key: cord-0841709-a6j1qz2u authors: TRAiN study group,; Bretthauer, M. title: Randomized Re-Opening of Training Facilities during the COVID-19 pandemic date: 2020-06-24 journal: nan DOI: 10.1101/2020.06.24.20138768 sha: c888bdd99329597dc0350e575e25bb6aec79687f doc_id: 841709 cord_uid: a6j1qz2u BackgroundClosed training facilities during the Covid-19 pandemic may negatively impact peoples health and wellbeing. We investigated SARS-CoV-2 virus transmission, Covid-19 and SARS-CoV-2 antibodies attributable to training facilities. MethodsWe randomised members aged 18 to 64 without relevant comorbidities at five training facilities in Oslo, Norway, to access or no access to their facility. Facilities opened May 22, 2020 for the training arm, applying physical distancing (1 meter for floor exercise, 2 meters for high-intensity classes) and enhanced hand and surface hygiene. We compared SARS-CoV-2 RNA status by self-administered naso-, oropharyngeal and sputum sampling after 14 days; clinical disease through electronic patient records after 21 days; and SARS-CoV-2 antibody status by dried-blood self-sampling after one month. (ClinicalTrials.gov number NCT04406909) Findings3,764 individuals were randomised; 1,896 in the training and 1,868 in the no-training arm. In the training arm, 81.8% trained at least once, and 38.5% trained [≥]six times. Of 3,016 individuals who returned the SARS-CoV-2 RNA tests (80.5%), there was one positive test. The positive individual was randomised to training, but had not used the training facility, and the workplace was identified as transmission source. There were no outpatient visits or hospital admissions due to Covid-19 in either group. Eleven individuals in the training arm (0.8% of tested) and 27 in the no-training arm (2.4% of tested) tested positive for SARS-CoV-2 antibodies (p=0.001). InterpretationProvided good hygiene and physical distancing measures, there was no increased transmission of SARS-CoV-2 at training facilities. FundingNorwegian Research Council, grant no. 312757 Governments and health policy makers around the world have been taking preventive measures against COVID-19 exceeding previous pandemics (1) . Social distancing such as increasing distance between individuals (minimum 1 or 2 meters) is of paramount importance to contain spread of COVID-19. Many countries have closed or restricted access to schools, stores, restaurants, and work places to achieve social distancing (2) . While increased social distancing between individuals may involve little disturbance for daily life, closures of schools, recreational activities and work places have potentially large consequences for education, health and wellbeing, and personal and societal economy. Thus, it is important to test social distancing measures properly, to gain knowledge about their negative consequences and their impact on preventing virus spread (3) . Due to the uncertainty of contagiousness, immunity, morbidity and mortality of COVID-19, it is unclear how to resume activities without risking increased spread of COVID-19. Training and exercise is important for health and wellbeing. In many countries, training facilities and gyms are an important part of training and exercise for individuals, and for population health. In Norway, by governmental emergency law, all training facilities and gyms have been closed since March 12, 2020 (4,5). Surveys have indicated that Norwegians have a more sedentary lifestyle and exercise less than before the restrictions (6) . It is important to restrict unnecessary closure of training facilities to prevent societal downsides of the epidemic and negative effects on health and wellbeing. Many countries have introduced general rules for social distancing (1 meter distance, avoid body contact and greetings), and hygiene measures (hand wash and disinfection). These All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint 4 measures have been widely accepted and followed. However, there is little scientific evidence available about the benefits and harms of closing training facilities for COVID-19 as a preventive measure of virus spread. We hypothesised that the risk of SARS-CoV-2 transmission in training facilities with good hygiene and social distancing measures would be low, and thus safe to re-open to ensure health and wellbeing. This report describes the randomized testing of re-opening training facilities with close monitoring of COVID transmission and disease activity to understand the impact of training facilities closure for COVID-19. All training facilities in Norway have been closed by governmental restriction during COVID-19 since March 12, 2020. For the purpose of the trial, five training facilities in Oslo, operated by three professional companies in Norway, opened their premises to participants randomized to training for the period of the trial. The training facilities were SATS Sjølyst and CC Vest (SATS Norway Inc., Oslo, Norway), STOLT Stovner and Rommen (STOLT Trening Inc, Oslo, Norway), and EVO Bryn (EVO Fitness Group Inc, Oslo, Norway). Facilities which did not participate in the project remained closed, and participants in the notraining control arm did not have access. All members of the five participating training facilities age 18 years or older who are not at increased risk for severe COVID-19 disease per criteria by the Norwegian Institute of Public Health, were eligible for participation. The criteria for high risk are at least one of the following: age 65 years or older; cardiovascular disease including hypertension; diabetes (https://www.fhi.no/nettpub/coronavirus/fakta/risikogrupper/, accessed May 15, 2020). All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint Eligible individuals were approached by email by their training facilities through member lists. Individuals signed up for the study through a secure website at the University of Oslo. Co-morbidities were assessed by self-assessment. A direct contact telephone line and email address to the study team was established for interested individuals in case of uncertainty of their medical status related to comorbidities, and other questions. All eligible individuals were informed about the nature of the trial, and provided consent before randomization. We randomized eligible individuals to either current practice which was no access (notraining arm), or to access (training arm) with mitigation measures as described by the "Norwegian guidelines for Hygiene and Social Distancing in Training Facilities during the COVID-19 Pandemic", available at https://t-i.no/wpcontent/uploads/2020/04/Bransjestandard-for-sentre.pdf. In brief, the following measures recommended in the guideline were implemented at all facilities during the trial period: Avoidance of handshake and other body contact; 1 meter distance between individuals at all times; 2 meter distance for high intensity activities such as spinning, workout classes; provision of disinfectants at all work stations; cleaning requirements of all equipment after each use by member; regular cleaning of facilities by personnel; and access control by entrance personnel to ensure distance measures and avoid overcrowding. Changing rooms were open, but showers and saunas remained closed. Staff was present during all opening hours. Lids on trash cans were removed. Individuals were advised to stay home if they had any COVID-19 related symptoms. No masks were required, but members were advised to avoid touching their eyes, nose and mouth. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 June 24, 2020. All individuals in both groups were mailed a home-test kit including two swabs and a tube with virus transport medium for SARS-CoV-2 RNA. The tests were analysed with a commercially available real-time SARS-CoV-2 RT-PCR test (Cobas®, Roche Diagnostics Inc.) at the Department of medical microbiology, Oslo University Hospital. Participants were instructed to sample from the oropharynx, nose and saliva according to national guidelines (https://www.fhi.no/globalassets/dokumenterfiler/rapporter/2020/saliva-sample-for-testingsars-cov-2-infection-memo-2020.pdf) after median two weeks of training access in the training arm (June 8 or 9) and deliver the test to their training facility. Dedicated study personnel provided onsite collection of all tests, and helped sampling for those who did not want to self-sample, at all five facilities on June 8 and 9, 2020. The facilities remained open All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint for individuals in the training arm for an additional 7 days after testing, but remained closed for the no-training arm. We also offered SARS-CoV-2 RNA testing to all training facility employees who were in contact with individuals in the intervention arm. A commercially available rapid self-sampling kit, for dried blood spots for testing for SARS-CoV-2 antibodies, will be mailed to all participants between 3 and 4 weeks after study start, and will be mailed back in a prepaid envelope. The antibody testing will be performed at Oslo University Hospital. On June 15, 2020 (three weeks after study start), we retrieved all admissions and outpatient contacts for all somatic diagnoses (ICD-10 coding); ICU admissions, ventilator treatment, and death for all individuals in both arms from the trial area hospital databases for electronic patient records. Norway has a public, single-payer hospital system with full coverage of data for all individuals. Thus, our data are 100% complete. For individuals with diagnoses which may relate to COVID-19, we contacted physicians at the respective hospitals for details to investigate if the contact was related to COVID-19. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 June 24, 2020. Participants were followed for the primary and secondary endpoints as described above. The primary analytic approach of the trial follows the intention-to-treat (ITT) principle. We All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint compared the differences in event rates for the trial endpoints between the arms by chi-square test. Due to small numbers, we did not perform significance testing for all diagnosis subgroups (table 2) . Analyses were performed using Stata Statistical Software release 16 (StataCorp LLC, College Station, TX.) The study was approved by the Regional Ethical Committee of South-East Norway and the data protection officers at participating sites. All individuals provided written informed consent before enrolment. An Independent Data Safety and Monitoring Committee (DSMB) was established to ensure adequate handling of all data and trial participants. The trial area was the city of Oslo with surrounding municipalities with a total population of 821,000. One week before the start of the trial, the facilities approached members between 18 and 64 years living in the trial area by email and asked for interest in participating in the trial. In total, 3,938 individuals signed up for the trial online and provided written consent. Before randomization and intervention start, the study team found 174 individuals ineligible (10 were not members at the participating facilities; 55 were outside the eligible age range; 9 fulfilled one or more exclusion comorbidities; seven were employees at the participating facilities; and 73 withdrew consent. Thus, 3,764 individuals were randomized and included in the analyses; 1,896 in the training arm and 1868 in the no-training arm (figure 1). Table 1 displays participant characteristics and shows that the arms were well-balanced. There were 1,929 female and 1,835 male participants, the majority was between 20 and 50 years old. COVID-19 in Oslo during the trial All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 June 24, 2020. Among individuals randomized to training, a large majority (81.8%) trained at least once at the facility, and 38.5% trained six times or more (table 1) . After the two-week trial period, 88.7% of the individuals randomized to training and 71.4% of those randomized to no-training performed the SARS-CoV-2 PCR test, for a total of 3,016 tests. There was one positive test; in an individual randomized to the training arm (table 2) Clinical disease All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint During the trial period, a total of 106 outpatient contacts for somatic disease were registered for 106 (2.8%) trial participants in the hospitals serving the trial area (table 2) . There were six hospital admissions of trial participants; 4 in the training arm and 2 in the no-training arm. Five of the registered hospital contacts or admissions were unrelated to any COVID-19 associated condition or symptom. One patient was admitted with pulmonary embolism. We contacted the attending