medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
1
Title: Impact of COVID-19 and Effects of Vaccination with BNT162b2 on Patient-Reported
2
Health-Related Quality of Life, Symptoms, and Work Productivity Among US Adult Outpatients
3
with SARS-CoV-2
4
5
Authors: Manuela Di Fusco1*, Xiaowu Sun2, Mary M. Moran3, Henriette Coetzer2, Joann M.
6
Zamparo3, Laura Puzniak3, Mary B. Alvarez4, Ying P. Tabak2, Joseph C. Cappelleri5
7
8
Affiliations:
9
1
Pfizer Inc., Health Economics and Outcomes Research, New York, NY, USA
10
2
CVS Health, Woonsocket, RI, USA
11
3
Pfizer Inc., MDSCA Vaccines, Collegeville, PA, USA
12
4
Pfizer Inc., Field Medical Outcomes and Analytics, New York, NY, USA
13
5
Pfizer Inc., Statistical Research and Data Science Center, Groton, CT, USA
15
*
Corresponding author:
16
Manuela Di Fusco
17
Health Economics and Outcomes Research
18
Pfizer, Inc., New York, NY, USA
19
Manuela.difusco@pfizer.com
14
20
21
22
23
1
NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
24
ABSTRACT
25
Background: Although there is extensive literature on the clinical benefits of COVID-19
26
vaccination, data on humanistic effects are limited. This study evaluated the impact of SARS-
27
CoV-2 infection on symptoms, Health Related Quality of Life (HRQoL) and Work Productivity
28
and Impairment (WPAI) prior to and one month following infection, and compared results
29
between individuals vaccinated with BNT162b2 and those unvaccinated.
30
Methods: Subjects with ≥1 self-reported symptom and positive RT-PCR for SARS-CoV-2 at
31
CVS Health US test sites were recruited between 01/31/2022-04/30/2022. Socio-demographics,
32
clinical characteristics and vaccination status were evaluated. Self-reported symptoms, HRQoL,
33
and WPAI outcomes were assessed using questionnaires and validated instruments (EQ-5D-5L,
34
WPAI-GH) across acute COVID time points from pre-COVID to Week 4, and between
35
vaccination groups. Mixed models for repeated measures were conducted for multivariable
36
analyses, adjusting for several covariates. Effect size (ES) of Cohen's d was calculated to
37
quantify the magnitude of outcome changes within and between vaccination groups.
38
Results: The study population included 430 subjects: 197 unvaccinated and 233 vaccinated with
39
BNT162b2. Mean (SD) age was 42.4 years (14.3), 76.0% were female, 38.8% reported prior
40
infection and 24.2% at least one comorbidity. Statistically significant differences in outcomes
41
were observed compared with baseline and between groups. The EQ-Visual analogue scale
42
scores and Utility Index dropped in both cohorts at Day 3 and increased by Week 4, but did not
43
return to pre-COVID levels. The mean changes were statistically lower in the BNT162b2 cohort
44
at Day 3 and Week 4. The BNT162b2 cohort reported lower prevalence and fewer symptoms at
45
index date and Week 4. At Week 1, COVID-19 had a large impact on all WPAI-GH domains:
46
the work productivity time loss among unvaccinated and vaccinated was 65.0% and 53.8%, and
2
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
47
the mean activity impairment was 50.2% and 43.9%, respectively. With the exception of
48
absenteeism at Week 4, the BNT162b2 cohort was associated with statistically significant less
49
worsening in all WPAI-GH scores at both Week 1 and 4.
50
Conclusions: COVID-19 negatively impacted HRQoL and work productivity among mildly
51
symptomatic outpatients. Compared with unvaccinated, those vaccinated with BNT162b2 were
52
less impacted by COVID-19 infection and recovered faster.
53
54
Keywords: COVID-19, SARS-CoV-2, HRQoL, WPAI, Quality of Life, COVID-19 symptoms,
55
BNT162b2, humanistic
3
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
56
BACKGROUND
57
The impact of the COVID-19 pandemic on the sustainability of quality of life of patients has
58
been reported globally [1-4]. The prolonged multisystem symptoms associated to SARS-CoV-2
59
infection can negatively affect daily activities, ability to work, and social interactions, leading to
60
poor health-related quality of life (HRQoL) [1-4].
61
The introduction of COVID-19 vaccination has significantly impacted the COVID-19 response,
62
and evidence regarding the efficacy, safety and effectiveness of vaccination is extensive [5].
63
However, there is limited research on the potential benefits of vaccination on physical, mental,
64
social, emotional functioning and economic well-being. Most of the studies assessing humanistic
65
outcomes of COVID-19 infection have been limited to inpatients [1, 2, 6] were conducted
66
outside of the US or focused on specific disease states and organ-specific functions [7-9].
67
Leveraging a US national retail pharmacy SARS-CoV-2 test database and using validated
68
patient-reported outcome measures (PROMs), this study assessed COVID-19 symptoms,
69
HRQoL and WPAI prior to through one month following SARS-CoV-2 infection in outpatients,
70
and compared results between unvaccinated individuals and those vaccinated with BNT162b2.
71
72
METHODS
73
Study Design and Participants
74
The source population consisted of individuals testing for SARS-CoV-2 at one of ~5,000 CVS
75
Health test sites across the US. As part of the registration process for scheduling a SARS-CoV-2
76
test at CVS Health, individuals are required to complete a screening questionnaire including
77
demographics, symptoms, comorbidities, and vaccination status. The screening variables and
4
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
78
RT-PCR test results are loaded in an analytic dataset, where ~80-90% of test results are reported
79
within 2-3 days. Leveraging this analytic platform, this study was designed as a prospective
80
survey-based patient-reported outcomes study targeting adults ≥18 with a positive RT-PCR test
81
result and self-reporting at least one symptom. These individuals were emailed an invitation as
82
soon as the test results became available, no later than 4 days from testing. The email invitation
83
directed the potential participants to an e-consent website to learn about the study, survey
84
schedule and informed consent. Figure 1 summarizes the study design. Recruitment of
85
participants was carried out between 01/31/2022 and 04/30/2022 (Ct.gov NCT05160636).
86
87
Data Sources and Variables
88
Baseline characteristics and symptoms
89
Baseline characteristics of the participants were obtained via the CVS Health pre-test screening
90
questionnaire. These included self-reported demographics, comorbidities (including
91
immunocompromised status), COVID-19 vaccination history, social determinants of health
92
including the Social Vulnerability Index (SVI), work and/or residency in a high-risk or
93
healthcare setting, and symptoms. The list of baseline COVID-19 symptoms was based on the
94
CDC [10].
95
96
Exposure groups
97
Immunocompetent participants were considered fully vaccinated with BNT162b2 if they self-
98
reported receipt of 2 doses of BNT162b2 ≥ 14 days of SARS-CoV-2 testing. They were
99
considered partially vaccinated if reporting receipt of a single dose and boosted if reporting
100
receipt of 3 doses. Participants self-reporting an immunocompromising condition and receipt of
5
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
101
3 doses were considered fully vaccinated (i.e., 3-dose primary series completion); if reporting 4
102
doses, they were considered boosted. Participants were considered unvaccinated if they did not
103
report any COVID-19 vaccine dose prior to testing. Heterologous schedules were excluded.
104
105
HRQoL
106
To assess HRQoL, we used the validated EQ-5D-5L questionnaire [11, 12]. On the day of
107
enrollment, consented participants completed the EQ-5D-5L questionnaire twice, using two
108
versions: a modified version where all the questions were past tense to retrospectively assess pre-
109
COVID-19 baseline QoL, and the standard version in present tense to assess current QoL. To
110
minimize responder bias, the order of administration of the two versions was random.
111
Subsequent completion was requested at one month (short-term study design in Figure 1). The
112
EQ-5D-5L results at each time point were converted into the Utility Index (UI) using the US-
113
based weights by Pickard et al [12, 13].
114
115
Work Productivity and Activity Impairment
116
To measure impairments in both paid work and unpaid work, we used the Work Productivity and
117
Activity Impairment General Health V2.0 (WPAI:GH) measure [14, 15]. Participants were asked
118
to complete this questionnaire twice, seven days after their RT-PCR test: once referencing seven
119
days prior to COVID-19 symptom onset and an additional assessment referencing the past seven
120
days. Similar to the EQ-5D-5L, subsequent completion of the WPAI was requested at one month
121
(Figure 1). Four WPAI scores were computed at each time point: percent of worktime missed
122
(absenteeism), percent of impairment while working (presenteeism), percent of work
6
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
123
productivity loss (considering both absenteeism and presenteeism), and percent of activity
124
impairment. Only employed subjects were included for work productivity analyses.
125
126
Post-COVID 19 Symptoms and Vaccination Status Update
127
To supplement the pre-test screening questionnaire and enable the collection of on-going or new
128
symptoms after the acute phase, participants were sent an additional survey four weeks following
129
the test asking to complete a checklist of COVID-19 related symptoms based on the CDC list
130
[16], To confirm vaccination status, participants’ subsequent responses to vaccination date
131
questions were compared with their index responses; if responses did not match, the information
132
was queried and adjudicated, and the latest information was typically used.
133
134
Statistical Analysis
135
Descriptive statistics were used to analyze participant characteristics at baseline. Continuous
136
variables were described using means and standard deviations. Categorical variables were
137
reported using number and percentage distributions. For continuous variables, t-tests were used
138
to test difference in means and Wilcoxon tests were used to test difference in medians. For
139
categorical variables, chi-square tests were used to test differences between groups When cell
140
frequency was less than 5, Fisher’s exact tests were used for 2-by-2 tables and Freeman-Halton
141
tests for r-by-c tables [17, 18]. P values were all two-sided and not adjusted for multiplicity.
142
Mixed models for repeated measures (MMRM) [19] were used to estimate the magnitude of
143
COVID-19 impact on HRQoL and WPAI over time. Models included variables of time, self-
144
reported SARS-CoV-2 vaccination status, and interaction of time by vaccination status, as well
145
as covariates of participant pre-COVID-19 symptom onset score, sociodemographic
7
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
146
characteristics (age, sex, regions, social vulnerability, race/ethnicity, high risk occupations),
147
previously tested positive for COVID-19, severity of acute illness (number of symptoms reported
148
on index date), and immunocompromised status. Assessment time was fitted as a categorical
149
covariate and a repeated effect (repeated by subject). Least squares mean (LS mean) and
150
standard errors of PRO scores for each time point of assessment were calculated. Per guidelines,
151
no adjustment was made for missing data when scoring the EQ-5D-5L UI and WPAI [11, 15].
