Document text
Page 1 of 40Statistical analy sis plan
Avon Community Acquired Pneumonia Study (Avon CAP):
A Pan- pandemic Acute Lower Respiratory Tract Disease
Surveillance Study
Sponsor Project Reference Number 2020 -2991
ISRCTN registry number
IRAS ID Number 283899
REC Number 20/EE/0157
Pfizer reference number WI255886 -1
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073634
Page 2 of 40Contents
CHAPTER 1 –Study introduction and conduct ....................................................................................... 5
1. Rationale for study ......................................................................................................................... 5
2. Summary ......................................................................................................................................... 5
3. Schedule of activities ...................................................................................................................... 7
3.1 Consented Arm ........................................................................................................................ 7
3.2 Non -consented arm ................................................................................................................. 8
4. Research methods .......................................................................................................................... 9
4.1 Study design ............................................................................................................................. 9
4.2 Patient selection .................................................................................................................... 10
4.2.1 Screening inclusion criteria ................................................................................................ 10
4.2.2 Screening exclusion criteria ................................................................................................ 11
4.2.3 Study inclusion/exclusion criteria ...................................................................................... 11
4.2.4 UAD and respiratory pathogen control group ................................................................... 11
4.2.5 COVID -19 VE TND analysis group ....................................................................................... 12
5. Data management ........................................................................................................................ 12
5.1 Data sourc es........................................................................................................................... 12
5.2 NHS Databases ....................................................................................................................... 13
5.3 University of Bristol Data ...................................................................................................... 13
5.4 Pseudonymisation ................................................................................................................. 13
5.5 Data Flow Diagram ................................................................................................................ 14
5.6 Data col lection ....................................................................................................................... 14
5.6.1 Consented participants....................................................................................................... 14
5.6.2 Patient outcome ................................................................................................................. 16
5.7 Surveillance participants (non -consented) ........................................................................... 16
5.7.1 Admission data collection .................................................................................................. 16
5.8 Patient Outcome measures ................................................................................................... 18
Chapter 2 –COVID -19 Test Negative Design Analysis for Pfizer BNT162b2 ........................................ 20
1. Introduction .................................................................................................................................. 20
2. Objectives ..................................................................................................................................... 21
2.1 Primary ................................................................................................................................... 21
2.2 Secondary ............................................................................................................................... 21
2.3 Exploratory ............................................................................................................................. 21
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073635
Page 3 of 403. Selection criteria for inclusion in the analysis dataset ................................................................ .22
3.1 Inclusion Criteria .................................................................................................................... 22
3.2 Exclusion Criteria ................................................................................................................... 22
4. Test Negative Design .................................................................................................................... 22
4.1 Detection of cases .................................................................................................................. 24
4.2 Defini tion of Test- Negative Controls ..................................................................................... 25
4.3 Primary exposure of interest ................................................................................................ .25
4.4 Patient Sociodemographic Characteristics, Clinical History, Health Behaviours and
Lifestyle, and History of Vaccination ........................................................................................... 26
4.4.1 Patient Sociodemographic Characteristics, Clinical History, and Health Behaviours and
Lifestyle ........................................................................................................................................ 26
5. Statistical Analysis ........................................................................................................................ 32
5.1 Per protocol analysis population .......................................................................................... 32
5.2 Estimated Crude (Unadjusted) VE ......................................................................................... 33
5.3 Estimating Adjusted VE .......................................................................................................... 33
5.4 Ha ndling missing data ............................................................................................................ 33
5.5 Analysis Timings ..................................................................................................................... 33
5.6 Hypothesis Testing ................................................................................................................. 34
5.7 Sample Size Determination ................................................................................................... 34
References (chapter 2) ......................................................................................................................... 35
Appendix 1 -Sample size requirements ................................................................................................ 37
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073636
Page 4 of 40List of Abbreviations
Abbreviation Definition
A&E Accident and emergency
ARI Acute respiratory illness
BNT162b2 Pfizer/BioNtech COVID -19 vaccine
BRI Bristol Royal Infirmary
CAP Community -acquired pneumonia
CHF Congestive heart failure
COPD Chronic obstructive pulmonary disease
COVID Coronavirus disease
CRF Case report form (eCRF = electronic CRF )
CSA Clinical study agreement
CT Computed tomography
CVA Cerebrovascular accidents
CXR Chest x -ray
GCP Good Clinical Practices
ICH International Council for Harmonisation
ICU Intensive care unit
IRB/EC Independent Review Board/ Ethics Committee
LRTD Lower respiratory tract disease
MRI Magnetic resonance imaging
NP Nasopharyngeal
NSTEMI Non -ST-elevation myocardial infarction
OP Oropharyngeal
PCV7 7-valent pneumococcal conjugate vaccine
PCV13 13-valent pneumococcal conjugate vaccine
PCV20 20-valent pneumococcal conjugate vaccine
PPV23 23-valent pneumococcal polysaccharide vaccine
RSV Respiratory syncytial virus
SAP Statistical analysis plan
STEMI ST-elevation myocardial infarction
TND Test Negative Design
UAD Urinary antigen detection assay
VE Vaccine effectiveness
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073637
Page 5 of 40CHAPTER 1 –Study introduction and conduct
1. Rationale for study
Accurate incidence rates of acute LRTD and its disease subsets, such as pneumonia and LRTI, remain
elusive and the impact of COVID -19 on respiratory disease burden is unclear. Accurate incidence rates
of vaccine -preventable infection are required to assess the potential population -level impact of
vaccination recommendations.
This population- based multi -hospital, active prospective surveillance is designed to determine
population- based incidence rates of hospitalized adults ≥18 of age with community -acquire d LRTI
(including CAP) in Bristol, England. The involved Bristol hospitals nearly completely capture hospital
admissions among residents of a well delineated geographic region allowing for calculation of
population- based incidence rates of LRTI. Study data derived from surveillance activities will fully
enumerate the number of acute LRTD cases in this region.
LRTD cases will be offered participation in the consented portion of this study involving enhanced
testing for pneumococcal ,RSV, and SARS -CoV-2 infection. This will allow for more complete
characterization the incidence of these infections than standard of care (SOC) testing alone.
Additionally, real world vaccine effectiveness (VE) estimates for COVID -19vaccines are needed to
demonstrate their effec t in general populations as well as in risk groups outside of the clinical setting .
These canbe achieved using a test negative design (TND) case control study nested within the ongoing
Bristol LRTD surveillance network .
2. Summary
This study will undertake surveillance of lower respiratory tract disease (LRTD) in adults in a defined
geographical area and will recruit a subset of the patients identified into the embedded study to
provide extra samples and data, as described in the following chapters .
To calculate disease burden accurately and describe acute LRTD and its subsets, this study will record
all cases of adults hospitalised with acute LRTD and its subgroups within the defined geographical area
who meet study criteria. This will provid e data for calculation of disease incidence, burden and
outcome analysis in addition to other epidemiological analyses, including identifying risk factors for
disease and poor outcome.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073638
Page 6 of 40Individuals who decline consent for access to medical data will have data fields from medical records
where information is recorded as part of SOC collected to determine features of clinical presentation
and disease, including co-morbidities, routine healthcare tests and outcomes (e.g. requirement for
intensive care and/or org an support, mortality).
The data collected during the non-consented surveillance activity will be combined with the data
obtained from the consented arm to provide a comprehensive dataset of patients with acute LRTD.
In this way, all patients and disease will be captured and described, allowing for a complete and
accurate estimation of disease incidence and burden.
Additionally, VE analyses will be undertaken using data collected in the study. Analyses are described
in full in each of the chapters followin g this introductory chapter.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073639
Page 7 of 403. Schedule of activities
3.1 Consented Arm
Sample collection and interview activities shown in bold below will be undertaken for consented
participants only.
Table 1. Comprehensive List of Procedures by Visit for Consented -Arm, Including Visit Windows.
