Skip to main content
Download PDF
  • Systematic Review
  • Open access
  • Published:

The effectiveness of non-exposure to incarceration in preventing COVID-19 and mitigating associated events: a systematic review and meta-analysis

BMC Public Health volume 25, Article number: 206 (2025) Cite this article

Abstract

Background

For a long time, the penalty of imprisonment has been studied and criticized as ineffective in achieving the goals of resocialization and rehabilitation of offenders, and studies have associated incarceration with increased prevalence of disease. In response to the COVID-19 pandemic, the World Health Organization recommended decarceration as a prevention measure. The aim of this review was to analyze the effectiveness of non-exposure to incarceration in preventing COVID-19 and mitigating associated events.

Methods

We conducted a systematic review and meta-analysis of observational studies comparing the adult general population (GP) and incarcerated population (IP).

Results

We identified 1,334 publications without duplicates and extracted data from 22 studies. We found that COVID-19 incidence was 61% lower in the GP (RR = 0.39 [0.34, 0.45], p < 0.0001). Non-exposure to incarceration was associated with lower age- and sex-adjusted mortality (RR = 0.36, [0.27, 0.49], p < 0.0001). We did not find standardized data on age-adjusted case fatality. The hospitalized GP was older and showed a higher rate of obesity than the hospitalized IP; however, no statistically significant differences were found between the populations for admission to intensive care (RR = 0,91 [0.74, 1.13], p = 0.41) and hospital mortality (RR = 0.81 [0.54, 1.23], p = 0.32). Prevalence of the use of invasive mechanical ventilation was 23% lower in the GP (RR = 0.77 [0.70, 0.84, p < 0.0001).

Conclusion

Non-exposure to incarceration can be a strategy for preventing the spread of COVID-19 and reduces COVID-19 mortality in younger populations. Despite differences in age distribution and presence of comorbidities among the hospitalized GP and IP, we did not find any statistically significant differences between the two populations across most of the hospital-related outcomes. These findings should be interpreted with caution because it was not possible to determine a cause-and-effect relationship between the COVID-19 outcomes and exposure to incarceration.

Registration

PROSPERO CRD42023446610.

Peer Review reports

Background

For a long time, the penalty of imprisonment has been studied and criticized as ineffective in achieving the goals of resocialization and rehabilitation of offenders [1,2,3,4,5]. Despite high cost of imprisonment worldwide, estimated at 62.5 billion dollars per year [5], prisons can adversely affect prisoner health due to limited access to natural lighting and ventilation, poor diet, sedentarism and idleness, restricted or lack of access to health services, violence, overcrowding, inadequate access to water, and poor sanitary condition. Prisons can therefore constitute a hotbed of disease [6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23].

Prisons can also serve as a setting where important health services can be accessed, as the wide-ranging human rights framework requires governments to ensure minimum prison standards, guaranteeing that people who are incarcerated are provided health care that is equivalent to that available in the community [18, 19]. However, studies show that health of people in prison is poorer than that of the general community [17, 19, 22, 23] and that exposure to incarceration doubles the risk of premature death [22]. Furthermore, evidence shows a correlation between incarceration rates and disease prevalence [7, 15, 17].

These criticisms highlight the need to place the implementation of alternative measures to imprisonment on the public agenda, both as a way of achieving the penal system’s goals of resocialization and rehabilitation and to rationalize the use of the penalty of imprisonment, guaranteeing rights and improving the health of vulnerable populations [7, 18].

The COVID-19 pandemic and the prison situation around the world once again made prisons a major public health concern. Given that the virus spreads from person-to-person, primarily due to the release of droplets when the person infected sneezes, coughs, or talks, the World Health Organization issued a set of recommendations for prison establishments, including decarceration [24]. Decarceration is defined as the depopulation of prisons, which includes liberating imprisoned people and using alternative measures to incarceration, such as conversion of prison sentences to fines, electronic monitoring and community service [18]. The latter was found to be one of the main strategies adopted by governments for the control and prevention of COVID-19 in prisons in 2020 [25].

The aim of this review was to analyze the effectiveness of non-exposure to incarceration in preventing COVID-19 and mitigating associated events.

Methods

We conducted a systematic review and meta-analysis of observational studies on this topic. This methods section follows the guidance provided in the Cochrane Handbook for Systematic Reviews of Interventions version 6.4 [26]. This review is registered in PROSPERO (International prospective register of systematic reviews; reference code CRD42023446610) and was conducted following the recommendations of the PRISMA checklist (Preferred Reporting Items for Systematic Review and Meta-Analysis) [27].

We included quantitative observational studies that compare COVID-19 outcomes between the general population (GP) and incarcerated populations (IPs) aged 18 years and over from December 2019 (when the first case of COVID-19 appeared) to July 2023 (the date of this review’s protocol). No restrictions were applied to country, language and methodological quality.

We performed searches for observational studies conducted in the following databases: MEDLINE/PubMed, Scopus, Embase, Web of Science, Portal BVS. The search strategies were designed using the descriptors and keywords best suited to each database, as shown in Board S1 of Supplementary Material 1. We also searched for relevant additional studies contained in the included study reference lists.

The efficacy of non-exposure was measured based on GP data. GP was defined as the general population of the country, state or municipality under study. We excluded studies that presented data on populations that may have a history of incarceration or contact with prisons, such as prison staff, health professionals, family members, incarcerated adolescents and/or children, as well as other groups housed in shelters, orphanages, asylums, and the homeless were excluded, unless they included disaggregated data on the groups of interest. In this case, only the data on the groups of interest were extracted.

