Document text
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 1 of 70Study information
Title A Non-Interventional Post -Approval Safet y Study of
Pfizer-BioNTech COVID -19 Vaccine in the United States
Protocol number C4591009
Protocol version identifier 1.0
Date 19 August 2021
EU Post-Authoriz ation Study
(PAS) register numberStudy to be registered
Active substance Pfizer-BioNTech c oronavirus disease 2019 (COVID- 19)
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 severe acute respiratory
syndrome coronavirus 2 (SARS-CoV-2).
Medicinal product Pfizer-BioNTech CO VID-19 Vaccine
Research question and
objectivesThe research question is “What isthe incidence (or birth
prevalence )of safetyevents of interest among individuals
vaccinated with (or exposed in utero to) Pfizer-BioNTech
COVID-19 Vaccine compared with individuals who have
not received (or not exposed in utero to) any vaccination
for COVID -19in the United States ?”
The primary objectives areas follows :
To estimate the relative risk (RR) of safet y events of
interest following receipt of at least 1dose of Pfizer -
BioNTech COVID -19 Vaccine within the overall
study population
To estimate the RR of safety events of interest
following receipt of at least 1dose of Pfizer -
BioNTech COVID -19 Vaccine in pregnant women, in
immunocompromised individuals, and in individuals
with a history of COVID -19
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077270
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 2of 70To estimate the birth prevalence and odds ratio of
birth outcomes among pregnant women vaccinated
with the Pfizer -BioNTech COVID- 19 vaccine
compared to unvaccinated pregnant women.
The secondary objectives are as follows:
To describe the proportion of individuals receiving at
least 1dose and a complete dose series of Pfizer -
BioNTech COVID -19 Vaccine, within the overall
study population, in pregnant women, in
immunocompromised individuals, and in individuals
with a history of COVID -19
To describe —among individuals who receive a first
dose of Pfizer-BioNTech COVID- 19 Vaccine —the
timing and ty pe of second dose of COVID -19 vaccine
(Pfizer-BioNTech COVID- 19 Vaccine or other
COVID-19 vaccine), within the overall study
population, in pregnant women, in
immunocompromised individuals, and in individuals
with a history of COVID -19
To describe baseline characteristics (demographics
and comorbidities) of individuals who receive at least
1 dose of Pfizer-BioNTech COVID- 19 Vaccine and
those with no record of COVID -19 vaccination of any
type, within the overall study population, in pregnant
women, in immunocompromised individuals, and in
individuals with a history of COVID -19
Author Alison Kawai, ScD
RTI Health Solutions
Waltham, Massachusetts
Jeffrey Brown, PhD
Department of Population Medicine
Harvard Medical School & Harvard Pilgrim Health Care
Institute
Boston, Massachusetts
Cynthia de Luise, MPH, PhD
Risk Management and Safety Surveillance Research
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077271
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 3of 70Pfizer Inc.
New York, New York
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.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077272
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 4of 701.TABLE OF CONTENTS
1.TABLE OF CONTENTS.......................................................................................................4
2. LIST OF ABBREVIAT IONS................................ ................................ ................................ 6
3. RESPONSIBLE PARTI ES................................ ................................ ................................ ....9
4.ABSTRACT ................................ ................................ ................................ ......................... 11
5.AMENDMENTS AND UP DATES................................ ................................ ..................... 19
6.MILESTONES ................................ ................................ ................................ ..................... 20
7.RATIONALE AND BACKGROUND ................................ ................................ ................ 22
8. RESEARCH QUESTION AND OBJECTI VES................................ ................................ .24
9. RESEARCH METHODS................................ ................................ ................................ ....25
9.1. Study design................................ ................................ ................................ ............25
9.2. Setting ................................ ................................ ................................ ...................... 26
9.2.1. Study population ................................ ................................ ......................... 26
9.2.1.1. G eneral population, immunocompromised individuals,
and individuals with history of COVID -19................................ .......26
9.2.1.2. Pregnant women ................................ ................................ ........28
9.2.1.3. Cohorts for analy ses of pregnant women ................................ ..29
9.2.2. Follow -up................................ ................................ ................................ ....32
9.3. Variables ................................ ................................ ................................ .................. 36
9.3.1. Vaccine exposures ................................ ................................ ...................... 36
9.3.2. Outcomes ................................ ................................ ................................ ....37
9.3.2.1. Safet y events of interest ................................ ............................ 37
9.3.2.2. Outcome identification and validation ................................ ......39
9.3.3. Covariates ................................ ................................ ................................ ...39
9.3.3.1. Potential confounding variables ................................ ................ 40
9.3.3.2. Variables for identify ing subcohorts ................................ .........44
9.3.3.3. Subgroup anal ysis................................ ................................ .....45
9.4. Data sources ................................ ................................ ................................ ............45
9.4.1. CVS Health, Aetna ................................ ................................ ..................... 47
9.4.2.HealthCore ................................ ................................ ................................ ..47
9.4.3. HealthPartners ................................ ................................ ............................. 48
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077273
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 5of 709.4.4. Humana.......................................................................................................48
9.4.5. Optum Research Database ................................ ................................ ..........48
9.5. Study size................................ ................................ ................................ ................ 48
9.6. Data management ................................ ................................ ................................ ....50
9.6.1. Data collection tools (DCTs)................................ ................................ ......50
9.6.2. Record retention ................................ ................................ .......................... 51
9.6.3. Data oversight ................................ ................................ ............................. 51
9.7. Data analy sis................................ ................................ ................................ ...........53
9.7.1. Descriptive anal ysis................................ ................................ .................... 53
9.7.2.Measures of disease frequency and association ................................ ..........54
9.7.2.1. Measures of disease frequency ................................ .................. 54
9.7.2.2. Measures of association................................ ............................ 54
9.7.3. Sensitivity analysis................................ ................................ ..................... 57
9.7.3.1. Exposure misclassification................................ ........................ 57
9.7.3.2. Risk intervals................................ ................................ .............59
9.8. Quality control................................ ................................ ................................ .........59
9.9. Strengths and limitations of the research methods ................................ .................. 60
9.10. Other aspects ................................ ................................ ................................ .........61
10. PROTECTI ON OF HU MAN SUBJECTS ................................ ................................ ........62
10.1. Patient information................................ ................................ ................................ 62
10.2. Patient consent ................................ ................................ ................................ .......63
10.3. Institutional review board/Independent ethics committee (IEC).......................... 63
10.4. Ethical conduct of the study ................................ ................................ .................. 63
11. MANAGEMENT AND R EPORTING OF ADVERSE EVENTS/ADVERSE
REACTIONS................................ ................................ ................................ ...................... 63
11.1. Structured data analy sis................................ ................................ ......................... 63
11.2. Human review of unstructured data ................................ ................................ ......64
12. PLANS FOR DI SSEMINATING AND COMMUNI CATING STUDY RESUL TS........65
13. REFERENCES ................................ ................................ ................................ .................. 66
14. LIST OF TABLES ................................ ................................ ................................ .............70
15. LIST OF FI GURES................................ ................................ ................................ ...........70
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077274
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 6of 702. LIST OF ABBREVIATIONS
Abbreviation Definition
ACIP Advisory Committee on Immunization Practices
AE adverse event
AEM adverse event monitoring
BEST Biologics Effectiveness and Safety
BLA Biological License Application
BNT162b2 Pfizer-BioNTech COVID-19 vaccine
CDC Centers for Disease Control and Prevention
CI confidence i nterval
COVID-19 c oronavirus disease 2019
CPT Current Procedural Terminology
CTS Clinical Trial Services
DCT data collection tool
DP data partner
ETL Extract, Transf ormation, Load
EUA Emergency Use Authorization
EU PAS Register European Union Electronic Register of Post-A uthorisation Studies
FDA Food and Drug Administration
FISMA Federal Information Security Management Act
FU follow-up
GPP Guidelines for Good Pharmacoepidemiology Practices
GVP Good Pharmacovigilance Practices
HHR Humana Healthcare Research
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077275
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 7of 70Abbreviation Definition
HIPAA Health Insurance Portability and Accountability Actof 1996
HPHCI Harvard Pilgrim Health Care Institute
ICD-9-CM International Classification of Diseases, 9th Revision, Clinical
Modification
ICD-10-CM International Classification of Diseases, 10th Revision, Clinical
Modification
ICD-10-PCS International Classification of Diseases, 10th Revision, Procedure
Coding Sy stem
IEC Independent Ethics Committee
IMEDS Innovation in Medical Evidence Development and Surveillance
IRB institutional review board
LNP lipid nanoparticle
mRNA messenger RNA
NIS non-interventional study
NIST National Institute of Standards and Technology
PASS post-authorization safety study
PCORnet The National Patient-Centered Clinical Research Network
PHI protected health information
Q (1-4) 1st, 2nd, 3rd, or 4thQuarter
QA quality assurance
QC quality control
RR relative risk
RTI-HS RTIHealth Solutions
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077276
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 8of 70Abbreviation Definition
SAP statistical analysis plan
SARS-CoV-2 s evere acute respiratory syndrome coronavirus 2
SCDM Sentinel Common Data Model
SCRI self-controlled risk interval design
STROBE Strengthening the Reporting of Observational Studies in
Epidemiology
TBD to be determined
TORCH toxoplasmosis , other (syphilis, varicella -zoster, parvovirus B19),
rubella, cy tomegalovirus, and herpes infections
US United States
VSD Vaccine Safet y Datalink
YRR Your Reporting Responsibilities
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077277
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 9of 703.RESPONSIBLE PARTIES
Principal Investigators and Contributors to the Protocol
Name, degree (s) Job title Affiliation Address
Cynthia de Luise,
MPH, PhDSenior Director,
EpidemiologyPfizerInc. 235 E 42nd Street, Mail
Stop 219/8/01; Office 61
New York, NY 10017
USA
Alison Kawai, ScD Senior Research
EpidemiologistRTI Health Solutions 307 Waverley Oaks Rd,
Suite 101
Waltham, MA, US 02452
Jeffrey Brown, PhD Associate Professor in
the Department of
Population MedicineHarvard Medical School
& Harvard Pilgrim
Health Care Institute401 Park Drive, Suite
401 Boston, MA 02215
Catherine Johannes,
PhDSenior Director,
EpidemiologyRTI Health Solutions 307 Waverley Oaks Rd,
Suite 101
Waltham, MA, US 02452
J Bradley Layton, PhD Senior Research
EpidemiologistRTI Health Solutions 3040 East Cornwallis Rd,
PO Box 12194
Research Triangle Park,
NC US 27709
Candace Fuller, PhD Research Scientist in the
Department of
Population MedicineHarvard Medi cal School
& Harvard Pilgrim
Health Care Institute401 Park Drive, Suite
401 Boston, MA 02215
Alicia Gilsenan, PhD,
FISPEVice President,
EpidemiologyRTI Health Solutions 3040 East Cornwallis Rd,
PO Box 12194
Research Triangle Park,
NC US 27709
Brian Calingaert, MS Director, Epidemiology
AnalysisRTI Health Solutions 3040 East Cornwallis Rd,
PO Box 12194
Research Triangle Park,
NC US 27709
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077278
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 10of 70Data Research Partner Coordinating Investigators
Name, degree (s) Job title Affiliation Address
Cheryl N McMahill -
Walraven, MSW, PhDDirector, Safety
Surveillance &
CollaborationCVS Health , Clinical
Trial Services (CTS)1425 Union Meeting
Road –U21n
Blue Bell, PA 19422 -
0031
Audrey Djibo, PhD Epidemiologist CVS Health, Clinical
Trial Services (CTS)1425 Union Meeting
Road –U21n
Blue Bell, PA 19422 -
0031
Aziza Jamal -Allial, PhD Senior Epidemiologist HealthCore 480 Pleasant Street
Suite A100
Watertow n, MA 02472
Kevin Haynes, PharmD,
MSCEPrincipal Scientist HealthCore 480 Pleasant Street
Suite A100
Watertow n, MA 02472
Pamala A. Pawloski,
PharmD., BCOP, FCCPSenior Research
InvestigatorHealthPartners Institute 295 Phalen Blvd;
MS41200F
St. Paul, MN 55130
Vinit Nair, BPharm,
MS, RPhPrincipal & Director,
Government Research
& ConsortiumsHumana 515 W. Market St
Louisville, KY 40202
Ryan Seals, PhD Epidemiologist Optum 1325 Boylston Street,
Suite 1100
Boston, MA 02215
Jessica Franklin, PhD Principal Consultant,
EpidemiologyOptum 1325 Boylston Street,
Suite 1100
Boston, MA 02215
Note: Data research partner coordinating investigators have reviewed and contributed to this protocol.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077279
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 11of 704.ABSTRACT
Title: A Non -Interventional Post -Approval Safet y Study of Pfizer-BioNTech COVID -19
Vaccine in the United States
Version and Date:Version 1.0, 19August 2021
Main authors: Alison Kawai, ScD , RTI Health Solutions ; Jeffrey Brown, PhD, Department
of Population Medicine, Harvard Medical School & Harvard Pilgrim Health Care Institute ;
Cynthia de Luise, MPH, PhD , Risk Management and Safet y Surveillance R esearch, Pfizer
Inc.
Rationale and b ackground
Severe acute respiratory syndrome coronavirus 2 ( SARS-CoV-2), the cause of coronavirus
disease 2019 ( COVID-19), has resulted in a global pandemic . On 11December 2020, Pfizer -
BioNTech COVID -19 Vaccine was auth orized for emergency use by the Food and Drug
Administration (FDA) to prevent COVID -19 in individuals aged 16 years and older in the
United States (US) . On 10May 2021, Pfizer -BioNTech COVID-19 Vaccine was authorized
for emergency use in children 12-15 years of age in the US. As of 07May 2021, Pfizer-
BioNTech has initiateda Biological License Application (BLA) for marketing approval of
the vaccine for the prevention of COVID-19 in individuals aged 1 6years and older.
Post-authorization observational s tudies using real -world data are needed to assess the
association between Pfizer- BioNTech COVID -19 Vaccine and predetermined safet y events
of interest in individuals administered the vaccine in the general population and in
subpopulations of interest (e.g. ,pregnant women, immunocompromised individuals, and
individuals with a history of COVID -19). This protocol describes a proposed observational
study of safety events of interest occurring in recipients of Pfizer -BioNTech COVID-19
Vaccine using data from c laims and electronic health records (where available) from data
research partners participating in the Sentinel Sy stem. The safety events of interest in this
study are based on those included in COVID- 19 vaccine rapid cycle analysis in the FDA ’s
Biologics Effectiveness and Safety (BEST) Sy stem and the Centers for Disease Control and
Prevention (CDC) Vaccine Safet y Datalink (VSD ),with the addition of vaccine- associated
enhanced respiratory disease, immune hemoly tic anemia, and thrombotic events with
thrombocytopenia. Pregnancy safety outcomes (spontaneous abortion, stillbirth, and preterm
birth, major congenital malformations, and s mall size for gestational age ) will also be
assessed in this study. Additional safety events of interest may be added as new evidence
develops during the pandemic and the data sources permit . The proposed non- interventional
studyis designated as a post-authorization safety study (PASS)and is a commitment to the
US FDA.
Research question and objectives
The research question i s, “What is the incidence or (or birth prevalence) of safet y events of
interest among individuals vaccinated with (or exposed in utero to) Pfizer -BioNTech
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077280
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 12of 70COVID-19 Vaccine compared with individuals who have not received (or not exposed in
utero to) an y vaccination for COVID -19 in the United States?”
The primary objectives areas follows :
To estimate the relative risk (RR) of safet y events of interest following receipt of at least
1 dose of Pfizer-BioNTech COVID-19 Vaccine within the overall study population
To estimate the RR of safety events of interest following receipt of at least 1dose of
Pfizer-BioNTech COVID-19 Vaccine in pregnant women, in immunocompromised
individuals, and in individuals with a history of COVID -19
To estimate the birth prevalence an d odds ratio of birth outcomes among pregnant
women vaccinated with the Pfizer -BioNTech COVID- 19 vaccine compared to
unvaccinated pregnant women
The secondary objectives are as follows:
To describe the proportion of individuals receiving at least 1 dose and a complete dose
series of Pfizer -BioNTech COVID-19 Vaccine, within the overall study population, in
pregnant women, in immunocompromised individuals, and in individuals with a history
of COVID -19
To describe —among individuals who receive a first dose of Pfizer-BioNTech COVID-19
Vaccine—the timing and ty pe of second dose of COVID- 19 vaccine (Pfizer -BioNTech
COVID-19 Vaccine or other COVID -19 vaccine), within the overall study population, in
pregnant women, in immunocompromised individuals, and in individua ls with a history
of COVID -19
To describe baseline characteristics (demographics and comorbidities) of individuals who
receive at least 1dose of Pfizer -BioNTech COVID-19 Vaccine and those with no record
of COVID -19 vaccination of any type, within the over all study population, in pregnant
women, in immunocompromised individuals, and in individuals with a history of
COVID-19
Study design
This is a retrospective cohort study comparing vaccinated individuals with unexposed
individuals who have not received an yCOVID-19 vaccine during a concurrent time .
