mild 
intermittent asthma
oJ45.4x, Moderate 
persistent asthma
oJ45.5x, Severe 
persistent asthma
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073114
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 77of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
oJ45.9x, Other and 
unspecified as thma
Bleeding diathesis or condition 
associated with prolonged bleeding:
ICD-9-CM codes:
o286.x, Coagulation 
defects
o289.8x, Other 
specified diseases of 
blood and blood -
forming organs
o287, 287.x, 287.xx,  
Purpura and other 
hemorrhagic 
conditions
ICD-10-CM codes:
oD65, Disseminated 
intravascular 
coagulation
oD66, Hereditary  
factor VIII 
deficiency
oD67, Hereditary  
factor IX deficiency
oD68, D68.x, D68.xx, 
Other coagulation 
defects
oD69, D69.x, D69.xx, 
Purpura and other 
hemorrhagic 
conditions
Cancer:
ICD-9-CM codes :
o140.x -149.x, 
Malignant neoplasm 
of lip, oral cavit y, 
and phary nx
o150.x -159.x, 
Malignant neoplasm 
of digestive organs 
and peritoneum
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073115
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 78of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
o160.x -165.x, 
Malignant neoplasm 
of respiratory  and 
intrathoracic organs
o170.x -176.x, 
Malignant neoplasm 
of bone, connective 
tissue, skin, and 
breast
o179.x -189.x, 
Malignant neoplasm 
of genitourinary  
organs 
o190.x -199.x, 
Malignant neoplasm 
of other unspecified 
sites
o200.xx -208.xx, 
Malignant neoplasm 
of lymphatic and 
hematopoietic tissue
o209.0x -209.3x, 
Malignant
neuroendocrine 
tumors
o230.x -234.x, 
Carcinoma in situ of 
digestive organs
ICD-10-CM codes:
oC00-C75, C00.x -
C75.x, C00.xx -
C75.xx, C7A., 
C7A.x, C7A.xx, 
C7B., C7B.x, 
C7B.xx, Malignant 
neoplasms, stated or 
presumed to be 
primary (of specified 
sites), and cert ain 
specified histologies, 
except 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073116
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 79of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
neuroendocrine, and 
of lymphoid, 
hematopoietic and 
related tissue 
oC76-C80, C76.x -
C80.x, C76.xx -
C80.xx, Malignant 
neoplasms of ill-
defined, other 
secondary  and 
unspecified sites
oC81-C96, C81.x -
C96.x, C81.xx -
C96.xx, Maligna nt 
neoplasms of 
lymphoid, 
hematopoietic and 
related tissue 
Cardiovascular conditions (e .g., heart 
failure, coronary  artery disease [CAD], 
cardiomyopathies):
ICD-9-CM codes:
o428.xx, Heart failure
o414.01, 429.2, 411.1, 
413.9, 414.11, 
414.12, 414.05, 
414.02, 414.04, 
414.03, 414.06, 
414.07, 414.2, 
411.81, 411.89, CAD
o425.xx, 
Cardiomy opathy
ICD-10-CM codes:
o150.x, 150.xx,  Heart 
failure
oI24.0, I24.8, I24.9, 
I25.10, I25.110, 
I25.111, I25.118, 
I25.119, I25.41, 
I25.42, I25.700, 
I25.701, I25.708, 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073117
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 80of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
I25.709, I25.710, 
I25.711, I25.718, 
I25.719, I25.720, 
I25.721, I25.728, 
I25.729, I25.730, 
I25.731, I25.738, 
I25.739, I25.750, 
I25.751, I25.758, 
I25.759, I25.760, 
I25.761, I25.768, 
I25.769, I25.790, 
I25.791, I25.798, 
I25.799, I25.810, 
I25.811, I25.812, 
CAD
oI42.x, 
Cardiomy opathy
Chronic kidney  disease/dialy sis:
ICD-9-CM codes:
o283.11, Hemoly tic-
uremic syndrome
o403, 403.x, 403.xx, 
Hypertensive chronic 
kidney disease
o404, 404.x, 404.xx, 
Hypertensive heart 
and chronic kidney  
disease
o440.1, 
Atherosclerosis of 
renal artery
o442.1, Aneurysm of 
renal artery
o572.4, Hepatorenal 
syndrome
o274.1, Gouty  
nephropath y, 
unspecified
o710, Systemic lupus 
erythematosus
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073118
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 81of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
o710.2, Sicca 
syndrome
o580, 580.x, 580.xx, 