The primary outcomes were: COVID-19 incidence; age- and sex-adjusted COVID-19 mortality rate; hospital mortality and COVID-19 hospital and intensive care unit (ICU) admission rates. The secondary outcomes were as follows: use of invasive mechanical ventilation and length of hospital stay.

The US Centers of Disease Control and Prevention estimates that the risk of death from COVID-19 among adults aged 65 and over is more than 60 times higher than in people aged 18 to 29 years [8]. Studies [28,29,30] also show that men are at higher risk of death from COVID-19 than women. In addition, people with comorbidities show higher risk of death from complications and death from COVID-19. Gender and age differences among the GP and IP were explored to determine whether it was necessary to adjust COVID-19 mortality and case fatality rates. We did not control for underlying health conditions as a potential confounding factor. These results were analyzed when presented by the studies and are presented in Supplementary Material 3.

The results of the systematic searches were independently assessed by two reviewers to select studies that met the eligibility criteria, consisting of the following stages: screening of titles, screening of abstracts and screening of the full-text versions of articles. In each stage, any disagreements between the reviewers were resolved by a third reviewer. To facilitate the screening process, we used Rayyan, a web-tool and mobile app for systematic reviews [31].

For data extraction we adapted the models found in the Review Manager Web (RevMan Web) knowledge base, a program used to manage data and conduct meta-analyses [32]. The data were extracted and analyzed jointly by two reviewers and revised by a third reviewer. For each included study, the reviewers extracted the following information: publication details, method, study period, number of participants in each of the two groups, location, population characteristics, definition and measurement of incarceration exposure, results, unadjusted and adjusted effect measures, covariables and limitations.

Risk of bias was assessed using The Risk Of Bias In Non-randomized Studies - of Exposure (ROBINS-E) tool developed to assess the risk of bias in observational epidemiological studies of exposure effects. ROBINS-E includes seven domains of bias: due to confounding; in exposure measurement; in selection of study participants; due to post-exposure interventions; due to missing data; in the measurement of outcomes; and in selection of reported results [33].

The tool helps the researcher define confounding factors and exposure measures in the risk assessment planning phase to establish whether there are enough study characteristics or outcomes to perform a comprehensive assessment of risk of bias and evaluate sensitivity [33]. We defined the following confounding factors: age, comorbidity and prison regime (open, semi-open, closed). Exposure was measured based on length of incarceration.

The following measures of effect were used: risk ratio (RR) for dichotomous outcomes (COVID-19 incidence, sex- and age-adjusted COVID-19 mortality, ICU admission rate, use of invasive mechanical ventilation and hospital mortality); and difference between means for the continuous outcomes (mean age on hospital admission).

The effects were measured adopting a significance level of 0.05 and 95% confidence interval (95%CI). Data conversion (for example, confidence interval to standard deviation, rate to number of events) was performed when necessary, using Excel®. Results reported as rates, where conversion to number of events was not possible, were included in the quantitative synthesis as rate ratios [34].

Central tendency and dispersion were measured using R-4.3.1. The meta-analyses were performed using RevMan Web, version 7.6.0. For dichotomous data we used the Mantel-Haenszel method with a random-effects model (M-H, Random). The meta-analyses of the continuous outcomes and measures of association reported by the studies (relative risk or rate ratio) were performed using the inverse variance method with a random-effects model (IV, Random).

In addition to the quantitative analysis of clinical and methodological heterogeneity (tau-squared and chi-squared), we also performed a statistical analysis of inconsistency (I²) [35]. Publication bias was assessed using a funnel plot and Egger’s test for outcomes investigated by a minimum of 10 studies [36].

We performed subgroup analysis to determine whether the results of the analyses changed when the studies were grouped according to methodological characteristics and region. We also considered incarceration rates in the countries where the studies were carried out to test for correlations between rates and health outcomes, which have previously been reported in the literature [7, 15, 17].

Subgroup analysis stratified by age was included in the protocol but common subgroups were not found. However, we performed subgroup analyses stratified by region (Latin America and the Caribbean, North America and Europe) and incarceration rate; type of study; application STROBE checklist (The Strengthening the Reporting of Observational Studies in Epidemiology Statement: guidelines for reporting observational studies) [38]; and control for age in the two populations. The subgroups by region and incarceration rate were grouped as shown in Board S2 of Supplementary Material 1.

For sensitivity analysis we used an alternative method of meta-analysis (fixed-effect model instead of a random-effects model) and the leave-one-out method. The GRADE (Grading of Recommendations Assessment, Development and Evaluation) [37] system was used to rate quality of evidence across critical outcomes, classified as “high”, “moderate”, “low” and “very low”.

Results

Description of studies

We identified 1,334 publications after excluding 1,578 duplicates. A total of 250 publications were selected after title and abstract screening, 28 of which were considered eligible. Six of these were excluded because they did not include the outcomes investigated by this review or were not observational studies, resulting in a final sample of 22 studies (Fig. 1).

Fig. 1
figure 1

PRISMA flow diagram

Full size image

A summary of the study characteristics is shown in Table S3 of the Supplementary Material 1. Of these, 15 (68%) were from the United States (US) [38,39,40,41,42,43,44,45,46,47,48,49,50,51,52], 16 (73%) were cohort studies [38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53] and only three applied the STROBE checklist [42, 47, 50]. Six studies [54,55,56,57,58,59] (27%) were case series presenting data on the two populations but whose aim was not to compare them.