Exposed and unexposed individuals will be matched (in a ratio of at least 1:1) within data
source on age, state (if feasible, or broader geographic region if not feasible), calendar time,
and propensit y score (for anal ysis in the overall population, immunocompromised
individuals, and individuals with a history of COVID -19) or on maternal age, state (if
feasible), and estimated pregnancy start(for analysis in pregnant women) . In analysis of
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077281
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 13of 70pregnant women, c onfounding will be addressed through propensit y score matching or
through the inclusion of propensity scores in exposure -outcome regression models .
The study period will start on the date that Pfizer -BioNTech COVID -19 Vaccine was granted
Emergency Use Authorization (EUA) in the US ( 11December 2020) and will end a
minimum of 3 years after this date . The study will include vaccinations received under both
the EUA and the BLA (once approved).
Population
The source population for this study will be health plan enrollees from 5 data research
partners that contribute data from claims and electronic health records to the Sentinel
System: CVS Health /Aetna, HealthCore/Anthem, HealthPartners, Humana, and
Optum/UnitedHealthcare .
Individuals of all ages will be included in the descriptive anal ysis of Pfizer -BioNTech
COVID-19 Vaccine utilization . Safety analysis is planned to be limited to individuals within
the age ranges approved (either under the EUA or the BLA) to receive Pfizer-BioNTech
COVID-19 Vaccine, with age -basedeligibility criteria changing over the study period as the
ages approved for use change. However, if the proportion of Pfizer -BioNTech COVID -19
Vaccinerecipients thatfallsoutside of t he ages approved for use is greater than 1% , then
safety analyses will include individuals of all ages who have received the vaccine at any time
during the stud y period.
Individuals will be eligible for the stud y if they have continuous medical and pharma cy
insurance coverage for at least 12 months before the index date (defined in the next
paragraph) . Women will be eligible to be included in analysis of the pregnant population if
they were pregnant for at least 1 day during the study period (regardless o f the timing of
estimated pregnancy startrelative to the stud y start date ). Analysis of congenital
malformations, preterm birth, and small size for gestational age will be limited to
pregnancies ending in a live birth.
For exposed individuals in the general population, immunocompromised individuals, and
individuals with a history of COVID -19, the index date is the date of receipt of the Pfizer -
BioNTech COVID -19 Vaccine ,and this date is the start of follow -up(FU); eachdoseof
Pfizer-BioNTech COVID-19 Vaccine will contribute separate index dates . For each
individual selected to be an unexposed match , an indexdate will be assigned to a randomly
selected date in close temporal proximity (e.g., within the same calendar month) to the
vaccination date of their exposed match.
For analy ses of spontaneous abortion, stillbirth, and preterm birth, each dose of Pfizer -
BioNTech COVID -19 Vaccine will contribute separate index dates . For each individual
selected to be an unexposed match, the index date w ill beset to the equivalent of the
gestational age (in day s) at the time of vaccination of their exposed match . For analy ses of
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077282
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 14of 70small size for gestational age and congenital malformations, the index date in exposed and
unexposed individuals is the estima ted pregnancy start date .
The following subpopulations will be identified for descriptive and comparative safet y
analysis: individuals with immunocompromising conditions, pregnant women, and
individuals with a history of COVID -19. Additional subgroup analysis will be conducted b y
agegroup(<18, 18-64, ≥65 years).
Variables
Safety events
Safety events of interest will be identified in claims and electronic health records (where
available, as not all data research partners will have access to electronic health records) using
predefined algorithms based on diagnosis codes, with procedure and/or pharmacy dispensing
codes as appropriate . Algorithms for select outcomes that may be susceptible to substantial
misclassification may be validated through clinician review of medical records or patient
profiles (i.e., listings of codes in data from claims or electronic health records in
chronological order) to estimate the positive predictive values. The determination of whether
each outcome may be susceptible to substantial misclassification will be informed by clinical
expert opinion and review of prior validation studies, if available.
The following safet y events of interest (referred to as “general safet y events”)will be
assessed in the g eneral population, immunocompromised individuals (e.g., individuals with
immunodeficiencies, immunosuppressant medication use, human immunodeficiency virus or
other immunocompromising conditions, or receipt of organ or bone marrow transplant) ,
individuals with a history of COVID -19, and pregnant women :
Neurologic: acute disseminated encephalom yelitis, Bell’s palsy, convulsions,
encephalomy elitis/encephalitis, Guillain Barr ésyndrome, narcoleps y, transverse m yelitis
Cardiac: acute myocardial infarction , myocarditis/pericarditis
Hematologic: deep vein thrombosis, disseminated intravascular coagulation, immune
hemolytic anemia, immune thrombocy topenia, pulmonary embolism, thromboembolic
events associated with thrombocy topenia, thrombotic thrombocy topenic purpur a, venous
thromboembolism, hemorrhagic stroke, ischemic stroke
Respiratory : acute respiratory distress sy ndrome, vaccine-associated enhanced respiratory
disease
Other system: anaphylaxis, appendicitis, Kawasaki disease, multisy stem inflammatory
syndrome
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077283
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 15of 70The following pregnancy safety outcomes will be assessed in pregnant women or their
infants:
Spontaneous abortion (spontaneous pregnancy loss before 20 completed weeks gestation)
Stillbirth (fetal deaths at or after 20 completed weeks gestation)
Preterm bi rth(live birth before 37 completed weeks gestation)
Major congenital malformations
Small size for gestational age
Other emergent safet y events of interest may be added as the understanding of the safet y
profile of Pfizer -BioNTech COVID -19 Vaccine evolves and if the data sources permit their
assessment . For general safet y events, risk windows will be defined for outcomes that have a
hypothesized increased risk during specific time periods following vaccination . For other
general safet y events, patients will be followed for outcomes for a maximum of 1 y ear.
Vaccine exposures
Exposures to Pfizer -BioNTech COVID -19 Vaccine will be identified in data from claims and
electronic health records via pharmacy dispensing and/or procedure codes. In analyses of
pregnant women, exposures occurring within 28 day s before the estimated pregnancy start or
during pregnancy will be considered. Whereexisting linkages with immunization registries
are available for use in research studies within the appropriate data research partner
databases ,the immunization registry data will also be used to assess exposure . The
completeness of data on COVID-19 vaccines will be assessed during the monitoring phase of
the study. Based on the level of completeness of data on COVID -19vaccine exposures and
its anticipated impact on comparative risk estimates, alternative stud y designs (eg. self-
controlled anal yses and/or linkages to immunization registries willbe considered.
Covariates
Covariates will be identified in data from claims and electronic health records (where
available, as not all data research partners will have access to electronic health records) using
administrative health plan enrollee data or codes for diagnoses (with procedures or
medications, as appropriate) .The following potential variables will be identified in relation
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077284
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 16of 70to the index date (i.e., cohort entry date), to be included in descriptive analy sis and to be
considered as potential confounders in analysis of general safet y events.
Demographics (on the index date, unless otherwise noted) : age, sex, geographic region
(using the latest information available as of the index date) ,and race/ethnicity (if
feasible)
Date of Pfizer -BioNTech COVID- 19 Vaccine (categorized as appropriate, e.g., b y year or
month)and dose of vaccine received (1 or 2)
Medical history :
Comorbidities (in the 12 months before or on the index date , unless otherwise noted):
history of anaphylaxis(not including the index date) , history of allergies, diabetes
(type1,type2, gestational diabete s in current pregnancy ), hypertension,
cardiovascular disease, cerebrovascular diseases, chronic respiratory disease, chronic
kidney disease, chronic liver disease, cancer, epilepsy, autoimmune disorders,
influenza and other respiratory infections (including COVID -19),
immunocompromising conditions, gastrointestinal infections, and obesity (capture
anticipated to be incomplete)
Pregnancy status (on the index date)
Medications and non –COVID-19 vaccinations (in the 12 months before or on the
indexdate), including vaccines administered concomitantly with Pfizer -BioNTech
COVID-19 Vaccine
Healthcare utilization (in the 12 months before or on the index date) : any healthcare
encounter (including telehealth encounters, if feasible );hospitalizations; emergency
department visits; cancer screening(s); skilled nursing facility , nursing home, or
extended care facility stay; other preventive health careservices, as appropriate ; and
COVID-19 tests
For comparative analysis of pregnancy safety outcomes, the following potential confounders
and descriptive variables will be identifiedin relation to the index date .
Demographics (on the index date, unless otherwise noted) : maternal age, geographic
region(using the latest available information on index date) , race/ethnicit y (if feasible)
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077285
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 17of 70Medical history :
Comorbidities (in the 12 months before or on the index date): diabetes mellitus (type
1, type 2), hypertension, connective tissue disorders, thy roid disorders, heart disease,
epilepsy and mood disorders, asthma, liver disease, kidney disease, cancer
Obesity(in the 12 months before or on the index date; capture anticipated to be
incomplete since obesit y is not routinely documented via diagnosis or procedure
codes in claims data )
Alcohol use and smoking (in the 12 month s before or on the index date; capture
anticipated to be incomplete)
Reproductive history (in all available data) : gravidity, parity, spontaneous abortions
in previous pregnancies, and pregnancy terminations in previous pregnancies ( capture
anticipated to be incomplete)
Pregnancy complications (recorded during the pregnancy ): multiple pregnancy ,
gestational diabetes , preeclampsia/eclampsia , TORCH infections (toxoplasmosis,
other [syphilis, varicella- zoster, parvovirus B19 ], rubella, cy tomegalovirus, and
herpes infections) ; except for multiple pregnancy , only information recorded up to
and including the index date will be used to identify potential confounders
Teratogenic medications (from 28 day s before pregnancy up to and including the end
of pregnancy ); only information up toand including the index date will be used to
identifypotential confounders
Non–COVID-19 vaccinations, including those administered concomitantly with
Pfizer-BioNTech COVID-19 Vaccine (from 28 day s before pregnancy up to and
including the end of pregnancy ); only information up toand including the index date
will be used to identify potential confounders
Data sources
This study will use data from 5data research partners , including data from 4 national insurers
(CVS Health/Aet na, HealthCore/Anthem, Humana, and Optum/UnitedHealthcare) and
1regional insurer (HealthPartners) . Each data research partner is a participant in the FDA
Sentinel Sy stem. This study will use the research eligible population within the most
recently available database at each data research partner at the time of analysis . These data
sources capture longitudinal medical care information on outpatient medication dispensings,
vaccine administrations, and inpatient and outpatient diagnoses and procedures . The data
sources also capture member demographic and health plan enrollment information . Each
data research partner can request access to full- text medical records for outcom e validation
for a subset of participants. Where available, maternal data will be linked with infant data to
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077286
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 18of 70identify outcomes to be assessed in infants (small size for gestational age and major
congenital malformations).
Study size
The size of the expose d population will depend on the use of Pfizer-BioNTech COVID-19
Vaccine,and the size of the comparator population will depend on the proportion of the
source population that comprises unvaccinated individuals over time in the data sources. The
precision of comparative risk estimates will depend on the background rate and the duration
of the risk interval for each safet y event of interest . For example, for Bell’s palsy
(background rate of 25.2per 100,000 person -years and 28days at risk per dose), with
1,000,000 exposed individuals, we estimate an 86% probability that the upper bound of the
observed RRwould be below 2.0, assuming a 1 :1 ratio between vaccinated and comparator
person-time and that the true RR is 1.0.
Data anal ysis
All analyses will initially be conducted separatel y within the data from each data source .
Pooled analy sis ofRRand odds ratio estimates from all data sources will be conducted using
privacy-preserving summary -level data sets or,if this is not feas ible, meta -analysis.
Descriptive anal ysis will report on utilization of Pfizer -BioNTech COVID -19 Vaccine during
the overall study period and in sequential increments of time (to assess vaccine uptake and
patterns of exposure over time) . Characteristics (demographics, comorbidities, and other
potential covariates) of the matched and unmatched cohorts will be shown in a table.
Vaccinated individuals will be matched to concurrent unexposed comparators (in a ratio of at
least 1:1) within data source on age , state (if feasible, or broader geographic region if not
feasible), and calendar time –specific propensity scoresfor analysis in the overall study
population, immunocompromised individuals, and individuals with a history of COVID -19.
In analyses of pregnant women, pregnant women who are vaccinated will be matched to
pregnant concurrent unexposed comparators (in a ratio of at least 1:1) on maternal age , state
(if feasible, or broader geographic region if not feasible) ,and pregnancy start. Confounding
will be addressed through propensity score matching or through the inclusion of propensit y
scores in exposure -outcome regression models .
In each data source, crude measures of incidence or prevalence of the study outcomes with
associated 95% confidence interva ls (CIs) will be estimated within the matched exposed and
unexposed cohorts.
Cox models or Poisson regression will be used to estimate RRsand 95% CIs for general
safety events in the overall study population, immunocompromised individuals, individuals
with a history of COVID -19 and pregnant women, and to estimate RRs and 95% CI s for
spontaneous abortion, stillbirth, and preterm birth . For small size for gestational age and
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077287
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 19of 70major congenital malformations, logistic regression will be used to estimate odds ratios and
95%CIs.
To address the potential for misclassification of unexposed status due to incomplete capture
of vaccination exposures in claims data, sensitivity anal yses will incorporate a self -controlled
risk interval design (for outcomes with well -defined onset and risk intervals of no longer than
42 days)ora cohort design with historical comparatorsin a period before the introduction of
COVID-19 vaccines (for outcomes with gradual onset and/or risk intervals longer than 42
days). Additional se nsitivity analyses will consider alternative risk intervals for safet y events
forwhich the risk interval is not well characterized .
Milestones
The anticipated start of data collection is quarter 2 ( Q2)2022, and the end of data collection
is anticipated to be by 30 June2025. A monitoring report is planned for Q3 2022, an interim
study report for Q3 2023, and a final study report no later than Q4 2025, depending on the
extent of validation and/or the need for external linkages.
5. AMENDMENTS AND UPDATES
None
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077288
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 20of 706.MILESTONES
Below is a proposed schedule of milestones.
Milestone Planned date Description of milestone
Registration in the EU
PAS RegisterTo be
determined
(TBD)To be registered before the start of data collection
Start of data collection,
estimatedQ2 2022 Start of data collection is the planned date for the initial data
extraction for the purpose of the monitoringanalysis
Monitoring analysis
reporta,bQ3 2022 Vaccine counts and proportions of individuals in the databases who
were vaccinated, within the overall study population, in
immunocompromised individuals, in individuals with a history of
COVID-19, and in subgroups defined by age (< 18, 18-64, 65 years
and older)
Interim study report Q3 2023 Vaccine counts and proportions of individuals in the databases who
were vaccinated, within the overall study population, in
immunocompromised individuals, in individuals with a history of
COVID-19, and in subgroups defined by age (< 18, 18-64, 65 years
and older)
Distribution of characteristics among exp osed and unexposed
individuals within the overall study population, in
immunocompromised individuals, and in individuals with a history
of COVID -19
Incidence rates of safety events of interest, overall, without regard
to exposure status ,in the overall stu dy population, in
immunocompromised individuals, and in individuals with a history
of COVID -19
For select safety events of interest that have signaled in other
studies or vaccine safety surveillance systems
(e.g.,myocarditis/pericarditis) , additional analysis describing
incidence by exposure status and by select covariates of interest
may be reported
End of data collection 30 June 2025 End of data collection is the planned date on which the analytical
data set w ill first be completely available
The analytical data set is the minimum set of data required to
perform the statistical analysis for the study objectives
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077289
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 21of 70Milestone Planned date Description of milestone
Final study report Q3 2025cDescriptive analysis of vaccine utilization in the overall study
population, in immunocompromised individuals, in individuals with
a history of COVID -19, and in pregnant women
Comparative safe ty analysis in the overall study population, in
immunocompromised individuals, in pregnant women, and in
individuals with a history of COVID- 19
a.Includes only data research partners with a data lag of < 6months.
b.Monitoring counts will not incorporate enro llment or any other study eligibility criterion.
c.Report may be delayed to Q4 2025, depending on the extent of validation and/or the need for external
linkages.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077290
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 22of 707.RATIONALE AND BACKGR OUND
The severe acute respiratory syndrome coronavirus 2 (SARS -CoV-2), the cause of
coronavirus disease 2019 ( COVID-19), has resulted in a global pandemic (WHO, 2020 ). As
of 03August 2021, approximately 200million cases of COVID -19 have been reported
globally, with over 3 5million cases and 618,407 deaths reported in the United States (US)
(CSSE, 2021 ). To date, the incidence of COVID -19 has continued to rise, largel y affecting
the elderl y and middle-aged individuals, with a disproportion of cases occurring in racial and
ethnic minority populations (Lee et al., 2021 ), and with worsening clinical sequelae linked to
increasing age and comorbid condi tions (e.g., cardiovascular disease, active cancer, obesity ,
diabetes, and chronic lung disease)( CDC Covid Response Team, 2020 ; Dorjee et al., 2020 ;
Gupta et al., 2020).