Acute 
glomerulonephritis
o581.x, 581.xx, 
Nephrotic s yndrome
o582, 582.x, 582.xx, 
Chronic 
glomerulonephritis
o583, 583.x, 583,xx, 
Nephritis and 
nephropath y, not 
specified as acute or 
chronic
o591, Hydronephrosis
o593.3, Stricture or 
kinking of ureter
o592, Calculus of 
kidney
o592.1, Calculus of 
ureter
o590.9, Infection of 
kidney, unspecified
o584.x, Acute kidney  
failure
o585.x, Chronic 
kidney disease
o588.x, 588.xx, 
Disorders resulting 
from impaired renal 
function
o587, Renal sclerosis, 
unspecified
o753.1x, Cy stic 
kidney disease
o753.2, 753.2x, 
Obstructive defects 
of renal pelvis and 
ureter
ICD-10-CM codes:
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073119
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 82of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
oD59.3, Hemoly tic-
uremic syndrome
oI12.x, Hy pertensive 
chronic kidney  
disease
oI13.x, I13.xx, 
Hypertensive heart 
and chronic kidney  
disease
oI70.1, 
Atherosclerosis of 
renal artery
oI72.2 Aneury sm of 
renal artery
oK76.7, Hepatorenal 
syndrome
oM10.30-M10.39, 
M10.30x-M10.37x, 
Gout due to renal 
impairment
oM32.14, Glomerular 
disease in s ystemic 
lupus erythematosus
oM32.15, Tubulo -
interstitial 
nephropath y in 
systemic lupus 
erythematosus
oM3504, Sicca 
syndrome with 
tubulo-interstitial 
nephropath y
oN00.x-N07.x, N08, 
Glomerular diseases
oN13.1, N13.2, 
N13.3x, Obstructive 
and reflux uropath y
oN14.x, Nephropathy
oN15.x, Other renal 
tubulo-interstitial 
diseases
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073120
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 83of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
oN16, Renal tubulo-
interstitial disorders 
in diseases classified 
elsewhere
oN17.x, N18.x, N19, 
Acute kidney  failure 
andchronic kidney  
disease
oN25.x, N26.x, 
N25.xx, Other 
disorders of kidney  
and ureter
oQ61.02, Q61.11x, 
Q61.2-Q61.9, Cy stic 
kidney disease
oQ62.x, Q62.xx, 
Congenital 
obstructive defects of 
renal pelvis and 
congenital 
malformation of 
ureter
COPD/interstitial lun g disease:
ICD-9-CM codes:
o491.9, Unspecified 
chronic bronchitis
o492.8, Other 
emphysema
o491.x, 491.xx, 
Chronic bronchitis
o493.2, Chronic 
obstructive asthma, 
unspecified
o496, Chronic airway  
obstruction, not 
elsewhere classified
o516, 516.x, 516.xx, 
Other alveolar and 
parietoalveolar 
pneumonopathy
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073121
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 84of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
o515, 
Postinflammatory  
pulmonary  fibrosis
o518.x, 518.xx, Other 
diseases of lung
o714.81, Rheumatoid 
lung
ICD-10-CM codes:
oJ41.x Simple and 
mucopurulent 
chronic bronchitis
oJ42, Unspecified 
chronic bronchitis
oJ43.x, Emphysema
oJ44.x, Other COPD
oJ80, J81.x, J82.xx, 
J84.xx, J84.xxx, 
Other respiratory  
diseases principall y 
affecting the 
interstitium 
oM05.10, Rheumatoid 
lung disease with 
rheumatoid arthritis 
of unspecified site
Diabetes mellitus (ie, Type 2 diabetes):
ICD-9-CM codes:
o250.xx, Diabetes 
mellitus
ICD-10-CM codes:
oE11.x, E11.xx, 
E11.xxx, Ty pe 2 
diabetes mellitus
Down syndrome:
ICD-9-CM codes:
o758.x, Down 
syndrome
ICD-10-CM codes:
oQ90.x, Down 
syndrome
Sickle cell disease:
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073122
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 85of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
ICD-9-CM codes:
o282.xx, Sickle -cell 
disease
ICD-10-CM codes:
oD57, D57.x, D57.xx, 
D57.xxx, Sickle -cell 
disorders
HBV:
ICD-9-CM codes:
o70.33, Chronic viral 
hepatitis B without 
mention of hepatic 
coma with hepatitis 
delta
o70.32, Chronic viral 
hepatitis B without 
mention of hepatic 
coma without 
mentionof hepatitis 
delta
o70.3, Viral hepatitis 
B without mention of 
hepatic coma, acute 
or unspecified, 