Fifteen of the selected studies [38, 42,43,44, 46,47,48,49, 52, 54,55,56,57,58,59] (68%) were national level and mean study period was 263 days (SD 197). Fourteen studies [39, 40, 45, 48,49,50,51,52,53, 57,58,59] (64%) presented data only from 2020, five [38, 40, 44, 51, 54] (23%) did not report the size of the GP and 3 [38, 44, 51] (14%) did not mention the size of the IP. Only three of the included studies [39, 42, 45] reported comorbidities in the two populations.

Risk of bias in the included studies

Length of incarceration was considered the most suitable method for measuring exposure. However, none of the studies controlled for this measure, and thus ROBINS-E [29] recommended discontinuing the assessment, indicating high risk of bias. For this reason, the risk of bias assessment was not presented as a traffic light plot of the domain-level judgement for each individual outcome, as recommended by in the PRISMA checklist [27].

Synthesis of results

Below we present the summary of findings table and grading of quality of evidence.

COVID-19 incidence

Eighteen studies [38, 39, 41, 44, 46,47,48,49,50,51,52,53,54,55,56,57,58,59] reported number of COVID-19 cases and we considered COVID-19 incidence regardless of the measurement and counting method used (clinical and laboratory, new and accumulative cases). Since 13 studies [41, 46,47,48,49,50, 52, 53, 55,56,57,58,59] presented the study population, meaning it was possible to conduct a dichotomous analysis, the meta-analyses were performed separately.

The synthesis of the results of these studies reveal that COVID-19 incidence was 61% lower in the GP than in the IP (RR = 0.39 [0.34, 0.45], p < 0.00001, I²=100%) (Fig. 2). This statistically significant difference was maintained in the sensitivity analysis undertaken using the fixed-effects model and leave-one-out method.

Fig. 2
figure 2

Forest plot of COVID-19 incidence

Full size image

The tests for differences between subgroups suggest statistically significant effects for the following subgroups: control for age in the two populations (p = 0.01), type of study (p = 0.02), region (p = 0.0005). However, there was high unexplained heterogeneity among the studies within each subgroup (for all subgroups I² = 100%). The validity of the treatment effect estimate for each subgroup is therefore uncertain. Tests for differences between subgroups are presented in Supplementary Material 2. Figure S2.1 presents the test for age control between the two populations; Figure S2.2 presents the test by study type; and Figure S2.3 presents the test by region.

As Richardson et al. [60] suggest, we consider that it is unlikely that the analysis of subgroups: incarceration rate and STROBE checklist will produce any useful findings since there is a large difference between the number of studies involved in each subgroup.

The COVID-19 incidence rates reported in the five studies that presented this data [38, 39, 44, 51, 54] were analyzed as a continuous outcome, with the results showing that non-exposure to incarceration was associated with a reduced the risk of getting COVID-19 by 65% (RR = 0.35 [0.20, 0.63], p = 0.0004, I² = 100%). Figure S2.4 in Supplementary Material 2 presents the meta-analysis of the COVID-19 incidence rate from studies that did not report the study population size. Subgroup analysis was not performed due to the low number of studies.

It is important to highlight however that only Hernando et al. [54], in a study conducted in a European country, found that COVID-19 incidence was higher among the GP than in the IP, except in the period July 1, 2021-September 30, 2021, when incidence was higher among the IP than in the GP (1282.98/100,000 population versus 757.14/100,000 population).

Age- and sex-adjusted COVID-19 mortality

Six studies [42, 43, 45, 46, 48, 52] presented data on proportion of older persons, showing that people age over 65 years accounted for 18.0% of the GP and only 3.0% of the IP (RR = 3.92 [3.35, 4.59], p < 0.00001, I² = 100%). Figure S3.1 in Supplementary Material 3 presents the analysis of the number of people over 65 in each population. Statistical significance and heterogeneity were not affected in the sensitivity analysis performed using the fixed-effects model and leave-one-out method.

Seven studies [39, 42, 43, 45, 46, 48, 52] presented data on sex, showing that 48.9% of the GP and 92.8% of the IP were male (RR = 0.53 [0.53, 0.53], p < 0.00001, I² = 98%) (S3, Fig. 3.2 presents the analysis of number of men in each population). Statistically significant differences and the presence of heterogeneity were maintained in the sensitivity analysis performed using the fixed-effects model and leave-one-out model.

Since risk of death from COVID-19 is higher in older age groups and men and the two populations have different demographic characteristics, especially in terms of age and sex distribution, we believe that the age and sex-adjusted COVID-19 mortality and case fatality rates are more precise than the unadjusted rates for comparing deaths from COVID-19 among the GP and IP. Seven studies [43, 46, 48, 52, 55, 58, 59] reported deaths from COVID-19 for each population; however, only five studies [40, 43, 46, 48, 52] adjusted mortality rates for sex and age. None of the studies adjusted COVID-19 case fatality rates.

The results of the meta-analysis of the studies that adjusted mortality rates for age and sex show that risk of death was 64% lower among the GP than in the IP (RR = 0.36, [0.27, 0.49], p < 0.00001, I² = 96%) (Fig. 3). This statistically significant difference and the presence of heterogeneity were maintained in the sensitivity analysis performed using the fixed-effects model and leave-one-out model. We did not perform a subgroup analysis due to the small number of studies.

