A scoping review of the impacts of COVID-19 physical distancing measures on vulnerable population groups

By August 10, 2022, the COVID-19 pandemic had resulted in approximately 586.5 million cases and 6.4 million deaths worldwide, with several countries among the hardest hit. Governments deployed multiple non-pharmaceutical policy instruments to curb transmission, notably physical (social) distancing measures such as lockdowns, school closures, and restrictions on gatherings, as defined by WHO. Although these measures were intended to break transmission chains, they also carried costs and unintended consequences that could be particularly severe for vulnerable populations. Prior research examined the effectiveness and impacts of distancing measures in general populations, but there was a lack of consolidated evidence on how these measures affected vulnerable groups and insufficient understanding of targeted “ringfencing” measures designed to protect them (e.g., restricting visitors to nursing homes and prisons). This study conducted a scoping review to synthesize global evidence on the negative impacts of physical distancing on vulnerable populations and to identify ringfencing measures implemented to mitigate these impacts.

The paper notes that while multiple systematic reviews have evaluated the effectiveness and impacts of physical distancing measures in general populations, there has been no systematic consolidation focused specifically on vulnerable groups and the design and implementation of ringfencing measures for them. The broader literature also debates the effectiveness of school closures and conditions for safe reopening (e.g., hygiene, masking, distancing, testing, tracing, ventilation, cohorting), indicating that with appropriate protocols, reopening did not necessarily increase transmission in many settings. Prior work has highlighted socio-economic disparities in the feasibility and consequences of distancing (e.g., digital divides, inability to work from home, crowded housing), suggesting disproportionate burdens on disadvantaged groups. This review addresses these gaps by collating evidence across 49 countries and 11 vulnerable population categories.

Design: Scoping review following PRISMA guidelines to examine (i) which physical distancing measures were implemented and how they negatively impacted vulnerable populations, and (ii) what ringfenced measures were designed and implemented to protect them. Data sources and search strategy: Initial comprehensive searches were conducted from March to April 2021 across ten databases: PubMed, Scopus, Web of Science, ProQuest, ProQuest Coronavirus Research Database, Embase, ERIC, LITCOVID/OVID, Cochrane Database of Systematic Reviews, and WHO COVID-19 Global literature on coronavirus disease. Search strings combined COVID-19 terms with physical distancing and policy terms (see Table 4). An updated targeted search in June 2022 identified additional studies from 2021 onwards; duplicates were removed. Eligibility criteria: Inclusion—studies focusing on vulnerable populations; examining physical distancing measures from public policy and/or legal perspectives; peer-reviewed articles (empirical, conceptual, reviews), policy briefs, reports, editorials, commentaries, perspectives, and letters; jurisdictional unit of analysis; quantitative, qualitative, or mixed methods; full-text English publications from Nov 2019 to Jun 2022. Exclusion—studies without impacts on vulnerable populations; purely clinical studies without policy/law dimensions; pre-Nov 2019; non-English or inaccessible full texts. Screening and selection: Titles/abstracts screened against inclusion/exclusion criteria informed by a vulnerability framework (Table 3). The first author screened all records; the third author independently screened more than half, iterating to achieve <10% discrepancies; disagreements resolved among all authors. Full texts of eligible records were retrieved for assessment. PRISMA flow: Databases—31,827 records identified; 19,286 duplicates removed; 12,561 titles/abstracts screened; 198 full texts assessed; 19 excluded. Other methods—7,989 records identified; 2,586 duplicates removed; 5,393 titles/abstracts screened; 5,392 excluded; 87 full texts assessed; 1 excluded. Total included studies: 265 (198 from 2021 searches; 68 from 2022 searches). Overall, 39,816 records screened. Data extraction and synthesis: A predefined template guided extraction. All authors piloted extraction on the first 10% of articles for consistency; the first author extracted remaining data; others validated a random subset. Data were managed in Excel 16. Framework synthesis (five stages: familiarisation, framework selection, indexing, charting, mapping/interpretation) grouped findings by vulnerable population categories, enabling structured analysis across heterogeneous study designs. Context: Studies spanned 49 countries across five continents. Vulnerable groups covered: children/students (n=96), low-income populations (n=58), older people (n=37), victims of domestic violence (n=16), people with disabilities (n=15), migrant workers (n=14), refugees (n=14), people from sexual and gender minorities (n=11), ethnic minorities (n=10), sex workers (n=9), people in prison (n=7), and five studies on vulnerable groups in general.