Procedure / Assessment aScreening/
Enrolment Visit bFinal Assessment/
Vital Status (Data
Collection only)Convalescent Visit
(“Serology Subset” Only)e
Visit 1 Visit 2 Visit 3
Day 1 Day 30 Day 45
Visit WindowWithin 48 hours of
admission Day 30 to 45 Day 22 to 60c
Screening (surveillance) X
Clinical symptoms X
Medical Historyc X X X
Vaccination History X X
CRB-65 and pneumonia severity
(PSI) scoreX
Informed Consent X
Eligibility confirmation for
embedded studyX
Patient Interview, including
approved questionnairesX X
Collect urine specimen X
Collect upper respiratory tract
sample for SARS -CoV-2, RSV and
other pathogensdX Xd
Collect blood specimeneX X
Obtain residual (scavenged)
standard care specimens from
clinical laboratoryX X X
Record SOC respiratory specimen
testing results X
Record SOC blood culture results X
Record SOC chest imaging results X
Final acute LRTD Illness Diagnosis X
Cardiac complications X
Vital Status/Mortality X
Hospitalization duration,
readmission, & level of careX
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073640
Page 8 of 40Procedure / Assessment aScreening/
Enrolment Visit bFinal Assessment/
Vital Status (Data
Collection only)Convalescent Visit
(“Serology Subset” Only)e
Visit 1 Visit 2 Visit 3
Day 1 Day 30 Day 45
Collect Research -related Injuries
(RRIs) X X X
Abbreviations: SOC = standard of care; UAD = urine antigen detection assay; RSV= respiratory syncytial virus;
RRI = research -related injury; CRB -65 = Confusion, Respiratory rate, Blood pressure, 65 years of age and older.
a. Sample collection and interview activities shown in bold above will be undertaken for consented
participants only.
b. All participants in the consented enhanced diagnostic testing will be have confirmation of their eligibility
confirmed and be subsequently enrolled at Visit 1 / Day 1.
c. At Visit 2 and 3, relevant changes to medical history since the last visit will be doc umented.
d. Respiratory samples will be collected for RSV and other respiratory pathogen testing if such testing has
not already been ordered or completed as part of standard -of-care testing. Details on the type and
collection process for samples will be inc luded in the laboratory manual . Another respiratory swab will be
collected at the convalescent visit ,only if the participant experienced a new ARIafter hospital discharge .
e. For those that consent to participate in the Serology subset, a blood sample will be collected for serologic
testing. A remnant of an appropriate blood sample from standard of care testing can be used if
appropriate for this use per laboratory manual specifications. For this subset of patients, an additional
convalescent visit will take place approximately 22 –60 days after enrolment for collection of a
convalescent blood sample. However, effort should be made to schedule this visit as close to day 42 as
possible. If subject has another enrolment qualifying acute LRTD event(s) prior to th eir convalescent visit,
they will have acute serology specimen taken at each enrolment and only one convalescent serology visit
will be completed 42 days after last acute specimen was taken.
3.2 Non -consented arm
Collection of SoC data will be done for all identified LRTD events in study hospitals and entered into
the surveillance databases hosted by the Trusts.
Table 2. Comprehensive List of Information to be collected for non -consented arm.
Procedure / AssessmentFirst Data Collection Second Data Collection
Admission Outcomes and Results
Day 1 Day 30
Visit WindowWithin 48 hours of
admission Day 30 to 45
Surveillance X
Clinical symptoms X
Medical HistoryaX X
Vaccination History X
CRB-65 and pneumonia severity
(PSI) scoreX
Record SOC respiratory specimen
testing results X
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073641
Page 9 of 40Procedure / AssessmentFirst Data Collection Second Data Collection
Admission Outcomes and Results
Day 1 Day 30
Record SOC blood culture results X
Record SOC chest imaging results X
Final acute LRTD Illness Diagnosis X
Cardiac complications X
Vital Status/Mortality X
Hospitalization duration,
readmission, & level of careX
Abbreviations: SOC = standard of care; UAD = urine antigen detection assay; RSV= respiratory syncytial virus;
RRI = research -related injury; CRB -65 = Confusion, Respiratory rate, Blood pressure, 65 years of age and older.
a. At Visit 2, relevant changes to medical history since the last visit will be documented.
4. Research methods
4.1 Study design
Adults with LRTD will be screened using population -level surveillance at study hospitals, and collection
of SOC data will be performed on all LRTD events, including from SOC laboratory tests . Patients
presenting during the recruitment period of the study, with documented or suspected COVID -19 will
fulfil study eligibility criteria, and therefore all references to LRTD also encompass documented or
suspected COVID -19 cases who may not otherwise qualify as LRTD. LRTD patients will be offered
participation in the enhanced diagnostic testing portion of this study with informed consent , which
will involve collection of urine, respiratory and,in some cases ,blood samples . These samples will be
used for study -specific testing and, if necessary, for COVID -19, pneumococcus ,and RSV tests if not
available from SOC records for any reason. A short patient questionnaire on COVID -related risk
behaviours will also be administered . The pneumococcal testing will include serotype to allow
estimation of the proportion of the burden that is potentially vaccine preventable –either by the
currently available PCV13 or the anticipated PCV20 , which is currently in the final phases of clinical
development. Information about the additional pneumococcal ,SARS -CoV-2 and RSV infection testing
will be integrated with the population -level surveillance data to allow for more accurate population -
based estimates of vaccine -preventable pneumococcal and COVID -19 and RSV-related LRTD
incidence. The epidemiologic data generated from the study will serve as the baseline for future
vaccine effectiveness studies –either for current and possibly future SARS -CoV-2 vaccines as well as
expected PCV20 and investigational RSV vaccine sthat arealso currently under development at Pfizer.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073642
Page 10 of 40An additional group of approximately 400 contemporaneous UAD and viral testing control
participants without acute LRTD , comparable by age group and season to subject sin the enhanced
testing group, will be enrolled. These participants will be identified as adults who are:
1.Patients attending outpatient clinics in participating hospitals in specialties other than respiratory
or cardiology OR
2.Patients currently admitted at participating hospitals on non-medical wards with no current acute
respiratory or cardiovascular conditions.
Control participants will provide urine for UAD and BinaxNOW testing and an upper respiratory swab
sample for RSV/ respiratory pathogen testing. The control population is used in every UAD study to
determine serotype and population specific cut -points for UAD positivity. For most adult populations
these cut-points have remained stable, but for study rigor, Pfizer policy is to obtain controls in each
population.
4.2 Patient selection
4.2.1 Screening inclusion criteria
Patients must meet all the following inclusion criteria to be eligible for enrolment :
1.Aged ≥18 years of age
2.Patients with illness with following 2 characteristics:
a.Acute illness (i.e., present for 2 8days or less); AND
b.Evidence of acute LRTD:
i.Patients with current or suspected COVID -19 or previous proven COVID -19 within last 28
days OR
ii. Clinical or radiologic diagnosis of pneumonia or an acute LRTI OR
iii. New onset or worsening of ≥2 of following 8 LRTD symptoms or clinical findings:
1.fever (>38.0°C) or hypothermia (<35.5°C) before or within 24 hours of enrolment ;
2.pleuritic chest pain;
3.cough (including nocturnal only);
4.sputum production or purulence;
5.dyspnea (shortness of breath) including orthopnea or on exertion only;
6.tachypnea (respiratory rate ≥20/min) documented by healthcare professional ;
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073643
Page 11 of 407.abnormal auscultatory findings suggestive of LRTD (e.g., crepitations/ rales or evidence of
pulmonary consolidation including dullness on percussion, bronchial breath sounds,
wheezing, or egophony);
8.radiologic finding that is consistent with LRTD, including pneumonia, and/or acute
congestive heart failure (e.g., pleural effusion, increased pulmonary density due to infection,
the presence of alveolar infiltrates (multilobar, lobar or segmental )containing air
bronchograms, or interstitial oedema).
4.2.2 Screening exclusion criteria
Patients meeting any of the following criteria will not be included in the study:
1.Any patient who develops signs and symptoms of LRTD after being hospitalized for ≥48 hours
(either at current hospital, another transferring hospital, or a combination of these), unless
admitted with current , previous proven ,or suspected COVID -19 infection.
2.Previously enrolled participants readmitted ≤7 days after discharge for their study qualifying
admission, unless admitted with current , previous proven ,or suspected COVID -19 infection
3.At the time of enrolment, an LRTD -related diagnosis has been excluded or another diagnosis
confirmed (for example, patient was found to have fever and tachypnoea due to an
intraabdominal process such as cho lecystitis)
4.2.3 Study inclusion/exclusion criteria
To participate in the enhanced diagnostic testing, individuals must meet all of the following:
1.Meet all screening inclusion criteria in section 5.2.2
2.Meet none of screening exclusion in section 5.2.3
3.Informed consent document signed and dated by patient, or the requirements for patients
unable to provide consent have been fulfilled (as detailed below)
4.2.4 U AD and respiratory pathogen control group
4.2.4.1 Control group inclusion criteria
Individuals must meet all of the following inclusion criteria:
1.Age 18 years and older.
2.Informed consent document signed and dated by patient
3.Individuals who are willing and able to provide urine and respiratory swab
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073644
Page 12 of 404.2.4.2 Control group exclusion criteria
Individuals presenting with any of the following will not be included in the control group :
1. Individuals who are investigational site staff members or relatives of those site staff member or
participants who are Pfizer employees directly involved in the con duct of the trial.
2. Individuals with suspicion of pneumonia or other respiratory infectious diseases or documented,
concomitant infectious disease.