Pfizer and BioNTech have developed a novel messenger RNA(mRNA)vaccine against
SARS-CoV-2 for the prevention of COVID -19 (BNT162b2) . This is a lipid nanoparticle–
formulated, nucleoside -modified mRNA vaccine that directs the host cell to produce the
SARS-CoV-2 spike protein, which is expressed on the host cell surfac e and induces
neutralizing antibod y and cellular immune responses (Lamb, 2021). The BNT162b2 vaccine
is administered intramuscularl y in a 2-dose regimen with a recommended interval between
doses of 21 days (Lamb, 2021).
Pfizer is conducting Phase 1/2/3, randomized, placebo- controlled, o bserver-blind,
dose-finding, vaccine candidate –selection, and efficacy studies among healthy individuals
(NCT04368728) . The Phase 3 clinical trial was initiated on 27July 2020, with a target
enrollment of 43,998 individuals; e fficacy and safety results from this ongoing multinational
trial were reported in December 2020 for 43,448 individuals aged 16 years orolder who
received injections (21,720 with BNT162b2 and 21,728 with placebo) and showed that a
2-dose regimen was 95% effective in preventing COVID-19 (Polack et al., 2020 ). Safety
was assessed in 37,706 participants with a median of at least 2 months of data available after
the second dose ,with most ad verse events (AEs) being transient reactogenicity events; the
most commonly solicited AEs werepain at the injection site (84.1%), fatigue (62.9%),
headache (55.1%), muscle pain (38.3%), chills (31.9%), joint pain (23. 6%),and fever
(14.2%)(Pfizer and BioNTech, 2020). The incidence of serious AEs was similar in the
vaccine (0.6%) and placebo (0.5%) groups (Polack et al., 2020). An imbalance was noted in
the occurrence of Bell ’s palsy,with 4 cases in the vaccine group and no cases in the placebo
group, but the frequency was not in excess of that expected in the general population (Pfizer
and BioNTech, 2020 ).
Efficacy and safet y results for the Phase 3 clinical trial were reported in May 2021 for 2 ,260
adolescents 12 to 15 years of age (1,131 with BNT162b2 and 1,129 with placebo) (Frenck et
al., 2021). No COVID -19 cases with onset of 7 or more day s after dose 2occurred among
recipients of BNT162b2 , and 18cases occurred among recipients of placebo , for an observed
vaccine efficac y of 100% (95% confidence interval [ CI], 78.1%-100%). Adverse events
were mainl y transient ,mild to moderate reactogenicity events such as injection site pain (in
79%-86% of stud y participants), fatigue (in 60 %-66% of participants), and headache (in
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077291
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 23of 7055%-65% of participants) . There were no vaccine -related serious AEs,and severe AEswere
very rare (in 0.2% of BNT162b2 recipients) .
On 11December 2020, Pfizer -BioNTech COVID-19 Vaccine was authorized for emergency
use by the US Food and Drug Administration (FDA) to prevent COVID -19 in individuals 16
years of age or older (Pfizer, 2020 )and on 12 December 2020, the Advisory Committee on
Immunization Practices (ACI P) issued an interim recommendation for use of the vaccine in
individuals aged 16 years orolder(Oliver et al., 2020). On 10May 2021, Pfizer -BioNTech
COVID-19 Vaccine was authorized for emergency use in children 12 -15 years of age (FDA,
2021), and on 12 May 2021, the ACIP issued an interim recommendation for the use of the
vaccine in children 12 -15 years of age (Wallace et al., 2021 ). As of 07May 2021, Pfizer-
BioNTech has initiated the submission of a Biological L icense Application (BLA) approval
authorization of Pfizer-BioNTech COVID -19 Vaccine for the prevention of SARS -CoV-2
infectionin individuals aged 1 6years orolder.
Because of the relatively short prelicensure per iod and limited number of participants in
clinical studies, monitoring of the safet y of the vaccine will be needed in the U Sin
populations large enough to detect rare possible AEsand with follow- up long enough to
evaluate the full safet y profile. The clinical study NCT04368728 did not include certain
subgroups of individuals for whom safety data about the vaccine is needed . These groups
include pregnant women, immunocompromised individuals, and individuals with a history of
COVID-19 (ClinicalTrials.gov NCT04368728, 2021 ). An ongoing clinical study
(NCT04754594) is evaluating the safet y, tolerability , and immunogenicit y of Pfizer-
BioNTech COVID -19 Vaccine in pregnant women, but the study is limited to healthy women
with uncomplicated pregnancies who were 24 to 34 weeks pregnant at the time of enrollment
(ClinicalTrials.gov NCT04754594, 2021 ). Post-authorization observational studies using
real-world data are needed to a ssess the association between Pfizer- BioNTech COVID -19
Vaccine and predetermined safet y events of interest in individuals administered the vaccine
in the general population of all ages and in subpopulations of interest (e.g., pregnant women,
immunocompromised individuals, and individuals with a history of COVID -19).
This protocol describes a proposed observational study of safety events of interest occurring
in recipients of Pfizer -BioNTech COVID -19 Vaccine using data from claims and electronic
health reco rds(where available) from data research partners participating in the Sentinel
System. The safety events of interest in this study arepartiallybased on those included in
COVID-19 vaccine safety surveillance in the FDA Biologics Effectiveness and Safet y
(BEST) S ystem (Wong et al., 2021 )and the Centers for Disease Control and Prevention
(CDC) Vaccine Safety Datalink (VSD) (Shimabukuro, 2021), with the addition of vaccine-
associated enhanced respiratory disease, immune hemoly tic anemia, and thrombotic events
with thrombocy topenia. Pregnancy safety outcomes (spontaneous abortion, stillbirth, and
preterm birth, major congenital malformations, and small size for gestational age) will also
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077292
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 24of 70be assessed in the stud y. Additional safet y events of interest may be added as new evidence
develops during t he pandemicand the data sources permit .
The proposed non-interventional study is designated as a p ost-authorization safet y study
(PASS) and is a commitment to the US F DA.
8.RESEARCH QUESTION AND OBJECTIVES
The main research question is, “ What is the inciden ce(or birth prevalence) of safet y events
of interest among individuals vaccinated with (or exposed in utero to) Pfizer -BioNTech
COVID-19 Vaccine compared with individuals who have not received (or not exposed in
utero to) any vaccination for COVID -19 in the United States?”
Primary objectives
To estimate the relative risk (RR) of safet y events of interest following receipt of at least
1 dose of Pfizer-BioNTech COVID-19 V accine within the overall study population
To estimate the RR of safety events of interest following receipt of at least 1dose of
Pfizer-BioNTech COVID-19 Vaccine in pregnant women, in immunocompromised
individuals, and in individuals with a history of COVID -19
To estimate the birth prevalence and odds ratio of birth outcomes among pregnant
women vaccinated with the Pfizer -BioNTech COVID- 19 vaccine compared to
unvaccinated pregnant women
Secondary objectives
To describe the proportion of individuals receiving at least 1 dose and a complete dose
series of Pfizer -BioNTech COVID-19 Vaccine, within the overall study population, in
pregnant women, in immunocompromised individuals, and in individuals with a history
of COVID -19
To describe —among individuals who receive a first dose of Pfizer -BioNTech COVID-19
Vaccine—the timing and ty pe of second dose of COVID- 19 vaccine (Pfizer -BioNTech
COVID-19 Vaccine or other COVID -19 vaccine), within the overall study population, in
pregnant women, in immunocompromised individuals, and in individuals with a history
of COVID-19
To describe baseline characteristics (demographics and comorbidities) of individuals who
receive at least 1dose of Pfizer -BioNTech COVID-19 Vaccine and those with no record
of COVID -19 vaccination of any type, within the overall study population, in pregnant
women, in immunocompromised individuals, and in individuals with a history of
COVID-19
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077293
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 25of 709.RESEARCH METHODS
9.1.Study design
This study will use a retrospective cohort design to compare the incidence or birth prevalence
of safety events of interest in individuals who have received Pfizer-BioNTech COVID -19
Vaccine to the incidence or birth prevalence in individual s who have no record of any
COVID-19 vaccine during a concurrent time. Vaccinated individuals will be matched to
concurrent unexposed comparators (in a ratio of at least 1: 1) within data source on age, state
(if feasible, or broader geographic region if not feasible ), calendar time , and propensity score
for analysis in the overall study population, in immunocompromised individuals, and in
individuals with a history of COVID -19. In pregnant women, those who are vaccinated will
be matched to pregnant concurrent unexposed comparators (in a ratio of at least 1: 1) within
data source on maternal age , state (if feasible, or broader geographic region if not feasible) ,
and estimated pregnancy start. In analysis of pregnant women, c onfounding will be
addressed through propensity score matching, or through the inclusion of propensity scores
in exposure -outcome r egression models .
The safety events in this study vary with respect to the onset of the event (well -defined
versus gradual) and duration of risk window (short versus long) . The strength of the cohort
design is thatit can handle a wide range of safety events with respect to these characteristics
(Baker et al., 2015 ). Contemporaneous rather than historical unexposed comparators will be
usedbecause both health care-seeking behaviors and health careutilization have changed from
the prepandemic period , which may impact outcome ascertainment . Further, COVID -19,
which will be used to identify the subpopulation of individuals with history of COVI D-19,
had been in existence for a relativel y short period of time in the U Sbefore the start of the
study period (i.e., prior to when Pfizer-BioNTech COVID- 19 Vaccine became available ).
A main limitation of the cohort design with concurrent unexposed comparators is that it may
be subject to misclassification of unexposed status, since many vaccinations occurring
outsidetraditional medical care settings without health insurance reimbursement (e.g., mass
vaccination campaigns by public health officials) may not be captured in administrative
claims dat a. To address this limitation , sensitivity analysis will incorporate a self-controlled
risk interval (SCRI ) designfor events with a well- defined onset and a short risk in terval
(Baker et al., 2015 )(defined for the purposes of this study as ≤ 42 days after vaccination). If
feasible, additional sensitivity analysis may incorporate a cohort design with historical
unexposed comparators for events with gradual onset and/or a long risk interval , except for
vaccine-associated enhanced respiratory disease(see Section9.7.3). Additionally , the
completeness of exposure information will be assessed during the monitoring phase . If
exposure information is deemed incomplete, the primary study design may be switched to the
SCRI and cohort design with historical comparators, or linkage with data from immunization
registries may be conducted (if feasible) . Another limitation of the proposed approach is that
outcomes may be misclassified in data from claims and electron ic health records. To address
this limitation, algorithms for select outcomes that may be susceptible to substantial
misclassification maybe validated through medical records or review of patient profiles
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077294
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 26of 70(i.e.,chronological listing of a patient’ s codes in data from claims and/or electronic health
records) (see S ection 9.3.2.2). The determination of which outcomes may be susceptible to
substantial misclassification will be informed by clinical expert opinion and review of
validation studies if available .
9.2. Setting
The study period will start on the date that Pfizer -BioNTech COVID -19 Vaccine was granted
EUAin the US ( 11December 2020) and will end a minimum of 3 years after this date . The
study will include both vaccinations received under the EUA and under the BLA (once
approved).
The source population for this study will be health plan enrollees from 5data research
partnersthat contribute data from claims and electronic health records to the Sentinel
System: CVS Health /Aetna, HealthCore/Anthem, HealthPartners, Hu mana, and
Optum/UnitedHealthcare . The data sources are described in more detail in S ection 9.4.
9.2.1.Study population
Individuals of all ages will be included in the descriptive anal ysis of Pfizer -BioNTech
COVID-19 Vaccine utilization . Safety analysis is planned to be limited to individuals within
the age-approved population for Pfizer -BioNTech COVID-19 Vaccine, with age -based
eligibility criteria changing over the study period as the age s approved for use of the vaccine
change. However, if the proportion of Pfizer -BioNTech COVID-19 Vaccinerecipients that
fall outside of the approved age rangeis greater than 1% , then safet y analyses will include
individuals of all ages who have received the vaccine at an y time during the study period.
Additional e ligibility requirements for the general population and subpopulations of interest
are described below .
9.2.1.1.General population, immunocompromised individuals, and individuals with
history of COVID -19
To be included in the general population, population of immunocompromised individuals
(individuals with immunodeficiencies, immunosuppressant medication use, human
immunodeficiency virus and other immunosuppressing conditions, and receipt of organ or
bone marrow transplant) , and population of individuals with history of COVID -19, patients
must have continuous medical and pharmacy insurance coverage for at least 12 months
before each index date , (as defined in Table 1and in Table 3in Section 9.2.1.3).
9.2.1.1.1. Cohorts for analyses of general population, immunocompromised i ndividuals,
and individuals with history of COVID-19
Because Pfizer -BioNTech COVID-19 Vaccine is currentl y recommended in a series of
2doses given at least 3 weeks apart, a separate exposed cohort will be formed for each of the
2doses. A separate unexposed cohort will be formed for each of the 2 exposed cohorts to
serve ascomparator cohorts. If approved in the US, third dose sof the Pfizer -BioNTech
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077295
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 27of 70COVID-19 Vaccine will be included in the study using methods to be described in the
statistical ana lysis plan (SAP).
The criteria for the cohorts for anal yses in the general population, immunocompromised
individuals, and individuals with history of COVID -19 is described in Table 1.
Table 1. Cohort definitions for analyses of general population,
immunocompromised individuals, and individuals with history of
COVID-19
Cohort Inclusion criteria Index date Exclusion criteria
Dose 1
exposedRecord of a first dose of Pfizer -
BioNTech COVID -19 Vaccine during
study periodDate of vaccination Record of any
COVID-19 vaccine
before the index date
Dose 2
exposedRecord of a second dose of Pfizer -
BioNTech COVID -19 Vaccine during
study periodDate ofvaccination Record of COVID-19
vaccine other than
Pfizer-BioNTech
COVID-19Vaccine
before the index date
Dose 1
unexposed No record of any COVID -19 vaccine
as of the index date (i.e., cohort entry)
and matched individually to
individuals in the dose1 exposed
cohort on time -specific propensity
score and state (if feasible, or broader
geographic region if not feasible)Randomly selected date in
a period w ithin close
temporal proximity to the
date of vaccination in
corresponding exposed
individuals (e.g., within the
same calendar month)Record of any
COVID-19 vaccine
before the index date
Dose 2
unexposed No record of any COVID -19 vaccine
as of the index date (i.e., cohort entry)
and matched individually to
individuals in the dose2exposed
cohort on time -specific propensity
score and state (if feasible, or broader
geographic region if not feasible)Randomly selected date in
a period w ithin close
temporal proximity to the
date of vaccination in
corresponding exposed
individuals (e.g. , within the
same calendar month)Record of any
COVID-19 vaccine
before the index date
Individuals in the dose1and dose2 cohorts will be matched to unexposed individuals (in a
ratio of at least 1: 1) on age, state (if feasible, or broader geographic region if not feasible) ,
and calendar time –specific propensity score, which is described in Section 9.7.2.2.1.1 .
Individuals who receive 2 doses of the vaccine can contribute to both the dose1 exposed and
dose2exposed cohorts. Individuals in the dose 1unexposed and the dose2unexposed
cohortsmay also contribute to the exposed cohorts if they subsequently receive Pfizer-
BioNTech COVID -19 Vaccine. Conversel y, individuals in the dose 1cohort maybe eligible
for the unexposed cohort sbefore they receive their first dose of Pfizer-BioNTech COVID-19
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077296
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 28of 70Vaccine. However, an individual womanmay contribute only once tothe dose1unexposed
cohort and once to thedose2unexposed cohort.
As appropriate , data from the dose1and dose2cohorts and their matched unexposed
comparators will be pooled to obtain RRestimates corresponding to receipt of at least 1dose
of Pfizer-BioNTech COVID-19 Vaccine.
9.2.1.2.Pregnant women
To be eligible for anal ysis of the pregnant population, women must have been pregnant for at
least 1 day during the study period (regardless of the timing of estimated pregnancy start
relative to the study start date) and had a pregnancy outcome (e.g., live birth, stillbirth,
spontaneous abortion, ectopic pregnancy ) recorded in the data sources during the study
period. Pregnant women will be required to be continuously enrolled in their health plan
from 12 months before the index date (as defined in Table 3in Section 9.2.1.3) until the end
of pregnancy . An algorithmic approach (to be describedin the SAP) will be used to identify
pregnancies via their outcomes (e.g., live birth, stillbirth, spontaneous abortion, termination,
ectopic pregnancy )in women of reproductive age . The algorithm will use diagnosis and/or
procedure codes toidentifythe final pregnancy outcome of each pregnancy episode, as well
as the start and end dates of pregnancy .
In pregnant women, safety events assessed in the general population (hereafter referred to as
“general safety events”; see Table 6in Section 9.3.2.1for list of outcomes), as well as
pregnancy safety outcomes (spontaneous abortion, stillbirth, preterm birth, major congenital
malformations, and small size for gestational age), will be assessed .