Fig. 3
figure 3

Forest plot of age- and sex-adjusted COVID-19 mortality

Full size image

Two studies [43, 48] reported mortality by age group and sex but were not included in the quantitative synthesis because they used different age groups. Both studies found that the age distribution of deaths shifted downward for the IP. Nowotny et al. [43] found that 39% of deaths among the IP were in the 50–64 year group, compared to only 20.6% in the GP, while Toblin [48] reported that 36% of deaths among the IP were in the 55–64 year age group, compared to only 13% in the GP. Similarly, Marquez et al. [40] observed that COVID-19 was associated with a 4.2-year reduction in life-expectancy among the IP, compared to only 1.5 years in the GP.

COVID-19 hospitalization

The hospitalization rate was reported by two studies conducted in Spain [54] and the US [42]. Hernando et al. [54] found that the COVID hospitalization rate was higher among the GP when compared to IP in five follow-up periods. The mean hospitalization rates across the periods were 13.6% (SD = 0.58) for the GP and 6.22 (SD = 0.40) for IP. Montgomery et al. [42] reported hospitalization rates of 49.7% for the GP and 63.5% for IP (p < 0.001).

Only one study [42] presented hospitalization data by age group. In age groups under 75 years, the risk of hospitalization was lower in the GP (< 25 years: RR = 0.69 [0.53, 0.91], p = 0.008; 25–44 years: RR = 0.58 [0.55, 0.62], p < 0.00001; 45–54 years: RR = 0.62 [0.59, 0.66], p < 0.00001; 55–64 years: RR = 0.71 [0.69, 0.74], p < 0.00001; 65–74 years: RR = 0.82 [0.79, 0.85], p < 0.00001). No statistically significant difference was found between the populations in the 75 and over age group (RR = 0.94 [0.88, 1.01], p = 0.07).

Three studies [39, 42, 45] undertaken in the US used hospital databases to compare the two populations. Subgroup and sensitivity analyses were not performed due to the small number of studies describing these outcomes.

Figure 3.3 (S3) shows that the mean age on admission to hospital among the GP was 7 years older than in the IP (mean difference = 7.67 [7.22, 8.11], p < 0.00001, I² = 0%). Figure 3.4 (S3) presents the risk of obesity was higher in the hospitalized GP (RR = 1.13 [1.04, 1.22], p = 0.004, I² = 0%). The results also show that risk of other comorbidities was lower among the GP. However, we did not find any statistically significant differences. The results are presented in Supplementary Material 3. Figures S3.5 shows the analysis of the number of people with hypertension hospitalized for COVID-19 in each population; Figure S3.6 shows the analysis of number of people with diabetes hospitalized for COVID-19; Figure S3.7 shows the analysis of number of people with asthma/COPD hospitalized for COVID-19; Figure S3.8 shows the analysis of number of smokers hospitalized for COVID-19.

Mean length of hospital stay was reported by two studies [42, 45]. Altibi [45] reported that the average length of hospital stay was 7 days for the GP and 9 days for the IP, while Montgomery et al. [42] found that it was 8 and 9 days, respectively. Time from admission to in-hospital death was reported by Altibi et al. [45]; however, no statistically significant differences were found between the populations (mean difference = -2.50, [-13.36, 8.36], p = 0.65).

The ICU admission rate was lower among the GP; however, this difference was not statistically significant (Fig. 4). Prevalence of use of IMV was 23% lower in the GP (Fig. 5). Risk of in-hospital death was lower in the GP; however, this difference was not statistically significant (Fig. 6).

Fig. 4
figure 4

Forest plot of ICU admission rate

Full size image
Fig. 5
figure 5

Forest plot of use of IMV for COVID-19

Full size image
Fig. 6
figure 6

Forest plot of COVID-19 hospital mortality

Full size image

In contrast to the above findings, Montgomery et al. [42] found that age-adjusted hospital mortality was higher among the IP (RR = 1.47, [1.28, 1.69]). In the same direction, in the age- and sex-adjusted models, Altibi et al. [45] found that the likelihood of in-hospital death was greater among the IP (OR = 2.38 [1.37, 4.12]).

Quality of evidence and missing outcomes

The quality of the evidence from the studies was classified as low (Table 1), primarily due to the high risk of bias in the studies. Inconsistency was classified as serious or very serious for most results due to the high heterogeneity of the analyses.

Table 1 Summary of findings and grading of quality of evidence
Full size table

The lack of publication bias found in the analysis of the funnel plot of COVID-19 incidence was confirmed by Egger’s test (p = 0,654); however, the results reveal large heterogeneity in the studies. Figure S2.5 in Supplementary Material 2 shows the funnel plot. We did not perform a publication bias risk assessment for the other outcomes due to the small number of studies. Risk of bias was therefore considered high. It is also likely controlling for confounding factors would yield different results. Most of the studies did not control for age, meaning that the age-adjusted secondary outcomes included in the protocol are missing.

The results of the analyses of the dichotomous data were presented in preference to the results of the continuous data analysis to enable the calculation of absolute risk; however, both analyses were undertaken and no statistically significant differences between the populations and in heterogeneity were found.

Discussion

Our review investigated the effects of non-exposure to incarceration on the prevention of COVID-19 and mitigation of associated events. In short, the findings reveal significantly better health outcomes among the GP, showing that COVID-19 incidence can be up to 60% lower in this population. Despite statistically significant differences, the results of the subgroup analyses are inconclusive due to the high heterogeneity in the subgroups.

Given that COVID-19 disproportionately affects men and people aged over 65 [28,29,30, 61], differences in distribution of demographic characteristics is an important confounding factor and adjustment for age and sex is essential when comparing populations. Individuals in the GP are three times more likely to be over 65 years old and half as likely to be male, with the findings showing that when COVID-19 mortality was adjusted for age and sex risk of death was 64% lower among the GP than in the IP. We did not find standardized data on case fatality.