3. Individuals residing in any long -term care facilities (for example, nursing homes orrespite care
faciliti es).
4. Individuals with known bronchial obstruction or a history of post -obstructive pneumonia.
(Chronic obstructive pulmonary disease (COPD) is permissible, provided there has not been an
exacerbation within the 3 months prior to enrolment. )
5. Individuals wi th primary lung cancer or another malignancy metastatic to the lungs.
6. Individuals with fever (measured temperature of ≥38.0° C measured by a healthcare provider).
7. Individuals with significant immunosuppressive disease such as leukaemia.
8. Individuals with either pneumococcal conjugate vaccine (PCV) and/or pneumococcal
polysaccharide vaccine (PPV) administration within the past 30 days
4.2.5COVID -19 VE TND analysis group
For the purposes of the primary COVID VE analysis, t he WHO definition of ARI will be used and is
defined as a nacute respiratory infection that occurred within the last 10 days and required
hospitalization with history of fever or measured fever of ≥ 38 C° AND cough. Because many COVID -
19 patients do not necessarily have fever (~15%) or cough (~15%) (Garg et al 2020 ), we will vary this
definition in sensitivity analyses.
5. Data management
5.1 Data sources
Data will be record edon an eCRF for the study, relating to outcomes of:
Radiology tests
Microbiology tests
Clinical parameters regarding hospital admission (including co -morbidities, symptom
type and duration, admission observations, etc)
Vaccination status
Patient treatment (including requirement for ventilation, renal support, etc)
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073645
Page 13 of 40Patient outcomes (including hospital length of stay, requirement for intensive care,
haemofiltration and complications arising from their LRTD).
Consent will be obtained to record which GP practi ce the participant is registered with, in order to
use this to generate a denominator for incidence calculation as well as for administration of the
COVID -risk behaviour questionnaire which the patient will complete with the research team member.
5.2 NHS Databases
A site-specific study database will be built within the NHS IT domain at each participating NHS site,
using the REDCap database management software . The REDCap database will be programmed to
mark identifiable fields (and therefore restrict their access or download), and to calculate fields such
as length of hospital admission, survival (up to 30 days following admission), age at admission. These
fields will be exported in the pseudonymised dataset as opposed to the specific date, thereby
removing identifiers and aggregating data.
5.3 University of Bristol Data
Data from each participating NHS Trust will be imported and held in two separate databases:
1.A REDCap database within a secure University IT domain -contain ingonly pseudonymised data
2.A password protected database containing identifiable data, with restricted user access on a
bespoke server
This will create a single unifying research dataset, allowing for analysis of data to meet the study
objectives, whilst enabling data security measure s to protect data and participants.
5.4 Pseudonymisation
Each patient with an admission that may meet eligibility will be assigned a surveillance number within
the database ,assigned by the research team. The database will be cleaned using the NHS number to
identify individuals who have more than one qualifying admission, both within one NHS Trust and
across participating study sites. These individuals will be identified in the pseudonymised database
using the surveillance number of their first study eligible admission. The NHS number will
subsequently be deleted following processing at the end of the study. This will provide a method of
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073646
Page 14 of 40ensuring that there is an accurate calculation of disease incidence with in a defined geographical area
with multiple NHS hospitals providing acute care, whilst maintaining a pseudonymised database. A
data flow diagram is included below .
5.5 Data Flow Diagram
5.6 Data collection
5.6.1 Consented participants
V1 data collection
Patients meeting screening criteria will have basic demographic and clinical data collected on an e -
CRF after consent. This will involve collection of the following data at visit 1:
Patient details
Eligibility checks – inclusion/ exclusion crit eria components
Date of hospitalization, length of stay in hospital
Demographics (age, gender, race/ethnicity ), socioeconomic status estimated by postcode,
vaccination ( influenza/ PCV13/PPV23 and date of last administration ), smoking status,
alcohol/drug use
Details of present illness (symptoms, date of onset, vital signs on admission to hospital , used
antibiotics in the 14 days prior to admission )
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073647
Page 15 of 40Standard -of-care test results
oRoutine biochemistry and haematology test results on admission, including C-reactive
protein and NT -proBNP
oBlood cultures
oRespiratory microbiology testing, including bacterial ( e.g., PCR and culture) and viral
testing ( e.g., COVID19, RSV and influenza)
oPneumococcal testing (e.g. , BinaxNOW urine test),
oAntibiotic resistance results for pneumococcal isolates
oSputum Gram stain results
oCOVID -19 testing data collection will include tests from up to 14 days prior to
hospitalisation based on patient self -report and medical records
Hospitalization data:
oAdmission hospital
oDates an d time of admission/discharge
oICU stay (yes/no)
oNumber of days in ICU
oMechanical ventilation and days of ventilator use
oRequirement for new/increased haemofiltration
oNon -invasive ventilation requirement and days of usage
New York Heart Association (NYHA) Heart Failure Classification, Pneumonia Severity (CRB65
and PSI scores)
Relevant Medical History and Major Comorbidities
Rockwood Frailty Score and Charlson Comorbidity Index
COVID risk behaviours questionnaire including household structure, inclusion in social
bubble, mask wearing (see appendix 1)
V2 data collection
Final standard of care/ clinical diagnosis of acute LRTD illness
Vital status at Day 30
Occurrence of cardiov ascular events within 30 days of illness onset (pneumonia only):
onon-ST elevation myocardial infarction (NSTEMI)
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073648
Page 16 of 40oST-elevation myocardial infarction (STEMI)
ocerebrovascular accident
onew episode of atrial fibrillation or other clinically significant arrhythmia
odeep venous thrombosis or pulmonary embolism
onew or worsening congestive heart failure
ocardiovascular- related death
5.6.2 Patient outcome
The follow -up information collected 30 days after enrolment or when the patient is discharged from
hospital (whichever occurs later) :
Record final clinical/standard -of-care diagnosis for qualifying acute LRTD illness:
o CAP –radiologically or clinically confirmed, acute bronchitis/LRTI, exacerbation of
underlying chronic respiratory disease, LRTI not otherwise specified, congestive cardiac
failure, empyema/lung abscess, non -infective process, and non-respiratory infecti on-
related diagnosis .
Vital status at Day 30 after enrolment:
odeceased, not recovered, recovered, recovered with sequelae ,recovery ongoing,
unknown
Cardiac complications through Day 30 after enrolment:
oST-and non -ST elevation myocardial infarction, cereb rovascular accident, new episode
of atrial fibrillation or other arrythmia, venous thromboembolism, cardiovascular -
related death
5.7 Surveillance participants (non -consented)
5.7.1 Admission data collection
Patients meeting screening criteria will have basic demographic and clinical data collected on an e -
CRF. This will involve collection of the following data at admission:
Patient details
Eligibility checks –inclusion/ exclusion criteria components
Date of hospitalization
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073649
Page 17 of 40Demographics (age, gender, race/ethnicity, socioeconomic status estimated by postcode,
vaccination (influenza/ PCV13/PPV23/COVID -19 and date of last administration ), smoking status,
alcohol/drug use)
Details of present illness (symptoms, date of onset, vital signs on admission to hospi tal)
oUsed antibiotics in the 14 days prior to admission
Standard -of-care test results
oRoutine biochemistry and haematology test results on admission, including C -reactiv e
protein and NT -proBNP, and blood group
Hospitalization data:
oAdmission hospital
oDates and time of admission
New York Heart Association (NYHA) Heart Failure Classification, Pneumonia Severity (CRB65 and
PSI scores)
Relevant Medical History and Major Comorbidities
Rockwell Frailty Score and Charlson Comorbidity Index for each patient base d on history of
chronic medical and immunocompromising conditions (Table 2).
5.7.1.1 Vaccination history
Vaccination history, including date(s) of vaccination and product(s) given, will be obtained from
medical records collected from relevant healthcare providers (e.g., primary care, public health
department), pharmacies, and any local, or national immunization registries for each enrolled patient
as follows:
Pneumococcal vaccine (e.g., 23-valent pneumococcal polysaccharide vaccine or 13-valent
pneumococcal conjugate vaccine) receipt in the last 5 years. History of vaccination with any other
newly- licensed or investigational pneumococcal vaccine (e.g., 20-valent or 15-valent pneum ococcal
conjugate vaccine which are under development by Pfizer, Inc. and Merck & Co., respectively) will
also be collected (if applicable) and categorized as ever received vsnever received in the last 5 years.
Influenza vaccine receipt in the year prior to enrollment .
Any COVID -19 vaccine receipt at any time .