without mention of 
hepatitis delta
o70.2, Viral hepatitis 
B with hepatic coma, 
acute or unspecified, 
without mention of 
hepatitis delta
ICD-10-CM codes:
oB18.0, B18.1, 
Chronic viral 
hepatitis B
oB19.1, B19.1x, 
Unspecified viral 
hepatitis B
HCV:
ICD-9-CM codes:
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073123
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 86of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
o70.7, Unspecified 
viral hepatitis C 
without hepatic coma
o70.71, Unspecified 
viral hepatitis C with 
hepatic coma
o70.54, Chronic 
hepatitis C without 
mention of hepatic 
coma
ICD-10-CM codes:
oB18.2, Chronic viral 
hepatitis C
oB19.2x, Unspecified 
viral hepatitis C
HIV:
ICD-9-CM codes:
o42, HIV disease
o79.53, HIV t ype 2 
ICD-10-CM codes:
oB20, HIV disease
oB97.35, HIV t ype 2 
as the cause of 
diseases classified 
elsewhere
Hyperlipidemia
ICD-9-CM codes:
o272.0x, Pure 
hypercholesterolemia
o272.1x, Pure 
hyperglyceridemia
o272.2x, Mixed 
hyperlipidemia
o272.4x, 
Hyperlipidemia, 
NOS
ICD-10-CM codes:
oE78.0-E78.5, 
E78.0x, E78.4x, 
Hyperlipidemia
Hypertension:
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073124
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 87of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
ICD-9-CM codes:
o401.1, Benign 
essential 
hypertension
o401.9, Essential 
hypertension, NOS
o405.1, Benign 
secondary  
hypertension
o405.9, Secondary  
hypertension, NOS
o997.91, 
Hypertension, NOS
ICD-10-CM codes:
oH35.03x, 
Hypertensive 
retinopathy
oI10, I11.x -I16.x, 
I13.xx, Hy pertensive 
diseases
oI67.4, Hy pertensive 
encephalopath y 
diseases
Liver disease:
ICD-9-CM codes:
o571, 571.x, 
Alcoholic fatt y liver
o572, 572.x, Hepatic 
encephalopath y
o573.x, Other disorder 
of liver
o570, Acute and 
subacute necrosis of 
liver
ICD-10-CM codes:
oK70.x, K70.xx, 
Alcoholic fatt y liver
oK71.x, K71.xx, 
Toxic liver disease
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073125
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 88of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
oK72.xx, Hepatic 
failure, not elsewhere 
classified
oK73.x, Chronic 
hepatitis, not 
elsewhere specified
oK74.x, K74.xx, 
Fibrosis and cirrhosis 
of liver
oK75.x, K75.xx, 
Other inflammatory  
liver diseases
oK76.x, K76.xx, 
Other diseases of 
liver
oK77, Liver disorders 
in diseases classified 
elsewhere
Neurological disease:
ICD-9-CM codes:
o780.97, Altered 
mental status
o780.93, Memory  loss
o781.8, Neurologic 
neglect syndrome
o797, Senility  without 
mention of ps ychosis
oV62.89, Other 
psychological or 
physical stress, not 
elsewhere classified
o799.5x, Signs and 
symptoms involving 
cognition
o780.99, Other 
general symptoms
o780.4, Dizziness and 
giddiness
o781.1, Disturbances 
of sensation of smell 
and taste
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073126
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 89of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
oV41.5, Problems 
with smell and taste
o368.16, 
Psychophysical 
visual disturbances
o307.9, Other and 
unspecified special 
symptoms or 
syndromes, not 
elsewhere classified
o300.9, Unspecified 
nonpsychotic mental 
disorder
o300.9, Unspecified 
nonpsychotic mental 
disorder
o308.9, Unspecified 
acute reaction to 
stress
o307.9, Other and 
unspecified special 
symptoms or 
syndromes, not 
elsewhere classified
oV62.85, Homicidal 
ideation
oV62.84, Suicidal 
ideation
o799.24, Emotional 
lability
o799.23, 
Impulsiveness
o799.29, Other signs 
and symptoms 
involving emotional 
state
oV40.39, Other 
specified behavioral 
problem
ICD-10-CM codes:
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073127
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 90of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
oR41, R41.x, R41.xx, 
Other symptoms and 
signs involving 