physician who after chart review ruled out that the condition was related to COVID-19. Thus, no trial participants in any of the two arms had hospital admissions or outpatient visits for COVID-19 ( figure 1, table 2 ). Out of 91 employees who worked at the training facilities during the trial period and agreed to provide data, 83 (91.2%) were tested for SARS-CoV-2. None were positive. Our trial showed no virus transmission or increase in COVID-19 disease related to opening of trainings facilities providing good hygiene and social distancing routines. The difference in SARS-CoV-2 test positivity between the training and no-training arms was 0.05% (one versus zero cases), well below the predefined non-inferiority margin of 1%. By emergency law, all training facilities were closed in Norway during the pandemic. The closure was reasoned by the assumption that training activity at the facilities would increase the risk of virus transmission between members of the facilities and thus COVID-19 disease among members, staff and the community. However, basic hand hygiene and social distancing measures by securing 1 to 2 meters distance between individuals are well-proven All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 June 24, 2020. The primary concern with COVID-19 is clinical disease, measured as hospital admission, need for ventilator support, and death. As a surrogate, positivity for SARS-CoV-2 RNA is often used. However, high SARS-CoV-2 RNA test positivity in individuals or groups of individuals is not necessarily a surrogate for seriousness of COVID-19 disease in a population, because SARS-CoV-2 infected individuals who do not themselves become seriously ill or who do not transmit the disease to others who become seriously ill, may All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint 13 contribute to achieve immunity in the population and thus contain the disease. Therefore, we measured both SARS-CoV-2 positivity and incidence of COVID-19 disease to understand the relationship of the surrogate outcome with the clinically significant disease outcomes. As our results show, there was no increase in COVID-related disease due to the opening of training facilities. risk factors, who were those who participated in the trial. We believe our trial population is representative of many users of training facilities and the results may thus be applied to other regions and countries (7) . However, it is unclear if our findings also apply to areas with higher COVID-19 incidence rates. Our sample size was based on estimates from prevalence testing in the community for COVID-19 activity. Most individuals in community testing had clinical signs or symptoms suspicious for COVID-19. Thus, in accordance with recent evidence from population sampling in Iceland (7), we assumed considerably higher SARS-CoV-2 rates in our sampling of individuals with no symptoms. This was not confirmed, as our observed rate in the trial was similar to those in the community. Compliance with SARS-CoV-2 testing was slightly higher in the training arm (89%) than in the no-training arm (71%). However, both compliance rates are high and we consider them All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint 14 satisfactory to provide valid results. Disease endpoints in the trial were gathered through complete hospital registries and thus are not prone to self-reporting bias. We did not have any missing data for clinical disease. Finally, the number of individuals who withdrew consent after randomization to the no-training group was small, and there were no participants who withdrew after intervention start. It is important to perform randomized implementation and de-implementation of societal measures with large potential harms and burden for individuals and the population at large. Our study showed that it is feasible to apply rigorous randomized testing of public health measures during an ongoing disease outbreak. We have demonstrated that such testing is doable, and demonstrated that it is safe to re-open training facilities in Norway. The Norwegian government indeed allowed re-opening of all facilities as of June 15, 2020, provided the hygiene and social distancing measures applied in the trial can be followed. We are currently planning new randomized testing towards normal activity at Norwegian training facilities, according to the principles of rapid-cycle randomized implementation for health care services (8, 9) . Medical Microbiology at Oslo University Hospital for virus testing, the hospitals in the Oslo region for their support with the clinical data, and all staff at the participating training facilities for their engagement and enthusiasm to establish this trial on very short notice. Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe A fiasco in the making? As the coronavirus pandemic takes hold, we are making decisions without reliable data Trust, threats, and consequences of the COVID-19 pandemic in Norway and Sweden: A comparative survey Spread of SARS-CoV-2 in the Icelandic Population Creating a Learning Health System through Rapid-Cycle, Randomized Testing Improving cancer screening programs No reuse allowed without permission. (which was not certified by peer review) 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 In addition to the authors mentioned above, the TRAiN study group comprised the following All rights reserved. No reuse allowed without permission.(which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint All rights reserved. No reuse allowed without permission.(which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint All rights reserved. No reuse allowed without permission.(which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint 18 1 Times trained at training facility during study period (data were available from four of the five facilities) All rights reserved. No reuse allowed without permission.(which was not certified by peer review) 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 June 24, 2020. (which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint 20 Figure 1 : Study flowchart and graphical abstract All rights reserved. No reuse allowed without permission.(which was not certified by peer review) 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 June 24, 2020. . https://doi.org/10.1101/2020.06.24.20138768 doi: medRxiv preprint