We did not perform a meta-analysis of hospitalization rate due to the small number of studies with relevant data. The two studies [42, 54] reported conflicting results, with Hernando [54] finding that the hospitalization rate was higher among the GP and Montgomery [42] observing higher rates in IP. This study [42] showed that not being incarcerated was a protective factor for COVID-19 hospitalization across all age groups below 75.

Only three studies, each of which conducted in the US, assessed hospital outcomes using hospital databases. The findings show that mean age at hospital admission was seven years older and risk of obesity was 13% higher among the hospitalized GP than in the hospitalized IP. Use of IMV was 23% lower among the GP than in the IP. The meta-analysis of these outcomes presented low heterogeneity (I²=0%). No statistically significant differences were found in ICU admission and hospital mortality between the two populations. Heterogeneity was high for both these outcomes. The results for these hospital outcomes should be interpreted with caution due to the small number of studies and/or high heterogeneity.

Our findings suggest that non-exposure to incarceration can be a protective factor against infection and death from COVID-19, which is consistent with other reviews that compared the GP and IP [62, 63]. These findings reinforce theories of the link between certain health conditions and crime, with studies [21, 22, 64] showing that not committing crime is a protective factor against premature death and disease.

However, the main limitation of the present review is that the lack of quantitative data on length of exposure to incarceration and control for confounding hinders the assessment of cause-and-effect. The studies reported the following limitations: underreporting of COVID-19 cases and deaths in prisons because facilities are not included in health surveillance systems; use of different sources of information for the two populations (such as different public databases), resulting in reporting bias, bias related to integrity and imprecision and inconsistency of results; and lack of data on the demographic characteristics of the IP in public databases.

The quality of evidence from the studies was considered low, primarily due to the high risk of bias in the studies caused by failure to control for exposure to incarceration, the small number of studies included in the meta-analyses and heterogeneity of results. The latter can be explained by methodological diversity in data treatment and the use of secondary data from different sources. Heterogeneity of results was lower in studies that used the same source of data to compare the two populations.

Despite the considerations for interpretating results with large statistical heterogeneity [26], we opted to perform meta-analysis because, to the best of our knowledge, the studies represent the best available data. It is also important to mention that ethical aspects and issues related government disclosure of prison data can hamper research [65]. In addition, despite the chronic health problems in prisons around the world and evidence of spillover into communities [15, 17, 66], the evidences of the health of people exposed to incarceration remains timid and there is little evidence regarding the benefits of not being exposed to prison.

It is worth noting that some studies [21, 23, 64] have established a link between crime and health, suggesting that committing crime can lead to poor health and that the latter can also lead to crime. We therefore believe that public safety and public health are inextricably linked.

It is also worth highlighting that, as mentioned in the studies, the surveillance of the health of the general population and prison population is tasked to different agencies. This suggests the need to create a standardized health surveillance system tailored specifically to prison settings. Our study is interdisciplinary, highlighting the need for an interdisciplinary approach to promote effective health surveillance in prisons around the world.

Future studies comparing COVID-19 outcomes between incarcerated populations and the general population should control for confounding factors and length of incarceration to facilitate the determination of cause-and-effect relationships.

Conclusion

The findings of this systematic review and meta-analysis reveal that non-exposure to incarceration can be associated with reduced risk across COVID-19 outcomes. It may also be concluded that decarceration, which includes the use of alternative measures to imprisonment, can be effective in preventing COVID-19 and mitigating associated events. However, decarceration policies need to focus on ensuring freed prisoners access to health services and better health outcomes.

Data availability

The datasets utilized and/or analyzed in the current study are available from the corresponding author on reasonable request.

Abbreviations

PROSPERO:

International prospective register of systematic reviews

PRISMA:

Preferred Reporting Items for Systematic Review and Meta-Analysis

GP:

Population not exposed to incarceration

IP:

Population exposed to incarceration

ROBINS-E:

Risk Of Bias In Non-randomized Studies – of Exposure RevMan Web: Review Manager Web

STROBE:

The Strengthening the Reporting of Observational Studies in Epidemiology Statement guidelines for reporting observational studies

GRADE:

Grading of Recommendations Assessment, Development and Evaluation

US:

United States

CI:

Confidence interval

MD:

Mean difference

OR:

Odds ratio

RR:

Risk ratio

DP:

Standard deviation

COPD:

Chronic obstructive pulmonary disease

ICU:

Intensive care unit

IMV:

Invasive mechanical ventilation

References

  1. Baratta A. Criminología Y Sistema penal: compilación in memoriam. Reimpr. Montevideo: Ed. B de F; 2006. p. 459. (Memoria Criminológica).

    Google Scholar 

  2. National Research Council. The growth of incarceration in the United States: exploring causes and consequences. Washington, D.C.: National Academies; 2014.

    Google Scholar 

  3. Garland D. Org. Mass imprisonment: social causes and consequences. London; Thousand Oaks, Calif: SAGE; 2001. p. 184.

    Book  Google Scholar 

  4. Wacquant L. O lugar da prisão na nova administração da pobreza. Novos estud – CEBRAP. 2008;80:9–19.

    Article  Google Scholar 

  5. van Zyl Smit D. Vereinte Nationen, organizadores. Handbook of basic principles and promising practices on alternatives to imprisonment. New York; 2007. 80 p. (Criminal justice handbook series).

