key: cord-0324242-abst7zi0
authors: Marquez, C.; Kerkhoff, A.; Schrom, J.; Rojas, S.; Black, D.; Mitchell, A.; Wang, C.-Y.; Pilarowski, G.; Ribeiro, S.; Jones, D.; Payan, J.; Manganelli, S.; Lemus, J.; Jain, V.; Chamie, G.; Tulier-Laiwa, V.; Petersen, M.; DeRisi, J.; Havlir, D.
title: COVID-19 symptoms and duration of direct antigen test positivity at a community testing and surveillance site, January 2021-2022
date: 2022-05-23
journal: nan
DOI: 10.1101/2022.05.19.22274968
sha: 023fcc1d7a74185df8ab84a7cce99af87045d141
doc_id: 324242
cord_uid: abst7zi0

Importance: Characterizing clinical symptoms and evolution of community- based SARS Co-V-2 infections can inform health practitioners and public health officials in a rapidly changing landscape of population immunity and viral variants. Objective: To characterize COVID-19 symptoms during the Omicron period compared to pre-Delta and Delta variant periods and assess the duration of COVID-19 BinaxNOW rapid antigen test positivity during the Omicron variant surge. Design, Setting, and Participants: This public health surveillance study was undertaken between January 2021- January 2022, at a walk-up community COVID-19 testing site in San Francisco, California. Testing with BinaxNOW rapid antigen tests was available regardless of age, vaccine status, or symptoms throughout. Main Outcomes and Measures: We characterized the prevalence of specific symptoms for people with a positive BinaxNOW test during the Omicron period and compared it to the pre-Delta and Delta periods. During the Omicron period, we examined differences in symptoms by age and vaccine status. Among people returning for repeat testing during Omicron period, we estimated the proportion with a positive BinaxNOW antigen test between 4-14 days from symptom onset or since first positive test if asymptomatic. Results: Of 63,277 persons tested, 18,301 (30%) reported symptoms and 4,568 (25%) tested positive for COVID-19. During the Omicron period, 41.6% (3032/7283) of symptomatic testers tested positive, and the proportion reporting cough (67.4%) and sore throat (43.4%) was higher than during Delta and pre-Delta periods. Congestion was higher during Omicron (38.8%) than during the pre-Delta period and loss of taste/smell (5.3%) and fever (30.4%) were less common. Fevers and myalgias were less common among persons who had received boosters compared to unvaccinated people or those who received the primary series. Five days after symptom onset, 31.1% of people with COVID-19 stated their symptoms were similar or worsening. An estimated 80.2% of symptomatic re-testers remained positive five days after symptom onset and 60.5% after ten days. Conclusions and Relevance: COVID-19 upper respiratory tract symptoms were more commonly reported during the Omicron period compared to pre-Delta and Delta periods, with differences by vaccination status. Antigen test positivity remained high after 5 days, supporting guidelines requiring a negative test to shorten the isolation period.

The study was conducted under a public health surveillance program that was reviewed 158 by the UCSF Committee on Human Research and determined to be exempt and waived 159 from IRB oversight. All participants provided informed consent in their preferred 160 language prior to survey administration and COVID-19 testing. 161 162

All participants or their caretakers completed a structured electronic survey capturing 164 socio-demographics, vaccination status, symptoms (including date-of-onset and 165 trajectory [e.g., improving, unchanged, worsening]). Trained laboratory assistants 166 performed a bilateral anterior nasal swab for COVID-19 testing using the BinaxNOW 167 COVID-19 Ag Card (Abbott Diagnostics) and a bilateral nasal swab for sequencing. 14,15 168

We performed genotyping of all isolates as previously described. 16 Full genome 169 sequences are available through GISAID and previously described for the pre-Delta 170 period (majority being Epsilon 17 and Alpha variants) and for the Omicron period. 16 

During the Omicron period, community health workers reminded people testing positive 172 to repeat their test 5 days after symptom onset or their initial test to assess candidacy 173 for shortening isolation per California Department of Public Health guidelines. 18 Clients 174 received a text message reminding them of this option. 175 176 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. prevalence differed between the Omicron compared to the Delta and pre-Delta periods. 186

Finally, during the Omicron period, we also assessed symptom prevalence by age and 187 vaccination status, and determined whether self-reported symptoms improved, 188 worsened, or remained unchanged over days following the test. 189

