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 170Start 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 170Fibromyalgia
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 170To 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 170Individuals 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 170Immunization 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 170Fibromyalgia
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 170Non-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
VTE714.9, 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]