  6. Dolan K, Wirtz AL, Moazen B, Ndeffo-mbah M, Galvani A, Kinner SA, Courtney R, McKee M, Amon JJ, Maher L, Hellard M, Beyrer C, Altice FL. Global burden of HIV, viral hepatitis, and tuberculosis in prisoners and detainees. Lancet. 2016;388(10049):1089–102.

    Article  PubMed  Google Scholar 

  7. Stuckler D, Basu S, McKee M, King L. Mass incarceration can explain population increases in TB and multidrug-resistant TB in European and central Asian countries. Proc Natl Acad Sci USA. 2008;105(36):13280–5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Burki T. Prisons are in no way equipped to deal with COVID-19. Lancet [Internet]. 2020;395(10234):1411–2.

    Article  CAS  PubMed  Google Scholar 

  9. Dourado JLG, Alves RSF. Panorama Da saúde do homem preso: dificuldades de acesso ao atendimento de saúde. Bol - Acad Paulista De Psicologia. 2019;39:47–57.

    Google Scholar 

  10. Heard C. Towards a health-informed approach to penal reform? Evidence from ten countries. London: Institute for Criminal Policy Research; 2019.

    Google Scholar 

  11. Minayo M, Constantino P. Deserdados sociais: condições de vida e saúde dos presos do estado do Rio De Janeiro. Rio de Janeiro: Fiocruz; 2015.

    Google Scholar 

  12. Reinhart E, Chen DL. Association of Jail Decarceration and Anticontagion policies with COVID-19 Case Growth Rates in US counties. JAMA Netw Open 1o De setembro de. 2021;4(9):e2123405.

    Article  Google Scholar 

  13. Sanchez A, Simas L, Diuana V, Larouze B. COVID-19 in prisons: an impossible challenge for public health? Cadernos De Saude Publica. 2020; 36(5).

  14. Reinhart E, Chen DL. Carceral-community epidemiology, structural racism, and COVID-19 disparities. Proc Natl Acad Sci U S A. 2021; 118(21).

  15. Nosrati E, Kang-Brown J, Ash M, McKee M, Marmot M, King LP. Incarceration and mortality in the United States. SSM Popul Health. 2021;15:100827.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Braithwaite R, Warren R. The African American Petri Dish. J Health Care Poor Underserved. 2020;31(2):491–502.

    Article  PubMed  Google Scholar 

  17. Hooks G, Sawyer W. Mass Incarceration, COVID-19, and Community Spread. Prison Policy Initiative; 2020.

  18. Wang EA, Western B, Backes EP, Schuck J. National Academies of Sciences, Engineering, and Medicine (U.S.), decarcerating correctional facilities during COVID-19: advancing health, equity, and safety. Washington, DC: the National Academies; 2020. p. 146.

    Book  Google Scholar 

  19. Levy M. International Public Health and Corrections: models of care and harm minimization. Public health behind bars: from prisons to communities. New York: Springer; 2007. pp. 73–87.

    Chapter  Google Scholar 

  20. Ramaswamy M, Freudenberg N. Health Promotion in Jails and prisons: an Alternative Paradigm for Correctional Health Services. Public health behind bars: from prisons to communities. New York: Springer; 2007. pp. 229–48.

    Chapter  Google Scholar 

  21. Skinner GCM, Farrington DP. A systematic review and meta-analysis of premature mortality in offenders. Aggress Violent Beh. 2020;53:101431.

    Article  Google Scholar 

  22. Pridemore WA. The Mortality Penalty of Incarceration: evidence from a Population-based case-control study of Working-age males. J Health Soc Behav. 2014;55(2):215–33.

    Article  PubMed  Google Scholar 

  23. Semenza DC, Testa AM, Jackson DB, Vaughn MG. Incarceration and cardiovascular health: multiple mechanisms within an intersectional framework. J Criminal Justice. 2021;73:101759.

    Article  Google Scholar 

  24. World Health Organization. Preparedness, prevention and control of COVID-19 in prisons and other places of detention. Interim Guidance [Internet]. World Health Organization; 2020. http://www.euro.who.int/__data/assets/pdf_file/0019/434026/Preparedness-prevention-and-control-of-COVID-19-in-prisons.pdf?ua=1.

  25. Martins ELC, de Oliveira GLA, Constantino P. Government actions for COVID-19 control and prevention in prisons: a scoping review. Cien Saude Colet. 2022;27(12):4435–50.

    PubMed  Google Scholar 

  26. Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, Welch V. Cochrane Handbook for Systematic Reviews of Interventions. version 6.4. Cochrane; 2023. https://www.training.cochrane.org/handbook

  27. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, Shamseer L et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021.

  28. Dessie ZG, Zewotir T. Mortality-related risk factors of COVID-19: a systematic review and meta-analysis of 42 studies and 423,117 patients. BMC Infect Dis 21 de agosto de. 2021;21(1):855.

    Article  CAS  Google Scholar 

  29. Harris M, Hart J, Bhattacharya O, Russell FM. Risk factors for SARS-CoV-2 infection during the early stages of the COVID-19 pandemic: a systematic literature review. Front Public Health. 2023;11:1178167.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Thakur B, Dubey P, Benitez J, Torres JP, Reddy S, Shokar N, Aung K, Mukherjee D, Dwivedi AK. A systematic review and meta-analysis of geographic differences in comorbidities and associated severity and mortality among individuals with COVID-19. Sci Rep. 2021;11(1):8562.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Ouzzani M, Hammady H, Fedorowicz Z, Elmagarmid A. Rayyan—a web and mobile app for systematic reviews. Syst Rev. 2016;5(1):210.