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073650
Page 18 of 405.8 Patient Outcome measures
The follow -up information should be collected 1 month after enrolment or when the patient is
discharged from hospital (whichever occurs later). At day 30 after admission, the following will be
assessed and recorded in the CRF:
Microbiological investigation results
oBlood cultures.
oRespiratory microbiology testing, including bacterial ( e.g., PCR and culture) and viral
testing ( e.g., COVID -19, RSV and influenza),
oPneumococcal testing ( e.g., BinaxNOW urine test),
oAntibiotic resistance results for pneumococcal isolates
oSputum Gram stain results
Hospitalization data:
oDates of discharge
oICU stay (yes/no)
oNumber of days in ICU
oMechanical ventilation and days of ventilator use
oRequirement for new/increased haemofiltration
oNon -invasive ventilation requirement and days of usage
Vital status at Day 30 after enrolment:
oDeceased, not recovered, recovered, recovered with sequelae, recovery ongoing,
unknown
oDate of death if under 30 days from admission
Record final clinical/standard -of-care diagnosis for qualifying acute LRTD illness:
o CAP –radiologically or clinically confirmed, acute bronchitis/LRTI, exacerbation of
underlying chronic respiratory disease, LRTI not othe rwise specified, congestive cardiac
failure, empyema/lung abscess, non -infective process, and non -respiratory infection -
related diagnosis.
Cardiac complications through Day 30 after enrolment:
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073651
Page 19 of 40oST-and non -ST elevation myocardial infarction, cerebrovascular accident, new episode
of atrial fibrillation or other arrythmia, venous thromboembolism, cardiovascular -
related death
Hospital related adverse events (e.g., falls in hospital and hospital -acquired infections)
References (chapter 1)
Garg S, Kim L, Whitaker M, et al. Hospitalization Rates and Characteristics of Patients Hospitalized with
Laboratory -Confirmed Coronavirus Disease 2019 -COVID -NET, 14 States, March 1-30, 2020. Mmwr
2020;69:458 -64.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073652
Page 20 of 40Chapter 2 –COVID -19 Test Negative Design Analysis for Pfizer BNT162b2
1. Introduction
The UK began vaccinating its adult population in December 2020, using a Department of Health
defined risk-based strategy that prioritized vaccination based onage, comorbidity and key worker
status. Initially, only the Pfizer/Biontech mRNA COVID -19 vaccine BNT162b2 was used. BNT162b2 is
a nucleoside modified messenger ribonucleic acid vaccine that encodes the full-length, membrane -
anchored spike (S) glycoprotein of SARS -CoV-2 with two introduced proline mutations to lock it i n the
prefusion conformation (Kariko et al. 2008; Pardi et al. 2015, Wrapp et al. 2020). BNT162b2 showed
an acceptable safety profile in a Phase 1/2 study (Walsh et al 2020) and, in a Phase 3 trial, was
tolerable and demonstrated 95% efficacy against COVI D-19 (Polack et al. 2020). The vaccine currently
has temporary authorisation for supply under MHRA regulation 174, and data confirming the
effectiveness of the vaccine outside of the clinical trial setting are needed. Additionally, real world
vaccine effec tiveness (VE) estimates for BNT162b2 are needed to demonstrate itseffect in general
populations as well as in risk groups. Thiscan be achieved using test negative design (TND) case
control analys is. For this analys is, cases are individuals hospitalized with acute respiratory tract
infection (ARI) and testedpositive for SARS -CoV-2 up to 14 days prior to admission or on admission
to hospital . Controls are those in whom SARS -CoV-2was not detected in the same timeframe. Almost
all data needed to conduct thisanalys is are already being collected in this study, including COVID -19
disease and vaccination status from standard of care records alongside other medical history and
current illness details. To allow for more complete multivariable adjustment for poten tial confounding
differences between the cases and controls, additional information on COVID -19-related behavioural
risk factors will be collected from participants using a standardised questionnaire, such as mask use
and inclusion in a social bubble.
Residual nasopharyngeal or other respiratory specimens tested for COVID -19 as part of standard of
care will be saved prior to being discarded for destruction. Swabs will be retained frozen until such
time as they are selected and sent for variant determination which may include whole genome
sequencing (WGS). WGS will occur to identify variants of interest, to identify prevalent circulating
strains , and to assess VE against specific variants. WGS will not be used for patient care.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073653
Page 21 of 402. Objectives
2.1 Primary
To estimate the effectiveness of 2 doses of BNT162b2 vaccine (i.e. fully vaccinated) against
hospitalization for ARIdue to SARS -CoV-2 infection .
2.2Secondary
1.To describe the effectiveness of 1 dose of BNT162b2 vaccine (i.e., partially vaccinated)
agains t hospitalization for ARIdue to SARS -CoV-2 infection.
2.To describe the effectiveness of ≥1 dose of BNT162b2 vaccine (i.e., ever vaccinated) against
hospitalization for ARI due to SARS -CoV-2 infection .
3.To evaluate the effectiveness of BNT162b2 against ARI h ospitalization stratified by prevalent
or important viral strains.
4.To evaluate the effectiveness of BNT162b2 against severe hospitalization -related outcomes
(e.g., ICU admission, mechanical ventilation, and death).
2.3 Exploratory
1.To describe the effectiveness of BNT162b2 against hospitalization for any ARI stratified by
various patient characteristics (e.g., age group , sex, race/ethnicity, chronic medical
conditions, history of SARS -CoV-2 infection, long term care facility residence, pregnancy
status, receipt of influenza vaccine , time since vaccination (i.e., durability ), and time
between doses among those who received 2 doses ).
2.To calculate the incidence rate reduction of BNT162b2 against ARI hospi talization due to
SARS -CoV-2 infection and all ARI hospitalization calculate das VE * background incidence.
3.To describe the proportion of patients with ARI where SARS -CoV-2 was identified.
4.To summarize the proportion and characteristics of hospitalized ARI and SARS -CoV-2
patients who receive 0, 1, or 2 doses of BNT162b2 .
5.To summarize the time between administration of the first and second dose of BNT162b2
among patients who received 2 doses .
6.To summarize time since vaccination with BNT162b2 (most -recent d ose) from illness onset .
7.To describe demographic , clinical , and laboratory characteristics (i.e., viral strain ) and
disease severity of any BNT162b2 vaccine failures .
8.To describe COVID -19 disease severity for vaccinated and unvaccinated cases .
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073654
Page 22 of 403. Selection criteria for inclusion in the analysis dataset
This analysis can fulfil its objectives only if appropriate participants are included . The following criteria
will be used to select individuals for analysis dataset:
3.1 Inclusion Criteria
Participants must meet all the following inclusion criteria to be included in the analysis dataset :
1.Age 18 years or older.
2.Admitted to hospital for ARI* at a participating site.
3.Have a result from NP or nasal swab and all relevant data available in the study d atabase .
* For the purposes of the primary COVID VE analysis, t he WHO definition of ARI will be used and is defined as an acute
respiratory infection that occurred within the last 10 days and required hospitalization with history of fever or measured
fever of ≥38 C° AND cough. Because many COVID-19 patients do not necessarily have fever (~15%) or cough (~15%)
(Garg et al 2020) , we will vary this definition in sensitivity analyses.
3.2 Exclusion Criteria
Participants will be excluded from the analysis dataset if any of the following criteria apply:
1.Received SARS -CoV-2-directed antiviral treatment within the past 30 days, or COVID -19
monoclonal antibody therapy or COVID -19 convalescent serum therapy within the past 90
days prior to collection of required study -related procedures for the detection of SARS -CoV-
2 (i.e., NP or nasal swab).
2.Previous enrolment in this study within the past 30 days. Patients can contribute >1 ARI
event to the study if a subsequent ARI event for the same patient occurred >30 days after
the previous event.
3.Receipt of any COVID -19 vaccine apart from BNT162b2
4. Test Negative Design
A test negative design (TND) will be used, with patients identified on the basis of a clinical case
definition (e.g. acute respiratory infection). Patients are then tested for a vaccine aetiologic agent,
and VE is estimated by comparing the odds of vaccination among patients testing positive for vaccine -
type aetiology compared to those testing negative. Adjusting can be undertaken for potential
confounding factors. TND studies are considered a robust type of observational study for evaluating
VE against inf ectious respiratory diseases (De Serres et al 2013, Jackson et al 2013, Lipsitch et al 2016,
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073655
Page 23 of 40Sullivan et al 2014, Foppa et al 2016, Orenstein et al 2007). The main advantages of the TND are that
it helps avoid bias due to (unmeasured) healthcare -seeking behaviour and its ease of access to a series
of controls that are representative of the source population. Previous research has shown that this
simplicity does not necessarily come at the cost of validity (De Serres et al 2013, Jackson et al 2013,
Lipsitch et al 2016, Sullivan et al 2014, Foppa et al 2016, Orenstein et al 2007, Schwartz et al 2017).