Overall: Physical distancing measures, while effective for transmission control, had substantial unintended negative consequences for vulnerable populations, including prolonged loneliness, mental distress, unemployment and income loss, food insecurity, widened inequalities, and disrupted access to social support and health services. Evidence covered 265 studies in 49 countries. By group (selected highlights): - Older people (n=37): Distancing mitigated infections and deaths, but increased risks of cardiovascular, autoimmune, and neurocognitive disease due to delayed care and reduced hospital service utilisation; worsened mental health and loneliness. Effective measures in LTC facilities included early screening, detection, contact tracing, and remote monitoring; ICT supported social connectivity and wellbeing. - Children/students (n=96): School closures led to learning disruption, decreased social interactions, mental health issues, and widened disparities by socio-economic status due to digital divides, lack of conducive home environments, and loss of school-based services. Remote learning strained teacher capacity and increased workloads. With robust health/safety protocols (masking, distancing, screening, cohorting, ventilation, reduced class sizes, staggered schedules, testing/tracing, quarantining, hybrid options), school reopening often did not increase transmission. Supports included printed materials, TV/radio content, targeted instruction, devices/connectivity assistance, helplines, food programs, and education funds/stimulus. - Low-income populations (n=58, 19 countries): Barriers to strict adherence due to overcrowded housing, lack of sanitation and clean water, inability to work from home, absence of job protection/paid leave, reduced employer-sponsored health coverage, limited access to government aid, and resource constraints for testing/isolation/tracing. Lockdowns were less effective in low-income contexts. Mental distress was prevalent. Context-specific measures required: intensive screening/testing/tracing/isolation; provision of affordable masks, water, sanitiser/soap; decongestion; expanded economic support (unemployment insurance, paid medical leave, reemployment services); strengthened government-community communication; NGO-led community efforts (food aid, information dissemination, contact tracing facilitation, quarantine support). - Migrant workers (n=14): Experienced job/income loss, food insecurity, crowded living arrangements with high exposure risk, limited access to healthcare, housing instability, and mental stress/loneliness. Often excluded from social benefits. Example: Singapore implemented extensive testing, quarantine facilities, employer subsidies to pay salaries, connectivity/phone cards, and free medical care; NGOs provided multilingual information and counselling. - People in prison (n=7): Institutional vulnerability and overcrowding heightened outbreak risks; increased isolation exacerbated mental health issues and contributed to unrest/riots in several countries. Mitigations included decarceration (home confinement/community supervision), rapid detection/quarantine, enabling virtual/video visits (e.g., tablets), and regular information sessions to reduce anxiety. - People with disabilities (n=15): Reduced social connections and disrupted access to healthcare, medications, rehabilitation, and transport; adults with mobility disabilities faced barriers to services; children with disabilities encountered remote learning challenges due to lack of specialised technologies/services and exacerbated behavioural issues; supports included device loans, financial assistance, tailored remote services, and caregiver engagement. - Sex workers (n=9): Reduced access to STI testing/treatment and ART; limited access to social safety nets; economic hardship and shifts to online work; community-based organisations provided food, financial aid, PPE/hygiene supplies, ART distribution via community-led systems, and information via messaging apps. - Victims of domestic violence (n=16): Increased incidence and reporting (with likely under-reporting due to shutdowns of support channels and transport barriers); children at higher risk when homebound; responses included public information campaigns, expanded helplines/online support, exemptions to travel to shelters, use of hotels as shelters, and extended accommodations and counselling by educational institutions. - Refugees (n=14): Overcrowded conditions impeded distancing/self-isolation; disruptions to food, sanitation, and medical services; exacerbated mental health issues (including retraumatisation); measures included mask distribution and education, mobility limits, sectoring/quarantine areas, rapid detection/isolation, and tele-mental health/outreach by social workers. - Ethnic minorities (n=10): Greater difficulty adhering to distancing due to overcrowded housing and adverse employment; worsened financial concerns and mental health; language barriers impeded understanding/compliance; actions included translated/visual materials and special education supports for non-English learners. - People from sexual and gender minorities (n=11): Disrupted access to essential medical services including HIV testing/treatment; increased mental health burdens due to isolation from supportive networks and stigma; telehealth interventions and NGO-led supports (online counselling, chats, material aid) were utilised. Quantitative synthesis landmarks: - Records screened: 39,816; studies included: 265; countries covered: 49. - Distribution of studies: children/students (96), low-income (58), older people (37), domestic violence (16), disabilities (15), migrant workers (14), refugees (14), sexual and gender minorities (11), ethnic minorities (10), sex workers (9), people in prison (7), and 5 general-vulnerable-group studies. Cross-cutting: Mental distress common across groups; distancing reduced utilisation/access to essential non-COVID health services; effectiveness of ringfenced measures was discussed but empirical evaluations remain limited.