    Article  PubMed  PubMed Central  Google Scholar 

  32. The Cochrane Collaboration. Review Manager Web (RevMan Web). 2024. https://www.revman.cochrane.org

  33. Higgins J, Morgan R, Rooney A, Taylor K, Thayer K, Silva R, Lemeris C. Akl. Risk Of Bias In Non-randomized Studies - of Exposure (ROBINS-E). 2023. https://www.riskofbias.info/welcome/robins-e-tool

  34. Higgins J, Li T, Deeks J. Choosing effect measures and computing estimates of effect. Cochrane Handbook for Systematic Reviews of Interventions. Cochrane; 2023. https://www.training.cochrane.org/handbook.

  35. Deeks J, Higgins J, Altman D. Analysing data and undertaking meta-analyses. Cochrane Handbook for Systematic Reviews of Interventions. Cochrane; 2023. https://www.training.cochrane.org/handbook.

  36. Page M, Higgins J, Sterne J. Chapter 13: Assessing risk of bias due to missing results in a synthesis. Cochrane Handbook for Systematic Reviews of Interventions. Cochrane;, von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344-9. PMID: 18313558.

  37. GRADEpro GDT. GRADEpro Guideline Development Tool [Software]. McMaster University and Evidence Prime; 2024.

  38. Hershow R, Burnett J, Nicolae L, Marquez N, Everett M, Tyagi E, Williams S. COVID-19 Burden in Adult Correctional or Detention facilities and the surrounding communities, January 1, 2020-July 20, 2021. J Correctional Health Care. 2023;29(4):241–6.

    Article  Google Scholar 

  39. Abdalbary M, Kakani E, Ahmed Y, Shea M, Neyra JA, El-Husseini A. Characteristics and outcomes of prisoners hospitalized due to COVID-19 disease. Clin Nephrol Abril De. 2022;97(4):232–41.

    Article  CAS  Google Scholar 

  40. Marquez NM, Littman AM, Rossi VE, Everett MC, Tyagi E, Johnson HC, Dolovich SL. Life expectancy and COVID-19 in Florida state prisons. Am J Prev Med. 2022;62(6):949–52.

    Article  PubMed  PubMed Central  Google Scholar 

  41. Marusinec R, Brodie D, Buhain S, Chawla C, Corpuz J, Diaz J, Durbin M, Moss N, Okada R, Sanchez Y, Watkins-Tartt K, Yette E, Chitnis AS. Epidemiology of Coronavirus Disease 2019 at a County Jail-Alameda County, California, March 2020-March 2021. J Public Health Manag Pract 1o De fevereiro de. 2022;28(1):50–9.

    Article  Google Scholar 

  42. Montgomery MP, Hong K, Clarke KEN, Williams S, Fukunaga R, Fields VL, Park J, Schieber LZ, Kompaniyets L, Ray CM, Lambert LA, D’Inverno AS, Ray TK, Jeffers A, Mosites E. Hospitalizations for COVID-19 among US people experiencing incarceration or homelessness. JAMA Netw Open. 2022;5(1):e2143407.

    Article  PubMed  PubMed Central  Google Scholar 

  43. Nowotny K, Metheny H, LeMasters K, Brinkley-Rubinstein L, Age. COVID-19 mortality in the United States: a comparison of the prison and general population. Int J Prison Health Published Online June. 2022;23. https://doiorg.publicaciones.saludcastillayleon.es/10.1108/IJPH-08-2021-0069.

  44. Towers S, Wallace D, Walker J, Eason JM, Nelson JR, Grubesic TH. A study of SARS-COV-2 outbreaks in US federal prisons: the linkage between staff, incarcerated populations, and community transmission. BMC Public Health. 2022;22(1):482–482.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Altibi AM, Pallavi B, Liaqat H, Slota AA, Sheth R, Al Jebbawi L, George ME, LeDuc A, Abdallah E, Russell LR, Jain S, Shirvanian N, Masri A, Kak V. Characteristics and comparative clinical outcomes of prisoner versus non-prisoner populations hospitalized with COVID-19. Sci Rep. 2021;11(1):6488–6488.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Marquez N, Ward JA, Parish K, Saloner B, Dolovich S. COVID-19 incidence and mortality in federal and state prisons compared with the US Population, April 5, 2020, to April 3, 2021. JAMA. 2021;326(18):1865–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  47. Nowotny KM, Seide K, Brinkley-Rubinstein L. Risk of COVID-19 infection among prison staff in the United States. BMC Public Health. 2021;21(1):1036–1036.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Toblin RL, Hagan LM. COVID-19 case and mortality rates in the Federal Bureau of Prisons. Am J Prev Med. 2021;61(1):120–3.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Wallace D, Eason JM, Walker J, Towers S, Grubesic TH, Nelson JR. Is there a temporal relationship between COVID-19 infections among Prison Staff, incarcerated persons and the larger community in the United States? Int J Environ Res Public Health. 2021;18(13):6873.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Jimenez M, Cowger T, Simon L, Behn M, Cassarino N, Bassett M. Epidemiology of COVID-19 among incarcerated individuals and staff in Massachusetts jails and prisons. JAMA Netw Open. 2020;3(8):e2018851.

    Article  PubMed  PubMed Central  Google Scholar 

  51. Lemasters K, McCauley E, Nowotny K, Brinkley-Rubinstein L. COVID-19 cases and testing in 53 prison systems. Health Justice. 2020;8(1).