Specifically, in observational settings, the TND is less susceptible to bias caused by differences in
healthcare -seeking behavior among cases and controls (De Serres et al 2013, Jackson et al 2013,
Lipsitch et al 2016, Haber et al 2015). This key premise of the TND is based on the idea that individuals
with a propensity to seek care when ill may be more likely to receive recommended vaccines and to
exhibit other behaviour (s) that reduces the risk of a given vaccine -preventable illness. Such
healthcare -seeking behaviour could confound the association between vaccination and the
development of infection in traditional cohort or case -control studies. Moreover, healt hcare -seeking
behavior is difficult, if not impossible, to measure directly. Thus, healthcare -seeking behaviour will, in
general, be an unmeasured confounder in traditional cohort or case -control studies that could
threaten the validity of VE estimates. By conditioning on patients presenting to healthcare providers
with the same clinical syndrome, the TND helps avoid bias due to (unmeasured) healthcare -seeking
behaviour.
Further, to prevent selection bias and obtain valid results from a case -control study, controls must be
representative of the source population (Miettinen et al 1976, Rothman et al 2017). As such, controls
should be individuals who, theoretically, would have been identified as cases had they acquired the
outcome, condition, or etiology of interest. If a control had had the disease, would they have been
likely to be enrolled as a case? (Aschengrau et al 2014) . Test-negative controls are derived from
patients presenting with similar clinical syndromes as the cases , for example acute respirator y
infection, and differ only on the specific pathogen, serotype or strain identified as aetiologic agent.
Given appropriate sensitivity and specificity of the test used to distinguish cases from controls,
(Lipsitch et al 2016, Orenstein et al 2007) the TND provides some inherent reassurance that controls
emerge from the same source population as cases, and that the primary difference between cases
and controls is which particular pathogen, serotype, or strain was identified as the causative agent of
the dis ease syndrome.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073656
Page 24 of 40More than thirty years ago, Broome et al. first applied the TND approach and used patients infected
by non-vaccine -type invasive pneumococcal disease to serve as controls in their analysis of
pneumococcal vaccine effectiveness (Broome et al 1980). The TND has since been extensively applied
to evaluate the effectiveness of influenza vaccine and is an extension of the same logic (De Serres et
al 2013, Jackson et al 2013, Lipsitch et al 2016, Sullivan et al 2014, Foppa et al 2016, Orenstein et al
2007, Schwartz et al 2017). Another recent study extended this design to an evaluation of 13 -valent
pneumococcal conjugate vaccine effectiveness against hospitalized community -acquired pneumonia
caused by vaccine serotypes (McLaughlin et al 2018 ). The current proposed evaluation of BNT162b2
will closely resemble these previous study designs.
Like all observational designs, the TND still requires assessment for bias and confounding that may
exist in the absence of randomized participation and blinded follow -up. As such, these studies will
collect a data for each participant to assess and control for potential confounding.
4.1 Detection of cases
SARS -CoV-2 will be detected by molecular techniques (i.e.nucleic acid amplification tests, NAAT) from
biological specimens collected from nasopharyngeal (NP) or nasal swabs . This will be undertaken via
collecting of results of SoC tests or from research specimen (if not undertaken as part of SoC testing).
Cases will be defined as patients who meet selection criteria and:
1.Test positive for SARS- CoV-2 via NAAT performed at hospital admission or study enrollment,
OR
2.Tested positive by NAAT from samples collection ≤14 days prior to hospital admission or
study enrollment.
Patients who have positive results for SARS -CoV-2 from samples taken ≥72hours after hospital
admission/enrollment will be considered to have nosocomial COVID -19 and will be excluded from the
primary analysis. The impact of excluding these patients on VE estimates will be examined in
sensitivity analyses. Laboratory confirmation is defined by NAAT, however, results from viral culture,
direct or indirect fluorescent antibody staining, or rapid antigen testing will be collected and analyzed
insensitivity analyses.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073657
Page 25 of 404.2 Definition of Test -Negative Controls
All other patients who met study inclusion criteria (e.g., at least one NP or nasal swab that was tested
for SARS -CoV-2 using NAAT) but for whom SARS -CoV-2 is not identified from NAAT will serve as test -
negative controls. This approach mimics the definitio n of test-negative controls that is commonly used
in TND studies of influenza (De Serres et al 2013, Jackson et al 2013, Lipsitch et al 2016, Sullivan et al
2014, Foppa et al 2016, Orenstein et al 2007, Schwartz et al 2017) and pneumococcal vaccines
(McLau ghlin et al 2018) .
4.3 Primary exposure of interest
The primary exposure of interest is history of vaccination with BNT162b2 vaccine . As an observational
study design, vaccination with BNT162b2 is not part of the study procedures, rather it would be given
as part of the national vaccination campaign. The dates of administration and confirmation that
BNT162b2 vaccine was given, which will be recorded in the medical records , will be included in the
study CRF data. Patients who received newly licensed or investigational SARS -CoV-2 vaccine sother
than Pfizer’s BNT162b2 will be excluded from all analyses .
BNT162b2 vaccination status will be also captured as part of the patient interview and can be self-
reported . Self- reported (only) vaccination status will be analyzed in sensitivity analyses.
Vaccination status will be defined as follows:
1.Fully vaccinated (primary objective) :defined as 2 doses of BNT162b2 vaccine received with
≥7 days between receipt of the 2nddose and ARI symptom onset . This group will serve as the
‘exposed’ group evaluated in the primary objective. Patients who received only 1 dose or 2
doses of BNT162b2 vaccine with <7 days between receipt of the 2nddose and ARI symptom
onset willbe excluded from this analysis. In sensitivity analyses, VE will also be calculated for
2 doses of BNT162b2 received with ≥14 days between receipt of the 2nd dose and ARI
symptom onset.
2.Partially vaccinated (secondary objective) :defined as 1 dose (only) of BNT162b2 vaccine
received with ≥14 days between receipt of the 1stdose and ARI symptom onset . This group
will serve as the ‘exposed’ group as a secondary endpoint. Patients who received 2 doses or
1 dose of BNT162b2 vaccine with <14 days between ARI symptom onset and receipt of the
1st dose will be excluded from this analysis.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073658
Page 26 of 403.Ever vaccinated (secondary objective): defined as ≥1 dose of the same BNT162b2 vaccine
received with ≥14 days between receipt of the 1st dose and ARI symptom onset . Patients
who received 1 dose of BNT162b2 vaccine received with <14 days between receipt of the
1st dose and ARI symptom onset will be excluded from this analysis. This group will serve as
the ‘exposed’ group as a secondary endpoint.
4.Never va ccinated defined as never received BNT162b2 vaccine. This group will serve as the
reference exposure group (i.e., ‘unexposed’ group) in all VE analyses.
4.4 Patient Sociodemographic Characteristics, Clinical History, Health Behaviours and Lifestyle,
and History of Vaccination
4.4.1 Patient Sociodemographic Characteristics, Clinical History, and Health Behaviours and
Lifestyle
Sociodemographic and clinical characteristics for each participant will be collected and described.
Importantly , these characteristics will be used to assess and control for potential confounding in
adjusted VE models. Factors assessed for confounding will include individual -level variables related to
participant sociodemographic characteristics (i.e., time and site of enrollment, age, sex, race,
ethnicity, and socioeconomic status using index of multiple deprivation [IMD] which uses postcode of
residence to assign a score of 1-10), clinical history and disease severity (i.e., presence of chronic
medical or immunoco mpromising conditions, history of SARS -CoV-2 infection, body mass index (BMI),
nursing home residency or other healthcare facility exposure in the past 3 months, and antibiotic use
in previous 14 days), health and lifestyle behaviors (e.g., smoking status , weekly exposure to children
<5 years of age, current drug abuse, and workplace exposure to COVID -19), and vaccination history
(e.g., influenza vaccine in last year and pneumococcal vaccine (s)in last 5 years). These are factors that
we have either found to be important covariates in previous work, have been identified in other risk
factor literature, or are variables that may be associated with the exposure as well as outcome (i.e.
prior positive SARS -CoV-2 PCR test, etc.) (Kumar et al 2020, Petrilli et al 2020, Popkin et al 2020, Singu
et al 2020, Tartof et al 2020, Webb et al 2020, Wu et al 2020, Zhou et al 2020, CDC website , Lancet
editorial 2020) .Section 5.6 of chapter 1 of this statistical analysis plan describes the data collection in
detail.
