key: cord-0967776-swob5i9e authors: Gordon, C. L.; Trubiano, J. A.; Holmes, N. E.; Chua, K. Y.; Feldman, J.; Young, G.; Sherry, N. L.; Grayson, M. L.; Kwong, J. C. title: Staff to staff transmission as a driver of healthcare worker infections with COVID-19 date: 2020-12-30 journal: nan DOI: 10.1101/2020.12.25.20248824 sha: a3d1a869301a25e9351e8e8e108500e39a9420db doc_id: 967776 cord_uid: swob5i9e Objectives: To investigate the COVID-19 infections among staff at our institution and determine the interventions required to prevent subsequent staff infections. Design: Retrospective cohort study Participants and setting: Staff working at a single tertiary referral hospital who returned a positive test result for SARS-CoV-2 between 25 January 2020 and 25 November 2020. Main outcome measures: Source of COVID-19 infection. Results: Of 45 staff who returned a positive test result for SARS-CoV-2, 19 were determined to be acquired at Austin Health. Fifteen (15/19; 79% [95% CI: 54-94%]) of these were identified through contact tracing and testing following exposures to other infected staff and were presumed to be staff-staff transmission, including 10 healthcare workers (HCWs) linked to a single ward that cared for COVID-19 patients. Investigation of the outbreak identified the staff tearoom as the likely location for transmission, with subsequent reduction in HCW infections and resolution of the outbreak following implementation of enhanced control measures in tearoom facilities. No HCW contacts (0/204; 0% [95% CI: 0-2%]) developed COVID-19 infection following exposure to unrecognised patients with COVID-19. Conclusions: Unrecognised infections among staff may be a significant driver of HCW infections in healthcare settings. Control measures should be implemented to prevent acquisition from other staff as well as patient-staff transmission. Austin Health is a tertiary hospital in Victoria, Australia that operates >900 beds and includes 64 state-wide transplantation and clinical specialty services. During the pandemic, several wards 65 were repurposed as dedicated areas to cohort patients with COVID-19 and were designated 66 "COVID wards". Maximum patient occupancy for COVID wards was reduced from standard 67 capacity to minimise density quotients for shared rooms, while maintaining sufficient staffing 68 to accommodate the increased time requirements for changes of PPE and additional cleaning. 69 Use of PPE was in accordance with state (DHHS) guidelines 9, 10 and included a "PPE spotter" 70 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. ; https://doi.org/10.1101/2020.12.25.20248824 doi: medRxiv preprint tested for SARS-CoV-2 at baseline and day 11 after last exposure to an infected individual, 95 with additional testing performed if they became symptomatic. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. ; https://doi.org/10.1101/2020.12.25.20248824 doi: medRxiv preprint Ward A is a 32-bed general hospital ward repurposed in March 2020 to provide care to 121 patients with confirmed COVID-19, including eight single occupancy rooms, and capacity 122 reduced from 4 to 1-2 patients per room in six shared patient rooms. The ward's Heating, 123 Ventilation and Air Conditioning (HVAC) system was modified to supply 100% fresh air 124 intake and exhaust return air from the ward outside the building to avoid recirculation. Air 125 flow was adjusted to achieve a net negative pressure differential of 3-4 Pa to establish inward 126 air movement to the ward from adjoining corridors. At the time of the detection of the first 127 staff case on Ward A, the PPE worn by staff entering Ward A included P2/N95 respirators (a 128 change made prior to the formal updates in state and national guidelines in response to 129 increasing HCW cases at other healthcare institutions and aligned with the decisions at those 130 institutions at the time 1, 9 ), face shield, isolation gown, and gloves. Most nursing and cleaning 131 staff worked exclusively on Ward A and were experienced in caring for patients infected with 132 COVID. Ward A shared the same PPE donning room and staff bathroom with another 133 adjacent COVID ward (Ward B), but had separate doffing areas (located immediately prior to 134 exiting each ward) and dedicated staff tearooms outside the clinical areas. 135 The first case identified on Ward A returned a positive PCR test to SARS-CoV-2 in late July 137 2020 after developing mild respiratory symptoms ( Figure 3 and Table 1 ). Testing of contacts 138 identified Case 2, who had tested negative on two occasions in the previous week after 139 exposure to a positive HCW on another ward in mid July. Case 2 acknowledged having 140 respiratory symptoms when interviewed with an interpreter, and had worked six shifts during 141 the infectious period. All staff who had worked on Ward A during the infectious period were 142 tested upon detection of Case 2, and again at day 7 and 11 after last exposure on the ward. 143 Additional upstream source contacts were also identified and tested. Due to the likelihood of 144 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint the Ward A HCW contacts did not use the staff tearoom at the same time as Case 2, and 155 included medical and allied health staff who worked the same hours on Ward A but who used 156 different break facilities, and night-shift nursing staff who did not work the same or is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. ; https://doi.org/10.1101/2020.12.25.20248824 doi: medRxiv preprint the exposure period with colleagues who worked on Ward A, and was subsequently linked to 170 the Ward A outbreak. 