key: cord-0991940-lpdvpg1y
authors: Abbas, Mohamed; Nunes, Tomás Robalo; Cori, Anne; Cordey, Samuel; Laubscher, Florian; Baggio, Stephanie; Jombart, Thibaut; Iten, Anne; Vieux, Laure; Teixeira, Daniel; Perez, Monica; Pittet, Didier; Frangos, Emilia; Graf, Christophe E.; Zingg, Walter; Harbarth, Stephan
title: Explosive nosocomial outbreak of SARS-CoV-2 in a rehabilitation clinic: the limits of genomics for outbreak reconstruction
date: 2021-08-27
journal: J Hosp Infect
DOI: 10.1016/j.jhin.2021.07.013
sha: b57f16f40b68fff9acbcdca05d762f3475a9a14b
doc_id: 991940
cord_uid: lpdvpg1y

BACKGROUND: Nosocomial outbreaks of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are frequent despite implementation of conventional infection control measures. We performed an outbreak investigation using advanced genomic and statistical techniques to reconstruct likely transmission chains and assess the role of healthcare workers (HCWs) in SARS-CoV-2 transmission. METHODS: We investigated a nosocomial SARS-CoV-2 outbreak in a university-affiliated rehabilitation clinic, involving patients and HCWs, with high coverage of pathogen whole genome sequences (WGS). We estimated the time-varying reproduction number from epidemiological data (Rt) and produced a maximum likelihood phylogeny to assess genetic diversity of the pathogen. We combined genomic and epidemiological data into a Bayesian framework to model directionality of transmission. We performed a Case-control study to investigate risk factors for nosocomial SARS-CoV-2 acquisition in patients. FINDINGS: The outbreak spanned from March 14 to April 12, 2020 and involved 37 patients (31 with WGS) and 39 employees (31 with WGS) of whom 37 are HCWs. We estimated a peak R(t) between 2.2 – 3.6. The phylogenetic tree showed very limited genetic diversity, with 60/62 (96.7%) isolates forming one large cluster of identical genomes. Despite the resulting uncertainty in reconstructed transmission events, our analyses suggest that HCWs (one of whom was the index Case) played an essential role in cross-transmission, with a significantly larger fraction of infections (p < 2.2e-16) attributable to HCWs (70.7%) than expected given the number of HCWs cases (46.7%). The excess of transmission from HCWs was larger when considering infection of patients (79.0%; 95%CI 78.5% - 79.5%), and especially frail patients (Clinical Frailty Scale >5: 82.3%; 95%CI 81.8% - 83.4%). Furthermore, frail patients were found to be at greater risk for nosocomial COVID-19 than other patients (adjusted OR 6.94; 95%CI 2.13 – 22.57). INTERPRETATION: This outbreak report highlights the essential role of HCWs in SARS-CoV-2 transmission dynamics in healthcare settings. Limited genetic diversity in pathogen genomes hampered the reconstruction of individual transmission events, resulting in substantial uncertainty in who infected whom. However, our study shows that despite such uncertainty, significant transmission patterns can be observed.

7 the HUG Department of Human Resources on HCW shifts. Patient trajectories were extracted from 124 the Electronic Health Records (EHR). 125

For the case-control study, we collected data retrospectively from the EHR for the control 126 group, using the same RedCAP TM eCRF as for the FOPH cohort study. We created additional data 127 collection instruments in order to capture relevant risk factors based on an informal literature search 128 for both cases and controls. Among the variables collected were the Clinical Frailty Scale (CFS) [12] , 129 hypoalbuminaemia (and hypoprealbuminaemia) as a proxy for malnutrition [13] , current smoking, 130 age, body mass index, comorbidities, the Charlson Comorbidity Index [14] , and the Cumulative 131

Illness Rating Scale-Geriatric (CIRS-G) [15] . The CIRS-G was dichotomised as a binary variable with a 132 cut-off at 15 points [16] . 133

We produced an epidemic curve using dates of symptom onset; where these were unavailable (e.g. 135 lack of symptoms), we imputed them with the median difference between date of symptom onset 136 and date of nasopharyngeal swab. 137

All COVID-19 cases in the outbreak were confirmed by RT-PCR on nasopharyngeal swabs. Three 139 diagnostic methods were used for routine screening: the Cobas 6800 SARS CoV2 RT-PCR (Roche, 140 Switzerland) , the BD SARS-CoV2 reagent kit for BD Max system (Becton, Dickinson and Co, USA), and 141 an in-house method based on the Charité assay [17] . We performed SARS-CoV-2 WGS using either 142 an unbiased high-throughput sequencing method or an amplicon-based sequencing method in order 143 to produce RNA sequences. Full details of the microbiological methods, including the WGS and the 144 sequence assembly can be found in the Supplement. Switzerland, during the same period. 153