152
Missing data at each time were not imputed. All available data were included in the analysis.
153
154
Cohen’s d, or a variation of it, was calculated to assess the magnitude of score change from
155
baseline within the BNT162b2 vaccinated cohort and, separately, the unvaccinated cohort, as
156
well as the difference between BNT162b2 and unvaccinated cohorts [20, 21]. Specifically,
157
within-cohort effect size (ES) was calculated as mean change from baseline to follow-up, divided
158
by the standard deviation of change scores from baseline to follow-up. Between-cohort ES was
159
calculated as the difference in mean changes from baseline between cohorts, divided by the
160
pooled standard deviation of change scores. When calculating model based ESs, the numerators
161
were the predicted mean change from the model for within-cohort ESs, and predicted differences
162
from the model for between-cohort ESs. Denominators were the corresponding observed
163
standard deviations. Values of 0.2, 0.5, and 0.8 standard deviation (SD) units represent small,
164
medium, and large ES, respectively. These cut-off estimates have been widely used to establish
165
important differences in HRQoL studies [22]. As such, we considered the magnitude of
166
(standardized) effect sizes of at least 0.20 SD units as important or meaningful differences in
167
gauging the magnitude of within-patient change and between-group differences. All data
168
obtained were collected and analyzed with SAS Version 9.4 (SAS Institute, Cary, NC). The
8
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
169
study followed the Strengthening the Reporting of Observational Studies in Epidemiology
170
(STROBE) reporting guideline [23].
171
172
RESULTS
173
Baseline characteristics
174
A total of 39,889 eligible candidates were outreached. Of those, 676 consented and completed
175
the first survey, for a consent rate of 1.7%. Compared with individuals in the CVS Health
176
analytic dataset who did not participate in our study, the study sample was over-represented by
177
women and Caucasians, with slightly more individuals vaccinated and with comorbidities
178
(Supplemental Table 1). The final study population included 430 subjects (Figure 2). 100%
179
completed the EQ-5D-5L questionnaire at pre-COVID-19 baseline and at Day 3, and 77.0%
180
completed it at Week 4. The WPAI-GH questionnaire was completed by 88.1% of the
181
participants at pre-COVID-19 baseline, 88.1% at Week 1 and 76.9% at Week 4.
182
The sociodemographic characteristics of the baseline participants are shown in Table 1. Overall,
183
the mean (SD) age was 42.4 (14.3), 76% were female, 68.6% Caucasian, 58.7% from Southern
184
US. There were 24.2% participants who reported ≥1 comorbidities, including 4.4% with
185
immunocompromising conditions and 39% reported a previous COVID-19 infection.
186
About 46% (197) were unvaccinated and 54% (233) were vaccinated with BNT162b2; of those,
187
respectively 140 (60%) and 93 (40%) received 2 and 3 doses. Compared with unvaccinated,
188
BNT162b2 participants were comparable with respect to gender, working and living settings, and
189
comorbidities, slight older with mean age 43.7 vs. 40.9 (p=0.049); living in less vulnerable area
190
with lower mean social vulnerability index (0.40 vs. 0.49, P<<0.001); and slight differences in
9
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
191
race/ethnicity and region. In the vaccinated group, mean (SD) time since vaccination before
192
infection was 186 (105) days.
193
At index date, the most reported acute symptoms were respiratory and systemic. BNT162b2
194
vaccinated participants reported fewer overall acute COVID-19 symptoms on average than
195
unvaccinated participants, mean 5.1 vs. 5.6, P=0.034 (Table 1). Directionally, the proportions of
196
all systemic and GI-related symptoms were numerically lower in the BNT162b2 cohort. Relative
197
to unvaccinated, those vaccinated with BNT162b2 reported significantly fewer symptoms of
198
fever (30.5% vs. 47.2% P<0.001), chills (42.9% vs. 57.4%, P=0.003), muscle or body aches
199
(49.4% vs. 59.4%, P=0.038), and diarrhea (15.9% vs. 25.9%, P=0.010), but more congestion or
200
runny nose (80.7% vs. 68.0%, P=0.003).
201
202
Post-COVID-19 symptoms
203
At Week 4, the mean number of symptoms was statistically lower in the BNT162B2 cohort (2.5
204
vs. 3.7, p=0.002). The overall prevalence decreased over time too, especially fever, cough,
205
headache, fatigue, diarrhea, muscle pain; however, ~70% of participants still reported at least 1
206
post-COVID-19 symptom. Directionally, the proportions of all symptoms were numerically
207
lower in the BNT162b2 cohort. Symptoms of worsening after physical or mental activities
208
(10.3% vs. 20.6%), general pain/discomfort (11.4% vs. 19.4%), change in smell or taste (10.9%
209
vs. 20.6%), headache (16.0% vs. 25.2%), sleep problems (20.0% vs. 29.7%), mood changes
210
(7.4% vs. 14.8%), memory loss (6.3% vs. 17.4%) and diarrhea (3.4% vs. 11.0%) were
211
statistically significant (P < 0.05) (Table 2).
212
213
Health-Related Quality of life
10
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
214
Utility Index scores
215
Mean pre-COVID-19 baseline UIs did not differ between the BNT162b2 and unvaccinated
216
cohorts, respectively 0.924 and 0.918 (P=0.547). COVID-19 infection had a detrimental effect
217
on the HRQoL of participants, especially during the acute episode (Day 3). In both the
218
BNT162b2 and the unvaccinated cohorts, UIs were lower at Day 3 and Week 4 relative to pre-
219
COVID-19. While UIs improvement was observed over time, the UI did not return to pre-
220
COVID levels at Week 4 (Table 3).
221
The BNT162B2 cohort was less impacted than the unvaccinated cohort, at both Day 3 and Week
222
4. After controlling for pre-COVID baseline score and other covariates, the least-square estimate
223
UI scores at Day 3 were, respectively 0.77 and 0.84 in the unvaccinated and BNT162B2 cohorts
224
(Table 4). Moderate ESs of, respectively, 0.64 and 0.49 were observed from baseline. At Week
225
4, the least-square estimate UI scores were, respectively, 0.86 and 0.90. Small-to-moderate ESs
226
of, respectively, 0.38 and 0.13 were observed from baseline. The differences between the two
227
groups were statistically significant (P<0.05). (Table 4) Small-to-medium ESs between cohorts
228
were observed and were 0.36 and 0.32 for Day 3 and Week 4, respectively. (Table 4,
229
Supplemental Figure 1 and 2)
230
231
EQ-VAS
232
The pattern of EQ-VAS scores was similar to that observed for UIs. Mean pre-COVID-19
233
baseline EQ-VAS were similar for the BNT162b2 and unvaccinated cohorts, respectively 86.9
234
and 87.8 (P=0.414) (Table 3). Similar to the UIs, the pre-COVID EQ-VAS were rated relatively
235
high by the participants, indicating a generally healthy cohort. The least-square estimate EQ-
236
VAS scores for the BNT162b2 and unvaccinated cohorts were, respectively, 76.2 and 72.6 at
11
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
237
Day 3 and 85.0 and 81.6 at Week 4. After controlling for pre-COVID-19 baseline score and
238
other covariates, the least-square estimates of change from pre-COVID-19 baseline in EQ VAS
239
for the BNT162B2 and the unvaccinated cohort were -11.1 and -14.8, respectively on Day 3, and
240
-2.3 and -5.7, respectively at Week 4. COVID-19 had a large adverse impact on EQ-VAS with
241
an ES of -0.89 for BNT162B2 cohort and -0.86 for Unvaccinated cohort on Day 3, and small ES
242
(-0.22) for BNT162B2 cohort and approaching medium ES (-0.42) for Unvaccinated cohort at
243
Week 4. BNT162B2 cohort was associated with 3.6 (P=0.013) on Day 3 and 3.4 (P=0.016) at
244
Week 4 less drop in EQ VAS than the Unvaccinated cohort. The ESs between cohorts were
245
small yet relevant, being 0.25 and 0.28 for Day 3 and Week 4, respectively (Table 4,
246
Supplemental Figure 1 and 2).
247
248
EQ-5D-5L dimensions
249
The health status of the study participants according to the dimensions of EQ-5D-5L is reported
250
in Figure 3 and Supplemental Table 2. In both groups, at Day 3, over half of the cohort reported
251
problems in usual activities, pain/discomfort and anxiety/depression, while the vast majority
252
reported no or slight problems in mobility and self-care. At Week 4, the vast majority continued
253
to report no or slight problems with mobility, self-care, as well as for usual activities; most
254
reported no, slight or moderate problems with pain/discomfort and anxiety/depression.
255
BNT162b2 cohort had lower mean responses across all 5 domains at both Day 3 and Week 4
256
relative to unvaccinated.
257
258
Work Productivity and Activity Impairment
12
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
259
Approximately 65% of participants reported being currently employed at baseline (155 in the
260
BNT162b2 cohort and 129 unvaccinated), and were eligible to complete the absenteeism,
261
presenteeism and work-productivity loss questions. At Week 1, COVID-19 had a large impact on
262
all four WPAI-GH domains for both the unvaccinated and BNT162b2cohort. The mean time loss
263
due to absenteeism was, respectively, 65.0% and 45.6%; the mean time loss due to presenteeism
264
was, respectively, 46.8% and 38.4%; the mean time of work productivity loss was 65.0% and
265
53.8%, and the mean time of activity impairment was 50.2% and 43.9%. All within-cohort ESs
266
were > 0.8, which are considered large effects (Table 3). After controlling for pre-COVID-19
267
baseline score, and other covariates, the BNT162b2 cohort was associated with less worsening in
268
WPAI-GH scores. Small-to-medium ESs were observed for work-related scores (absenteeism -
269
0.50, presenteeism -0.26, and work productivity loss -0.32) between the BNT162b2 cohort and
270
the unvaccinated cohort (Table 4). At Week 4, the mean time loss dropped across all four
271
domains. The time loss due to absenteeism dropped substantially; the change from baseline in
272
absenteeism was not found to be statistically significant between the BNT162b2 cohort and the
273
unvaccinated cohort. Small-to-medium ESs were observed for presenteeism (-0.38) work
274
productivity loss (-0.29), and activity impairment (-0.34) between the BNT162b2 cohort and the
275
unvaccinated cohort (Table 4, Supplemental Figure 3 and 4).