171 In response to the outbreak, staff break room policies were revised to apply stricter controls 173 to mitigate the potential risk of transmission (Table 2) . Following the return to work of 174 HCWs after furlough, ward occupancy was increased again to capacity with patients with Of the other nine staff cases presumed to be acquired at our institution, eight were HCWs. 184 The non-clinical staff member likely acquired COVID-19 when using the same office as a 185 HCW who had returned to work after completing a secondment assisting with the outbreak 186 response at a residential aged care facility. This HCW, who presumably acquired COVID-19 187 while on secondment, had used the office for non-clinical duties two days prior to developing 188 symptoms, and had vacated the office immediately prior to handing over to the non-clinical 189 staff member. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. ; https://doi.org/10.1101/2020.12.25.20248824 doi: medRxiv preprint remaining five HCW cases outside of Ward A were identified from testing of contacts after 195 investigation of staff cases, and were presumably linked to these other staff. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. ; https://doi.org/10.1101/2020.12.25.20248824 doi: medRxiv preprint involvement of Case 8 in the cluster, who had no contact with patients infected with COVID-221 19, but who used the Ward A staff tearoom, points towards transmission in this space. In 222 contrast, staff working on Ward A who did not use the tearoom did not test positive in the 223 outbreak investigation. Thirdly, the incidence of HCW infections significantly reduced after 224 changes were made to the protocols around use of staff tearooms, without any change in PPE, 225 ventilation or patient occupancy of the COVID ward. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. ; https://doi.org/10.1101/2020.12.25.20248824 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. ; https://doi.org/10.1101/2020.12.25.20248824 doi: medRxiv preprint Elimination - • Limited time in tearoom to 15 minutes for consumption of food/drink and encouraged remainder of break to be spent elsewhere while wearing a face mask • Reinforced designation of break rooms to specific wards, with mixing of staff working across different clinical areas discouraged • Removed excess furniture and placed markings/signs to reinforce physical distancing • Signs placed on break room doors to indicate maximum occupancy • Improved record keeping and auditing of break room use to monitor adherence to organisation policy • Emphasised importance of donning face masks as soon as finished eating/drinking . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. ; https://doi.org/10.1101/2020.12.25.20248824 doi: medRxiv preprint a 4 results were initially "indeterminate" but were negative on repeat testing of the same sample and subsequent samples b Includes 9 cases due to exposure to a household contact, 1 case related to overseas travel, 4 cases unknown exposure but who were working from home and had not been on site c No healthcare contact with confirmed or suspected cases; unknown community exposure or exposure in other healthcare facilities d Includes 4 staff who did not work at Austin Health in the 2 weeks prior to their infectious period, 4 staff who cared for patients with COVID-19 at other healthcare facilities but not at Austin Health, and 1 staff member who cared for COVID-19 patients in precautions at Austin Health but was tested as part of an outbreak investigation involving multiple staff and patients at another healthcare facility. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. ; https://doi.org/10.1101/2020.12.25.20248824 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. ; https://doi.org/10.1101/2020.12.25.20248824 doi: medRxiv preprint Table 1 ). Staff-staff linkages among staff diagnosed with COVID-19. HCWs involved in the Ward A outbreak are shown in black and numbered, with Case 2 presumed to be the primary case. Linkages are coloured based on contact assignment through contact tracing following each exposure. Case 10 was identified while the Ward A outbreak was still considered "active", but was not thought to be linked to the other staff. Two additional HCWs (shown in grey) who also worked in aged care facilities with active outbreaks have been included as the initial presumed source for subsequent staff infections. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Probable Close contact of a confirmed COVID-19 case in the community (with preceding or synchronous infection) In quarantine due to exposure from an active outbreak in a community setting Possible From a community local government area with high prevalence (defined as >200 active cases per million population = DHHS definition when initiating "hot spot" asymptomatic community testing in July 2020) Works in or visited a location with an active outbreak, but not in quarantine A household member in quarantine due to close contact or exposure in an active outbreak No known contacts or risk factors From a community local government area without high prevalence (defined above) . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted December 30, 2020. ; https://doi.org/10.1101/2020.12.25.20248824 doi: medRxiv preprint A 277 hospital-wide response to multiple outbreaks of COVID-19 in health care workers: lessons 278 learned from the field Monitoring approaches for health-care workers during the COVID-19 pandemic Test and quarantine for 14 days, plus test day 11 post exposure ModerateTest and furlough for 14 days, plus test day 11 post exposure CasualTest at baseline and day 11 post exposure Very lowNo contact with source but worked in same area. No action required. No contact with source and worked in a different area. No action required.