  52. Saloner B, Parish K, Ward JA, DiLaura G, Dolovich S. COVID-19 cases and deaths in federal and state prisons. JAMA. 2020;324(6):602–3.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Getaz L, Wolff H, Goncalves L, Togni G, Stringhini S, Pala K, Iten A, Guessous I, Kaiser L, Chappuis F, Baggio S. SARS-CoV-2 seroprevalence study after the first wave among persons living and working in an overcrowded Swiss prison. Int J Prison Health. 2022.

  54. Hernando V, Vázquez I, Díaz A, Gómez-Pintado P, Martínez-Aznar C, Acín E. COVID-19 pandemic in prisons in Spain: characteristics of cases and implemented control measures, March 2020-June 2022. Public Health. 2023;218:45–52.

    Article  PubMed  Google Scholar 

  55. Rice WM, Chudasama DY, Lewis J, Senyah F, Florence I, Thelwall S, Glaser L, Czachorowski M, Plugge E, Kirkbride H, Dabrera G, Lamagni T. Epidemiology of COVID-19 in prisons, England, 2020. Emerg Infect Dis. 2021;27(8):2183–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  56. Carabellese F, Petroni G, Ferracuti S, Ferorelli D, Solarino B, Vernaglione S, et al. The protection of health in prison during the covid-19 pandemic. Rassegna Italiana Di Criminologia. 2020;14(4):300–4.

    Google Scholar 

  57. Marmolejo L, Barberi D, Bergman M, Espinoza O, Fondevila G. Responding to COVID-19 in latin American prisons: the cases of Argentina, Chile, Colombia, and Mexico. Victmis Offenders. 2020;15(7–8):1062–85.

    Article  Google Scholar 

  58. Rapisarda S, Byrne J, Marmolejo L. An examination of COVID-19 outbreaks in south American prisons and jails. Victmis Offenders. 2020;15(7):1009–18.

    Article  Google Scholar 

  59. Ribeiro L, Diniz A. The Brazilian Penitentiary System under the threat of COVID-19. Victmis Offenders. 2020;15(7):1019–43.

    Article  Google Scholar 

  60. Richardson M, Garner P, Donegan S. Interpretation of subgroup analyses in systematic reviews: a tutorial. Clin Epidemiol Global Health. 2019;7(2):192–8.

    Article  Google Scholar 

  61. Centers for Disease Control and Prevention (CDC). Risk for COVID-19 Infection, Hospitalization, and Death By Age Group. Centers for Disease Control and Prevention. 2023 https://archive.cdc.gov/#/details?url=https://www.cdc.gov/coronavirus/2019-ncov/covid-data/investigations-discovery/hospitalization-death-by-age.html

  62. Kim H, Hughes E, Cavanagh A, Norris E, Gao A, Bondy SJ, McLeod KE, Kanagalingam T, Kouyoumdjian FG, Blumberg S. PLoS ONE. 2022;17(5):organizador.

    Google Scholar 

  63. Williams DB, Spinks B, Williams D, Lewis R, Bull F, Edwards A. Effects of the COVID-19 pandemic on people experiencing incarceration: a systematic review. BMJ Open. 2024;14(4):e076451.

    Article  PubMed  PubMed Central  Google Scholar 

  64. Testa A, Semenza D, Testa A, Semenza D. Criminal offending and health over the life-course: a dual-trajectory approach. J Criminal Justice. 2020;68:101691.

    Article  Google Scholar 

  65. Simpson PL, Guthrie J, Butler T. Prison health service directors’ views on research priorities and organizational issues in conducting research in prison: outcomes of a national deliberative roundtable. Int J Prison Health. 2017;13(2):113–23.

    Article  PubMed  Google Scholar 

  66. Mallart F, Cunha MI. Introdução: as dobras entre o dentro e o fora. Tempo Soc. 2019;31(3):7–15.

    Article  Google Scholar 

Download references

Acknowledgements

We are grateful to the librarians from the Cecília Minayo Information and Documentation Center at the Jorge Careli Department of Violence and Health Studies (Claves/ENSP/Fiocruz), who helped us during the article search and screening process, and to the translator who translated the manuscript into English.

Funding

This work was supported by the Academic Excellence Program of the Coordination for the Development of Higher Education Personnel (PROEX-CAPES).

Author information

Authors and Affiliations

  1. National School of Public Health (ENSP)/Oswaldo Cruz Foundation (Fiocruz), Rio de Janeiro, Brazil

    Elida Lucia Carvalho Martins & Patrícia Constantino

  2. Department of Health Economics and Development, Ministry of Health, Distrito Federal, Brazil

    Gustavo Laine Araújo de Oliveira

Authors
  1. Elida Lucia Carvalho Martins
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Patrícia Constantino
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Gustavo Laine Araújo de Oliveira
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

ELCM was responsible for study conception, designing the search strategies, study screening and selection, data extraction and analysis, and writing and revising the manuscript. PC was responsible for study conception, designing the search strategies, study screening and selection, data extraction and analysis, and revising the manuscript. GLAO was responsible for study conception, designing the search strategies, study screening and selection, data extraction and analysis, and revising the manuscript.

Corresponding author

Correspondence to Elida Lucia Carvalho Martins.

Ethics declarations

Ethical approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Supplementary Material 3

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Martins, E.L.C., Constantino, P. & de Oliveira, G.L.A. The effectiveness of non-exposure to incarceration in preventing COVID-19 and mitigating associated events: a systematic review and meta-analysis. BMC Public Health 25, 206 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12889-024-20859-1

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12889-024-20859-1

Keywords

BMC Public Health

ISSN: 1471-2458

Contact us