The conduct of this rapid review was guided by the Rapid Review Guide for Health Policy and Systems [
7] published by the World Health Organization (WHO) and reported according to the standards of the PRISMA Checklist [
8] (Additional file
1: Appendix 1). A protocol for the review was registered in the PROSPERO database prior to the start of data abstraction (PROSPERO ID:
CRD42020181993). A prior review examining infection control guidelines for any form of viral respiratory illness in long-term care leveraged in this work was commissioned by the World Health Organization and posted on a pre-print server (Supplementary File 2) [
9]. The results reported here were shared with the Canadian Frailty Network (review commissioner) in the form of a summary report and one-page brief. Our paper has not been published in a peer-reviewed journal.
Search strategy and selection criteria
In order to accommodate the rapid timeline (10 working days) requested by the review commissioners, a combination of automated and manual search and title/abstract screening methods were used to gather evidence. A comprehensive literature search for the EMBASE database was developed by an experienced librarian and peer-reviewed by a second using the PRESS checklist [
10] (Additional file
1: Appendix 2). Grey literature sources (i.e. difficult to locate or unpublished) such as COVID-19-focused evidence gathering services (e.g. EPPI Mapper, COVID-END), as well as guideline producers/repositories (e.g. NICE guidance, ECRI), were hand searched for potentially relevant publications. All other information sources (MEDLINE, Cochrane library, pre-print servers, and clinical trial registries; the full list is available in Additional file
1: Appendix 3) were searched using a supervised machine learning approach, called Continuous Active Learning® (CAL®) [
11]. All sources were searched from inception up to July 31, 2020.
CAL® used two methods to gather citations. For sources where the entire archive can be gathered automatically (e.g. MEDLINE), the archive is searched and processed using a priori inclusion criteria established during protocol development. For archives that are only accessible using keyword searches, specific terms drawn from the previously developed literature search (e.g. terms related to COVID-19 and/or long-term care) were applied and the ensuing results were processed using the same inclusion criteria. CAL® identifies titles and abstracts most likely to meet the inclusion criteria based on an iterative process where search results are compared against relevant citations that have previously been identified. The search results from EMBASE and grey literature sources were compared against the CAL® results in order to remove duplicates. The results were then screened against the same inclusion criteria by a human reviewer.
All citations that were included by the title/abstract screening process were then passed on to full-text review. A standardized screening form based on the inclusion criteria was developed and calibrated using a pilot test of 10 full-text articles conducted by the review team. A single human reviewer evaluated the full text of each article using the standardized form, and all excluded full-text articles were screened again by a second, independent reviewer. There was full agreement between both reviewers on the excluded articles.
The inclusion criteria established for this rapid review used the PICOS/T framework as follows:
➢ Population: Individuals aged 60 years and above residing in long-term care facilities (e.g. nursing home, long-term care hospital/facility, skilled nursing facility, convalescent home, assisted living facilities). The definition of long-term care that was used is as follows: “Long-term care homes are home-based health care facilities designed for adults who need access to on-site 24-hour nursing care, frequent assistance with activities of daily living (i.e., eating, bathing, toileting, etc.) and monitoring for safety or well-being. They are also known as nursing homes, charitable homes, or municipal homes for the aged” [
12,
13].
➢ Interventions: Any form of infection prevention and control (IPAC), including but not limited to appropriate ventilation, cohorting equipment, communication, consulting/notifying health professionals, diagnostic testing, environmental cleaning/disinfecting surfaces, droplet precautions, education, access to hand hygiene/hand sanitizer, access to personal protective equipment (PPE; for patients and healthcare providers), policies for visitors, IPAC policies for staff/residents and designated IPAC staff, providing supplies, respiratory hygiene/cough etiquette, smoking cessation, social distancing/isolation/cohorting, surveillance/monitoring/evaluation, antiviral prophylaxis for staff/residents, early mobilization, restrictions on resident movement and transportation, restrictions on visitors, restrictions on travel for healthcare providers, and other long-term care facility staff. Only those measures used to prevent COVID-19, MERS, or SARS were included; measures related to control and prevention of other infections (e.g. vaccination for influenza, oral care to prevent bacterial pneumonia) were excluded. Additionally, interventions related to remuneration/compensation policies for long-term care facility staff, staffing models to maintain staff levels, policies on mixing of staff in long-term care facilities (cohorting), and policies on staff working in more than one long-term care facility were included.
➢ Comparator: One of the interventions listed above or no intervention
➢ Outcomes: Lab-confirmed respiratory infection [primary outcome], symptoms, secondary transmission (e.g. other patients, healthcare workers), goal concordant care, hospitalization, intensive care unit (ICU) admission, mortality
➢ Study designs: clinical practice guidelines (CPGs) and systematic reviews addressing any coronavirus, using the Cochrane definition of a systematic review [
14]. Primary human studies of all designs (e.g. experimental studies, quasi-experimental studies, and observational studies, excluding case series) that involved patients with COVID-19 only (not including SARS or MERS) were included.
➢ Time periods: All periods of time and duration of follow-up were included.
No other limitations were imposed on the search or study selection process. Both peer-reviewed and pre-print papers were eligible for inclusion, as were papers written in languages other than English.
Data collection
A standardized data collection form was developed to capture the following information from primary studies: study characteristics (e.g. duration of follow-up, study design, country of conduct, multi-centre vs. single site), patient characteristics (e.g. mean age, age range, comorbidities), intervention details (e.g. type of intervention, duration and frequency of intervention, timing of intervention), comparator details (e.g. comparator intervention, duration and frequency of intervention, timing of intervention), and outcome results (e.g. lab-confirmed respiratory infection [primary outcome], symptoms, secondary transmission, hospitalization, ICU admission, mortality) at the longest duration of follow-up. A separate standardized form was developed to capture relevant information from clinical practice guidelines, including the method used for the guideline, the intended scope of the guideline, and the specific recommendations and level of evidence for each. All data were collected by a single reviewer and verified by a second, independent reviewer.