We performed descriptive statistics with medians and interquartile ranges (IQR), and counts and 155

proportions, as appropriate. We performed a case-control study to identify risk factors for 156 nosocomial COVID-19 among patients. The control population was defined as patients who were 157 hospitalised for ≥ 5 days (to avoid immortal bias) in the rehabilitation clinic during the outbreak 158 period, and who had neither a positive RT-PCR for SARS-CoV-2 nor a clinical diagnosis of COVID-19 159 (presumed or confirmed). We randomly selected unmatched controls from this population at a ratio 160 of 2:1. Associations between patient characteristics and nosocomial COVID-19 were assessed using 161 the chi-squared test for categorical variables, and the Kruskall-Wallis test for continuous variables. 162

We performed univariable logistic regression to produce crude odds ratios (OR) and 95% confidence 163 intervals (95% CI). We built a multivariable logistic regression model using a forward fitting method 164 and selected the most parsimonious model as guided by the Akaike Information Criterion and 165 likelihood ratio tests [20] . We evaluated goodness-of-fit using the Hosmer-Lemeshow test. 166

We estimated the time-varying reproduction number (Rt) from the incidence of symptoms, 167 using the incidence and EpiEstim packages for the R statistical software [21] . Rt estimation 168 was performed under a "short" serial interval (time between onset of symptoms in an 169 infector/infectee pair) assumption, consistent with early isolation of cases in the hospital context. A 170 material). 172

We combined epidemiological and genetic data to reconstruct who infected whom using the R 174 

The funding source had no involvement in the writing of the manuscript or the decision to submit it 196 for publication. All authors had full access to the full data in the study and accept responsibility to 197 submit for publication. 198 199

The index case of the outbreak was a HCW who was detected on March 14, 2020; the first patient 201 was detected after 5 days in the same ward, and the last case on April 17, 2020. Therefore, we 202 defined the outbreak period as spanning from March 1, 2020 to April 19, 2020. In total, 37 patients 203 and 39 hospital employees (including 2 administrative staff) were involved in the outbreak. The 204 institution-level attack rate for patients was 21.2%. The epidemic curve is shown in Figure 1A . All five 205 wards of the clinic were involved in the outbreak. Characteristics of patients and employees are 206 summarised in Table 1 and Table 2 , respectively. 207

We implemented several infection control measures sequentially ( Figure 1C ). Staff with a 208 positive SARS-CoV-2 RT-PCR were immediately placed on mandatory sick leave for a minimal 209 duration of 10 days from the onset of symptoms. We did 2 ward-level point-prevalence screening 210 surveys of all negative patients (including asymptomatic) in affected wards, and we strongly 211 encouraged all staff to undergo testing. Universal screening on admission was performed starting 212

April 02, 2020. Throughout the outbreak period, RT-PCR results were obtained with an average turn-213 around time of approximately 8 hours. 214

We estimated that Rt rapidly declined from 2.17 (95% credible intervals CrI 1.43 -3.07) by March 21, 216 2020, with the mean Rt reaching <1 on March 28, 2020 ( Figure 1B ). Trends were similar when 217 assuming a longer serial interval, albeit with higher mean Rt estimates, at 3.60 (95% CrI 2.37 -5.08) 218 J o u r n a l P r e -p r o o f 11 initially, and <1 by March 28, 2020 (Supplementary Figure 1) . Confining patients to their room 219 decreased Rt, but not below 1; although this was achieved by closing the wards on the second floor, 220 the upper limit of the 95% CrI was above 1 (Supplementary Figure 2) . Pre-emptive contact 221 precautions and mandatory masking of patients outside of rooms further decreased Rt with the 222 upper limit of the 95% CrI <1. 223