The eligible populations and e xposure windows for anal yses in pregnant women differ by
safety event and are listed in Table 2. Only Pfizer-BioNTech COVID - 19 Vaccine doses
administered during the exposure windows will be considered for inclusion in the exposed
cohorts. Women who were vaccinated outside the exposure window ( e.g., women who were
pregnant during the study period but who were vaccinated after they gave birth )will not
contribute to the exposed cohorts but may contribute unexposed person- time. Women with
Pfizer-BioNTech COVID-19 Vaccine administrations within 4 weeks before the estimated
pregnancy start will be considered for inclusion in the exposed cohorts because of the
imprecision of estimating pregnancy start date in claims data and because this period may be
of etiologic interest for some of these s afety events. If published validation studies
suggestingthat gestational algorithms are precise become available before the start of the
study, thestart of the exposure window may be redefined to be closer to pregnancy start(to
be documented in the SAP).
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077297
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 29of 70Table 2.Eligible populations and exposure windows for analyses of pregnant
women
Event Eligible population Exposure window
General safety eventsaAll eligible pregnancies 4 weeks before estimated pregnancy start to end
of pregnancy
Pregnancy safety outcomes
Spontaneous abortion All eligible pregnancies 4 weeks before estimated pregnancy start to end
of pregnancy or 19-6/7 weeks of gestation ,
whichever is earlier
Stillbirth Pregnancies with gestational
age ≥20weeks4 weeks before estimated pregnancy start to end
of pregnancy
Preterm birth Live deliveries 4 weeks before estima ted pregnancy start to end
of pregnancy or 36-6/7 weeks of gestation ,
whichever is earlier
Major congenital
malformationsLive deliveries with linkage
to infant data available 4 weeks before estimated pregnancy start to
13-6/7 weeks of gestation (end of first trimester)
Small size for gestational
ageLive deliveries with linkage
to infant data available4 weeks before estimated pregnancy start to end
of pregnancy
a.Safety events of interest listed in Table 6in Section 9.3.2.1.
Analyses of major congenital malformations and small size for gestational age will require
pregnancies to be linked to infants who have health plan enrollment from birth until age
3months(or from birth until deathif the infant dies before age 3 months). Where available,
the Sentinel mother -infant linkage table (see Section 9.4)included in the most recentl y
approved ETL(Extract, Transform, Load) at the time of the data extraction will be usedfor
this study. Mother-infant linkage algorithms differ by data source and may usesubscriber
identification numbers and/or names and addresses; these algorithms will be described in the
SAP.
9.2.1.3.Cohorts for analyses of pregnant women
In pregnant women, different cohorts will be formed to assess specific groups of outcomes
(Table 3). For analysis of pregnant wome n (similar to the anal ytic approach in the general
population), separate cohorts will be formed for each dose of Pfizer -BioNTech COVID -19
Vaccine to assess general safet y events, spontaneous abortion, stillbirth, and preterm birth .
For analy sis of small size for gestational age and major cong enital malformations, separate
cohorts of women exposed to Pfizer -BioNTech COVID- 19 Vaccine and women not exposed
to any COVID-19 vaccine during the exposure window (without regard to dose number) will
be formed .
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077298
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study 20-May-2021
Page 30of 70Table 3.Cohort definitions for analyses of pregnant women
Cohort Inclusion criteria Index date Exclusion criteria
Analysis of general safety eventsa, spontaneous abortion, stillbirth, and preter m birth
Dose 1 exposed Record of a first dose of Pfizer -BioNTech COVID -19
Vaccine during the exposure windowbDate of vaccination Record of any COVID-19 vaccine other than
Pfizer-BioNTech COVID -19 Vaccine before the
index date ( or during the exposure window before
the index date in analyses of spontaneous abortion,
stillbirth, and preterm birth )
Dose 2 exposed Record of a second dose of Pfizer -BioNTech COVID -19
Vaccine during the exposure window .
For analysis of general safety events, the first dose is not
required to have been administered during the exposure
window.
For analysis of spontaneous abortion, stillbirth, and
preterm birth, t he first dose must also have been
administered during the exp osure window . Date of vaccination Record of any COVID-19 vaccine other than
Pfizer-BioNTech COVID -19 Vaccine before the
index date (or during the exposure window before
the index date in analyses of spontaneous abortion,
stillbirth, and preterm birth)
Dose 1
unexposed No record of any COVID-19 vaccine as of the index date
(or no record of any COVID-19 vaccine during the
exposure window before the index date in analyses of
spontaneous abortion and stillbirth )
Individually matched to w omen in dose1cohort on
maternal age and pregnancy startEquivalent of the
gestational age (in days)
at the time of the
exposed individual’ s
vaccinationcRecordof any COVID -19 vaccine as of the index
date
Dose 2
unexposed Same criteria as dose1unexposed cohort, but matched
individually to w omen in dose2exposed cohortSame criteria as dose1
unexposed cohortSame criteria as dose1unexposed cohort
Analysis of small size for gestational age and m ajor congenital m alformations
Exposed Record of at least 1dose of Pfizer -BioNTech COVID -19
Vaccine during the exposure windowEstimated pregnancy
startNo record of Pfizer -BioNTech COVID -19 Vaccine
during the exposure window
Unexposed No record of any COVID-19 vaccine during the exposure
window, individually matched to w omen in exposed
cohort on maternal age and pregnancy startEstimated pregnancy
startRecord of anyCOVID-19 Vaccine during the
exposure window
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077299
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study 20-May-2021
Page 31of 70Table 3.Cohort definitions for analyses of pregnant women
Cohort Inclusion criteria Index date Exclusion criteria
a.Safety events of interest listed in Table 6in Section 9.3.2.1.
b. Pregnant w omen who have received their first dose of the Pfizer -BioNTech COVID -19 Vaccine before the exposure window and their second dose of
Pfizer-BioNTech COVID -19 Vaccine during the exposure window may contribute (via their second doses) to the exposed cohorts if all other eligibility criteria
are met. For analyses of general safety events, their second dose could contribute to the dose2exposed cohort, whereas for analyses of spontaneous abortion,
stillbirth, and preterm birth, their second dose could contribute to the dose1exposed cohort(in this situation, the first dose w ould be excluded from analysis ).
c.Following exposed and unexposed w omen for the same duration will ensure that w omen have equal opportunity to be exposed durin g pregnancy .
Furthermore, starting follow -up at the same gestational age in exposed and unexposed w omen will ensure that both groups have a comparable at -riskperiod
with respect to gestational age .
Note: Individuals in the dose1unexposed and dose2unexposed cohorts may contribute to the dose 1and/or dose2 exposed cohorts if they subsequently
receive the Pfizer -BioNTech COVID -19 Vaccine during the exposure window . Conversely, pregnant women in the dose1cohort may be eligible to
contribute to the dose1unexposed or dose 2unexposed cohorts before they receive their first dose of Pfizer -BioNTech COVID -19 Vaccine. How ever, an
individual woman may contribute only once to the dose1unexposed cohort and once to the dose2unexposed cohort .
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077300
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 32of 709.2.2.Follow-up
Events that can define the start and end of follow -up are provided in Table 4. Follow-up will
end at the earliest of all possible events that define the end of follow -up.
Table 4.Events defining the start and end of follow -up
Safety event of
interestEvents defining the start
of follow-upEvents defining the end of follow -upa
General safety
eventsbIndex date (if the risk
interval starts on day
0),or
One day after the index
date(if the risk interval
starts on day 1; see
Table 6in
Section9.3.2.1for
outcome- specific risk
windows)End of the study period
End of data availability
Disenrollment from the health plan
Death
Occurrence of the safety event of interest
End of duration of the outcome -specific risk
window (maximum of 1year; see Table 6in
Section 9.3.2.1for outcome -specific risk
windows)
Receipt of a dose of the Pfizer -BioNTech
COVID-19 Vaccine or any other COVID -19
vaccinec
Pregnancy safety
outcomes
Spontaneous
abortionOne day after theindex
date (if the index date
is on or after estimated
pregnancy start) ,or
Estimated pregnancy
start End of pregnancy
20 weeks of gestation
Receipt of a dose of Pfizer -BioNTech COVID -19
Vaccine or any other COVID -19 vaccinec
Stillbirth One day after the index
date (if the index date
is on or after 20 w eeks
of gestation) ,or
20 weeks of gestationEnd of pregnancy
Receiptof a dose of Pfizer -BioNTech COVID -19
Vaccine or any other COVID -19 vaccine3
Preterm birth Index date (if the index
date is on or after 20
weeks of gestation) ,or
20 weeks of gestationEnd of pregnancy
36 6/7 weeks of gestation
Receipt of a dose of Pfizer -BioNTech COVID -19
Vaccine or any other COVID -19 vaccinec
Major congenital
malformationsdDay of birth End of study period
End of data availability
Disenrollment from the health plan
Death
Diagnosis of major congenital malformation
Age 1year
Small size for
gestational agedDay of birth Predefinedperiod shortly after birth (specific
period to be defined in the SAP)
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077301
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 33of 70Table 4.Events defining the start and end of follow -up
Safety event of
interestEvents defining the start
of follow-upEvents defining the end of follow -upa
a.The earliest of the listed events will mark the end of follow -up for each outcome. If follow -up ends due
to the occurrence of a particular safety event of interest, follow -up will continue for all other safety
events.
bSafety events of interest are listed in Table 6in Section 9.3.2.1.
c.If dose 2 is Pfizer -BioNTech COVI D-19 Vaccine, individuals will stop follow -up in the dose 1 cohort
and may start follow -up in the dose 2 cohort. When the risk interval for dose 1 overlaps w ith the risk
interval for dose 2 and the risk interval definition includes day 0, the date of rece ipt of dose 2 will be
included in follow -up of dose 1. Otherw ise, the date of receipt of dose 2 will be included in follow -up
of dose2.
d.Major congenital malformations and small size for gestational age will be assessed in the infant.
For illustrative purposes, Figure1depicts the timelines of hy pothetical patients to show the
concepts of eligibility for study cohorts and follow -upinanalysis of general safet y eventsin
the general population, immunocomprom ised individuals, orindividuals with history of
COVID-19. Figure2depicts the timelines of hy pothetical patients to illustrate eligibility for
the study cohorts and f ollow-up, using anal ysis of spontaneous abortion as an example.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077302
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 34of 70Figure1.Hypothetical patients to illustrate eligibility for study cohorts and follow- up
of general safety events in the general population, immunocompromised
individuals, or individuals with history of COVID -19
Note: If follow -up ends due to the occurrence of a particular safety event of interest, follow -up will continue for
all other safety events.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077303
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 35of 70Figure2. Hypothetical patients to illustrate eligibility for study cohorts and follow- up
of spontaneous abortion events in pregnant women
a.End of exposure window is 20 weeks of gestation or pregnancy end, whichever is earlier.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077304
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 36of 709.3.Variables
9.3.1.Vaccine exposures
The primary exposure of interest is Pfizer -BioNTech COVID -19 Vaccine(receipt of Pfizer -
BioNTech COVID -19 Vaccine versus no receipt of any COVID-19 vaccine) . Receipt of
Pfizer-BioNTech COVID-19 Vaccine will be identified in data from claims and e lectronic
health records (where available, as not all data research partners will have access to
electronic health records) via pharmacy dispensing and/or procedure codes ( Table 5). Where
existing linkages with immunization registries are available for use in research studies within
the appropriate participating research databases, immunization registry data will be
combined
with data from claims and electronic health r ecordsto identify vaccine exposures.
Receipt of other COVID -19 vaccines available in the US will be identified in a similar
manner. The vaccines in Table 5 are those that have received authorization in the US as of
the time of writing of this protocol. Other COVID -19 vaccines that are authorized during the
study period will be added, as needed . Rules to handle de -duplication of codes and/or
implausible spacing of doses (e.g., 2 codes for COVID-19 vaccine within 2days) will be
described in the SAP .
Table 5. Codes for COVID- 19 vaccines available in the U nited Statesas of 28July
2021
COVID-
19
Vaccine Role of variable Vaccine codes Vaccine
administration
codesNational Drug
Codes
Pfizer-
BioNTech
COVID-19
VaccinePrimary
exposure, used
to define
exposed cohorts 91300(severe acute respiratory
syndrome coronavirus 2 (SARS-
CoV-2) (coronavirus disease
[COVID- 19]) vaccine, mRNA /lipid
nanoparticle( LNP), spike protein,
preservative free, 30 mcg/0.3mL
dosage, diluent reconstituted, for
intramuscular use )001A (firstdose)
002A (second
dose)59267-1000-1
59267-1000-01
59267-1000-02
59267-1000-03
Moderna
COVID-19
VaccineExclusion
variable before
index date;
censoring
variable during
follow-up91301(severe acute respiratory
syndrome coronavirus 2 (SARS -
CoV-2) (coronavirus disease
[COVID- 19]) vaccine, mRNA/LNP,
spike protein, preservative free,
100mcg/0.5mL dosage, for
intramuscular use)0011A (first
dose)
0012A (second
dose)80777-273-10
80777-0273-10
80777-0273-99
Janssen
COVID-19
VaccineExclusion
variable before
index date;
censoring
variable during
follow-up91303(severe acute respiratory
syndrome coronavirus 2 (SARS -
CoV-2) (coronavirus disease
[COVID- 19]) vaccine, DNA, spike
protein, adenovirus type 26 (Ad26)
vector, preservative free, 5 x1010
viral particles/0.5mL dosage, for
intramuscular use)0031A (single
dose)59676-580-05
59676-0580-05
59676-580-15
Sources: AMA (2021); DailyMed (2021) .
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077305
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 37of 70If COVID -19 vaccines are administered without reimbursement from health insurers , there is
the potential that they will not be recorded in claims and electronic health records . This
situation may lead to misclassification of trul y exposed individuals as “unexposed ”
comparators, which will underestimate vaccine coverage rates and may bias comparative risk
estimates for the cohort design with concurrent unexposed comparators . The completeness
of exposure data will be assessed in monitoring analy ses before the final analy ses are
conducted by comparing study data with publicly available estimates of vaccine coverage
and/or estimates based on imm unization registry data from select states (if available); if the
coverage estimates differ meaningfully from the “benchmarking” estimates (based on to-be-
defined criteria in the SAP) , then modifications to the study approach may be considered . If
this happens, the SCRI and the cohort design with historical unexposed comparators may be
designated as the primary study designsand/or linkage to immunization registries may be
considered if feasible.
9.3.2.Outcomes
9.3.2.1.Safety events of interest
Safety events of interest to be assessed in the general population, immunocompromised
individuals, individuals with a history of COVID -19, and pregnant women , and their risk
interval definitions ,are listed in Table 6. Throughout the protocol, these safety events are
referred to as “general safety events
.”
Theseevents comprise outcomes being monitored in rapid- cycle analysis of COVID -19
vaccines in the FD A’s BEST Sy stem (Wong et al., 2021 )andthe CDC’s VSD (Wong et al.,
2021), with the addition of vaccine- associated enhanced respiratory disease(Munoz et al.,
2021), immune hemolytic anemia, and thromboembolic events with thrombocy topenia.
Other safety events of interest may be added as the under standing of the safety profile of
Pfizer-BioNTech COVID-19 Vaccine evolves and feasibility of their assessment permits in
the data sources.
Table 6.General safety events to be assessed in the general population,
immunocompromised individuals, individuals with a history of COVID -
19, and pregnant women
Organ
systemSafety event of interest Risk window (days following
receipt of Pfizer- BioNTech
COVID-19 Vaccine)a
Neurologic Acute disseminated encephalomyelitis 1-42
Bell’s palsy 1-42
Convulsions 1-42
Encephalomyelitis/encephalitis 1-42
Guillain-Barré syndrome 1-42
Narcolepsy 1-180
Transverse myelitis 1-42
Cardiac Acute myocardial infarction 1-28
Myocarditis/pericarditis 1-21b
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077306
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 38of 70Table 6.General safety events to be assessed in the general population,
immunocompromised individuals, individuals with a history of COVID -
19, and pregnant women
Organ
systemSafety event of interest Risk window (days following
receipt of Pfizer- BioNTech
COVID-19 Vaccine)a
Hematologic Deep vein thrombosis 1-28
Disseminated intravascular coagulation 1-28
Immune hemolytic anemia 1-42
Immune thrombocytopenia 1-42
Pulmonary embolism 1-28
Thromboembolic events associated w ith thrombocytopenia 1-28
Thrombotic thrombocytopenic purpura 1-28
Venous thromboembolism 1-28
Hemorrhagic stroke 1-28
Ischemic stroke 1-28
Respiratory Acute respiratory distress syndrome 1-28
Vaccine-associated enhanced respiratory disease 1-365
Other system Anaphylaxis 0-1
Appendicitis 1-42
Kawasaki disease 1-42
Multisystem inflammatory syndrome 1-42
a.Time interval following vaccination when patients will be follow ed for safety events of interest. Day 0
refers to the day of vaccination.
b.Sensitivity analysis will assess alternative risk interval definitions of 1 -7 and 1-14 days.
