Key papers comparing infection rates of COVID-19/SARS-CoV-2 in PEI versus the general population
| Study | Key results | Overall assessment of the quality of evidence | Infection rates in PEI vs general population |
| A large outbreak of COVID-19 in a UK prison, October 2020 to April 2021. Adamson et al41 | Crude attack rate in residents 12% (95% CI 9% to 15%). Period-incidence of 60.4 cases per 1000 population for residents, lower than that of general population | Low | ↓ |
| A study of SARS-CoV-2 outbreaks in US federal prisons: the linkage between staff, incarcerated populations and community transmission. Towers et al37 | Incarcerated population showed a comparative 4.32 risk ratio of per capita COVID-19 rates vs the general population (p≤.001). Significant correlation demonstrated between per capita rates in the outbreaks among the incarcerated population and the community, despite stoppage of visitation over the time period of the study. Significant difference in per capita rate demonstrated between levels of facility security level: high<minimum<medium<low Decarceration was significantly associated with a decrease in incarcerated per capita rates during the winter wave (p=.015) but not during the summer wave | Medium | ↑ |
| A time-series analysis of testing and COVID-19 outbreaks in Canadian federal prisons to inform prevention and surveillance efforts. Blair et al29 | 3% prevalence COVID-19 of total incarcerated population in comparison to 0.2% in general population | Low | ↑ |
| Adverse SARS-CoV-2-associated outcomes among people experiencing social marginalisation and psychiatric vulnerability: a population-based cohort study among 4,4 million people. Nilsson et al42 | Reduced risk of positive SARS-CoV-2 PCR test in PEI vs general population: aIRR 0.84 (95% CI 0.80 to 0.88) p<.0001. Reduced rate of testing in PEI OR 0.47 (0.46–0.48) p<.0001 | Medium | ↓ |
| Association between prison crowding and COVID-19 incidence rates in Massachusetts prisons, April 2020–January 2021. Leibowitz et al32 | COVID-19 incidence rate in incarceration facilities was 965/100 000 compared with 150/100 0000 person weeks in general population during study period, incidence lower in facilities that were less full and had higher percentage of people in single cells | Low | ↑ |
| Characteristics of persons with secondary detection of severe acute respiratory syndrome coronavirus ≥90 days after first detection, New Mexico. Hicks et al51 | When adjusted in multivariable model, staff or residents of incarceration facilities had higher rates of secondary SARS-CoV-2 detection (aOR 4.7 CI 1.8 to 12.1) | Low | ↑ |
| COVID-19 case and mortality rates in the Federal Bureau of Prisons. Toblin and Hagan26 | Crude case rate for BOP 11 710 per 100 000 and 2484 for general USA. Ratio of 4.7× more cases in incarceration facilities | Medium | ↑ |
| COVID-19 cases and deaths in federal and state prisons. Saloner et al27 | Case rate for PEI was 5.5× higher than in general population | Low | ↑ |
| COVID-19 cases and testing in 53 prison systems. Lemasters et al31 | 34 prison systems had higher case rates per thousand than general population | Low | ↑ |
| COVID-19 community spread and consequences for prison case rates. LeMasters et al34 | Mean active case rate of 427 per 100 000 in the incarcerated population compared with a rate of 215 per 100 000 in the general population. When community rates reached the threshold case rate of at least 50 per 100 000, there was an immediate increase in the COVID-19 case rate in incarceration facilities by 118.55 cases per 100 000 (95% CI −3.71 to 240.81). No significant difference between community COVID-19 rates in counties with and without an incarceration facility | Low | ↑ |
| COVID-19 in prisons: state health care contracting and the pandemic behind bars. Smith and Glidden39 | Lower mean COVID-19 diagnoses per 10 000 (333.20, range 0–1640) compared with the general population (COVID-19 diagnoses per 100 000 1255.32, range, 74.48–20 617.31). Incarcerated residents in states who provide at least some healthcare from Department of Correction staff (as opposed to purely privately contracted healthcare) showed significantly reduced COVID-19 diagnosis rate per 10 000 (b=−448.70, p=.01). Average expenditure on healthcare per incarcerated resident had no significant effect on COVID-19 rates or mortality | Low | ↓ |
| COVID-19 incidence and mortality in federal and state prisons compared with the US population, April 5, 2020, to April 3, 2021. Marquez et al28 | Crude case rate was 30 780/100 000 for PEI and 9350/100 000 for general population, incident ratio of 3.3 (95% CI 3.3 to 3.3) for incarcerated population | Low | ↑ |