cognitive functions 
and awareness
oR42, Dizziness and 
giddiness
oR43, R43.x, 
Disturbances of 
smell and taste
oR44, R44.x, Other 
symptoms and signs 
involving general 
sensations and 
perceptions
oR45, R45.x, R45. xx, 
Symptoms and signs 
involving emotional 
state
oR46, R46.x, R46. xx, 
Symptoms and signs 
involving appearance 
and behavior
Other immune deficiencies:
ICD-9-CM codes:
o279.x, 279.xx, 
Deficiency  of 
humoral immunity
o135, Sarcoidosis
o273.x, Disorders of 
plasma protein 
metabolism
ICD-10-CM codes:
oD80, D80.x, 
Immunodeficiency  
with predominantly  
antibody defects
oD81, D81.x, D81.xx, 
Combined 
immunodeficiencies
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073128
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 91of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
oD82, D82 .x, 
Immunodeficiency  
associated with other 
major defects
oD83, D83.x, 
Common variable 
immunodeficiency
oD84, D84.x, D84.xx, 
Other 
immunodeficiencies
oD86, D86.x, D86.xx, 
Sarcoidosis
oD89, D89.x, D89.xx, 
Other disorders 
involving the 
immune mechanism, 
not elsew here 
classified
Solid organ transplant:
CPT codes: 
o32850-32856, 
Transplantation of 
lung
o33930-33945, 
Transplantation of 
heart
o44132, 44133, 
47133, 47135, 
47140-47147, 
Transplantation of 
liver
o44135-44137, 44715, 
44720, 44721, 
Transplantation of 
intestine
o48160, 48550 -48552, 
48554, 48556, 
Transplantation of 
pancreas
o50300, 50320, 
50323, 50325, 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073129
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 92of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
50327, 50328, 
50329, 50340, 
50340, 50360, 
50365, 50370, 
50380, Renal 
transplantation
ICD-9-PCS codes:
o00.91 -00.93, 
Transplant from 
donor or cadaver
o37.51, Heart 
transplantation
o33.51, Unilateral 
lung transplantation
o33.52, Bilateral lung 
transplantation
o46.97, Transplant of 
intestine
o50.59, Other 
transplant of 
intestine
o52.82, 
Homotransplant of 
pancreas
o55.69, Other kidney  
transplant
ICD-10-PCS codes:
o02YA0Z0, 
02YA0Z1, 
Transplantation of 
heart
o0BYC0Z0, 
0BYC0Z1, 
0BYD0Z0, 
0BYD0Z1, 
0BYF0Z0, 
0BYF0Z1, 
0BYG0Z0, 
0BYG0Z1, 
0BYH0Z0, 
0BYH0Z1, 
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073130
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 93of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
0BYJ0Z0,  
0BYJ0Z1,  
0BYK0Z0,  
0BYK0Z1,  
0BYL0Z0, 
0BYL0Z1, 
0BYM0Z0, 
0BYM0Z1, 
Transplantation of 
lung
o0DY60Z0, 
0DY60Z1, 
Transplantation of 
stomach
o0DY80Z0, 
0DY80Z1, 
Transplantation of 
small intestine
o0DYE0Z0, 
0DYE0Z1, 
Transplantation of 
large intestine
o0FY00Z0, 0FY00Z1, 
Transplantation of 
liver
o0FYG0Z0, 
0FYG0Z1, 
Transplantation of 
pancreas
o0TY00Z0, 0TY00Z1, 
0TY10Z0, 0TY10Z1, 
Transplantation of 
kidney
VTE:
ICD-9-CM codes:
o415.1x, Pulmonary  
embolism and 
infarction
o451.x, 451.xx, 
Phlebitis and 
thrombophlebitis
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073131
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 94of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
o452, Portal vein 
thrombosis
o453.x, 453.xx, Other 
venous embolism 
and thrombosis
ICD-10-CM codes:
oI26, I26.x, I26.xx, 
Pulmonary  embolism
oI80, I80.x, I 80.xx, 
I80.xxx, Phlebitis 
and thrombophlebitis
oI81, Portal vein 
thrombosis
oI82, I82.x, I82.xx, 
I82.xxx Other 
venous embolism 
and thrombosis
Concurrent 
immunizationsCategorical variable:
Seasonal influenza 
Tetanus diphtheria 
and pertussis (Tdap or 
Td)
Chickenpox 
(Varicella)
Shingles (Herpes 
Zoster recombinant 
and/or live)
Human 
papillomavirus (HPV)  
Pneumococcal 
conjugate
Pneumococcal 
polysaccharide
Hepatitis A
Hepatitis B
Meningococcal 
conjugate 
(MenACWY) and 
serogroup B Description of immunization, 
immunization I D, lot number, and 
manufacturer code will be available.