276
277
DISCUSSION
278
The impacts of SARS-CoV-2 infection go beyond its clinical outcomes.
279
We found that mild acute infection can negatively impact the humanistic outcomes for up to four
280
weeks post infection. Shortly after infection, the UI and EQ-VAS HRQoL scores dropped from
281
pre-COVID, and over half of the study population reported problems in usual activities,
13
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
282
pain/discomfort and anxiety/depression. At Week 1, the work productivity and activity
283
impairment time loss were over 50%. At Week 4, both the HRQoL and WPAI scores improved,
284
although they did not return to pre-COVID levels. Individuals vaccinated with BNT162b2 were
285
less impacted and recovered faster than unvaccinated individuals. Multivariable analyses showed
286
that BNT162b2 was significantly associated with higher EQ-VAS and UI scores, less symptoms
287
and better WPAI scores, except for absenteeism at Week 4.
288
There is limited evidence measuring the health-related wellbeing of non-hospitalized individuals
289
affected by COVID-19 [1-4]. To our knowledge, this is the first report measuring the impact of
290
COVID-19 on the HRQoL and WPAI among a national sample of outpatients in the United
291
States. In contrast to our study, previous research that used EQ-5D scales to measure COVID-19
292
impact on the HRQoL reported mean UI scores ranging from 0.61 to 0.86 depending on the
293
hospitalization treatment and time since discharge [1, 2]. The EQ-VAS scores ranged from 50.7
294
to 70.3 [1, 2]. In our study, the HRQoL scores at Day 3 and Week 4 among outpatients with mild
295
disease are higher than those, likely due to the different study populations and periods. In a small
296
US study assessing the impact of COVID-19 on WPAI ~4 months post-infection among subjects
297
enrolled in clinical trials before the introduction of vaccines, 46% of the non-hospitalized
298
patients reported health-related impairment in daily activities [24]. Among the employed, 11.5%
299
missed work and 38.9% reported impairment at work due to health. In our study, all the WPAI
300
scores among unvaccinated at Week 1 are higher, and those at Week 4 similar or lower than
301
those, likely due to the different cut-off, study populations, periods and design.
302
HRQoL and WPAI studies are scarce in COVID-19 related vaccination research. To our
303
knowledge, this is the first report on the effects of BNT162b2 on these patient-centric outcomes.
14
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
304
These results indicate an additive benefit beyond vaccine effectiveness that should be explored
305
further.
306
Strengths of this study include the nationally representative real-world source population of
307
mildly symptomatic outpatients, the prospective collection of the primary outcomes via validated
308
instruments, and the representativeness of the employed population for work productivity
309
analyses (~65% of the cohort). The age distribution of study participants was comparable with
310
the non-enrolled tested population (p=0.076 for mean age, Supplemental Table 1) and with CDC
311
research in non-hospitalized adults: 91.6% were between 18 and 64 and 8.4% were 65 or older,
312
which was quite similar to Hernandez-Romieu [25], 92.7% and 7.3%, respectively. In the
313
vaccinated group, the time between vaccination and breakthrough infection (mean: ~6 months)
314
was consistent with literature on vaccine-induced duration of protection [5].
315
316
317
318
319
320
321
322
323
The study findings are subject to limitations. All the data analyzed was self-reported and may be
subject to error, missingness, recall bias, social desirability bias, and selection bias associated
with survey drop-out. Out of 430 participants completing Day 3 survey, 12% (51/430) missed
Week 1 and 23% (99/430) missed Week 4 survey. The drop in responses may partly be the result
of responders’ fatigue, and/or recovered cases not returning to follow-up surveys. Female and
older age (≥30 years) were found to be more likely to miss follow-up surveys. However, after
taking into account several variables, the model predicting missingness indicated that
missingness was not associated with vaccination status, nor with HRQoL in terms of EQ VAS on
Day 3 and its change from pre-COVID-19 baseline (
15
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
324
Supplemental Table 4).
325
The study population differed from non-enrolled tested outpatients in being predominantly
326
female, white, with a higher vaccination uptake, and slightly more comorbidities (Supplemental
327
Table 1). The female over-representation is in line with prior research indicating that women are
328
more likely to contribute to health research surveys [2]. In our study, ~24% of participants
329
reported at least one comorbidity, in contrast to 20% among the non-enrolled tested and ~35% in
330
Hernandez-Romieu (2021) [25]. Various models were fit to account for potential effects due to
331
sociodemographic factors and comorbidities. These adjusted effect sizes between BNT162b2 and
332
unvaccinated cohorts were similar and consistent with those calculated from observed data
333
(unadjusted effect sizes).
334
The pre-COVID baseline scores were slightly higher than US population norms. The mean EQ-
335
VAS and UI were 86.9 and 0.924 for the BNT162b2 cohort, and 87.8 and 0.918 for the
336
unvaccinated cohort, respectively. Cha et al (2019) reported a mean EQ-VAS of 84.6 for the U.S.
337
general population [12]. Jiang et al (2021) reported the US population norm of EQ-VAS as 80.4
338
and the mean EQ-5D-5L UI as 0.851 [26]. The healthy pre-infection status of the study
339
population and the potential for retrospective recall bias may partially explain the differences. A
340
modified EQ-5D-5L questionnaire was used to retrospectively measure pre-COVID-19 baseline
341
data; despite literature suggesting concordance between prospective and retrospective
342
measurements of EQ-5D-5L [27, 28], and a good correlation between assessment of baseline
343
before the index date and recall assessment of the baseline after index date, there is currently no
344
information regarding the recall application of the EQ-5D-5L for COVID-related studies.
345
The pre-COVID-19 values for abseenteism and presenteeism were 3.1% and 9.5% in the
346
BNT162b2 cohort, and were generally in line with Tundia et al (2015) [29], whom reported 4%
16
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
347
absenteeism and 10% presenteeism for the US general population. The reported values were
348
slightly higher among unvaccinated, respectively 11.7% and 9.4%.
349
350
There is currently no standard definition of minimal clinically important difference of PROs in
351
COVID-19 research. We used effect size (ES) of Cohen’s d to quantify the magnitude of score
352
change from baseline within the BNT162b2 vaccinated cohort and the unvaccinated cohort, as
353
well the difference between these two cohorts [20]. An effect size ES of 0 between groups
354
indicates that the average (typical) treated vaccinated person has a score that is no different from
355
the typical control person; equivalently, scores of the typical treated person are more favorable
356
than 50% of the individual scores in the control group, meaning no incremental benefit (a coin
357
toss as to which intervention is better). If the vaccinated cohort is presumed more effective than
358
the unvaccinated cohort, effect sizes thresholds of 0.2, 0.3, 0.5, and 0.8 (in absolute value)
359
indicate that, based on the standardized normal distribution, the score of the typical person in the
360
vaccinated cohort is more favorable than 58% (8% incremental benefit), 62% (12% incremental
361
benefit), 69% (19% incremental benefit), and 70% (29% increment benefit) of the scores from
362
individuals in the unvaccinated cohort. For example, from baseline to Week 1, the increase in the
363
absenteeism WPAI score of the typical person in the vaccinated cohort was less (more favorable)
364
than the corresponding change in 69% of individuals in the unvaccinated group (effect size = -
365
0.5). Depending on the type of outcome, the same type of effect size interpretation for between
366
cohorts can be given within cohort. For instance, from baseline to Week 4, the quality of life EQ-
367
VAS score of the typical person in the vaccine cohort at Week 4 was less (worse) than 59% of
368
the individual scores in the vaccine cohort at baseline (effect size = -0.22); the quality of life EQ-
17
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
369
VAS score of the typical person in the unvaccinated cohort at Week 4 was less (worse) than 66%
370
of the individual scores in the unvaccinated cohort at baseline (effect size = -0.42).
371
372
The study did not assess the impact on pediatrics, caregivers, long-term outcomes (e.g., “Long
373
COVID”), and the data was collected during Omicron predominance in the US. Therefore, these
374
findings may not be generalizable to prior or future variants, other countries, time periods and
375
populations that were excluded. COVID-19 sequalae can affect a substantial portion of patients,
376
with long-term consequences for their health, continuity of care and ability to work [1, 2].
377
Persistent symptoms and work impairment were reported ~4 months after infection among non-
378
hospitalized US patients enrolled in clinical trials [24]. Continued follow-up studies covering
379
longer time periods may inform whether the protection provided by COVID-19 vaccination
380
extends beyond the acute phase. Only generic PROMs were used in this study; COVID-19
381
disease-specific instruments are under development [30, 31], warranting research on their
382
implementation. The PROMs omitted questions on vaccine adverse events. The mean 6-month
383
interval between vaccination and breakthrough infection and a medical review ruled out cases of
384
residual symptoms from vaccination. Research on the impact of vaccine adverse events on
385
HRQoL is warranted.
386
Lastly, the study adopted an observational design, which is limited in establishing causal
387
relationships. Future studies using different data collection methods could corroborate the study
388
findings, including those with more rigorous study design.
389
Consistent with literature, our study found that COVID-19 is detrimental to mildly symptomatic
390
COVID-19 patients. The findings provide a meaningful contribution suggesting that the ability
18
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
391
of BNT162b2 to reduce adverse outcomes from COVID-19 disease can translate to lessening the
392
broad impact of COVID-19 and improvements in quality of life, work productivity and activity.
393
394
CONCLUSION
395
This study found that mild COVID-19 infection at a time of Omicron predominance adversely
396
impacted the HRQoL, daily activity and work productivity of patients. This detrimental effect
397
improved over time, although it persisted for at least one month post infection. Compared with
398
unvaccinated, those vaccinated with BNT162b2 were less impacted and recovered faster. These
399
findings advance research on COVID-19 associated humanistic outcomes and the potential effect
400
of BNT162b2 in lessening the loss of HRQoL, daily activity and work productivity due to
401
COVID-19. The results can inform the estimation of quality-adjusted life years and indirect cost
402
savings in health economic studies.
403
404
List of Abbreviations
405
CDC: Centers for Disease Control and Prevention; CI: Confidence interval; EQ-5D-5L:
406
EuroQoL Group 5 dimension and 5 level questionnaire; ES: Effect size; GH: General Health;
407
HRQoL: Health Related Quality of Life; MMRM: Mixed models for repeated measures; SD:
408
standard deviation; SE: standard error; STROBE: the Strengthening the Reporting of
409
Observational Studies in Epidemiology; UI: utility index; VAS: Visual analogue scale; WPAI:
410
Work Productivity and Impairment.