The following pregnancy safety outcomes will be assessed in pregnant women or their
infantsbased on diagnosis codes because clinical information, such asgestational age
estimates and ultrasound results, will not be available in electronic data :
Spontaneous abortion: s pontaneous pregnancy loss before 20 completed weeks gestation
Stillbirth: fetal deaths at or after 20 completed weeks gestation
Preterm birth : livebirth before 37 completed weeks gestation
Major congeni tal malformations: major congenital malformations will be identified in
live-born infants using code lists from the National Birth Defects Prevention Network
(NBDPN, 2021a ), whichdefines a major defect as a congenital abnormality that requires
medical or surgical treatment,has a serious adverse effect on health and development, or
has significant cosmetic impact( NBDPN, 2021b )
Small size for gestational age : less than 10th percentile of weight for gestational age,
based on diagnosis codes for small size for gestational age or combinations of diagnosis
codes for birthweight (in categories) and gestational age (in categories)
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077307
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 39of 709.3.2.2. Outcome identification and validation
Safety events of interest will be identified with a date of diagnosis indata from claims or
electronic health records (where available, as not all data research partners will have access
to electronic health records) using predefined algorithms based on codes for diagnoses (with
codes for procedures and/or treatmentsif appropriate for the outcome) . As possible,
validated algorithms will be used . Detailed algorithm definitions (including washout periods
to define incident events and the medical care settings in which safet y events will be
identified )and code lists will be included in the SAP.
Algorithms for select outcomes that are susceptible to substantial misclassification may be
validated. The determination of whether each outcome is susceptible to substantial
misclassification will be informed by clinical expert opinion and review of prior validation
studies,if available . For the outcomes selected for validation, c linician review of medical
records or patient profiles (i.e., listings of codes in data from claims or electronic health
records in chronological order ) will be conducted on a sample of cases to estimate the
positive predictive values of case-finding algorithms and to estimate the proportion of cases
that were accuratel y identified as occurring during the risk interval . To the greatest extent
possible, such reviews will be conducted without knowledge of vaccination status . Clinical
definitions of safety eventsof interest will be based on Brighton Collaboration definitions ,
where available , or other clinical definitions from published literature if medical record
review is implemented .
For safety events of interest selected for algorithm validation, a sampling strategy that
considers the rarity of events will be used to identify cases that will undergo clinician review .
If the outcome algorithm performs adequatel y (using pre definedcriteriato be defined in a
separate data validation plan ), then all algorithm -identified cases will be included in the final
analysis. If the algorithm does not perform adequately, then the validation results will be
used to inform or adjust RRsor odds ratios that wereestimated from electronic data (e.g., in
quantitative bias analyses). The sampling strategy , selection of safet y events for which
algorithms will be validatedand rationale for the selection, details on the methods for
validation (including criteria for determining whether algorithms have performed
adequatel y), and the plan for integrating the validation results into final analy sis will be
described separatel y in a data validation plan .
9.3.3.Covariates
Covariates will be identified in claims and electronic health records (where available, as not
all data research partner s will have access to electronic health records) using administrative
health plan enrollee data or codes for diagnoses (with procedures or medications, as
appropriate ). Detailed algorithms and code lists to identify each covariate will be included in
the SAP.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077308
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 40of 709.3.3.1.Potential c onfounding variables
The following variables will be used to describe the overall stud y population,
immunocompromised individuals, individuals with a history of COVID -19, and pregnant
womenand will be considered as potential confounding va riables to be included in
propensity score models foranalysis of general safety events.
Demographics will be evaluated at the index date of each patient, unless otherwise noted.
Age: Individuals of all ages will be included in the study ; for descriptive analyses,
age categories will be 0 -4, 5-11, 12-15, 16-20, 21-29, 30 -49, 50-64, 65-80, and
greater than 80 years.
Sex
Geographic region (zip code where available, state or census region ; evaluated using
the latest information available on the index date)
Race/ethnicit y:Data on race/ethnicit y are anticipated to be incomplete, and the
feasibility of including this variable in the analy ses will be assessed before study start.
Date of Pfizer-BioNTech COVID-19 Vaccine(categorized as appropriate, e.g., b y yearor
month)
Dose of vaccine received (1 or 2)
Comorbidities, identified in the 12 months before and including the index date (unless
otherwise noted) in claims or electronic health records using diagnosis codes (with
procedure and/or pharmacy dispensing cod es as appropriate )
History of anaphylaxis (not including the index date)
History of allergies
Diabetes mellitus (ty pe1, type2, or gestational diabetes in current pregnancy )
Hypertension
Cardiovascular disease
Cerebrovascular disease
Chronic respiratory disease
Chronic kidney disease
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077309
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 41of 70Chronic liver disease
Cancer
Epilepsy
Autoimmune disorders
Influenza and other respiratory infections (including COVID -19)
Gastrointestinal infections
Immunocompromising conditions , including the following ( to be defined in m ore
detail in the SAP):
Immunodeficiencies
Immunosuppressant medication use
Human immunodeficiency virus and other immunosuppressing conditions
Receipt of organ or bone marrow transplant
Obesity (to be identified with proxies using diagnosis and procedure codes; capture
anticipated to be incomplete )
Pregnancy status(on the index date) , identified using an algorithm as described in
Section 9.2.1.2
Medications in the 12 months before and including the index date , identified in data from
claims or electronic health records via procedure and/or pharmacy dispensing codes
Analgesics
Antibiotics
Antiviral medications
Corticosteroids
Nonsteroidal anti -inflammatory drugs
Psychotropics
Statins
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077310
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 42of 70Novel oral anticoagulants
Warfarin
Non–COVID-19 vaccinations (including those administered concomitantly with Pfizer-
BioNTech COVID -19 Vaccine) in the 12 months before or on the index date, identified
in data from claims or electronic health records with procedure and/or pharmacy
dispensing codes
Influenza
Pneumococcal disease
Diphtheria, tetanus, and pertussis
Polio
Measles, mumps, and rubella
Haemophilus influenzae ty pe b
Hepatitis B virus
Human papillomavirus
Meningitis
Rotavirus
Varicella
Herpes zoster
Healthcare utilization (including telehealth encounters, if feasible) in the 12 months
before or on the index date
Any healthcareencounter
Hospitalizations
Emergency department visits
Skilled nursing facilit y, nursing home ,or extended care facility stay
Cancer screening (s)
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077311
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 43of 70Other preventive health careservices, as appropriate
COVID-19 tests
The following variables will be used to describe the population of pregnant women and will
be considered as potential confounders to be included in propensity score models foranalysis
of pregnancy safety outcomes . Except where noted, o nly information up to and including the
index date will be used to ident ify potential confounders.
Demographics: maternal age( on the index date), geographic region (using the latest
information available as of the index date) , race/ethnicity (if feasible, on the index date)
Comorbidities, identified in the 12 months before and including the index date in claims
or electronic health records using diagnosis codes (with procedure and/or pharmacy
dispensing codes as appropriate ).
Diabetes mellitus (type 1, type 2)
Hypertension
Connective tissue disorders
Thyroid disorders
Heartdisease
Epilepsy and mood disorders
Asthma
Liver disease
Kidney disease
Cancer
Obesity, identified in the 12 months before or on the index date with proxies based on
diagnosis and procedure codes ( capture anticipated to be incomplete since documentation
via diagnosis or procedure codes is not routinel y done in claims data )
Alcohol use , identified in the 12 months before or on the index date with proxies based
on diagnosis and procedure codes ( capture anticipated to be incomplete)
Smoking, identified in the 12 months before or on the index date with proxies based on
diagnosis and procedure codes ( capture anticipated to be incomplete)
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077312
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 44of 70Reproductive history identified in all available data (capture anticipated to be incomplete
because data will be limite d by duration of enrollment before index date)
Gravidity , identified via diagnosis codes
Parity, identified via diagnosis or procedure codes for deliveries or C- sections
Spontaneous abortions in previous pregnancie s, identified via diagnosis codes
Pregnancy terminations in previous pregnancies , identified via diagnosis codes
Multiple pregnancy, identified during pregnancy via diagnosis codes . Information
recorded after the index date will be used to identify potential confounders .
Gestational diabetes, identified during pregnancy via diagnosis codes
Preeclampsia /eclampsia , identified during pregnancy via diagnosis codes
TORCH infections (toxoplasmosis, other [ syphilis, varicella-zoster, parvovirus B19],
rubella, cy tomegalovirus, and herpes infections) , identified during pregnancy via
diagnosis codes
Teratogenic medications from 28 day s before pregnancy up to and including the end of
pregnancy , identified via pharmacy dispensing and procedure codes
Vaccinations other than COVID- 19 vaccinations (including those administered
concomitantly with Pfizer -BioNTech COVID-19 Vaccine)from 28 day s before
pregnancy up to and including the end of pregnancy , identified via pharmacy dispensing
and procedure codes
9.3.3.2. Variables for identifying subcohorts
Three sub cohortswill be identified from the general population cohort in which descriptive
and comparative analyses of general safet yevents will be conducted: individuals with
immunocompromising conditions , pregnant women, and individuals with a history of
COVID-19. Comparative safet y analysesof pregnancy safety outcomes will also be
conducted in pregnant women. These 3subcohorts will be identified as follows:
Immunocompromising conditions will be identified using diagnosis, procedure, and
medication dispensing codes in all available databefore the index date for
immunodeficiencies, immunosuppressant medication use, human immunodeficiency
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077313
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 45of 70virus and other immuno suppressing conditions, and receipt of organ or bone marrow
transplant.
Pregnancy status will be identified using an algorithm as described in Section 9.2.1.2.
History of COVID -19 (in all available data ) will be identified in data from claims or
electronic health records via I CD-10-CM(International Classification of Diseases, 10th
Revision, Clinical Modification )diagnosis codes for COVID -19 (U07.1), personal
history of COVID -19 (Z86.16) ,or pneumonia due to coronavirus disease 2019 (J12.82) .
If feasible, a positive laboratory test result for SARS -CoV-2 will be incorporated into the
definition.
9.3.3.3.Subgroup analysis
Subgroup anal ysis will be conducted in the following age categories: 0 -17, 18-64, 65 years
and older.
In additional anal yses, the following safety events will bestudied among individuals of
specific ages if sufficient numbers of exposures are identified within these age groups :
Multisystem inflammatory syndromein children: ages 0-20 years
Convulsions: ages 0-4 years
Kawasaki disease: ages 0-4years
Myocarditis/pericarditis: ages 12-29 years
Additional subgroup analysis (e.g., stratified by other demographic variables or calendar
time) may be conducted for specific safet y events of interest that have signaled in other
studies or vaccine safet y surveillance s ystems(e.g.,myocarditis/pericarditis) .
Additionally , analysis to identify risk factors for postvaccine outcomes may be conducted for
specific safety events of interest (e.g., m yocarditis/pericarditis) .
These analy ses will be described in the SAP .
9.4.Data sources
This study will use data from 5data research partners , including data from 4 national insurers
(CVS Health/Aetna, HealthCore /Anthem, Humana, and Optum/UnitedHealthcare) and
1regional insurer (HealthPartners) . Each data research partner is a participant in the FDA
Sentinel Sy stem. The Sentinel Sy stem is an active surveillance s ystem that uses routine
querying and analy tical tools to evaluate electronic health caredata from a distributed data
network for monitoring the safet y of regulated medical products in the US, established under
the Sentinel I nitiative(Behrman et al., 2011 ; Platt et al., 2018 ). All of these data research
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077314
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 46of 70partnersupdate their curated Sentinel database multiple times per y ear. This study will use
the research -eligible population with in the most recentl y available database at each data
research partner at the time of anal ysis.
These data sources capture longitudinal medical care information on outpatient medication
dispensings, vaccine administrations, and inpatient and outpatient diag noses and procedures.
The data sources also capture member demographic and health plan enrollment information .
Each data research partner can request access to full -text medical records for outcome
validation for a subset of participants . All data resea rch partners are able to link to external
data sources (e.g.,state immunization registries)and can collect additional information via
surveys in at least a subset of members . As part of their participation in the Sentinel Sy stem,
3data research partner s(CVS Health, HealthCore, and Optum) maintain a mother -infant
linkage table to support studies of medication exposures during pregnancy . All of the
national insurers contribute claims data ,while the regional insurer (HealthPartners)
contributes data from both claims and electronic health records to the Sentinel database . As
all data research partners contribute data to the Sentinel Sy stem, this study will leverage the
Sentinel database and distributed querying infrastructure, including quality -checked a nd
curated data formatted to the Sentinel Common Data Model (SCDM) and the publicly
available Sentinel anal ytic tools(Curtis et al., 2012 ; Sentinel, 2018).
The data research partners use the SCDM ( Curtis et al., 2012 ; Sentinel, 2018) tostandardiz e
demographic and clinical data elements . Publicly available routine anal ytical tools
(i.e.,reusable, modular SAS programs) designed to be executed against the SCDM permit
rapid and standardized queries across data from different partners, including descriptive
analyses and complex methodologies (e.g., comparative anal yses).
Specific information in the SCDM includes, but is not limited to, the following t ypes of data:
Enrollment data, including 1record per covered individual per unique enro llment span .
The average enrollment length for patients across data sources in the Sentinel Sy stem is
similar to that in other claims databases of members with medical and pharmacy
coverage; approximately 25% of patients have over 3 y ears of enrollment, a nd patients
with chronic conditions such as diabetes and older members typicall y have longer than
average enrollment periods within these databases.
Individuals are assigned a unique identifier by their insurer that is linkable to other data
in the SCDM. Each record in the enrollment file indicates the patient identifier,
enrollment start and end dates, and whether the patient was enrolled in medical coverage,
pharmacy coverage, or both during that enrollment .
Demographic data, including birth date, sex, race/ethnicity , and zipcode of their most
recently recorded primary residence .
Outpatient pharmacy dispensing data, including the date of each vaccination or
prescription dispensing, the National Drug Code identifier associated with the dispensed
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077315
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 47of 70product, the nominal days ’supply, and the number of individual units (e.g., pills, tablets,
vials) dispensed . Products purchased over the counter or at some cash-onl y retail
locations selling prescription drug products (e.g., through the Walmart Prescription
Program) are not consistently captured.
Medical encounter data, including the health careprovider most responsible for the
encounter ,as well as the facility at which the encounter occurred and its zipcode.
Admission and discharge dates (if applicable) are also included ,in addition to the
encounter t ype (i.e., an ambulatory visit, emergency department visit, inpatient hospital
stay, nonacute inpatient stay , or otherwise unspecified ambulatory visit). Discharge
disposition (i.e., alive, expired, or unknown ) anddischarge status (i.e., where a patient
was discharged) are also included for acute and nonacute inpatient hospital stay s.
Diagnosis data, including the date of diagnosis, its associated encounter identifier,
admission date, provider identifier, and encounter ty pe. Diagnoses are recorded with
ICD-9-CM (International Classification of Diseases, 9th Revision, Clinical Modification)
and ICD-10-CMcodes. For acute and nonacute inpatient stay s, the SCDM includes both
principal and nonprincipal discharge diagnoses.
Procedure data, including the procedure date (e.g., date of vaccination), its associated
encounter identifier, admission date, provider identifier, and encounter t ype, are coded as
ICD-9-CM procedure and ICD-10-PCS (ICD-10 Procedure Coding S ystem)codes; CPT
(Current Procedural Terminology ) categories II, III, or IV codes; revenue codes and
Healthcare Common Procedure Coding S ystem levels II and III codes.
The following subsections include brief descriptions of each individual data source.
9.4.1.CVS Health, Aetna
Aetna, a CVS Health company , is one of the nation ’s leading health carebenefits companies,
currently serving 38 million people . Aetna became part of the Sentinel S ystem in 2008 .
Aetna’s SCDM captures longitudinal information on dispensed pre scriptions, inpatient and
outpatient diagnoses, inpatient and outpatient treatments and procedures , and outpatient
laboratory results. The healthcare experience for over 34 million individuals isavailable for
research,covering all ages, with median (ran ge) age of 45 (0-119) years(based on patients’
most recent available data).
9.4.2.HealthCore
HealthCore, Inc., became a participant in the Sentinel Sy stem in 2008 and contributes both
by submitting data to the Sentinel database and as a collaborator . As of February 2021, there
were 79million unique individuals with medical coverage and approximately 60million with
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077316
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 48of 70medical and pharmacy coverage available for research , covering all ages, with median ( Q1,
Q3) age of 40 (26, 57) years.
9.4.3.HealthPartners
HealthPartners is the largest consumer -governed nonprofit healthcare organization in the US,
providing care, insurance coverage, research, and education to its members and patients.
HealthPartners operates primarily in the Midwest and serves more than 1.8 million medical
and dental health plan members and more than 1.2 million patients, covering all ages, with
median (range) age of 39 (0-110) years. HealthPartners and its associated research team,
HealthPartners Institute,became a member of the Sentinel Sy stem in 2008 .
9.4.4.Humana
Humana Healthcare Research (HHR) is a health economics and outcomes research subsidiary
of Humana, which focuses on treatment effectiveness, drug safety , adherence, medical and
pharmacy benefit design, disease management programs, and other health careservices based
on the Humana health plan member population . Humana/HHRis an active collaborator and
data research partner in the Sentinel Sy stem. The Humana research -eligible database
represents geographic coverage for the entire US population (Puerto Rico excluded), and as
of 31March 2021 has 32.7 million unique individuals, covering all ages, with median ( Q1,
Q3) age of 67 (48 , 76) years.