We were able to obtain SARS-CoV-2 sequences for 62 cases (31 patients, 31 employees) and 225 generated a phylogenetic tree (Figure 2) . Interestingly, although some tree branches should be 226 interpreted with caution due to moderate bootstrap values reflecting the very high sequence 227 homology between isolates, some branch specific mutations (e.g.C5239T, C15324T or G29781T) 228

were observed that support some branching order. Sixty of 62 sequences (96.8%) formed a large 229 single cluster with clear segregation from the community sequences. We observed one large 230 subcluster (bootstrap 60%), which corresponds to a specific ward in the clinic. Sequences from 231 patients and HCWs were distributed similarly across all branches of the tree. The phylogenetic 232 analysis suggests importation of 2 cases (3.2%), which were H2030 (HCW) and H2013 (admin). We compared the proportion of infections attributed to HCWs (noted fHCW) to random 244 expectations assuming HCWs and patients were equally likely to seed new infections. We found that 245 fHCW was significantly higher than expected in all settings considered (Figure 4 , Supplementary Table  246 2), with a relative excess of transmission from HCWs ranging from to 31% to 76% more infections 247 than expected. The proportion of transmission from HCWs was larger when considering infection of 248 patients (79.0%; 95%CI 78.5% -79.5%), and especially frail patients (with a CFS > 5: 82.3%; 95%CI 249 81.8% -83.4%). 250

The sensitivity analyses did not show major changes in either the probabilities of 251 transmission pairs, or the overall structure of the consensus transmission tree, with the exception of 252 the model without any contact data, which, as expected, had worse resolution, and suggested fewer 253 transmission events by HCWs (Supplementary Figures 4-7) . All sensitivity analyses similarly showed 254 the infections from HCWs were more frequent than usual (all p-values < 2.2e-16), albeit with varying 255 effect sizes (Supplementary Figure 8) We describe here an explosive nosocomial outbreak of SARS-CoV-2 in a rehabilitation clinic involving 272 both patients and HCWs. Despite rich data and sequencing availability, including HCWs, and even 273 using advanced epidemiological and genomic analyses, we were unable to reconstruct who infected 274 whom in the outbreak with high confidence. Still, we show that HCWs played an undisputable role in 275 introducing SARS-CoV-2 into the facility, and were the main drivers of infection to patients and each 276 other. 277

It has been conventional wisdom, particularly during the first pandemic wave, that patients 278 posed a greater risk to HCWs than vice-versa, even with appropriate personal protective equipment. 279

This has also been suggested by a recent outbreak investigation in an acute care hospital [24] . 280

Although we do not dispute the fact that appropriate IPC measures are essential to protect HCWs, 281 our results suggest that in non-COVID settings, patients are more likely to become infected by HCWs 282 than vice-versa. HCWs did not only introduce SARS-CoV-2 in the clinic, but were also at the origin of 283 most super-spreading events. Conversely, direct patient-to-patient transmission in a setting such as 284 a rehabilitation clinic or LTCF does not appear to be a major driver of infection. 285

Our results suggest that in such closed settings, with a dense outbreak and a relatively 286 slowly evolving pathogen, genomic sequencing data offers little added value. Indeed, in this 287 outbreak, the large majority of isolates formed a single large cluster. Nevertheless, genomic 288 surveillance can still be useful, for example for the detection of novel variants of disease, and to 289 understand epidemic and evolutionary patterns at larger geographical and temporal scales [25, 26] . 290 J o u r n a l P r e -p r o o f 14 Although this outbreak occurred in a setting with high adherence to traditional IPC 291 measures, we have successfully managed to control the outbreak with non-pharmaceutical 292 interventions, without availability of real-time genomic data. Indeed, the Rt decreased to <1 after 293 closing the ward with highest number of infections. We are confident that the IPC measures were 294 responsible for the control of the outbreak, and not depletion of susceptibles. Indeed, given our 295 peak estimated Rt values, we would have expected an attack rate ranging between 83%-96% in the 296 absence of IPC measures, which is far greater than what we observed. 297

The findings from this study confirm previous concerns that HCWs play a major role in 298 initiating, amplifying, and sustaining outbreaks of nosocomial SARS-CoV-2 [6]. In this outbreak, 299 approximately 80% of patient infection events are attributable to HCWs. Furthermore, HCWs can 300 infect their peers in the work environment, but outside of direct clinical contact in places such as the 301 cafeteria, break rooms, or offices. Physical distancing guidelines can sometimes be difficult to adhere 302 to due to architectural constraints (small offices, for example). Practices such as car-sharing or 303 couchsurfing due to border restrictions may have also played a role. 304

We identified patient frailty as an important independent risk factor for nosocomial 305 acquisition of SARS-CoV-2, which has been suggested by previous reports [27] . This can stem from 306 biological phenomena whereby there is increased susceptibility to acquiring an infection (e.g. due to 307 immunosenescence), or from differences in contact patterns because frail patients require more 308 assistance for daily activities such as personal hygiene or dressing. Frailty in older patients with 309 COVID-19, however, has been associated with poorer outcomes in geriatric patients [27, 28] , and 310 this may be due to underlying biological reasons. For this reason, it is paramount that SARS-CoV-2 be 311 kept outside the walls of LTCFs and nursing homes [29] . 312