411
412
DECLARATIONS:
413
Ethics Approval and consent to participate
19
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
414
This study was approved by the Sterling IRB, Protocol #C4591034. Participation in the study
415
was voluntary and anonymous. Consent was obtained electronically via the CVS Health E-
416
Consent platform. Participants were informed of their right to refuse or withdraw from the study
417
at any time. Participants were compensated for their time.
418
Consent for publication
419
All authors have given their approval for this manuscript version to be published.
420
Availability of data and material
421
Aggregated data that support the findings of this study are available upon reasonable request
422
from the corresponding author MDF, subject to review. These data are not publicly available due
423
to them containing information that could compromise research participant privacy/consent.
424
Competing interests
425
MDF, MMM, JMZ, LP, MBA and JCC are employees of Pfizer and may hold stock or stock
426
options of Pfizer. XS and HC are employees of CVS Health and may hold stock of CVS health.
427
YPT was employee of CVS Health when current study was conducted.
428
Funding
429
This study was sponsored by Pfizer Inc.
430
Authors’ contributions
431
All named authors meet the International Committee of Medical Journal Editors (ICMJE) criteria
432
for authorship for this article. All authors contributed to study conception and design, data
433
acquisition, analysis, and interpretation, drafting and revising of the manuscript.
434
Acknowledgements
435
The authors acknowledge Alejandro Cane, Deepa Malhotra and Nancy Gifford (Pfizer
436
employees), Joseph Ferenchick, Shiyu Lin and Shawn Edmonds (CVS Health employees) for
20
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
437
specific contributions to this research project. Editorial support was provided by Laura Anatale-
438
Tardiff and Leena Samuel at CVS Health and was funded by Pfizer.
439
440
REFERENCES
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Poudel AN, Zhu S, Cooper N, Roderick P, Alwan N, Tarrant C, Ziauddeen N, Yao GL
(2021) Impact of Covid-19 on health-related quality of life of patients: A structured
review. PLoS One. 16(10):e0259164.
Nandasena H, Pathirathna M, Atapattu A, Prasanga P (2022) Quality of life of COVID 19
patients after discharge: Systematic review. PloS one. 17(2):e0263941.
Figueiredo EAB, Silva WT, Tsopanoglou SP, Vitorino DFdM, Oliveira LFLd, Silva
KLS, Luz HDH, Ávila MR, Oliveira LFFd, Lacerda ACR (2022) The health-related
quality of life in patients with post-COVID-19 after hospitalization: a systematic review.
Revista da Sociedade Brasileira de Medicina Tropical. 55.
Amdal CD, Pe M, Falk RS, Piccinin C, Bottomley A, Arraras JI, Darlington AS, Hofsø
K, Holzner B, Jørgensen NMH (2021) Health-related quality of life issues, including
symptoms, in patients with active COVID-19 or post COVID-19; a systematic literature
review. Quality of Life Research. 30(12):3367-81.
International Vaccine Access Center, Johns Hopkins Bloomberg School of Public Health,
World Health Organization. Results of COVID-19 Vaccine Effectiveness & Impact
Studies: An Ongoing Systematic Review. https://view-hub.org/sites/default/files/202204/COVID19_VE_and_Impact_Lit_Review_Methods.pdf Accessed August 15, 2022.
Arab-Zozani M, Hashemi F, Safari H, Yousefi M, Ameri H (2020) Health-related quality
of life and its associated factors in COVID-19 patients. Osong public health and research
perspectives. 11(5):296.
Petersen EL, Goßling A, Adam G, Aepfelbacher M, Behrendt C-A, Cavus E, Cheng B,
Fischer N, Gallinat J, Kühn S (2022) Multi-organ assessment in mainly non-hospitalized
individuals after SARS-CoV-2 infection: the Hamburg City Health Study COVID
programme. European Heart Journal. 43(11):1124-37.
McKay SC, Lembach H, Hann A, Okoth K, Anderton J, Nirantharakumar K, Magill L,
Torlinska B, Armstrong M, Mascaro J (2021) Health‐related quality of life, uncertainty
and coping strategies in solid organ transplant recipients during shielding for the
COVID‐19 pandemic. Transplant International. 34(11):2122-37.
Talman S, Boonman-de Winter L, De Mol M, Hoefman E, Van Etten R, De Backer I
(2021) Pulmonary function and health-related quality of life after COVID-19 pneumonia.
Respiratory Medicine. 176:106272.
Centers for Disease Control and Prevention. Updates - Symptoms of COVID-19.
https://www.cdc.gov/coronavirus/2019-ncov/symptoms-testing/symptoms.html Accessed
August 1 2021.
EuroQol Research Foundation. (2019) EQ-5D-5L User Guide, Version 3.0.
https://euroqol.org/publications/user-guides Accessed August 5, 2021.
21
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
Cha AS, Law EH, Shaw JW, Pickard AS (2019) A comparison of self-rated health using
EQ-5D VAS in the United States in 2002 and 2017. Quality of Life Research.
28(11):3065-9.
Pickard AS, Law EH, Jiang R, Pullenayegum E, Shaw JW, Xie F, Oppe M, Boye KS,
Chapman RH, Gong CL (2019) United States valuation of EQ-5D-5L health states using
an international protocol. Value in Health. 22(8):931-41.
Reilly MC, Zbrozek AS, Dukes EM (1993) The validity and reproducibility of a work
productivity and activity impairment instrument. Pharmacoeconomics. 4(5):353-65.
Reilly Associates. (2002) WPAI Scoring.
http://www.reillyassociates.net/WPAI_Scoring.html Accessed August 5, 2021.
Centers for Disease Control and Prevention. Long COVID or Post-COVID Conditions.
https://www.cdc.gov/coronavirus/2019-ncov/long-term-effects/index.html Accessed
August 15, 2022.
Rosner B (2015) Fundamentals of biostatistics. Eighth ed. Cengage learning, Boston, MA
Freeman G, Halton JH (1951) Note on an exact treatment of contingency, goodness of fit
and other problems of significance. Biometrika. 38(1/2):141-9.
Fitzmaurice GM, Laird NM, Ware JH (2012) Applied longitudinal analysis. John Wiley
& Sons, Hoboken, NJ
Cohen J (1988) Statistical power analysis for the behavioral sciences. 2nd ed. Lawrence
Erlbaum Assoc, Hillsdale, NJ.
McLeod LD, Cappelleri JC, Hays RD (2016) Best (but oft-forgotten) practices:
expressing and interpreting associations and effect sizes in clinical outcome assessments.
The American journal of clinical nutrition. 103(3):685-93. Erratum:
https://doi.org/10.3945/ajcn.116.148593.
Mouelhi Y, Jouve E, Castelli C, Gentile S (2020) How is the minimal clinically important
difference established in health-related quality of life instruments? Review of anchors
and methods. Health and Quality of Life Outcomes. 18(1):1-17.
STROBE Statement—Checklist of items that should be included in reports of cohort
studies. https://www.strobe-statement.org/download/strobe-checklist-cohort-studies-pdf
Accessed August 5, 2022.
Jacobson KB, Rao M, Bonilla H, Subramanian A, Hack I, Madrigal M, Singh U,
Jagannathan P, Grant P (2021) Patients with uncomplicated COVID-19 have long-term
persistent symptoms and functional impairment similar to patients with severe COVID19: a cautionary tale during a global pandemic. Clinical Infectious Diseases.
10.1093/cid/ciab103.
Hernandez-Romieu AC, Leung S, Mbanya A, Jackson BR, Cope JR, Bushman D, Dixon
M, Brown J, McLeod T, Saydah S (2021) Health care utilization and clinical
characteristics of nonhospitalized adults in an integrated health care system 28–180 days
after COVID-19 diagnosis—Georgia, May 2020–March 2021. Morbidity and Mortality
Weekly Report. 70(17):644.
Jiang R, Janssen M, Pickard AS (2021) US population norms for the EQ-5D-5L and
comparison of norms from face-to-face and online samples. Quality of Life Research.
30(3):803-16.
Lawson A, Tan AC, Naylor J, Harris IA (2020) Is retrospective assessment of healthrelated quality of life valid? BMC Musculoskeletal Disorders. 21(1):1-10.
22
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
28.
29.
30.
31.
Rajan SS, Wang M, Singh N, Jacob AP, Parker SA, Czap AL, Bowry R, Grotta JC,
Yamal JM (2021) Retrospectively Collected EQ-5D-5L Data as Valid Proxies for
Imputing Missing Information in Longitudinal Studies. Value Health. 24(12):1720-7.
10.1016/j.jval.2021.07.007.
Tundia N, Hass S, Fuldeore M, Wang LL, Cavanaugh T, Boone J, Heaton P (2015)
Validation and US population norms of health-related productivity questionnaire. Value
in Health. 18(3):A24.
Romano C, Fehnel S, Stoddard J, Sadoff J, Lewis S, McNulty P, Chan EK, Evans E,
Jamieson C, Slagle AF (2022) Development of a novel patient-reported outcome measure
to assess signs and symptoms of COVID-19. Journal of Patient-Reported Outcomes.
6(1):1-12.
Hughes SE, Haroon S, Subramanian A, McMullan C, Aiyegbusi OL, Turner GM,
Jackson L, Davies EH, Frost C, McNamara G (2022) Development and validation of the
symptom burden questionnaire for long covid (SBQ-LC): Rasch analysis. BMJ. 377.