Clinical characteristics describing the presence or absence of underlying immunocompromising and
chronic medical conditions will also be collected. History of prior SARS -CoV-2infection (prior to the
current testing period and documented through a positive test in their medical records) will be
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073659
Page 27 of 40assessed. History of immunocompromising conditions will include a history of human
immunodeficiency virus/acquired immune deficiency syndrome (HIV/AIDS), leukemia, lymphoma,
Hodgkin’s Disease, generalized malignancy (excluding skin cancer), diseases requiring treatment with
immunosuppressive drugs including long term corticosteroids or radiation therapy, nephrotic
syndrome, chronic renal failure (including end-stage renal disease), organ transplantation, multiple
myel oma, sickle cell disease, functional or anatomic asplenia, and immune deficiency or other
conditions consistent with an immunocompromised state. Other chronic medical conditions and
health behaviors to be evaluated will include: hypertension, obesity, chr onic heart disease (including
heart failure, coronary artery disease, cardiomyopathy, and pulmonary hypertension), diabetes
(including type 1, type 2, or gestational), stroke, asthma (moderate or severe persistent), chronic lung
disease (such as chronic obstructive pulmonary disease (COPD, including emphysema and chronic
bronchitis ), idiopathic pulmonary fibrosis, and cystic fibrosis), chronic kidney disease (with dialysis),
chronic liver disease (including cirrhosis), alcoholism, or cigarette smoking. A classical and updated
Charlson comorbidity index (cCCI (Charlson et al 1987 )and uCCI (Quan etc at 2011 )) will be calculated
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073660
Page 28 of 40TABLE 1. CATEGORIZATION OF VARIABLES DESCRIBING SOCIODEMOGRAPHIC CH ARACTERISTICS ,CLINICAL HISTORY ,HEALTH BEHAVIORS AN D LIFESTYLE ,AND HISTORY OF
VACCINATION
Variable d.f. Categories
Sociodemographic Characteristics
Time of enrollment TBD in 1-week intervals
Recruitment site -Bristol 2 Site 1 –Southmeads
Site 2 -BRI
Age group, years at enrollment 1 linear tail -restricted cubic splines
Sex 1 Female
Male
Index of multiple deprivation (IMD) Group
level1-10 (inc)
Race 6 White British
White other
Mixed
Black or African American
Asian
Other
Unknown
Clinical History and Disease Severity
History of prior SARS -CoV-2 infection 1 Yes
No
classical Charlson Comorbidity Index
(cCCI) ( Charlson et al 1987 )5 0
1
2
3
4
5 or higher
updated Charlson Comorbidity Index
(uCCI) (Quan et al 2011 )5 0
1
2
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073661
Page 29 of 40Variable d.f. Categories
3
4
5 or higher
Body mass index (BMI) e5 Underweight (<18.5)
Normal or healthy weight (18.5 −24.9)
Overweight (25.0 −29.9)
Obese, class 1 (30.0 −34.9)
Obese, class 2 (35.0 −39.9)
Obese, class 3 ( ≥40.0)
Immunocompromising condition 1 Yes
No
Hypertension 1 Yes
No
Chronic heart disease 1 Yes
No
Diabetes 1 Yes
No
Asthma 1 Yes
No
Chronic lung disease 1 Yes
No
Chronic kidney disease 1 Yes
No
Chronic liver disease 1 Yes
No
Alcoholism 1 Yes
No
Antibiotic use in 14 days prior to enrollment 1 Yes
No
Health Behavior and Lifestyle
Healthcare facility exposure in past 3 monthsd1 Yes
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073662
Page 30 of 40Variable d.f. Categories
No
Living in long -term care, assisted living, or
skilled nursing facility at time of enrollment1 Yes
No
Current drug abuse 1 None
Alcohol excess, IVDU, Marijuana, other smoked drugs
Smoking status 3 Current smoker
Ex-smoker
Non -smoker
Unknown
Weekly exposure to children <5 years of age 1 Yes
No
School -age children currently living at home
with you1 Yes
No
Employed in a healthcare setting 6 Hospital or emergency medical services (EMS)
Ambulatory care clinic (e.g., urgent care, outpatient clinic, physician’s office)
Nursing home or long -term care setting
Pharmacy
Home health care
Other health care setting
No, I don’t work in the health care setting
Required to leave home to go to work 3 Yes, I cannot work from home
Most of the time, but I work from home 1 –2 days per week
Occasionally, but I work from home 3 or more days per week
No, I work from home exclusively
Job resulted in being offered vaccination against
COVID -191 Yes
No
Travelled internationally in the last month 1 Yes
No
Able to practice currently recommended social
distancing measures (for your area) for
combatting COVID -19 4 Always
Usually
About half the time
Seldom
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073663
Page 31 of 40Variable d.f. Categories
Never
Wear a mask or facial covering when you are in
public?4 Always
Usually
About half the time
Seldom
Never
Social interaction with other not sheltering
(living) with you4 A great deal
A moderate amount
Occasionally
Seldom
Never
Other Vaccination History
Influenza vaccination within previous year 1 Yes
No
PPSV23 vaccination in previous 5 years 1 Received ≥1 dose PPSV23
Never received a dose of PPSV23
PCV13 vaccination in previous 5 years 1 Received ≥1 dose PCV13
Never received a dose of PCV13
d.f.=degrees of freedom; PCV13=13 -valent pneumococcal conjugate vaccine; PPSV23=23- valent pneumococcal polysaccharide vaccine; SDI=social deprivation
index; TBD=to be determined.
a. Other includes American Indian or Alaska Native, Native Hawaiian or other Pacific Islander, or other or unspecified racial categories.
b. BMI is calculated based on weight and height and is reported in kg/m2.
c. Includes dialysis centers or long -term care, skilled -nursing, assisted -living, rehabilitation, or other healthcare facilities.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073664
Page 32of 40–AvonCAP SAPTABLE 2. CLASSICAL AND UPDATED CHARLSON COMORBIDITY INDEX (CCI) SCORING
Comorbid conditionCharlson Comorbidity Index
cCCI uCCI
History of myocardial infarction 1 0
Congestive heart failure 1 2
Peripheral vascular disease 1 0
Cerebrovascular disease 1 0
Dementia 1 2
Chronic pulmonary disease 1 1
Rheumatic disease 1 1
Peptic ulcer disease 1 0
Mild liver disease 1 2
Diabetes without chronic complication 1 0
Diabetes with chronic complication 2 1
Hemiplegia or paraplegia 2 2
Renal disease 2 1
Any malignancy without metastasis 2 2
Leukemia 2
Lymphoma 2
Moderate or severe liver disease 3 4
Metastatic solid tumor 6 6
AIDS (excluding asymptomatic infection) 6 4
Maximum (summary) comorbidity score 33 24
AIDS= acquired immunodeficiency syndrome; cCCI= classical Charlson comorbidity index (Charlson et al 1987) ;
uCCI= updated Charlson comorbidity index (Quan etc at 2011 )
5. Statistical Analysis
5.1 Per protocol analysis population
The Per -Protocol Population will include all participants who:
1.Meet all inclusion and exclusion criteria,
2.Have a final diagnosis consistent with ARI *
3.Meet the definition of either case or test -negative control
4.Report SARS -Cov-2 vaccination or have a history available from government -issued COVID -19
vaccination card s, medical and billing records collected from relevant healthcare providers
(e.g., primary care, public health department), health -insurance providers, pharmacies, and
any local, state, or national immunization registries
5.Did not receive any other newly licensed or investigational SARS -CoV-2 vaccine or COVID -19
prophylactic agent other than Pfizer’s BNT162b2.
*Note: As defined by WHO as an acute respiratory infection that occurred within the last 10 days and required
hospitalization with history of fever or measured fever of ≥ 38 C° AND cough.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073665
Page 33of 40–AvonCAP SAP5.2 Estimated Crude (Unadjusted) VE
Odds of having received BNT162b2 (fully, ever, and partially vaccinated) for cases and test-negative
controls will be constructed and compared using ORs and 95% CIs. VE will be calculated as 1−OR
multiplied by 100%. Corresponding 95% CIs will be calculated using the Wald method.
5.3 Estimating Adjusted VE
In addition to constructing crude OR and VE estimates, logistic regression modeling to assess
BNT162b2 VE after adjustment fo r potentially confounding factors, including (but not limited to) age,
gender, nursing home residence, comorbidities, history of COVID -19 infection, and study week will be
performed. These are factors that either been found to be important covariates in previous work, have
been identified in other risk factor literature, or are variables that may be associated with the exposure
as well as outcome ( e.g., prior positive COVID -19 test )(Kumar et al 2020, Petrilli et al 2020, Popkin et
al 2020, Singu et al 2020, Tartof et al 2020, Webb et al 2020, Wu et al 2020, Zhou et al 2020, CDC
website, Lancet editorial 2020). Covariates will be entered in the logistic regression model in
backward stepwise manner. Only variable(s) that change the estimated OR for BNT16 2b2 by ≥10%
(i.e., confounder) (Mickey et al 1989) will remain in the final VE model. Corresponding 95% CIs will be
calculated using the Wald method. Mixed -effects models will be used to model group -level
socioeconomic variables as a random intercept . In addition to results from the final model, univariate
VE results will be presented for each independent variable that is assessed for potential confounding,
as the results from a fully adjusted model.