Findings demonstrate that universally applied physical distancing measures can disproportionately harm vulnerable populations via employment/income loss, food insecurity, disrupted access to essential health and social services, social isolation and worsened mental health, and learning losses that widen socio-economic gaps. These outcomes directly address the research questions by clarifying the nature and extent of negative impacts across 11 vulnerable groups and by cataloguing ringfenced measures implemented to mitigate harms. The review underscores the need for effective, accessible public communication tailored to socially isolated groups (e.g., prisons, non-native language communities) and for integrating ringfenced supports—such as telehealth, targeted financial assistance, education technology and services, and community-based outreach—into broader pandemic control strategies. Policymakers must strategically package measures to balance competing objectives (transmission control, economic stability, and equity) and ensure that interventions are context-specific, especially in low-resource settings where standard distancing may be less feasible or effective. The limited evidence on the effectiveness of ringfenced measures highlights a critical need for robust evaluations to inform optimal policy design and enhance social inclusion and health equity.

This scoping review consolidates global evidence on how COVID-19 physical distancing measures negatively impacted vulnerable populations and documents ringfenced strategies used to mitigate harms. The review fills a gap by centering vulnerable groups across 49 countries and 11 categories, revealing common adverse consequences—unemployment/income loss, heightened social isolation and mental health burdens, and disrupted access to essential non-COVID services. It highlights the intersectional nature of vulnerabilities (economic, health, social, institutional, and cognitive/communicative) and the exacerbation of pre-existing inequities. Policy implications include the need to: (a) leverage digital technologies for social connection and telehealth; (b) ensure continuity of essential health services; (c) provide comprehensive financial and social support packages (cash/in-kind assistance, unemployment benefits, reemployment services, and utility relief); (d) strengthen targeted public communication and outreach; (e) enhance the capacity of education systems for inclusive remote/hybrid learning; and (f) bolster public finance to sustain such measures and prepare for future emergencies. Future research should rigorously evaluate the impacts of ringfenced interventions and examine jurisdictional differences to guide equitable, context-appropriate policy mixes.

Two main limitations are noted: (1) Some relevant studies may have been missed due to heterogeneity in the definitions and conceptualisation of vulnerabilities and vulnerable populations. (2) The review did not analyse jurisdictional differences in the negative impacts experienced by the 11 vulnerable groups, which presents an avenue for future research.