23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
537
Figure 1. Study design a
538
539
540
a
QoL refers to the EQ-5D-5L survey
24
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
541
Figure 2. Study flow diagram
542
543
25
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
544
Table 1 Patient Characteristics on Index Day
Total, n
Age, years
Mean, SD
18-29
30-49
50-64
≥65
Gender
Female
Male
Race / Ethnicity
White or Caucasian (not
Hispanic or Latino)
Black or African American
Hispanic
Asian
Patient Refused
Other
CMS Geographic Region (n, %)
Region 1: ME, NH, VT, MA, CT,
RI
Region 2: NY, NJ, PR, VI
Region 3: PA, DE, MD, DC, WV,
VA
Region 4: KY, TN, NC, SC, GA,
MS, AL, FL
Region 5: MN, WI, IL, MI, IN,
OH
Region 6: NM, OK, AR, TX, LA
Region 7: NE, IA, KS, MO
Region 8 : MT, ND, SD, WY, UT,
CO
Region 9: CA, NV, AZ, GU
Region 10: AK, WA, OR, ID
U.S. Geographic Region
Northeast
South
Midwest
West
Previously Tested Positive
All
BNT162b2
Unvaccinated
430
233
197
42.4 (14.3)
43.7 (15.3)
40.9 (12.9)
87 (20.2%)
213 (49.5%)
94 (21.9%)
49 (21.0%)
100 (42.9%)
60 (25.8%)
38 (19.3%)
113 (57.4%)
34 (17.3%)
36 (8.4%)
24 (10.3%)
P-value a
0.049
0.011
12 (6.1%)
327 (76.0%)
103 (24.0%)
177 (76.0%)
56 (24.0%)
150 (76.1%)
47 (23.9%)
295 (68.6%)
166 (71.2%)
129 (65.5%)
20 (4.7%)
61 (14.2%)
22 (5.1%)
13 (3.0%)
19 (4.4%)
7 (3.0%)
35 (15.0%)
15 (6.4%)
5 (2.2%)
5 (2.2%)
13 (6.6%)
26 (13.2%)
7 (3.6%)
8 (4.1%)
14 (7.1%)
19 (4.4%)
10 (4.3%)
9 (4.6%)
11 (2.6%)
7 (3.0%)
4 (2.0%)
37 (8.6%)
22 (9.4%)
15 (7.6%)
156 (36.3%)
81 (34.8%)
75 (38.1%)
58 (13.5%)
31 (13.3%)
27 (13.7%)
82 (19.1%)
19 (4.4%)
56 (24.0%)
11 (4.7%)
26 (13.2%)
8 (4.1%)
1 (0.2%)
1 (0.4%)
0 (0.0%)
46 (10.7%)
1 (0.2%)
13 (5.6%)
1 (0.4%)
33 (16.8%)
0 (0.0%)
53 (12.2%)
254 (58.7%)
77 (17.8%)
49 (11.3%)
167 (38.8%)
29 (12.3%)
150 (63.6%)
42 (17.8%)
15 (6.4%)
89 (38.2%)
24 (12.2%)
104 (52.8%)
35 (17.8%)
34 (17.3%)
78 (39.6%)
0.966
0.026
0.009
0.005
0.589
26
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
All
545
BNT162b2
Unvaccinated
Work in healthcare
47 (10.9%)
29 (12.4%)
18 (9.1%)
Work in high-risk setting
44 (10.2%)
30 (12.9%)
14 (7.1%)
Live in high-risk setting
22 (5.1%)
12 (5.2%)
10 (5.1%)
Social vulnerability index
Mean (SD)
0.44 (0.22)
0.40 (0.22)
0.49 (0.21)
Median (Q1, Q3)
0.41 (0.27, 0.59) 0.37 (0.23, 0.55)
0.47 (0.31, 0.63)
Self-Reported Comorbidity
Asthma or Chronic Lung
34 (7.9%)
21 (9.0%)
13 (6.6%)
Disease
Cirrhosis of the liver
1 (0.2%)
1 (0.4%)
0 (0.0%)
Immunocompromised
19 (4.4%)
12 (5.2%)
7 (3.6%)
Conditions or Weakened
Immune System c
Diabetes
20 (4.7%)
13 (5.6%)
7 (3.6%)
Heart Conditions or
52 (12.1%)
30 (12.9%)
22 (11.2%)
Hypertension
Overweight or obesity
19 (4.4%)
12 (5.2%)
7 (3.6%)
At least 1 comorbidity
104 (24.2%)
61 (26.2%)
43 (21.8%)
Number of comorbidities,
0.34 (0.68)
0.38 (0.75)
0.28 (0.58)
Mean (SD)
Index day b acute COVID-19 symptoms
Systemic symptoms
Fever
93 (47.2%)
164 (38.1%)
71 (30.5%)
Chills
113 (57.4%)
213 (49.5%)
100 (42.9%)
Muscle or Body Aches
117 (59.4%)
232 (54.0%)
115 (49.4%)
Headache
140 (71.1%)
293 (68.1%)
153 (65.7%)
Fatigue
125 (63.5%)
266 (61.9%)
141 (60.5%)
Respiratory symptoms
29 (14.7%)
Shortness of Breath or
54 (12.6%)
25 (10.7%)
Difficulty Breathing
Cough
141 (71.6%)
309 (71.9%)
168 (72.1%)
Sore Throat
104 (52.8%)
238 (55.3%)
134 (57.5%)
22 (11.2%)
New/Recent Loss of Taste or
45 (10.5%)
23 (9.9%)
Smell
Congestion or Runny Nose
134 (68.0%)
322 (74.9%)
188 (80.7%)
GI symptoms
Nausea or Vomiting
31 (15.7%)
55 (12.8%)
24 (10.3%)
Diarrhea
51 (25.9%)
88 (20.5%)
37 (15.9%)
Number of acute COVID-19
5.3 (2.6)
5.1 (2.4)
5.6 (2.7)
symptoms, Mean (SD)
a
P value refers to the comparison between BNT162B2 and Unvaccinated.
P-value a
0.309
0.158
0.553
<0.001
<0.001
0.355
1.0000
0.422
0.320
0.588
0.422
0.294
0.138
<0.001
0.003
0.038
0.231
0.532
0.213
0.903
0.327
0.662
0.003
0.093
0.010
0.034
27
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
546
547
548
549
550
b
COVID-19 test nasal swab day
Immunocompromised conditions includes compromised immune system (such as from
immuno-compromising drugs, solid organ or blood stem cell transplant, HIV, or other
conditions), conditions that result in a weakened immune system, including cancer treatment,
and kidney failure or end stage renal disease
c
28
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
551
Table 2 Post-COVID-19 Symptoms at Week 4
Symptom
552
553
554
All
BNT162b2
P-value a
Unvaccinated
General symptoms
Tiredness or fatigue
71 (40.6%)
65 (41.9%)
136 (41.2%)
Symptoms that get worse after physical or
50 (15.2%)
18 (10.3%)
32 (20.6%)
mental activities
Fever
1 (0.3%)
0 (0.0%)
1 (0.7%)
General pain/discomfort
50 (15.2%)
20 (11.4%)
30 (19.4%)
Respiratory and cardiac
Difficulty breathing or shortness of breath 58 (17.6%)
26 (14.9%)
32 (20.6%)
Cough
86 (26.1%)
40 (22.9%)
46 (29.7%)
Chest or stomach pain
32 (9.7%)
14 (8.0%)
18 (11.6%)
Fast-beating or pounding heart (also
38 (11.5%)
17 (9.7%)
21 (13.5%)
known as heart palpitations)
Neurologic
Change in smell or taste
51 (15.5%)
19 (10.9%)
32 (20.6%)
Headache
67 (20.3%)
28 (16.0%)
39 (25.2%)
Dizziness on standing (lightheadedness)
45 (13.6%)
20 (11.4%)
25 (16.1%)
Difficulty thinking or concentrating
86 (26.1%)
43 (24.6%)
43 (27.7%)
(sometimes referred to as “brain fog”)
Pins-and-needles feeling
24 (7.3%)
10 (5.7%)
14 (9.0%)
Sleep problems
81 (24.5%)
35 (20.0%)
46 (29.7%)
Mood changes
36 (10.9%)
13 (7.4%)
23 (14.8%)
Memory loss
38 (11.5%)
11 (6.3%)
27 (17.4%)
Other
Diarrhea
23 (7.0%)
6 (3.4%)
17 (11.0%)
Joint or muscle pain
67 (20.3%)
29 (16.6%)
38 (24.5%)
Rash
11 (3.3%)
3 (1.7%)
8 (5.2%)
Changes in period cycles
28 (11.5%)
12 (9.4%)
16 (13.8%)
Number of post-COVID-19 symptoms, Mean
3.1 (3.6)
2.5 (3.0)
3.7 (4.1)
(SD)
0
100 (30.3%)
54 (30.9%)
46 (29.7%)
1-2
100 (30.3%)
59 (33.7%)
41 (26.5%)
3-5
61 (18.5%)
40 (22.9%)
21 (13.5%)
6-8
32 (9.7%)
9 (5.1%)
23 (14.8%)
≥9
37 (11.2%)
13 (7.4%)
24 (15.5%)
a
P-values of t-test for number of symptoms, chi-square or Fisher’s exact tests when any one cell
has an expected frequency less than 5 for individual symptoms and number of symptom
category comparing the BNT162b2 cohort and the unvaccinated cohort.