| COVID-19 infection among incarcerated individuals and prison staff in Lombardy, Italy, March 2020 to February 2021. Mazzilli et al35 | The study demonstrated a higher relative risk of COVID-19 infection in incarcerated residents than the general population (first wave: RR 1.30; 95% CI 1.06 to 1.58 second wave RR 3.91; 95% CI 3.73 to 4.09). A lower average weekly positivity rate per 100 individuals was noted in incarcerated individuals vs the general population however (first wave: 1.76 range, 0.00–10.68 vs 9.55 range, 1.21–37.50 second wave: 4.46 range, 0.00–17.92 vs 8.71 range, 1.16–20.71 | Low | ↑ |
| COVID-19 outbreak in a large penitentiary complex, April-June 2020, Brazil. Gouvea-Reis et al33 | Higher COVID-19 incidence rate in the case study incarcerated population vs the general population of the Brasilia region (1832 cases/100 000 persons vs 47 cases/100 000) Shorter mean serial case interval at 2.51 days (SD 1.21) in case study facility vs general Brazil population (figures for comparison not documented) | Low | ↑ |
| Epidemiology of coronavirus disease 2019 at a County Jail-Alameda County, California, March 2020–March 2021. Marusinec et al36 | Total incidence rate during the investigation period was 280/1000 which was 5.29× (95% CI 4.87 to 5.75) higher than Alameda county, younger, Hispanic/Latino and black people had higher percentage of positive tests | Low | ↑ |
| Epidemiology of COVID-19 among incarcerated individuals and staff in Massachusetts jails and prisons. Jiménez et al80 | Incidence of COVID-19 was 44.3/1000 for PEI, 2.91 times higher than Massachusetts general population and 4.8 times greater than USA general population, systems with higher testing rates had higher case rates, case incidents were higher among systems that released a lower proportion of their baseline population | Low | ↑ |
| Epidemiology of COVID-19 in prisons, England, 2020. Rice et al38 | Crude incidence in PEI in England was 988/100 000, compared with 935/100 000 in general population (not statistically different). Higher percentage of positive tests for black (6.4% vs 3.3%) and Asian (7.8% vs 7.5%) ethnic groups compared with the general population | Low | ⇆ |
| Health management in Italian prisons during COVID-19 outbreak: a focus on the second and third wave. Vella et al40 | Prevalence of SARS-CoV-2 infection among PEIranging from 0.19% to 1.94% (mean 1.02%, SD 0.51%). Authors state lower prevalence than Italian general population but data supporting this not presented. Time lag on average of 1–2 weeks between peaks of infection rates in the general population and the incarcerated population on cross-correlation time lag plot | Low | ↓ |
| SARS-CoV-2 seroprevalence in the adult detainees of the Paris area in 2021: a multicenter cross-sectional study. Mellon et al43 | 18.2% (95% CI 16.9 to 19.4) of incarcerated population, adjusted for age/sex, were seropositive over the entire study period. Over the week 08–14 February 2021 incarcerated population seropositivity was 18.4% (95% CI 16.8 to 20.1) compared with 20.6% (95% CI 16.6 to 24.9) in the general Paris population. Statistically significant factors independently associated with seropositivity in males=lower number of cigarettes per day (p<.0001) and higher number of inmates per cell (p=.0008. In females=younger age (p=.0002) and lower number of cigarettes per day (p=.0216) | Low | ↓ |
| Seroprevalence and risk factors for SARS-CoV-2 among incarcerated adult men in Quebec, Canada 2021: a cross-sectional study. Kronfli et al44 | 22% of participants were seropositive over the study period. This compared with 13.75% in the comparative general population sample of Montreal blood donors. Factors with a statistically significant association with seropositivity=time spent incarcerated (‘most time’: aPR, 1.47; 95% CI 1.01 to 2.12; ‘all time’: aPR, 2.17; 95% CI 1.53 to 3.07), employment during incarceration (aPR, 1.64; 95% CI 1.28 to 2.11), shared meal consumption during incarceration (‘with cellmates’: aPR, 1.46; 95% CI 1.08 to 1.97; ‘with sector’: aPR, 1.34; 95% CI 1.03 to 1.74), and incarceration post in-prison outbreak (aPR, 2.32; 95% CI 1.69 to 3.18) | Low | ↑ |
| Testing lags and emerging COVID-19 outbreaks in federal penitentiaries: a view from Canada. Blair et al53 | COVID-19 prevalence was 1.2% in incarceration facilities compared with 0.1% in general population, COVID-19 prevalence higher among women’s incarceration facilities | Low | ↑ |
aIRR, adjusted Incidence Rate Ratio; aOR, adjusted Odds Ratio; aPR, adjusted prevalence ratio; BOP, Bureau of Prisons; PCR, Polymerase Chain Reaction; PEI, people experiencing incarceration.