9.4.5. Optum Research Database
TheOptum Research Database is a proprietary research database that contains eligibility data
and medical claims and includes health plan members who are geographically diverse across
the US. The Optum Research Database comprisesapproximately 3% to 4% of the US
population, covering all ages, with median ( Q1, Q3) age of 36 (21, 51)years. Optum has
curated and quality -checked data formatted to the SCDM available for use and is a longtime
participant in the Sentinel Sy stem.
9.5.Study size
The size of the exposed population will depend on the use of Pfizer- BioNTech COVID-19
Vaccine,and the size of the comparator population will depend on the proportion of the
source population that comprises unvaccinated individuals over time in the data sources. The
precision of comparative risk estimates will depend on the background rate and the duration
of the risk interval for each safet y event of interest .
Assuming a matching ratio of 1: 1, Table 7presents the probability that the upper limit of the
95% CI for the observed RRwill be below 1.5, 2.0, 2.5, and 3.0 for assumed true RRs of 1.0,
1.2, 1.4 and study sizes ranging from 500,000 to 20,000,000 vaccinated individuals
(1,000,000 doses to 40,000,000 doses, under the assumption that each individual will receive
2 doses). The estimates in the table on the following page reflect a cohort analy sis. These
estimates are presented to cover a range of safet y events of interest with respect to r areness,
based on background rates in the general population.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077317
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 49of 70For example, for Bell’ s palsy, with 1,000,000 exposed individuals we estimate an 86 %
probability that the upper bound of the observed RR would be below 2.0, assuming a 1 :1
ratio between vaccinated and comparator person -time and that the true RRis 1.0.
Table 7.Study size calculations
Assumed
true RRSafety
event of
interestEstimated
back-
ground
rate per
100,000
person-
years
(Black et
al., 2021)Number of
individuals
vaccinatedProbability that the upper confidence limit of
RR will be below the following thresholdsa:
1.5 2.0 2.5 3.0
1.0 Guillain-
Barré
syndrome1.68 500,000 0.06 0.10 0.14 0.19
1,000,000 0.08 0.16 0.25 0.33
2,500,000 0.14 0.33 0.52 0.68
5,000,000 0.24 0.58 0.81 0.93
10,000,000 0.43 0.86 0.98 1.00
20,000,000 0.71 0.99 1.00 1.00
Bell’s palsy 25.2 500,000 0.24 0.58 0.81 0.93
1,000,000 0.43 0.86 0.98 1.00
2,500,000 0.80 1.00 1.00 1.00
5,000,000 0.98 1.00 1.00 1.00
10,000,000 1.00 1.00 1.00 1.00
20,000,000 1.00 1.00 1.00 1.00
Myocardial
infarction 208 500,000 0.95 1.00 1.00 1.00
1,000,000 1.00 1.00 1.00 1.00
2,500,000 1.00 1.00 1.00 1.00
5,000,000 1.00 1.00 1.00 1.00
10,000,000 1.00 1.00 1.00 1.00
20,000,000 1.00 1.00 1.00 1.00
1.2 Guillain-
Barré
syndrome1.68 500,000 0.04 0.08 0.11 0.15
1,000,000 0.05 0.11 0.19 0.26
2,500,000 0.07 0.22 0.39 0.56
5,000,000 0.11 0.38 0.66 0.85
10,000,000 0.17 0.65 0.92 0.99
20,000,000 0.30 0.91 1.00 1.00
Bell’s palsy 25.2 500,000 0.11 0.38 0.66 0.85
1,000,000 0.17 0.65 0.92 0.99
2,500,000 0.37 0.96 1.00 1.00
5,000,000 0.63 1.00 1.00 1.00
10,000,000 0.90 1.00 1.00 1.00
20,000,000 1.00 1.00 1.00 1.00
Myocardial
infarction 208 500,000 0.55 1.00 1.00 1.00
1,000,000 0.84 1.00 1.00 1.00
2,500,000 1.00 1.00 1.00 1.00
5,000,000 1.00 1.00 1.00 1.00
10,000,000 1.00 1.00 1.00 1.00
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077318
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 50of 70Table 7.Study size calculations
Assumed
true RRSafety
event of
interestEstimated
back-
ground
rate per
100,000
person-
years
(Black et
al., 2021)Number of
individuals
vaccinatedProbability that the upper confidence limit of
RR will be below the following thresholdsa:
1.5 2.0 2.5 3.0
20,000,000 1.00 1.00 1.00 1.00
1.4 Guillain-
Barré
syndrome1.68 500,000 0.03 0.06 0.09 0.12
1,000,000 0.03 0.08 0.14 0.21
2,500,000 0.04 0.13 0.28 0.44
5,000,000 0.04 0.22 0.49 0.73
10,000,000 0.05 0.40 0.79 0.95
20,000,000 0.07 0.67 0.97 1.00
Bell’s palsy 25.2 500,000 0.04 0.22 0.49 0.73
1,000,000 0.05 0.40 0.79 0.95
2,500,000 0.07 0.76 0.99 1.00
5,000,000 0.11 0.97 1.00 1.00
10,000,000 0.18 1.00 1.00 1.00
20,000,000 0.31 1.00 1.00 1.00
Myocardial
infarction 208 500,000 0.10 0.93 1.00 1.00
1,000,000 0.15 1.00 1.00 1.00
2,500,000 0.32 1.00 1.00 1.00
5,000,000 0.56 1.00 1.00 1.00
10,000,000 0.85 1.00 1.00 1.00
20,000,000 0.99 1.00 1.00 1.00
a. Estimates in this table assume a risk window duration of 42 days for Guillain -Barré syndrome and 28 days
for Bell’s palsy and myocardial infarction.
9.6.Data management
9.6.1. Data collection tools (DCTs)
As the anal yses will be based on secondary data, the only data collection tool that may be
applicable will be data abstraction forms that will be developed for the purpose of validation
of select outcomes if validation is implemented. Details of how data will be handled during
validation will be described in a validation plan thatwould be developed prior to
implementing validation in the data sources .
As used in this protocol, the term DCTshould 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 DCTis required and should be completed for each patient included in the chart validation
activities. The completed original DCTare the sole propert y of Pfizer and should not be
made ava ilable in any form to third parties, except for authorized representatives of Pfizer or
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077319
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 51of 70appropriate regulatory authorities, without written permission from Pfizer. Harvard Pilgrim
Health Care Institute (HPHCI) shall ensure that the DCTs shared with HPHCI are securel y
stored at HPHCI in encrypted electronic form and will be password protected to prevent
access by unauthorized third parties.
HPHCIhas ultimate responsibility for oversight of the collection and reporting of all clinical,
safety, and laborator y data entered on the DCTsand any other data collection forms (source
documents) and ensuring that they are accurate, authentic/original, attributable, complete,
consistent, legible, timely (contemporaneous), enduring, and available when required. The
DCTsmust be signed b y HPHCIor by an authorized staff member to attest that the data
contained on the DCTsare true. An y corrections to entries made in the DCTsor source
documents must be dated, initialed, and explained (if necessary ) and should not obscure the
original entry .
The source documents are the hospital or the ph ysician's chart. In these cases, data collected
on the DCTs must match those charts.
9.6.2.Record retention
To enable evaluations and/or inspections/audits from regulatory authorities or Pfi zer, HPHCI
as the coordinating c enter agrees to keep all study -related records, including programming
specifications, aggregate data reports submitted by data research partners , final study reports,
and any related materials . The records should be retained by HPHCI according to local
regulations or as specified in the research agreement with Pfizer , whichever is longer .
HPHCI must ensure that the records continue to be stored securely for so long as they are
retained.
If HPHCI becomes unable for any reason to continue to retain stud y records for the required
period, Pfizer should 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 completio n or discontinuation of
the study, unless HPHCI and Pfizer have expressly agreed to a different period of retention
via a separate written agreement . Records must be retained for longer than 15 years if
required b y applicable laws or regulations .
HPHCI must obtain Pfizer ’s written permission before disposing of any records, even if
retention requirements have been met.
9.6.3.Data oversight
HPHCI, located in Boston, Massachusetts, will serve as the coordinating c enter for the
proposed study . HPHCI staff or cont ractors will be responsible for writing and distributing
SAS programs that can be used to evaluate the data included in databases at participating
data research partners . The distributed network will allow data research partners to maintain
physical and o perational control of their data while allowing use of the data to meet the study
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077320
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 52of 70needs. HPHCI will maintain a secure, distributed ,querying web-based portal to enable
secure distribution of analy tic queries, data transfer , and document storage . The system will
meet all required state and federal security guidelines for health data (e.g., Federal
Information Securit y Management Act [FISMA], Health Insurance Portability and
Accountability Act of 1996) and will be specifically FISMAcompliant for FISMA secu rity
controls as specified in the National Institute of Standards and Technology (NIST) Special
Publication 800 -53 (NIST, 2020).
HPHCI brings expertise in conducting multisite evaluations using disparate electronic
healthcare data systems, including work with the Health Care S ystems Research Network,
the VSD, the National Institutes of Health, Health Care Sy stems Research Collaboratory ,
IMEDS(Innovation in Medical Evidence Development and Surveillance) , the Biologics and
Biosimilars Collective I ntelligence Consortium, PCORnet (the National Patient -Centered
Clinical Research Network), and the Sentinel S ystem. HPHCI will oversee all project
activities, including scientific leadership, management of the partnership, coordination of
activities with the data research partners and other participants, oversight of the project plan
and budgets, establishment of secure infrastructure used for collaborati on, and training
related to use of the data sources and associated analy tic tools. In collaboration with RTI
Health Solutions ( RTI-HS), HPHCI will also oversee all activities related to implementation
of any potential medical record reviews . HPHCI will develop standard operating procedures
and processes to guide any potential linkages to state registries or implementation of medical
chart reviews in collaboration with RTI -HS and the data research partners . The data research
partnerswill establish and ma intain the administrative, hardware, and software capabilities
and capacit y to respond to data requests in a timely manner . Data research partners will also
provide data science support with epidemiologic review.
Figure3summarizes the general analytic workflow . Based on the stud y design developed b y
the study team, the stud y coordinating center first submits through a secure portal a computer
program designed to meet the needs of the study. Next, the participating data research
partnersreceive and run the computer program behind their firewalls, using data that is
formatted to the SCDM. Then , the data research partners review the analysis results and
return them to the study coordinating center through a secure portal . The study coordinating
center then reviews and aggregates the results across the data research partners . In the final
step, the aggregated results are transferred to the study team.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077321
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 53of 70Figure3.General analytic workflow
9.7.Data analysis
All analyses will initially be conducted separatel y within the data from each data source .
Pooled analy sis of RR and odds
ratio estimates from all data sources will be conducted using
privacy-preserving summary -level data sets (e.g., risk set –level data sets) or if this is not
feasible, meta -analysis. Detailed methodology for summary and statistical anal yses of data
collected in this study will be documented in the SAP, which will be dated, filed, and
maintained by the sponsor . The SAP may modify the plans outlined in the protocol; any
major modifications of primary outcomedefinitions or their anal yses will be reflected in a
protocol amendment.
9.7.1.Descriptive analysis
Descriptive anal ysis will report on utilization of Pfizer -BioNTech COVID -19 Vaccine
during
the overall stud y period and during the stud y period, stratified in 12- week increments (to
assess vaccine uptake and patterns of exposure over time) . The proportion of individuals
receiving at least 1dose and a complete dose series of Pfizer -BioNTech COVID -19 Vaccine
will be estimated within the overall stud y population, in immunocompromi sed individuals, in
individuals with a history of COVID- 19, and in pregnant women . Among patients who
receive a first dose of Pfizer -BioNTech COVID -19 Vaccine, the proportion of patients will
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077322
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 54of 70be reported b y type of second dose of COVID -19 vaccine and time between the 2doses will
be described.
The number and proportion of patients with a convulsion event who have a history of
epilepsy will be reported separately in each of the matched exposed and matched unexposed
cohorts. The length of follow -up in the exposed and unexposed cohorts contributing to
analyses of vaccine -associated enhanced disease will be output . Characteristics
(demographics, comorbidities, and other potential covariates) of the matched and unmatched
cohorts will be shown in a table . No statistical tests are planned for this comparison, but the
balance of variables in the matched cohorts will be assessed using standardized differences or
other suitable methods .
Additional descriptive analy sis may be conducted for safety events that have signaled in
other studies or vaccine safet y surveillance s ystems(e.g.,myocarditis/pericarditis) . These
analyses will be described in the SAP .
9.7.2.Measures of disease frequency and association
All eligible individuals in each stud y cohort will be included in analysis of disease frequency
and measures of association. However, in anal ysis ofsomesafety events, individuals who
have experienced the outcome in the recent past will be excluded from the anal ysis. This
will be done to d istinguish between follow-up care for events that have happened in the past
from incident events occurring during follow -up. The washout periods for defining incident
events will depend on the outcome and will be specified in the SAP .
9.7.2.1. Measures of disease frequency
In each data source, crude measures of incidence (for all outcomes except congenital
malformations and small size for gestational age) or birth prevalence (for congenital
malformations and small size for gestational age) with associated 95% CIs will be estimated
within the matched exposed and unexposed cohort s. Prevalence will be estimated for major
congenital malformations because the outcome is identified after birth without the ability to
determine its true timing of onset during pregnancy . Prevalence will be estimat ed for small
size for gestational age because the outcome is identified at a single time point(atbirth).
9.7.2.2.Measures of association
For comparative anal yses of general safety events,spontaneous abortion, stillbirth, and
preterm birth, Cox models or Poisson regression will be used to estimate RRs and 95%CIs
within the matched cohorts .
For comparative anal ysis of small size for gestational age and major congenital
malformations, logistic regression will be used to estimate odds ratios and 95% CIswithin
the matched cohorts .
For comparative anal ysis of outcomes identified separately in dose1and dose2cohorts,
comparative risks will be estimated separatel y by dose number . If comparative risk estimates
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077323
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 55of 70are similar b y dose number, then data from the dose1exposed/dose1unexposed cohorts and
the dose2exposed/dose2 unexposed cohorts will be combined to obtain comparative risk
estimates associated with any dose of the vaccine .
For the dose1 comparative risk estimation, when a second dose of vaccine is administered
before the risk interval following dose1is complete, the risk interval will be truncated at the
time of dose2, and follow -up after dose2will be excluded from the dose 1 risk estimation .
In this situation, for the dose2comparative risk estimation, the risk interval will start on the
date of the second vaccine dose and will extend for the duration of the risk interval for
dose2.
Because each dose of vaccine within the same individual will be considered a separate
observation when combined in anal ysisand because some individuals may contribute to both
the exposed and unexposed cohorts , the correlation between dose1and dose2 will be
accounted for when estimating the variance ,using appropriate statistica l methods to be
detailed in the SAP .
9.7.2.2.1. Methods for addressing confounding
9.7.2.2.1.1. General population, immunocompromised individuals, and individuals with
a history of COVID -19
Analysis of nonpregnant populations will usematching on age, state (if feasible, or broader
geographic region if not feasible) ,andtime-specific propensit y scores within the data from
each data research partner to account for confounding. The propensity score is the predicted
probability of an individual being in the exposed cohort rather than in the corresponding
unexposed cohort, given a set of observed covariates.
Estimation of propensity scores will be performed for the dose1 and 2 cohorts combined, but
matching of exposed to unexposed individuals will be done separately for the dose1exposed
cohort and for the dose2exposed cohort, within narrow time periods to account for changing
predictors of vaccination over time , seasonality of circulating infections, and changes in
healthcare utilization over time . Matching will occur b yage and state (if feasible, or broader
geographic region if not feasible) within each time period and dose number . The matching
and propensit y score estimation process will be done first for the general population and then
will be repeated separately for immunocompromised individuals and for individuals with a
history of COVID -19. Propensity score estimation will be conducted within each data
research partner. The steps for propensit y score modeling and matching are as follows.
a.Within each data research partner, the study period will be divided into 1-month intervals
of calendar time (“time intervals”). All individuals who received a vaccine dose, either
dose1or dose2, during the considered month wouldcontribute vaccinated index dates
during the time interval—if a patient received 2doses during a calendar month, both
index dates will be included as independent observations in the propensity score model .