29
0.802
0.009
0.470
0.045
0.168
0.159
0.268
0.276
0.014
0.039
0.214
0.513
0.247
0.042
0.031
0.002
0.007
0.073
0.082
0.280
0.002
0.001
555
556
Table 3 Summary of EQ-5D-5L and WPAI-GH Scoresa and Their Changes from Baseline by Assessment Time
Score
n Mean (SD)
EQ-5D-5L
Visual analogue
scale (VAS)
Baseline g
233 86.9 (10.7)
Day 3
233 73.9 (15.6)
Week 4
171 82.9 (13.8)
Utility Index
(US weight)
0.924
Baselinee
233
(0.117)
0.820
Day 3
233
(0.193)
0.882
Week 4
176
(0.145)
WPAI-GH
Absenteeism
Baselinee
153 2.8 (11.8)
Week 1
Week 4
Presenteeism
Baselinee
Week 1
Week 4
Work
productivity
loss
Baselinee
Week 1
BNT162b2 Cohort
Change from Baseline b
n Mean (SD) P-value c ESwd
233 -13.0 (12.5) <0.001
171 -3.7 (10.7) <0.001
-1.04
-0.35
Unvaccinated Cohort
Difference in Change from
Baseline Between Cohorts
Score
Change from Baseline
n Mean (SD) n Mean (SD) P-value c ESwd Mean (SD) P-value e ESbf
197 87.8 (11.0)
194 71.8 (19.6) 194 -16.1 (17.1) <0.001 -0.94 3.1 (14.8)
151 80.9 (15.7) 151 -7.2 (13.7) <0.001 -0.53 3.5 (12.2)
197
233
176
-0.105
(0.159)
-0.039
(0.120)
<0.001
-0.66
197
<0.001
-0.33
155
0.918
(0.117)
0.762
(0.252)
0.838
(0.197)
0.414 h
0.033 0.21
0.011 0.28
0.547 h
197
155
-0.155
(0.228)
-0.074
(0.156)
<0.001 -0.68
<0.001 -0.47
129 11.7 (26.4)
0.050
(0.194)
0.035
(0.138)
-13.0
(39.0)
-0.27 9.0 (24.1)
0.007
0.26
0.023
0.25
<0.001 h
153 44.9 (38.3) 147 42.5 (38.8)
<0.001
1.09
129 66.7 (37.0) 127 55.5 (39.2) <0.001 1.42
0.006
-0.33
128 4.6 (17.2) 122
0.460
0.07
106 3.3 (12.2)
0.006
0.37
1.13
0.06
124 9.4 (20.4)
82 47.8 (34.2)
105 18.7 (23.4)
0.958 h
81 38.2 (35.9) <0.001 1.06 -5.1 (32.1) 0.268 -0.16
96 10.7 (24.1) <0.001 0.45 -9.3 (24.3) 0.006 -0.38
1.35
123 14.2 (24.7)
82 66.9 (32.4)
0.175 h
80 53.2 (35.3) <0.001 1.51 -6.4 (34.9) 0.203
1.4 (20.1)
153 9.5 (18.5)
123 41.7 (27.4) 119 33.0 (29.1)
125 11.7 (18.8) 119 1.4 (24.5)
152 10.5 (20.0)
123 56.6 (31.5) 118 46.7 (34.5)
<0.001
0.527
<0.001
102 -7.7 (28.1)
0.007
30
-0.18
557
558
559
560
561
562
563
564
565
566
Week 4
Activity
impairment
Baselinee
Week 1
Week 4
a
125 12.9 (20.7) 118
1.4 (27.5)
203 12.5 (22.2)
202 48.6 (29.8) 202 36.1 (31.9)
175 15.5 (21.7) 175 2.3 (26.3)
0.589
<0.001
0.242
0.05
105 19.8 (24.6)
96 7.7 (30.6)
0.016
0.25 -6.3 (29.0)
0.113
1.13
0.09
176 17.2 (27.2)
0.065 h
176 54.2 (32.7) 176 37.0 (37.5) <0.001 0.99 -0.9 (34.6) 0.801 -0.03
155 25.9 (28.5) 155 8.8 (33.8)
0.002 0.26 -6.4 (30.1) 0.053 -0.21
Score ranges: EQ-5D-5L VAS 0 to 100, EQ-5D-5L UI (the United States weights) -0.573 to 1; WPAI-GH (absenteeism,
presenteeism, work productivity loss, and activity impairment) 0 to 100 percent.
b
Baseline refers to pre-COVID-19 symptom onset.
c
P-value of t-test comparing mean score changes from baseline and 0 within BNT162b2 or Unvaccinated cohorts.
d
ESw refers to effect size for score changes from baseline within BNT162b2 or Unvaccinated cohorts.
e
P-value of t-test comparing mean score changes from baseline between BNT162b2 and Unvaccinated cohorts.
f
ESb refers to effect size for score changes from baseline between BNT162b2 and Unvaccinated cohorts.
g
Prior to symptom onset (pre-COVID).
h
P-values of t-tests comparing pre-COVID-19 baseline mean scores between BNT162b2 and Unvaccinated cohorts.
31
-0.22
567
Table 4 Least-Square Mean Estimate and 95% Confidence Interval for EQ-5D-5L and WPAI-GH Scores a
Variable
BNT162b2 Cohort
Unvaccinated Cohort
Score
Change from Baseline
Score
Change from Baseline
c
LSE (95% CI) LSE (95% CI)
PESw LSE (95% CI) LSE (95% CI)
PESwc
b
b
value
value
EQ-5D-5L
Visual analogue scale (VAS)
-11.1 (-14.1, <0.001 -0.89
8.2)
Day 3
76.2 (73.3, 79.2)
Week 4
85.0 (82.1, 87.9) -2.3 (-5.2, 0.6)
Utility Index (US weight)
0.842 (0.805,
Day 3
0.879)
0.903 (0.868,
Week 4
0.938)
WPAI-GH
Absenteeism
Week 1
45.6 (38.0, 53.1)
Week 4
5.3 (0.0, 10.7)
Presenteeism
Week 1
38.4 (30.3, 46.5)
Week 4
6.8 (0, 13.9) f
Work productivity loss
Week 1
53.8 (45.1, 62.4)
Week 4
8.6 (0.9, 16.2)
Activity impairment
Week 1
43.9 (37.7, 50.1)
0.119 -0.22
72.6 (69.7,
75.4)
81.6 (78.8,
84.4)
-0.077 (-0.115,
0.773 (0.736,
<0.001 -0.49
-0.040)
0.809)
-0.016 (-0.051,
0.859 (0.826,
0.369 -0.13
0.019)
0.892)
39.1 (31.6,
65.0 (57.5,
<0.001 1.01
46.7)
72.5)
-1.1 (-6.5, 4.2) 0.676 -0.06 2.2 (0.0, 7.1) f
50.2 (44.2,
56.2)
P-value d
ESbe
<0.001 -0.86
3.6 (0.8, 6.5)
0.013
0.25
-5.7 (-8.5, -2.9)
<0.001 -0.42
3.4 (0.6, 6.2)
0.016
0.28
-0.147 (-0.183, <0.001 -0.64
0.110)
-0.060 (-0.093, <0.001 -0.38
0.027)
0.069 (0.032,
0.107)
0.044 (0.013,
0.075)
<0.001
0.36
0.005
0.32
58.6 (51.1, 66.0) <0.001 1.49 -19.4 (-28.3, -10.6) <0.001
-0.50
-4.3 (-9.2, 0.6)
0.086 -0.15
42.4 (33.8,
65.0 (56.2,
<0.001 1.23
53.6 (44.8, 62.4) <0.001 1.52
51.0)
73.7)
-2.8 (-10.4, 4.9) 0.476 -0.10 17.0 (9.9, 24.0) 5.7 (-1.4, 12.7)
0.116 0.18
<0.001 0.91
LSE (95% CI)
-14.8 (-17.6, 11.9)
29.0 (20.8,
46.8 (38.6,
<0.001 1.00
37.4 (29.1, 45.6) <0.001 1.04
37.1)
55.1)
-2.7 (-9.7, 4.4) 0.458 -0.11 16.0 (9.4, 22.5) 6.5 (0.0, 13.0)
0.050 0.27
29.0 (22.8,
35.3)
Between Cohort Difference
35.4 (29.3, 41.4) <0.001 0.94
3.2 (-1.1, 7.4)
0.145
0.13
-8.4 (-16.7, -0.1)
0.047
-0.26
-9.2 (-14.7, -3.6)
0.001
-0.38
-11.2 (-19.9, -2.5)
0.012
-0.32
-8.4 (-14.5, -2.3)
0.007
-0.29
-6.3 (-12.4, -0.2)
0.044
-0.18
32
Week 4
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
11.0 (5.1, 16.9)
-3.9 (-9.7, 2.0)
0.194 -0.15
21.3 (15.7,
26.9)
6.4 (0.8, 12.0)
0.025
0.19
-10.3 (-15.6, -5.0)
<0.001
-0.34
Abbreviations: LSE = Least-Square Mean Estimate; CI = Confidence Interval.
a
Multivariate models include variables for time, vaccination status and interaction of time by vaccination status, as well as covariates
of participant pre-COVID-19 symptom onset score, sociodemographic characteristics (age, sex, regions, social vulnerability,
race/ethnicity, high risk occupations), previously tested positive for COVID-19, severity of acute illness (number of symptoms
reported on index date), and immunocompromised status. Parameter estimates are presented in Supplemental Table 3.
b
P-value refers to the comparison of lease-square mean estimates score changes from baseline and 0 within BNT162b2 or
Unvaccinated cohorts
c
ESw, within-cohort effect size, was calculated as the least square estimate of mean change from divided by the observed standard
deviation of change scores from baseline to follow-up.
d
ESb, between-cohort effect size, was calculated as the difference in least square estimates of mean changes from baseline between
cohorts, divided by the observed pooled standard deviation of change scores
e
P-value refers to the difference in lease-square mean estimates between BNT162b2 and Unvaccinated cohorts.
f
Lower limit of 95% CI was truncated from -2.7 to 0 for absenteeism and -0.3 to 0 for presenteeism at Week 4 because the valid range
is 0 to 100.
33
599
Figure 3 Mean Responses of EQ-5D-5L Dimensions by Timepoint
600
601
602
603
604
605
606
607
Mean dimension scores range from 1 for no problem to 5 for extreme / unable. The blue and red solid lines indicate that vaccinated
and unvaccinated were similar at the pre-COVID baseline. At Day 3 and Week 4 post-index date, vaccinated cohort was less impacted
(lower scores) than unvaccinated by COVID on anxiety/depression, pain/discomfort, and usual activities (dotted lines for Day 3,
dashed lines for Week 4).