The second analysis (n=942) sought to understand the proportion of participants 190 during the Omicron wave with a positive repeat COVID-19 BinaxNOW rapid antigen test 191

result. This analysis was restricted to 're-testers', defined as participants who had a 192 positive BinaxNOW rapid antigen test result on or after January 1, 2022, and had at 193 least one additional BinaxNOW rapid antigen result two or more days after their initial 194 positive test. For each day from 4 to 14 following symptom onset, or day since initial 195 positive test if asymptomatic, we estimated the proportion of persons who remained 196 positive by day. Participants whose second test was positive were assumed to be 197 positive each day between the positive tests. Test positivity on days between a positive 198 and a negative test was imputed in three different ways: (1) assuming a linearly 199 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) (Table 1A ). The proportion of symptomatic COVID-19 positive persons 233 reporting cough and sore throat was higher during the Omicron period relative to Delta 234 and pre-Delta periods, while fever, and loss of taste/smell decreased relative to Delta 235 and pre-Delta periods (Table 1A , Supplementary Figure 1 , p<0.05 for all comparisons). 236

Congestion was more common among symptomatic people testing positive during the 237

Omicron period compared to pre-Delta (p<0.001). The leading COVID-19 symptoms 238 among symptomatic people testing positive during the Omicron period (cough, sore 239 throat, and congestion) were also common among symptomatic people who tested 240 negative (Table 1, Supplementary Figure 1) . 241

During the Omicron period, 47.7% of symptomatic children (<12 years) with 242 COVID-19 reported only one symptom. Reporting a single symptom was less common 243 among adults (37.7%, p<0.001) and adolescents (31.8%, p<0.001) (Supplemental 244 Table 2 ). Among children with COVID-19, cough, fever, sore throat, and congestion 245 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 23, 2022. ;

were the most common symptoms. Loss of taste/smell was uncommon among children 246 (0.3%) compared to adolescents (5.8%, p<0.001) and adults (5.8%, p<0.001) ( Table  247 1B). 248

We also found differences in symptoms according to vaccination status (Table 3) . 249

Among boosted persons, congestion was more common, and fever was less common 250 compared to both unvaccinated people and vaccinated/non-boosted people (p<0.05 for 251 all comparisons; Table 1C ); myalgias were also less common among boosted persons 252 and in 5.9% were worsening (95%CI: 3.3%-9.7%). Among participants testing ten days 261 after symptom onset, symptoms were improving in 82.2% (95%CI: 56.6%-69.2%), were 262 similar in 17.8% (95%CI: 9.8%-28.5%) and were worsening in 0% (95%CI: 0%-4.9%). 263

Symptom trajectory was similar across vaccination status or age (Figure 1B, 1C) . 264

Among persons who undertook COVID-19 re-testing, an estimated 65% (95%CI: 62%-267 69%) people remained test-positive 5 days after symptom onset or 5 days from their 268 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 23, 2022. ; https://doi.org/10.1101/2022.05.19.22274968 doi: medRxiv preprint initial positive test if they were asymptomatic ( Figure 2A ). Among symptomatic persons, 269 an estimated 80% (95% CI: 76%-84%) and 35% (95%CI: 30%-40%) remained positive 270 on day 5 and 10 respectively. Among persons asymptomatic at testing, an estimated 271 49% (95% CI: 34-55%) and 19% (95% CI, 5%-42%) remained positive on day 5 and 10. and rapid antigen test positivity to guide safe return-to-work guidelines. 289 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 23, 2022. ;

In over 3,000 symptomatic people testing positive for COVID-19 during an 290

Omicron surge, upper respiratory tract symptoms (sore throat, congestion) were more 291 common than during pre-Delta and Delta periods. The loss of taste and smell -often 292 used by clinicians to help distinguish SARS Co-V 2 from other viral illness -was much 293 less common during the Omicron period (5.3%) compared to the pre-Delta (17.2%) and 294

Delta (20.5%) periods. This shift in symptomatology seen from pre-Delta to Omicron is 295 likely due to both the increase in population immunity during the Omicron surge and 296 biologic characteristics of the Omicron variant, which in vitro-studies suggest replicates 297 better in bronchial tissue versus deeper lung sites. 19 During the Omicron surge, 298 symptoms also varied by vaccine status; congestion was more common, and fever and 299 myalgia were less common among boosted persons compared to partially or 300 unvaccinated persons. 301