All individuals with at least 1 unexposed person -dayin the time interval will have an
unvaccinat ed index date randomly assigned during the time interval . In this context,
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077324
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 56of 70“unexposed person- day”refers to a day in which the patient has no record of COVID -19
vaccine on or before that day .
b.Within each time interval, the propensity to be vaccinated will be estimated among all
individuals with index dates within the time intervalusing logistic regression. T he
dependent variable for the logistic regression model will be vaccination status, and the
independent variables will be baseline covariates (i.e., individual characteristics as
described in Section 9.3.3). The model will combine dose 1and dose2index dates ,
which assumes that the factors influencing an individual ’s likelihood of being vaccinated
does not change between dose1and dose2.
c.Within each data research partner, the distribution of propensit y scores in each dose
cohort in each time interval will be plotted to evaluate the comparability of the
2exposure groups . Greater overlap of the propensity score distributions will indicate
greater exchangeability .
d.After the comparabilit y of the treatment groups is confirmed , unexposed individuals will
be matched on propensity scores to exposed individuals (in a ratio of at least 1:1) within
each data research partner. The matching will be done separately for the dose 1exposed
cohort and for the dose2exposed cohort and will be done by age, state (or geographic
region)within each time interval.
e.The matching procedure within each dose cohort will be executed chronologically from 1
time interval to the next time interval . An individual may only be selected once for the
dose1unexposed cohort and once for the dose 2unexposed cohort . Individuals who
match as an unvaccinated index date in 1 time interval will not be considered for the
same unvaccinated cohort (i.e., dose1unexposed cohort or dose2unexposed cohort) in
future time intervals . However, if an individual becomes vaccinated after being selected
as an unexposed match, he or she may be eligible for the dose1and/or dose 2exposed
cohorts.
f.The matched exposed and unexposed individuals from each time intervalwill be
combined into the overall matched anal ytic cohorts.
Further details on the matching process and estimation of propensity scores will be described
in the SAP.
9.7.2.2.1.2. Pregnant women
To minimize confounding b y seasonality and maternal age, within the data from each data
research partner, exposed pregnant women will be matched to unexposed pregnant women
on estimated pregnancy startdate (+/-14 days), state (if feasible, or broader geographic
region if not feasible) ,and maternal age . Propensity scores will be estimated within the
matched population and incorporated in toregression anal ysisfor exposure -outcome
associations (e.g.,through weighting or stratification) to address confounding b y other
variables. The matching and propensit y score esti mation process will be done first for the
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077325
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 57of 70overall population of pregnant women . The process will then be repeated for the subsets of
pregnant women eligible for anal yses of the different safet y events of interest (e.g.,
pregnancies surviving bey ond 20 completed weeks of pregnancy for analysis of stillbirth;
live births for analysis of preterm birth and small size for gestational age ; and live births that
can be linked to infants for anal ysis of major congenital malformations).
The steps for m atching and propensity score analy sis are as follow s.
a.Unexposed individuals will be matched on estimated pregnancy start date , state (or
broader geographic region, as feasible), and maternal age to exposed individuals within
each data research partner (ina ratio of at least 1:1). When applicable (when separate
dose1and dose2exposed cohorts are formed), the matching will be done separately for
the dose1exposed cohort and for the dose 2exposed cohort .
An individual may be selected for each unexposed cohort (i.e., dose1unexposed, dose2
unexposed cohort, or unexposed cohort) only once. However, if an individual becomes
vaccinated after being selected as an unexposed match, he or she may be eligible fo r the
exposed cohort(s) .
b. The propensit y to be vaccinated will be estimated among all individuals with eligible
index dates. The dependent variable for the logistic regression model will be vaccination
status,and the independent variables will be baseline covariates (i.e., individual
characteristics as described in Section 9.3.3). When applicable ( when separate dose1
and dose2exposed cohorts are formed), the model will combine dose 1and dose2index
dates.
c.The distribution of propensity scores in each cohort will be plotted to evaluate the
comparability of the 2exposure groups. Greater overlap of the propensit y score
distributions w ill indicate greater exchangeability .
d.After the comparabilit y of the treatment groups is confirmed, propensity scores will be
incorporated into regression modeling for exposure -outcome associations through
weighting or stratification.
Further details on the matching process andestimation of propensity scoreswill be described
in the SAP.
9.7.3.Sensitivity analysis
9.7.3.1. Exposure misclassification
During the early stages of the roll out of COVID -19 vaccines , many vaccinations may have
occurredoutsidetraditional med ical care settings without reimbursement from health
insurers. The potential for lack of recording of COVID -19 vaccines in claims and electronic
health records may lead to misclassification of exposed individuals as “unexposed”
individuals, which will underestimate vaccine coverage rates and bias comparative risk
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077326
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 58of 70estimates for the cohort design with concurrent unexposed individuals as comparators . To
address this potential bias in comparative risk estimates , sensitivity analysis will be
performed using a SCRIdesign or a cohort design with historical comparators, depending on
the safety event of interest.
9.7.3.1.1. Self-controlled risk interval design
Sensitivity analyses incorporating a SCRI design will be implemented for outcomes with a
well-definedonset and risk intervals no longer than 42 day s (specific outcomes to be named
in the SAP) in the overall study population, in pregnant women, in immunocompromised
individuals, and in individuals with a history of COVID -19. The SCRI design will not be
considered for pregnancy safety outcomes. Onl y vaccinated individuals will be included in
the SCRI analysis; the rate of a specific safet y event in a post -vaccination risk interval will be
compared with the rate in a control interval within the same individual.
For individuals who receive 2 doses of the vaccine, the risk interva l will combine person -
time in the risk intervals after the first dose and after the second dose . The control interval
will be outcome specific, with the duration and timing relative to vaccination specified in
more detail in the SAP . Control intervalswill be defined during specific periods following
vaccination (up to a maximum of 183 day s); prevaccination periods will not be used to avoid
bias due to health y vaccinee effects . Foroutcomes with risk intervals shorter than the
recommended spacing between dose1and dose2(e.g., anaph ylaxis), the control interval
may comprise person -time between the first and second dose s. For outcomes with risk
intervals longer than the recommende d spacing between dose 1and dose2, the control
interval will comprise person- time after the second dose . A washout period between the risk
and control intervals may be incorporated for safety events for which the risk interval is not
well characterized .
Because of the self -controlled nature of the design, bias of comparative risk estimates arising
from differences in the distribution of time -constant confounding factors between vaccinated
and unvaccinated individuals is avoided with the SCRI design. Furthermore, as the design
only includes vaccinated individuals, it avoids the potential for misclassification of
unexposed status due to incomplete capture of COVID -19 vaccinations in data from claims
or electronic health record s.
9.7.3.1.2. Cohort design with histori cal unexposed comparators
If feasible, sensitivity analysis of events not meeting criteria for the SCRI design
(i.e.,outcomes with gradual onset and/or risk intervals longer than 42 day s, except for
vaccine-associated enhanced respiratory disease; specific outcomes to be named in the SAP )
will incorporate historical unexposed individuals as comparators . The comparator cohorts
will be identified and followed in a time period before the introduction of COVID -19
vaccines. The index date in historical unexposed individuals will be 2or more yearsprior to
the date of vaccination in exposed individuals . Exposed and unexposed individuals will be
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077327
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 59of 70matched on age and propensity score, using similar methods as for the cohort desi gn with
concurrent unexposed comparators.
The use of historical unexposed individuals as comparator savoids the potential for
misclassification of unexposed status due to incomplete capture of COVID -19 vaccinations
in data from claims or electronic health records. The feasibility of this analy sis will depend
on the absence of trends in coding for each safet y event of interest over time in the historical
comparator period and the study period. Further details on the composition of the cohorts,
methods to a ddress confounding, and criteria used to determine whether this design is
feasible will be specified in the SAP .
9.7.3.2. R isk intervals
The study design approach proposed in this protocol requires that risk intervals be specified
correctly. If risk intervals aretoo long, comparative risk estimates may be attenuated.
Sensitivity analysis of myocarditis/pericarditis will be conducted using alternative risk
intervaldefinition s of 1-7 and 1-14 days.
For events for which the risk intervals are not well characterized (to be defined in the SAP),
descriptive analy ses of the timing of events relative to vaccination will be conducted. If they
are identified , temporal clusters will be used to define alternative risk intervals thatwill be
used in sensitivity analyses.
9.8.Quality control
The data research partners thatwill contribute data for this study are all participants in the
Sentinel Sy stem. The study will use the same data quality assurance (QA) procedures as the
Sentinel Sy stem and the same curated data sets used by the FDA to conduct Sentinel
analyses. The QAapproach assesses consistency with the SCDM, evaluates adherence to
data model requirements and definitions, evaluates logical relationships between data model
tables, and reviews trends in medica l and pharmacy services use within and across data
research partners . Full QAprocesses and details on the Sentinel database curation approach
are documented on the Sentinel website( Sentinel, 2021 ; Sentinel, 2017). The data curation
approach is consistent with guidance set forth b y the US FDA in its current recommendations
for data QA, Guidance for Industry and FDA Staff: Best Practices for Conducting and
Reporting Pharmacoepidemiologic Safety Studies Using Electroni c Healthcare Data , Section
IV.E Best Practices –Data Sources: Quality Assurance (QA) and Quality Control (QC) ,
published in May 2013 (FDA, 2013 ; Sentinel, 2017 ). This Guidance describes best practices
that particularl y apply to observational studies designed to assess the risk associated with a
drug exposure using electronic health caredata.
In addition to QAof data elements, HPHCI adopts standard SAS programming QAand
quality control (QC) processes used b y the Sentinel Sy stem to check SAS programs and
deliverables . Figure4illustrates the standard operating procedures for SAS programming
QAand QCin the Sentinel Sy stem.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077328
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 60of 70Figure4.Standard operating procedure for SAS programming quality assurance and
quality control in the Sentinel System
9.9.
Strengths and limitations of the research methods
A major strength of this study is that it will include a very sizeable source population in the
US, as the participating data research partners together collect data on more than 100 million
individuals . The use of secondary data will enable the efficient assessment of many safety
events of interest identified by the CDC ’s VSD and the FDA ’s BEST Initiative, in addition to
pregnancy safety outcomes ,while using robust study design and anal ytic approaches to
adjust for potential confounding . Moreover, the secondary use of administrative data
collected as part of routine medical care avoids selection bias that might occur in primary
data collection st udies, as a patient ’s inclusion in this study is not voluntary .
Nevertheless, t his study is subject to limitations arising from the use of secondary data and
the selected stud y designs . Limitations related to the data sources include the potential for
lack of recording in claims and electronic health records of COVID -19 vaccines administered
without reimbursement from health insurers. If the data appear to be substantially
incomplete in monitoring anal yses, then the primary study design may be reconsider ed. If
this happens, the SCRI and the cohort design with historical unexposed comparators may be
designated as the primary stud y designs; and/or linkage to immunization registries may be
considered. Additionall y, the use of data from claims and electroni c health records may lead
to some misclassification of outcomes (e.g., false positives and false negatives) . Some
events, such as spontaneous abortion, will be incompletely captured in existing databases .
Conversel y, validation studies of I CD-10-CM–basedalgorithms for man y of the safety
events of interest have been limited , and the accuracy of algorithms for many safety events of
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077329
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 61of 70interest are unknown. When possible, validated algorithms will be used, and outcomes that
are likely to be misclassified (based on prior validation studies and clinical expert input) may
be validated through review of medical records or claims profiles, depending on the safet y
event of interest .
A study design-related limitation is that any uncertainty regarding risk periods will lead to
misclassification and attenuation of risk estimates. Sensitivity analyses with alternative risk
intervals will be considered for outcomes for which the risk interval is not well characterized .
A limitation of the cohort design with concurrent unexposed comparators is the potential for
residual or unmeasured confounding because it is unlikely that the data sources will have
information on all potential confounders . To address potential confounding, the SCRI, which
automatically adjusts for time -invariant confounders, will be used as a secondary approach
where feasible . However, the SCRI is not well suited to study outcomes with gradual onset,
long risk intervals , or risk periods that are not well characterized .
A limitation specific to the coho rt design with concurrent unexposed comparators is that
unvaccinated individuals may become exposed to COVID -19 vaccine at any time during the
study; if this situation occurs frequentl y, the amount of unexposed person -time in the
unexposed comparator group will bereducedsubstantially , which will limit the precision of
comparative risk estimates and could potentiall y lead to substantial imbalances in seasonality
between exposed individuals and unexposed individuals, particularl y for outcomes with long
riskintervals. Forming 2 separate exposed cohorts by dose number and matching unexposed
to exposed at the time of each vaccine will minimize the loss of unexposed person -time due
to receipt of vaccine in these individuals between the first and second doses . Additionally ,
the sensitivity analyses with the historical unexposed comparator cohort and the SCRI design
will not be subject to this limitation. However ,it is anticipated that even with separate
matching of doses that follow -up time will be substantia lly longer in anal yses of vaccine -
associated enhanced respiratory disease in individuals in the vaccinated cohorts than inthe
unvaccinated comparator cohorts, since the risk interval is 365 day s long. Further, the SCRI
design and historical unexposed com parator cohort design are not feasible to study this
outcome.
9.10.Other aspects
Not applicable
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077330
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 62of 7010. PROTECTION OF HUMAN SUBJECTS
This study involves use of existing structured data and may also include human review of
unstructured data for the subset of patient ch arts thatmay be reviewed for validation purpose.
Each data research partner will obtain appropriate reviews and determinations from
respective institutional review boards ( IRBs)according to its site requirements or cede
authority to HPHCI’s IRB,if possible.
Data protection and privacy regulations will be observed in collecting, forwarding,
processing, and storing data from stud y participants.
10.1.Patient information
This study mainly involves data that exist in anonymized structured format and conta in no
patient personal information. If chart validation is required, during this component of this
study, data research partners will remove and redact all direct patient identifiers as delineated
in the Privacy Rule of HIPAA (Health Insurance Portability and Accountability Actof 1996).
A limited data set of protected health information (PHI)—includingdate of birth, date of
vaccination, date of death, visit date, and diagnosis date —may be collected . Dates related to
the individual (date of birth, date of death, visit date , and diagnosis date) are required in order
to investigate the safet y of COVID-19vaccines.
All parties will comply with all applicable laws, including laws regarding the implementation
of organizational and technical measures to ensur e protection of patient personal data. Such
measures will include omitting patient names or other directl y identifiable data in an y
reports, publications, or other disclosures, except where required b y applicable laws .
HPHCIwill maintain a secure web -based portal to enable secure data transfer and document
storage. The system will be FISMA compliant (FISMA Moderate Risk security controls, as
specified in the NIST Special Publication 800-53). The system will comply with relevant
FISMA, HIPAA, and NIST r equirements. A study identification number will also be used in
place of direct patient identifiers to minimize risk. Patient personal data will be stored at the
individual data research partner or at HPHCI in encry pted electronic form and will be
password protected to ensure that only authorized study staff have access. Each data
research partnerand HPHCI will implement appropriate technical and organizational
measures to ensure that the personal data can be recovered in the event of disaster . In the
event of a potential personal data breach, each data research partnerand HPHCI shall be
responsible for determining whether a personal data breach has in fact occurred and, if so,
providing breach notifications as required b y law.
To protect the ri ghts and freedoms of natural persons with regard to the processing of
personal data, when study data are compiled for transfer to Pfizer and other authorized
parties, an y patient names will be removed and will be replaced by a single, specific,
numerical c ode. 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
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077331
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 63of 70maintain high standards of confidentialit y and protection of patients’ perso nal data consistent
with the research agreement and applicable privacy laws.
10.2.Patient consent
As this study does not involve data subject to privacy laws according to applicable legal
requirements, obtaining informed consent from patients i s not required .
10.3. Institutional review board/Independent ethics committee (IEC)
Each data research partner ,as well as HPHC I,will followitslocal requirements and data
custodian requirements to access the data . As the coordinating center , HPHCI will seek
approval from its local IRB. There must be prospective approval of the study protocol,
protocol amendments, and other relevant documents (e.g., informed consent forms if
applicable) from the relevant IRBs/IECs. All correspondence with the IRB or independent
ethics commi tteeand applicable documentation will be retained as part of the study
materials. Copies of IRB/IEC approvals mustbe forwarded to Pfizer.
10.4.Ethical conduct of the study
Thisis a post-authorization study of vaccine safety and will comply with the definitio n of the
non-interventional (observational) study referred to in the I nternational Conference on
Harmonisation tripartite guideline Pharmacovigilance Planning E2E (ICH, 2004).
The study will be registered in the EU PAS Register (ENCePP, 2021) before data collection
commences.
The study will be conducted in accordance with legal and regulatory requirements, as well as
withscientific purpose, value ,and rigor and follow generally accepted research
practices described in Guidelines for Good Pharmacoepidemiology Practices (GPP) (ISPE,
2015)issued by theInternational Society for Pharmacoepidemiology and Good
Epidemiological Practice guidelines issued by the International Epidemiological Association
(IEA, 2007 ).
11.MANAGEMENT AND REPORTING OF ADVERSE EVENTS/ADVERSE
REACTIONS
This study involves data that exist as structured data by the time of study start . If validation
of algorithms for identifying outcomes is conducted ,the study may also involve human
review of unstructured data.
11.1.Structured d ata analysis
For the data that exist as structured data by the time of study start , 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 an y
individual . Thus, the minimum criteria for reporting an AE(i.e., identifiable patient,
identifiable reporter, a suspect product, and event) cannot be met.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077332
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 64of 7011.2.Humanreview of unstructured data
If validation is carried out , there will be human review of patient -level unstructured data;
unstructured data refer to verbatim medical data, including text -based descriptions and visual
depictions of 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
report AEswith explicit attribution to any Pfizer drug that appear in the reviewed
information (defined per the pa tient population and study period specified in the protocol) .
Explicit attribution is not inferred b y a temporal relationship between drug administration
and an AE butmust be based on a definite statement of causality by a healthcare provider
linking dru g administration to the AE.