5.4Handling missing data
Crude estimates of VE will be based on the observed, determinate SARS -CoV-2 test results and
BNT162b2 vaccination status. For adjusted VE evaluations, patients with all available covariates will
be included in the logistic regression model. If there is a substantial amount of missing data (>10%)
for any variables deemed necessary to include in our final analyses, sensitivity analysis will be
performed using multiple imputation for missing covariates (under the assumption of missing at
random) to understand the impact of excluding patients with missing information in adjusted models.
5.5Analysis Timings
The COVID -19 pandemic continues to cause substantial morbidity and mortality with new data
becoming rapidly available. As such, flexibility and a rapid response to the changing dynamics of
disease is needed. We plan to conduct informal interim analyses to inform decision making which may
be presented or published. Analyses for the primary, secondary, and exploratory outcomes will be
analyzed when available , in the form of ad hoc interim analyses, following LPLV of the study.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073666
Page 34of 40–AvonCAP SAP5.6Hypothesis Testing
The overall study type -I error is 5%. For the primary objective, hypothesis testing will be used to assess
if 2 doses of BNT162b2 are effective in preventing ARI requiring admission to hospital where SARS -
CoV-2 is identified. The primary null hypothesis (H0) vs the alternative hypothesis (H1) is H0: VE ≤ 20%
vsH1: VE> 20%, where VE is 1–OR multiplied by 100%, and the OR is equal to the odds of being fully
vaccinated with BNT162b2 (i.e., 2 doses of BNT162b2 with ≥7 days between receipt of the 2nddose
and ARI symptom onset )for ARI cases requiring hospitalization where SARS -CoV-2 is identified relative
to the odds of being fully vaccin ated with BNT162b2 for ARI cases requiring hospitalization where
SARS -CoV-2 is not identified. For the primary objective, BNT162b2 will be considered effective for
preventing ARI requiring hospitalization where SARS -CoV-2 is identified if the lower bound o f the 95%
confidence interval for the estimated VE is >20%.
5.7Sample Size Determination
This will be an event -driven study based on the number of cases identified. Study sample size is based
on the primary endpoint (BNT162b2 VE against ARI requiring hospitalization where SARS -CoV-2 is
identified). The required sample size will depend primarily on i) the proportion of all-cause ARI
requiring hospitalization caused by SARS -CoV-2 (which determines the number of cases identified and
the ratio of cases to controls in the primary analysis), ii) the average uptake of BNT162b2 in the study
population over the duration of the study, and iii) the assumed VE of BNT162b2 against ARI requiring
hospitalization where SARS -CoV-2 is identified. Sample size calculations were based on the following
fixed assumptions:
Two -sided, type -I error of 5%
90% power
Log(OR) following approximated normal distribution
Assumed true BNT162b2 VE varying from 70‒90% to prevent ARI requiring hospitalization was
modeled
5% of all -cause ARI episodes requiring hospitalization will test positive for SARS -CoV-2. A range of
5–30% was also modeled given the attack rate of COVID -19 may vary based on social distancing
and shelter- in-place measures, underlying levels of population immunity, and other factors.
Average BNT162b2 vaccine uptake in controls over the study period was allowed to vary in sample
size calculations (range: 10–90%) and will depend on potential future vaccination uptake scenarios
and timing of the conduct of the study. Final study enrollment size will also depend on the proportion
of enrolled patients excluded from the Per Protocol Population because i) vaccination records could
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073667
Page 35of 40–AvonCAP SAPnot be obtained, ii) they received a newly- licensed or investigational SARS -CoV-2 vaccine other than
BNT162b2 vaccine, or iii) they received BNT162b2 vaccine, but did not receive the full 2-dose schedule.
Appendix 1presents sample size calculations for various scenarios of BNT162b2 uptake and the
proportion of all-cause ARI where SARS -CoV-2 is identified. Depending on the uptake of BNT162b 2
and the proportion of ARI hospitalizations where SARS -CoV-2 is identified at the time of the study,
approximately 3,000 to 12,000 persons ≥18 years of age will be enrolled. A Statistics Center will
monitor BNT162b2 uptake among controls and the proportion of all-cause ARI where SARS -CoV-2 is
identifi ed to inform decisions on the sample size required to reach an effectiveness endpoint.
References (chapter 2)
Aschengrau A, Seage GR. Essentials of Epidemiology in Public Health. Third ed. Jones & Bartlett Learning; 2014.
Broome CV, Facklam RR, Fraser DW. Pneumococcal disease after pneumococcal vaccination: an alternative
method to estimate the efficacy of pneumococcal vaccine. N Engl J Med . Sep 04 1980;303(10):549 -52.
doi:10.1056/NEJM198009043031003
Butler DC, Petterson S, Phillips RL, Bazemore AW. Measures of social deprivation that predict health care
access and need within a rational area of primary care service delivery. Health Serv Res. Apr 2013;48(2 Pt
1):539 -59. doi:10.1111/j.1475 -6773.2012.01449.x
Charlson ME, Pompe i P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in
longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373 -83. doi:10.1016/0021 -
9681(87)90171 -8
De Serres G, Skowronski DM, Wu XW, Ambrose CS. The test -negative design: validity, accuracy and precision of
vaccine efficacy estimates compared to the gold standard of randomised placebo -controlled clinical trials. Euro
Surveill. Sep 12 2013;18(37)
Foppa IM, Ferdinands JM, Chaves SS, et al. The case test -negati ve design for studies of the effectiveness of
influenza vaccine in inpatient settings. Int J Epidemiol. Dec 1 2016;45(6):2052 -2059. doi:10.1093/ije/dyw022
Garg S, Kim L, Whitaker M, et al. Hospitalization Rates and Characteristics of Patients Hospitalized with
Laboratory -Confirmed Coronavirus Disease 2019 -COVID -NET, 14 States, March 1 -30, 2020. Mmwr
2020;69:458 -64.
Haber M, An Q, Foppa IM, Shay DK, Ferdinands JM, Orenstein WA. A probability model for evaluating the bias
and precision of influenza vaccine effectiveness estimates from case -control studies. Epidemiol Infect . May
2015;143(7):1417 -26. doi:10.1017/S0950268814002179
Jackson ML, Nelson JC. The test -negative design for estimating influenza vaccine effectiveness. Vaccine . Apr 19
2013;31(17):2165 -8. doi:10.1016/j.vaccine.2013.02.053
Kumar A, Arora A, Sharma P, Anikhindi SA, Bansal N, Singla V, Khare S, Srivastava A. Is diabetes mellitus
associated with mortality and severity of COVID -19? A meta -analysis. Diabetes Metab Syndr. 2020 Jul -
Aug;14(4):535 -545. doi: 10.1016/j.dsx.2020.04.044. Epub 2020 May 6.
Lipsitch M, Jha A, Simonsen L. Observational studies and the difficult quest for causality: lessons from vaccine
effectiveness and impact studies. Int J Epidemiol. Dec 01 2016;45(6):2060 -2074. doi:10.1093/ije/dyw124
McLaughlin J, Jiang Q, Isturiz R, Sings, H, Swerdlow D, Gessner B, Carrico RM, Peyrani P, Wiemken TL, Mattingly
WA, Ramirez JA, Jodar . Effectiveness of 13 -Valent Pneumococcal Conjugate Vaccine Against Hospitalization for
Community -Acquired Pneumonia in Older US Adults: A Test -Negative Design CID 2018 Oct 30;67(10):1498 -
1506. doi: 10.1093/cid/ciy312.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073668
Page 36of 40–AvonCAP SAPMickey RM, Greenland S. The impact of confounder selection criteria on effect estimation. Am J Epidemiol. Jan
1989;129(1):125 -37.
Miettinen O. Estimability and estimation in case -referent studies. Am J Epidemiol. Feb 1976;103(2):226 -35.
Orenstein EW, De Serres G, Haber MJ, et al. Methodologic issues regarding the use of three observational
study designs to assess influenza vaccine effectiveness. Int J Epidemiol . Jun 2007;36(3):623 -31.
doi:10.1093/ije/dym021
Petrilli CM, Jones SA, Yang J, et al. Factors associated with hospital admission and cr itical illness among 5279
people with coronavirus disease 2019 in New York City: prospective cohort study. BMJ. 2020;369:m1966.
Popkin BM, Du S, Green WD, Beck MA, Algaith T, Herbst CH, Alsukait RF, Alluhidan M, Alazemi N, Shekar M.