34
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
Supplemental Table 1 Patient Characteristics by Enrollment Status
Patient characteristics
Vaccinated
Yes
No
Missing
Age, years
Mean, SD
Median, Q1, Q3
Age Category
18-29
30-49
50-64
≥65
Gender, n (%)
Male
Female
Race / Ethnicity, n (%)
White or Caucasian (non-Hispanic or Latino)
Black or African American
Hispanic or Latino
Asian
Patient Refused
Other
US Geographic Region
Northeast
South
Midwest
West
Number of acute COVID-19 symptoms, Mean
(SD)
Number of comorbidities, Mean (SD)
≥1 comorbidity
Enrolled
N=676
Not-Enrolled
N=39,213
P value
<0.001
465 (68.8%)
211 (31.2%)
22,125 (56.4%)
17,087 (43.6%)
43.2 (14.7)
41 (31--55)
42.1 (15.5)
40 (29--54)
0.076
0.010
134 (19.8%)
317 (46.9%)
156 (23.1%)
69 (10.2%)
10,095 (25.7%)
16,706 (42.6%)
8,561 (21.8%)
3,850 (9.8%)
0.005
<0.001
181 (26.8%)
495 (73.2%)
17,843 (45.5%)
21,370 (54.5%)
<0.001
486 (71.9%)
32 (4.7%)
85 (12.6%)
35 (5.2%)
16 (2.4%)
22 (3.3%)
22,647 (57.8%)
4,162 (10.6%)
6,878 (17.5%)
2,551 (6.5%)
1,535 (3.9%)
1,439 (3.7%)
0.258
92 (13.6%)
402 (59.5%)
118 (17.5%)
64 (9.5%)
5.2 (2.5)
4,387 (11.2%)
24,305 (62.0%)
6,414 (16.4%)
4,105 (10.5%)
5.1 (2.5)
0.732
0.3 (0.7)
164 (24.3%)
0.3 (0.6)
7,677 (19.6%)
0.021
0.002
35
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
Supplemental Table 2 Summary of EQ-5D-5L Dimensions
Dimension
Pre-COVID-19 Baseline
Mobility
No problems
Slight problems
Moderate problems
Severe problems
Unable
Missing
Self-Care
No problems
Slight problems
Moderate problems
Severe problems
Unable
Missing
Usual Activities
No problems
Slight problems
Moderate problems
Severe problems
Unable
Missing
Pain / Discomfort
No
Slight
Moderate
Severe
Extreme
Missing
Anxiety / Depression
No
Slightly
Moderately
Severely
Extremely
Missing
Days 3
Mobility
All
BNT162b2
Unvaccinated
P-value a
0.841
396 (92.1%)
30 (7.0%)
4 (0.9%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
213 (91.4%)
18 (7.7%)
2 (0.9%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
183 (92.9%)
12 (6.1%)
2 (1.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
419 (97.4%)
11 (2.6%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
226 (97.0%)
7 (3.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
193 (98.0%)
4 (2.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
394 (91.6%)
23 (5.4%)
12 (2.8%)
1 (0.2%)
0 (0.0%)
0 (0.0%)
214 (91.8%)
11 (4.7%)
7 (3.0%)
1 (0.4%)
0 (0.0%)
0 (0.0%)
180 (91.4%)
12 (6.1%)
5 (2.5%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
307 (71.4%)
93 (21.6%)
27 (6.3%)
3 (0.7%)
0 (0.0%)
0 (0.0%)
172 (73.8%)
46 (19.7%)
13 (5.6%)
2 (0.9%)
0 (0.0%)
0 (0.0%)
135 (68.5%)
47 (23.9%)
14 (7.1%)
1 (0.5%)
0 (0.0%)
0 (0.0%)
230 (53.5%)
140 (32.6%)
49 (11.4%)
10 (2.3%)
1 (0.2%)
0 (0.0%)
124 (53.2%)
80 (34.3%)
26 (11.2%)
3 (1.3%)
0 (0.0%)
0 (0.0%)
106 (53.8%)
60 (30.5%)
23 (11.7%)
7 (3.6%)
1 (0.5%)
0 (0.0%)
0.560
0.904
0.614
0.395
0.131
36
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
No problems
Slight problems
Moderate problems
Severe problems
Unable
Missing
Self-Care
No problems
Slight problems
Moderate problems
Severe problems
Unable
Missing
Usual Activities
No problems
Slight problems
Moderate problems
Severe problems
Unable
Missing
Pain / Discomfort
No
Slight
Moderate
Severe
Extreme
Missing
Anxiety / Depression
No
Slightly
Moderately
Severely
Extremely
Missing
Week 4
Mobility
No problems
Slight problems
Moderate problems
Severe problems
Unable
343 (79.8%)
68 (15.8%)
17 (4.0%)
2 (0.5%)
0 (0.0%)
0 (0.0%)
193 (82.8%)
31 (13.3%)
7 (3.0%)
2 (0.9%)
0 (0.0%)
0 (0.0%)
150 (76.1%)
37 (18.8%)
10 (5.1%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
370 (86.0%)
47 (10.9%)
11 (2.6%)
2 (0.5%)
0 (0.0%)
0 (0.0%)
205 (88.0%)
21 (9.0%)
6 (2.6%)
1 (0.4%)
0 (0.0%)
0 (0.0%)
165 (83.8%)
26 (13.2%)
5 (2.5%)
1 (0.5%)
0 (0.0%)
0 (0.0%)
207 (48.1%)
132 (30.7%)
64 (14.9%)
21 (4.9%)
6 (1.4%)
0 (0.0%)
117 (50.2%)
75 (32.2%)
30 (12.9%)
9 (3.9%)
2 (0.9%)
0 (0.0%)
90 (45.7%)
57 (28.9%)
34 (17.3%)
12 (6.1%)
4 (2.0%)
0 (0.0%)
145 (33.7%)
183 (42.6%)
82 (19.1%)
17 (4.0%)
3 (0.7%)
0 (0.0%)
85 (36.5%)
99 (42.5%)
43 (18.5%)
4 (1.7%)
2 (0.9%)
0 (0.0%)
60 (30.5%)
84 (42.6%)
39 (19.8%)
13 (6.6%)
1 (0.5%)
0 (0.0%)
190 (44.2%)
135 (31.4%)
75 (17.4%)
26 (6.1%)
4 (0.9%)
0 (0.0%)
108 (46.4%)
79 (33.9%)
37 (15.9%)
8 (3.4%)
1 (0.4%)
0 (0.0%)
82 (41.6%)
56 (28.4%)
38 (19.3%)
18 (9.1%)
3 (1.5%)
0 (0.0%)
0.571
0.361
0.080
0.047
0.278
270 (81.3%)
49 (14.8%)
12 (3.6%)
1 (0.3%)
0 (0.0%)
149 (84.7%)
22 (12.5%)
5 (2.8%)
0 (0.0%)
0 (0.0%)
121 (77.6%)
27 (17.3%)
7 (4.5%)
1 (0.6%)
0 (0.0%)
37
medRxiv preprint doi: https://doi.org/10.1101/2022.08.31.22279264; this version posted September 6, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .
Missing
Self-Care
No problems
Slight problems
Moderate problems
Severe problems
Unable
Missing
Usual Activities
No problems
Slight problems
Moderate problems
Severe problems
Unable
Missing
Pain / Discomfort
No
Slight
Moderate
Severe
Extreme
Missing
Anxiety / Depression
No
Slightly
Moderately
Severely
Extremely
Missing
1 (0.0%)
1 (0.0%)
0 (0.0%)
308 (93.1%)
17 (5.1%)
5 (1.5%)
1 (0.3%)
0 (0.0%)
2 (0.0%)
167 (94.9%)
5 (2.8%)
4 (2.3%)
0 (0.0%)
0 (0.0%)
1 (0.0%)
141 (91.0%)
12 (7.7%)
1 (0.7%)
1 (0.7%)
0 (0.0%)
1 (0.0%)
232 (70.1%)
75 (22.7%)
20 (6.0%)
4 (1.2%)
0 (0.0%)
2 (0.0%)
132 (75.0%)
35 (19.9%)
8 (4.6%)
1 (0.6%)
0 (0.0%)
1 (0.0%)
100 (64.5%)
40 (25.8%)
12 (7.7%)
3 (1.9%)
0 (0.0%)
1 (0.0%)
170 (51.4%)
118 (35.6%)
37 (11.2%)
5 (1.5%)
1 (0.3%)
2 (0.0%)
95 (54.0%)
66 (37.5%)
13 (7.4%)
2 (1.1%)
0 (0.0%)
1 (0.0%)
75 (48.4%)
52 (33.5%)
24 (15.5%)
3 (1.9%)
1 (0.7%)
1 (0.0%)
167 (50.5%)
98 (29.6%)
51 (15.4%)
10 (3.0%)
5 (1.5%)
2 (0.0%)
92 (52.3%)
53 (30.1%)
25 (14.2%)
5 (2.8%)
1 (0.6%)
1 (0.0%)
75 (48.4%)
45 (29.0%)
26 (16.8%)
5 (3.2%)
4 (2.6%)
1 (0.0%)
0.068
0.152
0.097
0.605
a
P values of Freeman-Halton tests comparing BNT162b2 cohort and unvaccinated cohort,
excluding category of missing.