The requirements for reporting safet y events on the non -interventional study (NIS) adverse
event monitoring (AEM) Report Form to Pfizer Safet y are as follows:
All serious and nonserious AEs with explicit attribution to any Pfizer drug that appear in
the reviewed information must be recorded on the chart abstraction form and reported,
within 24 hours of awareness, to Pfizer Safety using the NI S AEM Report Form.
Scenarios involving drug exposure, including exposure during pregnancy , exposure
during breast feeding, medication error, overdose, misuse, extravasation, lack of efficacy ,
and occupational exposure associated with the use of a Pfizer product must be reported,
within 24 hours of awareness, to Pfizer Safety using the NI S AEM Report Form.
For the anal ysis of the subpopulation in pregnant women, data on the exposure to the
Pfizer-BioNTech COVID -19 Vaccine during pregnancy , as well as pregnancy safety
outcomes , will be included in the anal ytic dataset. For pregnant women whose c harts are
reviewed for outcome algorithm validation purposes, exposure during pregnancy cases
are not reportable unless associated with serious or nonserious adverse events.
For these AEs with an explicit attribution or scenarios involving exposure to a Pf izer 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 information known regarding
these AEs . No follow -up on related AEs will be conduct ed.
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 demographic fields w ill be completed, at the
very least, at least 1 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 accordance with phar macovigilance 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 .
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077333
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 65of 70Additionally , the onset/start dates and stop dates for “Illness,” “Study Drug,”and “Drug
Name”may be documented in month/y ear (mmm/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 members must complete the following Pfizer training requirements:
Your Reporting Responsibilities ( YRR)Training for Vendors Working on Pfizer Studies
(excluding interventional clinical studies and non -interventional primary data collection
studies with sites/investi gators).
These trainings must be completed by research staff members that will have access to copies
of medical records prior to the start of data collection . All trainings include a “Confirmation
of Training Certificate ”(for signature b y the trainee) a s 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 YRR training
materials.
12.PLANS FOR DISSEMINAT ING AND COMMUNICATING STUDY RESULTS
Results of analy sis and interpretation will be delivered in the form of reports . A monitoring
analysis report and an interim study report are planned for the first and second y ear of
follow-up. After the end of the third y ear of follow -up, the final report will be produced,
including the anal ysis and interpretation of each outcome including pregnancy safety
outcomes.
Study results will be published following guidelines, including those for authorship,
established by the International Committee of Medical Journal Editors (ICMJE, 2019 ).
When reporting results of this study , the appropriate Strengthening the Reporting of
Observational Studies in Epidemiology (STROBE) checklist will be followed ( von Elm et
al., 2008). Independent publication rights will be granted to the research team in line with
Section VIII.B.5., Publication of Study Results, of the European Medicines Agency ’s
Guideline on Good Pharmacovigilance Practices (GVP) Module VIII: Post -Authorisation
Safety Studies (EMA, 2017).
Communication via appropriate scientific venues will be considered.
In the event of an y prohibition or restriction imposed (e.g., clinical hold) by an applicable
competent authorit y in any area of the world, or if the investigator party responsible for
collecting data from the participant is aware of any new information that might influence the
evaluation of the benefits and risks of a Pfizer product, Pfizer should be informed
immediately .
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077334
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 66of 7013.REFERENCES
AMA. American Medical Association. COVID-19 CPT coding and guidance. 2021.
https://www.ama -assn.org/find -covid-19- vaccine-codes#. Accessed 28 July 2021.
Baker MA, Lieu TA, L i L, Hua W, Qiang Y, Kawai AT, et al. A vaccine study design
selection framework for the postlicensure rapid immunization safety monitoring
program. Am J Epidemiol. 2015;181(8):608 -18.
doi:http://dx.doi.org/10.1093/aje/kwu322 .
Behrman RE, Benner JS, Brown JS, McClellan M, Woodcock J, Platt R. Developing the
Sentinel Sy stem--a national resource for evidence development. N Engl J Med. 2011
Feb 10;364(6):498 -9. doi:http://dx.doi.org/10.1056/NEJMp1014427 .
Black SB, Law B, Chen RT, Dekker CL, Sturkenboom M, Huang WT, et al. The critical role
of background rates of possible adverse events in the assessment of COVID -19
vaccine safet y. Vaccine. 2021 Mar 6;39(19):2712 -8.
doi:http://dx.doi.org/10.1016/j.vaccine.2021.03.016 .
CDC Covid Response Team. Preliminary estimates of the prevalence of selected underly ing
health conditions among patients with coronavirus disease 2019 —United States,
February 12-March 28, 2020. MMWR Morb Mortal Wkly Rep. 2020 Apr
3;69(13):382 -6. doi:http://dx.doi.org/10.15585/mmwr.mm6913e2 .
ClinicalTrials.gov NCT04368728. Study to describe the safety , tolerability, immunogenicit y,
and efficacy of RNA vaccine candidates against COVID -19 in healthy individuals.
2021. https://clinicaltrials.gov/ct2/show/NCT04368728 . Accessed 21 April 2021.
ClinicalTrials.gov NCT04754594. Study to evaluate the safet y, tolerabilit y, and
immunogenicit y of SARS CoV -2 RNA vaccine candidate (BNT162b2) against
COVID-19 in healthy pregnant women 18 years of age and older. 2021.
https://clinicaltrials.gov/ct2/show/NCT04754594 . Accessed 19 May 2021.
CSSE. Center for Sy stems Science and Engineering at Johns Hopkins University .
Coronavirus COVID -19 global cases. 2021.
https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40
299423467b48e9ecf6. Accessed 6 July 2021.
Curtis LH, Weiner MG, Boudreau DM, Cooper WO, Daniel GW, Nair VP, et al. Design
considerations, architecture, and use of the Mini- Sentinel distributed data system.
Pharmacoepidemiol Drug Saf. 2012 Jan;21 Suppl 1:23 -31.
doi:http://dx.doi.org/10.1002/pds.2336 .
DailyMed. US National Library of Medicine. Label: Janssen COVI D-19 vaccine-
ad26.cov2.s injection, suspension. 14 July 2021.
https://daily med.nlm.nih.gov/dail ymed/drugInfo.cfm?setid=14a822ff- f353-49f9-a7f2-
21424b201e3b. Accessed 28 July 2021.
Dorjee K, Kim H, Bonomo E, Dolma R. Prevalence and predictors of death and severe
disease in patients hospitalized due to COVID -19: a comprehensive s ystematic
review and meta -analysis of 77 studies and 38,000 patients. PL oS One.
2020;15(12):e0243191. doi: http://dx.doi.org/10.137 1/journal.pone.0243191 .
EMA. European Medicines Agency . Guideline on good pharmacovigilance practices (GVP)
Module VIII -Post-authorisation safety studies (Rev 3). 2017.
https://www.ema.europa.eu/en/documents/scientific- guideline/guideline -good-
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077335
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 67of 70pharmacovig ilance-practices- gvp-module-viii-post-authorisation -safety-studies-rev-
3_en.pdf. Accessed 18 May 2021.
ENCePP. European Network of Centres for Pharmacoepidemiology and Pharmacovigilance.
The European Union Electronic Register of Post -Authorisation Studi es (EU PAS
Register). 2021. http://www.encepp.eu/encepp_studies/indexRegister.shtml. Accessed
18 May 2021.
FDA. US Department of Health and Human Services, Food and Drug Administration.
Guidance for industry and FDA staff. Best practices for conducting an d reporting
pharmacoepidemiologic safet y. 2013. https://www.fda.gov/media/79922/download .
Accessed 18 May 2021.
FDA. Food and Drug Administration. Pfizer -BioNTech COVID -19 Vaccine EUA Letter of
Authorization (12 -15 years of age). 10 May 2021.
https://www.fda.gov/media/144412/download . Accessed 19 May 2021.
Frenck RW, Jr., Klein NP, Kitchin N, Gurtman A, Absalon J, L ockhart S, et al. Safet y,
immunogenicit y, and efficacy of the BNT162b2 COVID -19 vaccine in adolescents. N
Engl J Med. 2021 May 27. doi:http://dx.doi.org/10.1056/NEJMoa2107456 .
Gupta S, Hay ek SS, Wang W, Chan L, Mathews KS, Melamed ML, et al. Factors associated
with death in criticall y ill patients with coronavirus disease 2019 in the US. JAMA
Intern Med. 2020 Jul 15. doi: http://dx.doi.org/10.10 01/jamainternmed.2020.3596 .
ICH. International Conference on Harmonisation of Technical Requirements for Registration
of Pharmaceuticals for Human Use. Pharmacovigilance planning. E2E. 2004.
https://database.ich.org/sites/default/files/E2E_Guideline.pdf . Accessed 18 May
2021.
ICMJE. International Committee of Medical Journal Editors. Recommendations for the
conduct, reporting, editing, and publication of scholarly work in medical journals.
December 2019. http://www.icmje.org/recommendations . Accessed 18 M ay 2021.
IEA. International Epidemiological Association. Good epidemiological practice: IEA
guidelines for proper conduct of epidemiological research. 2007.
https://ieaweb.org/IEAWeb/Content/I EA_Publications.aspx . Accessed 10 August
2021.
ISPE. International Society for Pharmacoepidemiology . Guidelines for good
pharmacoepidemiology practices (GPP). Revision 3. June 2015.
https://www.pharmacoepi.org/resources/policies/guidelines-08027 . Accessed 18 May
2021.
Lamb YN. BNT162b2 mRNA COVID-19 v accine: first approval. Drugs. 2021
Mar;81(4):495- 501. doi:http://dx.doi.org/10.1007/s40265 -021-01480-7.
Lee FC, Adams L, Graves SJ, Massetti GM, Calanan RM, Penman- Aguilar A, et al. Counties
with high COVID -19 incidence and relativel y large racial and ethnic minority
populations—United States, April 1 -December 22, 2020. MMWR Morb Mortal Wkly
Rep. 2021 Apr 2;70(13):483- 9. doi:http://dx.doi.org/10.15585/mmwr.mm7013e1 .
Munoz FM, Cramer JP, Dekker CL, Dudley MZ, Graham BS, Gurwith M, et al. Vaccine -
associated enhanced disease: case definition and guidelines for data collection,
analysis, and presentation of immunization safety data. Vaccine. 2021 Feb 23.
doi:http://dx.doi.org/10.1016/j.vaccine.2021.01.055 .
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077336
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 68of 70NBDPN. National Birth Defects Prevention Network. I CD10-CM and birth defects. 2021a.
https://www.nbdpn.org/icd10.php. Accessed 21 April 2021.
NBDPN. National Birth Defects Prevention Network. NBDPN guidelines for conducting
birth defects surveillance. Chapter 3: Case definition. 2021b.
https://www.nbdpn.or g/docs/SGSC_-_Ch3_Case_Definition_-
_final_draft_2016DEC20.pdf . Accessed 17 May 2021.
NIST. National Institute of Standards and Technology . Security and privacy controls for
information systems and organizations. NIST special publication 800 -53, revision 5.
2020. https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NI ST.SP.800- 53r5.pdf.
Accessed 18 May 2021.
Oliver SE, Gargano JW, Marin M, Wallace M, Curran KG, Chamberland M, et al. The
Advisory Committee on Immunization Practices' interim recommendation for use of
Pfizer-BioNTech COVID-19 vaccine— United States, December 2020. MMWR
Morb Mortal Wkly Rep. 2020 Dec 18;69(50):1922-4.
doi:http://dx.doi.org/10.15585/mmwr.mm6950e2 .
Pfizer, Inc. on behalf of Pfizer and BioNTech. Emergency use authorization (EUA) for an
unapproved product review memorandum [Pfizer- BioNTech COVID -19 Vaccine]. 20
November 2020. https://www.fda.gov/media/144416/download . Accessed 21 April
2021.
Pfizer and BioNTech. Food and Drug Administration. Vaccines and Related Biological
Products Advisory Committee Meeting, December 10, 2020. FDA briefing document.
Pfizer-BioNTech COVID- 19 Vaccine. 2020.
https://www.fda.gov/media/144245/download . Accessed 21 April 2021.
Platt R, Brown JS, Robb M, McClellan M, Ball R, Ngu yen MD, et al. The FDA Se ntinel
Initiative—an evolving national resource. N Engl J Med. 2018 Nov 29;379(22):2091 -
3. doi:http://dx.doi.org/10.1056/NEJMp1809643.
Polack FP, Thomas SJ, Kitchin N, Absalon J, Gurtman A, Lockhart S, et al. Safety and
efficacy of the BNT162b2 mRNA COVID -19 vaccine. N Engl J Med. 2020 Dec
31;383(27):2603 -15. doi:http://dx.doi.org/10.1056/NEJMoa2034577 .
Sentinel. Food and Drug Administration. Data quality review and characterization programs.
2 August 2021. https://www.sentinelinitiative.org/methods -data-tools/sentinel -
common-data-model/data -quality-review-and-characterization -programs . Accessed 9
August 2021.
Sentinel. Food and Drug Administration. Sentinel common data model. 2018.
https://www.sentinelinitiative.org/sentinel/data/distributed- database-common-data-
model. Accessed 11 March 2021.
Sentinel. Food and Drug Administration. Sentinel data quality assurance practices. 23 March
2017. https://www.sentinelinitiative.org/methods -data-tools/sentinel -common-data-
model/sentinel- data-quality-assurance -practices. Accessed 18 May 2021.
Shimabukuro T. Advisory Committee on Immunization Practices (ACI P) March 1, 2021:
COVID-19 vaccine safety update. 2021.
https://www.cdc.gov/vaccines/acip/meetings/downloads/slides -2021-02/28-03-01/05-
covid-Shimabukuro.pdf . Accessed 21 April 2021.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077337
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 69of 70von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, et al. The
Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)
statement: guidelines for reporting observational studies. J Clin Epidem iol. 2008
Apr;61(4):344 -9. doi:http://dx.doi.org/10.1016/j.jclinepi.2007.11.008 .
Wallace M, Woodworth KR, Gargano JW, Scobie HM, Blain AE, Moulia D, et al. The
Advisory Committee on Immunization Practices’ interim recommendation for use of
Pfizer-BioNTech COVID -19 vaccine in adolescents aged 12 –15 years—United
States, May 2021. MMWR Morb Mortal Wkly Rep. 2021;70.
WHO. World Health Organization. WHO Director -General's opening remarks at the media
briefing on COVID -19—11 March 2020. 2020.
https://www.who.i nt/dg/speeches/detail/who -director-general-s-opening-remarks-at-
the-media-briefing-on-covid-19---11-march-2020. Accessed 1 April 2021.
Wong HL, Zhou CK, Thompson D, Dimova R, Clarke T, Forshee R. Food and Drug
Administration. COVI D-19 vaccine safet y surveillance: active monitoring master
protocol. 2021. https://www.bestinitiative.org/wp -content/uploads/2021/02/C19 -
Vaccine-Safety-Protocol-2021.pdf . Accessed 5 March 2021.
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077338
Pfizer-BioNTech COVID -19 Vaccine
C4591009 NON -INTERVENTIONAL STUDY PROTOCOL
1.0, 19 August 2021
PFIZER CONFIDENTIAL
CT24-WI-GL02-RF02 3.0 Non-Interventional Study Protocol Template For Secondary Data Collection Study
20-May-2021
Page 70of 7014.LIST OF TABLES
Table 1. Cohort definitions for analy ses of general population,
immunocompromised individuals, and individuals with history of
COVID-19................................ ................................ ................................ 27
Table 2. Eligible populations and exposure windows for analy ses of
pregnant women ................................ ................................ ........................ 29
Table 3. Cohort definitions for analy ses of pregnant women ................................ .30
Table 4. Events defining the start and end of follow -up................................ .........32
Table 5. Codes for COVID -19 vaccines available in the United States as of
28July 2021................................ ................................ .............................. 36
Table 6. General safet y events to be assessed in the general population,
immunocompromised individuals, individuals with a history of
COVID-19, and pregnant women ................................ ............................. 37
Table 7. Study size calculations ................................ ................................ ..............49
15.LIST OF FIGURES
Figure1. Hypothetical patients to illustrate eligibility for study cohorts and
follow-up of general safety events in the general population,
immunocompromised individuals, or individual s with history of
COVID-19................................ ................................ ................................ 34
Figure2. Hypothetical patients to illustrate eligibility for study cohorts and
follow-up of spontaneo us abortion events in pregnant women ................ 35
Figure3. General anal ytic workflow ................................ ................................ .......53
Figure4. Standard operating procedure for SAS programming qualit y
assurance and qualit y control in the Sentinel Sy stem............................... 60
ANNEX 1. LIST OF STAND -ALONE DOCUMENTS
None
ANNEX 2. ADDITIONAL INFORMATION
Not applicable
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077339
Document Approval Record
Document Name:
Document Title:
Signed By: Date(GMT) Signing Capacity
!"! #$%&'(
)*&%&+$*&&
!! ,- .&'(
090177e197e62b74\Approved\Approved On: 26-Aug-2021 16:54 (GMT)
FDA-CBER-2021-5683-1077340