Individuals with obesit y and COVID -19: A global perspective on the epidemiology and biological relationships.
Obes Rev. 2020 Nov;21(11):e13128. doi: 10.1111/obr.13128. Epub 2020 Aug 26.
Quan H, Li B, Couris CM, et al. Updating and validating the Charlson comorbidity index and sc ore for risk
adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. Mar 15
2011;173(6):676 -82. doi:10.1093/aje/kwq433
Rothman KJ. Invited Commentary: When Case -Control Studies Came of Age. Am J Epidemiol. Jun 01
2017;185(11 ):1012 -1014. doi:10.1093/aje/kwx074
Schwartz LM, Halloran ME, Rowhani -Rahbar A, Neuzil KM, Victor JC. Rotavirus vaccine effectiveness in low-
income settings: An evaluation of the test -negative design. Vaccine . Jan 3 2017;35(1):184 -190.
doi:10.1016/j.vaccin e.2016.10.077
Singu S, Acharya A, Challagundla K, Byrareddy SN. Impact of Social Determinants of Health on the Emerging
COVID -19 Pandemic in the United States. Front Public Health. 2020 Jul 21;8:406. doi:
10.3389/fpubh.2020.00406. eCollection 2020.
Sullivan SG, Feng S, Cowling BJ. Potential of the test -negative design for measuring influenza vaccine
effectiveness: a systematic review. Expert Rev Vaccines . Dec 2014;13(12):1571 -91.
doi:10.1586/14760584.2014.96669
Tartof SY, Qian L, Hong V, et al. Obesity and Mortality Among Patients Diagnosed With COVID -19: Results From
an Integrated Health Care Organization. Ann Intern Med. 2020;173(10):773 -781.
Webb Hooper M, Napoles AM, Perez -Stable EJ. COVID- 19 and Racial/Ethnic Disparities. JAMA.
2020;323(24):2466 -2467.
Wu Z, McGoogan JM. Characteristics of and Important Lessons From the Coronavirus Disease 2019 (COVID -19)
Outbreak in China: Summary of a Repor t of 72314 Cases From the Chinese Center for Disease Control and
Prevention. JAMA. 2020;323(13):1239 -1242.
Zhou F, Yu T, Du R, et al. Clinical course and risk factors for mortality of adult inpatients with COVID -19 in
Wuhan, China: a retrospective cohort s tudy. Lancet. 2020;395(10229):1054 -1062.
The plight of essential workers during the COVID -19 pandemic. The Lancet. 2020 May 23;395(10237):1587.
doi: 10.1016/S0140 -6736(20)31200 -9.
WHO website –ARI definition:
https://www.who.int/influenza/surveillance monitoring/ili sari surveillance case definition/en/
Centers for Disease Control and Pre vention. COVID -19 and People with Certain Medical Conditions. 2021;
https://www.cdc.gov/cor onavirus/2019 -ncov/need -extra -precautions/people -with-medical- conditions.html .
Accessed March 2, 2021.
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073669
Page 37of 40–AvonCAP SAPAppendix 1-Sample size requirements for the final analysis population to detect BNT162b2 VE >20% assuming true VE=70 ‒90% with 90% power and type -I
error of 5% (2 -sided) under various BNT162b2 uptake scenarios ( 5to 30% of ARI hospitalization sdue to SARS- CoV-2)
Assume true VE=70%
5% of ARI is SARS -CoV-2 positive 15% of ARI is SARS -CoV-2 positive 25% of ARI is SARS -CoV-2 positive 30% of ARI is SARS -CoV-2 positive
BNT162
UptakeControls Cases Tot EvalTot
Enroll*Controls CasesTot
EvalTot
Enroll*Controls Cases Tot EvalTot
Enroll*Controls Cases Tot EvalTot
Enroll*
10 6769 356 7125 11875 2104 371 2475 4125 1171 390 1561 2602 938 402 1340 2233
20 3266 172 3438 5730 1022 180 1202 2003 573 191 764 1273 461 198 659 1098
30 2108 111 2219 3698 665 117 782 1303 377 126 503 838 304 130 434 723
40 1540 81 1621 2702 491 87 578 963 281 94 375 625 229 98 327 545
50 1213 64 1277 2128 392 69 461 768 228 76 304 507 187 80 267 445
60 1015 53 1068 1780 335 59 394 657 199 66 265 442 165 71 236 393
70 909 48 957 1595 308 54 362 603 187 62 249 415 157 67 224 373
80 905 48 953 1588 318 56 374 623 200 67 267 445 171 73 244 407
90 1174 62 1236 2060 435 77 512 853 287 96 383 638 251 108 359 598
Assume true VE=80%
5% of ARI is SARS -CoV-2 positive 15% of ARI is SARS -CoV-2 positive 25% of ARI is SARS -CoV-2 positive 30% of ARI is SARS -CoV-2 positive
BNT162
UptakeControls Cases Tot EvalTot
Enroll*Controls Cases Tot EvalTot
Enroll*Controls Cases Tot EvalTot
Enroll*Controls Cases Tot EvalTot
Enroll*
10 4945 260 5205 8675 1518 268 1786 2977 832 277 1109 1848 661 283 944 1573
20 2325 122 2447 4078 717 127 844 1407 396 132 528 880 315 135 450 750
30 1455 77 1532 2553 452 80 532 887 252 84 336 560 201 86 287 478
40 1024 54 1078 1797 321 57 378 630 181 60 241 402 146 63 209 348
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073670
Page 38of 40–AvonCAP SAP50 770 41 811 1352 245 43 288 480 140 47 187 312 114 49 163 272
60 608 32 640 1067 197 35 232 387 115 38 153 255 95 41 136 227
70 505 27 532 887 169 30 199 332 102 34 136 227 85 36 121 202
80 455 24 479 798 160 28 188 313 101 34 135 225 86 37 123 205
90 513 27 540 900 196 35 231 385 132 44 176 293 116 50 166 277
Assume true VE=90%
5% of ARI is SARS -CoV-2 positive 15% of ARI is SARS -CoV-2 positive 25% of ARI is SARS -CoV-2 positive 30% of ARI is SARS -CoV-2 positive
BNT162
UptakeControls Cases Tot EvalTot
Enroll*Controls Cases Tot EvalTot
Enroll*Controls Cases Tot EvalTot
Enroll*Controls Cases Tot EvalTot
Enroll*
10 4275 225 4500 7500 1294 228 1522 2537 698 233 931 1552 549 235 784 1307
20 1955 103 2058 3430 594 105 699 1165 322 107 429 715 253 108 361 602
30 1183 62 1245 2075 361 64 425 708 197 66 263 438 155 66 221 368
40 798 42 840 1400 245 43 288 480 135 45 180 300 107 46 153 255
50 568 30 598 997 176 31 207 345 98 33 131 218 78 33 111 185
60 417 22 439 732 132 23 155 258 74 25 99 165 60 26 86 143
70 313 16 329 548 101 18 119 198 59 20 79 132 49 21 70 117
80 241 13 254 423 83 15 98 163 51 17 68 113 43 18 61 102
90 212 11 223 372 82 14 96 160 56 19 75 125 50 21 71 118
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073671
Page 39of 40–AvonCAP SAPAssume true VE=90%
BNT162
UptakeCases Controls Tot EvalTot
Enroll*Cases Controls Tot EvalTot
Enroll*Cases Controls Tot EvalTot
Enroll*Cases Controls Tot EvalTot
Enroll*
10 224 22162 22386 37310 225 4275 4500 7500 227 2039 2266 3777 228 1295 1523 2538
20 102 10125 10227 17045 103 1955 2058 3430 104 934 1038 1730 105 594 699 1165
30 62 6116 6178 10297 62 1183 1245 2075 63 567 630 1050 64 361 425 708
40 42 4116 4158 6930 42 798 840 1400 43 383 426 710 43 245 288 480
50 30 2921 2951 4918 30 568 598 997 30 274 304 507 31 176 207 345
60 22 2131 2153 3588 22 417 439 732 23 203 226 377 23 132 155 258
70 16 1580 1596 2660 16 313 329 548 17 154 171 285 18 101 119 198
80 12 1194 1206 2010 13 241 254 423 14 122 136 227 15 83 98 163
90 10 992 1002 1670 11 212 223 372 13 115 128 213 14 82 96 160
*Assumes 40% of enrolled participants will be unevaluable (i.e., excluded from the Per Protocol Population because i) vaccina tion records could not be obtained, ii) they received a newly -
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073672
Page 40of 40–AvonCAP SAP
090177e1969c5f96\Approved\Approved On: 24-Mar-2021 17:28 (GMT)
FDA-CBER-2021-5683-1073673