38
Supplemental Table 3 Mixed Models for Repeated Measurements EQ-5D-5L and WPAI-GH Scores
Work productivity
EQ-5D-5L Utility Index
Absenteeism
Presenteeism
Activity impairment
(U.S. weights)
loss
Coeff (SE) P value Coeff (SE) P value Coeff (SE) P value Coeff (SE) P value Coeff (SE) P value Coeff (SE) P value
EQ VAS
Intercept
4.51 (5.36)
Vaccinated
BNT162B2BNT162b2 3.65 (1.46)
No
Reference
Assessment Time
Day 2-4 / Week 1
Reference
Week 4
9.04 (1.23)
Assessment Time * Vaccinated
Day 2-4 / Week 1 *
Reference
BNT162b2
Week 4 * BNT162B2 -0.25 (1.69)
Pre-COVID-19 Baseline
-0.22 (0.05)
Score
Age, years
18-29
Reference
30-49
-1.82 (1.45)
50-64
-2.84 (1.74)
≥65
-1.86 (2.29)
Gender
Male
4.07 (1.30)
Female
Reference
Race / Ethnicity
White (non-Hispanic) Reference
Black/African
2.17 (2.73)
American
Hispanic or Latino
0.65 (1.68)
0.401
0.038 (0.062)
0.546
53.6 (4.9) <0.001 41.4 (5.7) <0.001 57.7 (6.0) <0.001
0.013
0.069 (0.019) <0.001 -19.4 (4.5) <0.001 -8.4 (4.2)
Reference
Reference
Reference
0.047 -11.2 (4.4) 0.012
Reference
41.2 (4.9)
<0.001
-6.3 (3.1)
Reference
0.044
Reference
Reference
Reference
Reference
Reference
<0.001 0.086 (0.014) <0.001 -62.9 (3.4) <0.001 -30.9 (3.3) <0.001 -48.0 (3.6) <0.001 -28.9 (2.5)
Reference
Reference
Reference
Reference
22.6 (4.8) <0.001 -0.8 (4.4)
0.561
-4.0 (3.5)
0.257
<0.001 -0.177 (0.057) 0.002
-0.8 (0.0) <0.001 -0.7 (0.1) <0.001 -0.8 (0.1) <0.001
-0.7 (0.0)
<0.001
Reference
0.211 -0.011 (0.018) 0.532
0.102 -0.048 (0.021) 0.025
0.417 -0.025 (0.028) 0.367
Reference
3.9 (2.4) 0.102
10.8 (2.9) <0.001
1.7 (5.5) 0.760
Reference
0.065 4.1 (3.2) 0.204
0.074 7.3 (4.1) 0.075
0.964 -1.4 (7.9) 0.863
Reference
7.8 (2.9)
13.7 (3.5)
6.9 (4.7)
0.006
<0.001
0.139
0.002
-2.1 (2.2) 0.360 -6.8 (2.9)
Reference
Reference
0.019
-5.8 (3.1) 0.057
Reference
-9.8 (2.6)
Reference
Reference
Reference
0.028 (0.016)
Reference
0.080
Reference
Reference
Reference
2.8 (4.8)
Reference
0.884 -0.025 (0.020) 0.209
Reference
5.6 (3.0)
6.8 (3.8)
0.3 (7.4)
0.863
<0.001
<0.001
0.428
0.017 (0.033)
0.613
-1.8 (5.1)
0.726
0.8 (6.4)
0.899
0.1 (7.1)
0.994
-1.9 (5.6)
0.730
0.701
0.005 (0.021)
0.818
-1.9 (3.0)
0.530
4.3 (3.9)
0.268
4.8 (4.1)
0.243
-2.3 (3.4)
0.494
39
Asian
4.24 (2.54)
Patient refused
-0.30 (3.26)
Other
2.90 (2.81)
Region
Northeast
1.91 (2.00)
South
0.87 (1.51)
Midwest
Reference
West
0.49 (2.08)
Social Vulnerability Index
<0.25
Reference
≥0.25 and <0.5
1.21 (1.45)
≥0.5 and <0.75
0.93 (1.62)
≥0.75
-2.80 (2.14)
Previously tested
0.99 (1.13)
positive
Work in healthcare
-0.90 (1.80)
Work in high-risk
-2.83 (1.96)
setting
Live in high-risk setting 1.36 (2.53)
Immunocompromised
1.22 (2.61)
Number of Symptoms
-0.59 (0.23)
on index day
0.096
0.927
0.303
0.032 (0.031)
0.031 (0.040)
0.019 (0.035)
0.316
0.438
0.582
-1.7 (4.7)
-3.3 (6.1)
3.9 (4.8)
0.713
0.589
0.418
1.4 (5.8)
1.1 (8.4)
-6.9 (6.5)
0.807
0.897
0.288
2.4 (6.3) 0.705
-2.3 (9.0) 0.800
-3.1 (6.9) 0.661
0.9 (5.2)
-6.4 (7.0)
-7.5 (5.8)
0.858
0.359
0.198
0.341 0.015 (0.025) 0.534
0.568 -0.002 (0.019) 0.905
Reference
0.815 0.019 (0.026) 0.448
6.2 (3.3) 0.057 -1.9 (4.2)
-1.0 (2.6) 0.698 -2.4 (3.3)
Reference
Reference
-0.7 (3.4) 0.830 -0.8 (4.5)
0.648
0.466
-0.5 (4.5) 0.908
-2.4 (3.5) 0.507
Reference
0.857 -1.3 (4.8) 0.793
-2.3 (3.9)
-2.9 (3.0)
Reference
-3.7 (4.2)
0.551
0.328
0.403
0.568
0.192
Reference
0.034 (0.018)
0.020 (0.020)
0.009 (0.026)
0.060
0.321
0.740
Reference
Reference
1.3 (2.4) 0.579 -3.9 (3.0)
4.7 (2.8) 0.090 -4.8 (3.6)
2.3 (3.7) 0.531 -9.3 (4.7)
Reference
0.196 -4.6 (3.3) 0.158
0.180 -6.0 (3.9) 0.124
0.048 -7.8 (5.0) 0.121
Reference
-5.2 (2.9)
-3.6 (3.3)
-5.0 (4.3)
0.076
0.276
0.247
0.381
0.003 (0.014)
0.816
-1.9 (2.0)
0.333
-0.5 (2.6)
0.851
-2.7 (2.8) 0.329
-1.2 (2.3)
0.591
0.616 -0.018 (0.022) 0.424
3.1 (2.8)
0.272
0.8 (3.6)
0.818
-0.7 (4.0) 0.851
3.7 (3.6)
0.300
0.150 -0.006 (0.024) 0.790
4.7 (3.4)
0.166 10.3 (4.3)
0.017 13.3 (4.6) 0.004
5.7 (4.0)
0.155
0.591 -0.028 (0.031) 0.370
0.639 -0.018 (0.032) 0.560
3.2 (4.7) 0.502
13.1 (6.1) 0.033
4.8 (6.0)
9.9 (8.4)
0.431
0.243
7.2 (6.5)
7.9 (9.1)
0.268
0.385
12.1 (5.0)
5.5 (5.1)
0.017
0.282
0.009 -0.010 (0.003) 0.001
0.8 (0.4)
1.3 (0.5)
0.008
1.9 (0.5)
0.001
1.8 (0.4)
<0.001
0.048
0.374
40
Supplemental Figure 1 Least-Square Estimates and 95% Confidence Intervals of EQ-5D-5L Scores a
a
Score ranges: EQ-5D-5L VAS 0 to 100, EQ-5D-5L UI (the United States weights) -0.573 to 1.
41
Supplemental Figure 2 Summary Results of EQ-5D-5L scores across time periods a
Score ranges: EQ-5D-5L VAS 0 to 100, EQ-5D-5L UI (the United States weights) -0.573 to 1. Dots are the mean values and
whiskers are the 95% Confidence Intervals. The Pre-COVID values represent the pooled means with 95% Confidence Internals. The
Day 3 and Week 4 values are the least square estimate scores.
a
42
Supplemental Figure 3 Least-Square Estimates and 95% Confidence Intervals of WPAI-GH Scores a
a
WPAI-GH score (absenteeism, presenteeism, work productivity loss, and activity impairment) ranges from 0 to 100 percent.
43
Supplemental Figure 4 Summary Results of WPAI-GH scores across time periods a
a
WPAI-GH score (absenteeism, presenteeism, work productivity loss, and activity impairment) ranges from 0 to 100 percent. Dots are
the mean values and whiskers are the 95% Confidence Intervals. The Pre-COVID values represent the pooled means with 95%
Confidence Internals. The Week 1 and Week 4 values are the least square estimate scores.
44
45
Supplemental Table 4 Model Predicting the Missingness at Week 1 and Week 4
Intercept
Assessment Time
Week 4
Week 1
Vaccination status
BNT162b2
Unvaccinated
EQ VAS on Day 3
EQ VAS change from preCOVID-19 baseline to Day 3
Age, years
18-29
30-49
50-64
≥65
Gender
Male
Female
Race / Ethnicity
White (non-Hispanic)
Black/African American
Hispanic or Latino
Asian
Patient refused
Summary: % (n) / Mean (SD)
Total
Week 1 Missing Week 4 Missing
430
11.9% (51)
23.0% (99)
Coeff (SE)
-3.14 (1.19)
Model
P value
0.009
Odds Ratio
23.0% (99)
0.85 (0.12)
Reference
<0.001
.
2.3 (1.9, 2.9)
1.0
12.9% (30)
10.7% (21)
24.5% (57)
21.3% (42)
68.9 (19.1)
-18.2 (18.9)
71.7 (17.3)
-16.0 (16.6)
0.36 (0.26)
Reference
0.00 (0.01)
0.166
.
0.967
1.4 (0.9, 2.4)
1.0
1.0 (1.0, 1.0)
-0.01 (0.01)
0.311
1.0 (1.0, 1.0)
87
213
94
36
8.0% (7)
10.3% (22)
16.0% (15)
19.4% (7)
16.1% (14)
23.9% (51)
27.7% (26)
22.2% (8)
Reference
0.69 (0.36)
0.94 (0.40)
0.72 (0.52)
.
0.052
0.019
0.161
1.0
2.0 (1.0, 4.0)
2.6 (1.2, 5.6)
2.1 (0.8, 5.6)
103
327
4.9% (5)
14.1% (46)
16.5% (17)
25.1% (82)
-0.80 (0.32)
Reference
0.013
.
0.4 (0.2, 0.8)
1.0
295
20
61
22
13
9.8% (29)
20.0% (4)
11.5% (7)
18.2% (4)
30.8% (4)
19.7% (58)
35.0% (7)
27.9% (17)
27.3% (6)
30.8% (4)
Reference
0.67 (0.50)
0.21 (0.36)
0.37 (0.52)
0.94 (0.62)
.
0.184
0.565
0.480
0.128
1.0
2.0 (0.7, 5.3)
1.2 (0.6, 2.5)
1.4 (0.5, 4.0)
2.6 (0.8, 8.6)
430
430
233
197
72.9
(17.5)
-14.4
(14.8)
11.9% (51)
46
Other
Region
Northeast
South
Midwest
West
Social Vulnerability Index
<0.25
≥0.25 and <0.5
≥0.5 and <0.75
≥0.75
Previously tested positive
Work in healthcare
Work in high-risk setting
Live in high-risk setting
Immunocompromised
Number of Symptoms on index
day
19
15.8% (3)
36.8% (7)
1.22 (0.54)
0.024
3.4 (1.2, 9.8)
52
252
77
49
7.7% (4)
13.5% (34)
6.5% (5)
16.3% (8)
21.2% (11)
24.6% (62)
16.9% (13)
26.5% (13)
0.50 (0.48)
0.32 (0.37)
Reference
0.44 (0.47)
0.295
0.378
.
0.343
1.7 (0.6, 4.2)
1.4 (0.7, 2.8)
1.0
1.6 (0.6, 3.9)
98
164
120
48
167
47
44
22
19
12.2% (12)
9.8% (16)
12.5% (15)
16.7% (8)
15.6% (26)
12.8% (6)
13.6% (6)
9.1% (2)
10.5% (2)
22.4% (22)
19.5% (32)
25.0% (30)
31.3% (15)
26.9% (45)
19.1% (9)
22.7% (10)
27.3% (6)
15.8% (3)
Reference
.
1.0
-0.22 (0.33)
0.07 (0.35)
0.38 (0.45)
0.45 (0.25)
-0.24 (0.42)
-0.25 (0.43)
0.38 (0.54)
-0.33 (0.67)
0.496
0.850
0.397
0.069
0.567
0.559
0.484
0.623
0.8 (0.4, 1.5)
1.1 (0.5, 2.1)
1.5 (0.6, 3.5)
1.6 (1.0, 2.5)
0.8 (0.3, 1.8)
0.8 (0.3, 1.8)
1.5 (0.5, 4.3)
0.7 (0.2, 2.7)
5.3 (2.6)
5.4 (2.6)
5.0 (2.5)
-0.07 (0.05)
0.188
0.9 (0.8, 1.0)
47