Seasonal influenza:
CPT codes:
o90653, Influenza vaccine, 
inactivated (IIV), subunit, 
adjuvanted, for 
intramuscular use
o90724, Influenza virus 
vaccine
o90662, Influenza virus 
vaccine (IIV), split virus, 
preservative free, enhanced 
immunogenicit y via 
increased antigen content, 
for intramuscular use
o90662, Influenza virus 
vaccine (IIV), split virus, 
preservative free, enhanced 
immunogenicit y via 
increased antigen content, 
for intramuscular use
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073132
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 95of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
meningococcal 
(MenB)
Haemophilus 
influenza ty pe bo90694, Influenza virus 
vaccine, quadrivalent 
(aIIV4), inactivated, 
adjuvanted, preservative 
free, 0.5 mL dosage, for 
intramuscular use
o90756, Influenza virus 
vaccine, quadrivalent 
(ccIIV4), derived from cell 
cultures, subunit, antibiotic 
free, 0.5 mL dosage, for 
intramuscular use
o90674, Influenza virus 
vaccine, quadrivalent 
(ccIIV4), derived from cell 
cultures, subunit, 
preservative and antibiotic 
free, 0.5 mL dosage, for 
intramuscular use
o90688, Influenza virus 
vaccine, quadrivalent (IIV4), 
split virus, 0.5 mL  dosage, 
for intramuscular use
o90686, Influenza virus 
vaccine, quadrivalent (IIV4), 
split virus, preservative free, 
0.5 mL dosage, for 
intramuscular use
o90630, Influenza virus 
vaccine, quadrivalent (IIV4), 
split virus, preservative free, 
for intradermal use
o90682, Influenza virus 
vaccine, quadrivalent 
(RIV4), derived from 
recombinant DNA, 
hemagglutinin (HA) protein 
only, preservative and 
antibiotic free, for 
intramuscular use
090177e1962518d5\Approved\Approved On: 29-Jan-2021 20:00 (GMT)
FDA-CBER-2021-5683-1073133
BNT162b2 (COVID -19 vaccine)
C4591011NON-INTERVENTIONAL STUDY PROTOCOL
Version 1.0, 29 January 2021
PFIZER CONFIDENTIAL
Page 96of 170Appendix Table 1. Demographic and Clinical Characteristics Definitions
Variable Description Operational definition
o90672, Influenza virus 
vaccine, quadrivalent, live 
(LAIV4), for intranasal use
o90661, Influenza virus 
vaccine, trivalent (ccIIV3), 
derived from cell cultures, 
subunit, preservative and 
antibiotic free, 0.5 mL 
dosage, for intramuscular 
use
o90658, Influenza virus 
vaccine, trivalent (IIV3), 
split virus, 0.5 mL  dosage, 
for intramuscular use
o90656, Influenza virus 
vaccine, trivalent (IIV3), 
split virus, preservative free, 
0.5 mL dosage, for 
intramuscular use
o90654, Influenza virus 
vaccine, trivalent (IIV3), 
split virus, pr eservative -free, 
for int
…[truncated]