Our findings are concordant with a population-representative household study 302 conducted during the Omicron period in the United Kingdom (UK) 5 that found a 303 predominance of upper respiratory tract symptoms. Additionally, our data expands 304 existing knowledge on COVID-19 symptomatology by describing differences in 305 symptomatology by vaccine status and age-group. Assessing the ability of symptoms to 306 predict test positivity was outside the scope of our analysis. However, similar to findings 307 from the UK findings, 5 we found the background rate of common COVID-19 symptoms 308 during the Omicron period to be high among both people who tested both positive and 309 negative. These findings emphasizing the importance of ensuring low-barrier testing is 310 available regardless of the tester's symptom profile. 311 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) it is noteworthy that recovery varied and 31% of patients did not report perceived 321 improvement by day 5. During the Omicron surge, frontline workers were urged to 322 return to work to meet staffing needs in occupations such as transport, home health 323 care, and the food service industry. During surges of the magnitude of Omicron, 324 frontline workers can be urged to return to work to meet staffing needs in occupations 325 such as transport, home health care, and the food service industry. Despite not having 326 illness serious enough to merit hospitalization, many persons with Omicron still felt ill, 327 yet felt pressure by their employers to return to work. Safeguards routinely in place in 328 the formal sector in the United States, are often unavailable in the informal sector. 23,24 329

Pandemic planning must account for these health and economic pressures in future 330

Of persons with COVID-19 re-testing on Day 5, 80% of symptomatic people and 332 41% of asymptomatic people remained positive. This number is nearly identical to a 333 community-based sample of people seeking testing in Alaska 25 and similar to the 334 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) variants. In a predominantly boosted cohort from the National Basketball Association 337 (NBA), viral trajectories of Omicron were heterogenous, but the duration of the acute 338 phase (proliferation and clearance) was 9.9 days and similar to prior variants, including 339

Delta. 28,29 Additionally, in this cohort of predominantly healthy, boosted NBA players 340 with infections due to the Omicron variant, only 52% of people had a cycle threshold of 341 <30 5 days after symptom onset. 28 342

Interestingly, we found that COVID-19 test positivity remained high even 10 days 343 following symptom onset, where more than 30% or people with symptomatic COVID-19 344 during their initial still had a positive rapid antigen test. Additional epidemiologic studies 345 are needed to assess whether people remain infectious at this juncture. However, prior 346 studies show a strong correlation between rapid antigen positivity and viable virus, 7-9,30 347 even in the clearance phase and after 10 days of symptom onset 9,10,31 -suggesting 348 infectiousness beyond 10 days may be possible, though less common. 349

Positive rapid antigen tests correlate to having viable virus, and thus capture 350 persons with the highest degree of infectivity to others. 7-9,11 The CDC recommends that 351 people can leave isolation with a well-fitting mask after 5 days of symptoms if symptoms 352 are improving, regardless of repeat testing. 6 If a public health goal is to break chains of 353 SARS CoV-2 transmission, our data strongly support current California Department of 354

Public Health guidelines 18 that require a rapid antigen test to shorten the 10-day 355 isolation period. 356 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. analysis. Sensitivity analysis assumes participants whose second test is negative either 555 immediately became negative after their first test (lower bound) or remained positive 556 until the day before their second test (upper bound). Panel B shows results stratified by 557 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 23, 2022. ; https://doi.org/10.1101/2022.05.19.22274968 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 23, 2022. ; Figure 1 : Proportion of COVID-19 positive clients during Omicron period indicating improvement, no change, or worsening of symptoms at time of test, among symptomatic BinaxNOW positive participants testing within 14 days of symptom onset, A. Overall, B. by Age Group and C. and by Vaccination Status . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 23, 2022. ; Figure 2 : BinaxNOW positivity among 942 repeat testers with COVID-19 during the Omicron period by day of symptom onset (if symptomatic) or day since initial positive test (if asymptomatic). COVID-19 status between a participant's tests is inferred to be positive (if second test is positive) or with a linearly decaying probability of being positive (if second test is negative). Panel A shows overall, including symptomatic and asymptomatic people with dotted lines indicating upper and lower bounds of sensitivity analysis. Sensitivity analysis assumes participants whose second test is negative either immediately became negative after their first test (lower bound) or remained positive until the day before their second test (upper bound). Panel B shows results stratified by symptom status. Panel C show results stratified by self-report vaccination status as of their first test.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 23, 2022. ;

The COVID-19 Symptom to Isolation Cascade in 412