Our study has several strengths, including a comprehensive outbreak investigation, which 313 captured a high proportion of cases due to our aggressive testing strategy early in the first pandemic 314 wave. Widespread testing of patients and HCWs, especially asymptomatic, is now conventional 315 J o u r n a l P r e -p r o o f 15 wisdom, but at the time represented a veritable paradigm shift in the management of respiratory 316 viral outbreaks during the early phase of the pandemic. Another strength of our study is that we 317 included epidemiological and genomic data on HCWs who were part of the outbreak. Indeed, many 318 outbreak investigations have little data or genomic sequences of HCWs [6], yet we have 319 demonstrated that their role is essential. The data we have collected is of high quality as it was 320 mostly collected prospectively. Also, we were able to obtain genomic sequences of SARS-CoV-2 for 321 most (>80%) of all cases, which increases the robustness of our approach. Finally, we used 322 sophisticated modelling techniques, which combined epidemiological and genetic sequencing data in 323 order to reconstruct the outbreak and to provide insight into transmission patterns. 324

Nevertheless, several limitations should be considered when interpreting the results. First, 325 the estimates of serial interval and incubation periods that we selected can be challenged; however, 326 the results were not sensitive to a longer serial interval, which implies less timely control of the 327 outbreak. Second, not all cases were included in the outbreaker2 model due to lack of genomic 328 sequences (14 out of 76 cases); nonetheless, the model is designed and has been proven able to 329 identify unsampled cases in the transmission chains [23] . 330

In conclusion, we have shown that nosocomial outbreaks of SARS-CoV-2 in rehabilitation 331 clinics can spread very quickly in a population with a naïve immune system, and that both 332 introduction and spread of disease can be mediated by HCWs. This has long-term implications for 333 

genome sequencing to track SARS-CoV-2 transmission in nosocomial outbreaks

The ORION 388 statement: guidelines for transparent reporting of outbreak reports and intervention studies of 389 nosocomial infection

Prevention & control of healthcare-associated COVID-19 outbreaks

A global clinical 395 measure of fitness and frailty in elderly people

Hypoalbuminemia: Pathogenesis and Clinical Significance

A new method of classifying prognostic 399 comorbidity in longitudinal studies: development and validation

A manual of guidelines to score 402 the modified cumulative illness rating scale and its validation in acute hospitalized elderly 403 patients

Prospective comparison of 405 6 comorbidity indices as predictors of 1-year post-hospital discharge institutionalization

readmission, and mortality in elderly individuals

Detection of 2019 408 novel coronavirus (2019-nCoV) by real-time RT-PCR

Molecular Evolutionary Genetics Analysis 410 across Computing Platforms

Estimation of the number of nucleotide substitutions when there are strong 412 transition-transversion and G+C-content biases

A new look at the statistical model identification

A new framework and software to estimate time-416 varying reproduction numbers during epidemics

Bayesian reconstruction of 418 disease outbreaks by combining epidemiologic and genomic data

outbreaker2: a modular 421 platform for outbreak reconstruction

A SARS-CoV-2 Cluster in 423 an Acute Care Hospital. Annals of internal medicine

CoV-2 Spike Mutation D614G on Transmissibility and Pathogenicity

The Lancet. Genomic sequencing in pandemics

COVID-19: a retrospective 430 cohort study with focus on the over-80s and hospital-onset disease

Hospital Mortality in Older Patients With COVID-19: The COVIDAge Study

Can long-term care 435 facilities remain a coronavirus disease 2019 (COVID-19)-free bubble? An outbreak report

lines indicate the mean estimate of the proportion. A. All cases. B. Transmission to HCWs only. C

Transmission to patients only. D. Transmission to frail patients only

/Switzerland/GE2759/2020|EPI ISL 429214|2020-03-23 hCoV-19/Switzerland/GE1736/2020|EPI ISL

/Switzerland/GE8086/2020|EPI ISL 429213|2020-03-16 hCoV-19/Switzerland/GE6099/2020|EPI ISL

/Switzerland/GE8147/2020|EPI ISL 429210|2020-03-16 hCoV-19/Switzerland/GE0304/2020|EPI ISL

We would like to thank Rachel Goldstein (Geneva University Hospitals) for helping with data 340 collection, Frédéric Bouillot (Geneva University Hospitals) for providing us with HR data, Aurore 341Britan (Geneva University Hospitals) for data management. We would like to thank the team from