Disparities in cancer prevalence, incidence, and mortality for incarcerated and formerly incarcerated patients: A scoping review

Abstract Background Racial and ethnic minority status, structural racism, low educational attainment, and poverty are consistently associated with cancer disparities and with higher rates of incarceration. The objective of this scoping review is to conduct a qualitative synthesis of the literature on cancer prevalence, incidence, mortality, and disparities in these outcomes for incarcerated and formerly incarcerated patients, as this literature is fragmented and heterogenous. Methods This scoping review included Bureau of Justice Statistics reports and searched PubMed in May 2021 for all English language studies published between 1990 and 30 April 2021, that reported on cancer prevalence, incidence, or mortality for incarcerated or formerly incarcerated individuals in the United States. Results Twenty studies were selected. Data on cancer prevalence and incidence were scarce but suggested that incarcerated and formerly incarcerated patients have a similar overall risk of cancer diagnosis as the general population, but elevated risk of certain cancers such as cervical, lung, colorectal, and hepatocellular carcinoma for which effective prevention and screening interventions exist. Cancer mortality data in state and local jails as well as prisons were robust and suggests that both incarcerated and formerly incarcerated patients have higher cancer mortality than the general population. Conclusions Incarcerated and formerly incarcerated patients likely have a higher risk of dying from cancer than the general population, but important gaps in our knowledge about the extent and drivers of disparities for this population remain. Additional research is needed to guide interventions to reduce cancer disparities for patients experiencing incarceration.


| INTRODUCTION
Despite decades of progress in the prevention and treatment of cancer in the United States, disparities in cancer incidence and survival are significant. 1 Factors such as racial/ethnic minority status, structural racism, low educational attainment, and poverty have been consistently associated with a higher incidence of cancer, late diagnosis, and worse survival. 1 Nearly all of these disadvantaging factors converge upon one particularly vulnerable population: incarcerated and formerly incarcerated individuals. For instance, Black men are incarcerated at nearly six times the rate for White men, 2 and Black men have higher cancer mortality than White men for lung cancer (52 vs. 45 per 100,000 of Black and White men, respectively), colorectal cancer (22 vs. 16 per 100,000), and liver cancer (13 vs. 9 per 100,000), among other preventable cancer types. 3 The United States has a large and aging incarcerated population that faces an enormous rise in cancer risk. Nearly 2.1 million individuals were incarcerated in local, state, and federal jails and prisons in 2018. 4 The population of individuals incarcerated in US prisons aged 55 years or older (i.e., those at highest risk for cancer) grew 14-fold from 8853 in 1981 to 124,900 in 2010, and is expected to reach 400,000 by 2030. 5 Unsurprisingly, cancer deaths are rising in these prisons.
There are no comprehensive studies describing the relationship between incarceration and cancer incidence and mortality. Though often outsourced to private health companies, federal, state, and local governments are responsible for healthcare for individuals incarcerated in the nation's jails and prisons. As a result, healthcare data for incarcerated patients are fragmented and the population is absent from disparities analyses based upon Medicare, Medicaid, or commercial claims databases. Cancer diagnoses are reported to state cancer registries and thus included in analyses using these databases, but in the absence of time-consuming linkage to incarceration records, disparities related to incarceration status are largely invisible in these studies. A 2004 national survey found that incarcerated individuals had 22% higher odds of a cancer diagnosis compared to the general population, 6 while cancer caused 27% of deaths in state prisons between 2001 and 2016, with mortality rates rising 59% over that period (from 58 to 92 per 100,000 individuals incarcerated in state prisons). 7 However, data on cancer incidence and mortality rates, particularly compared to the general population, are limited in scope, geography, and applicability.
Cancer control efforts to address disparities for vulnerable populations experiencing incarceration are stymied by this complex literature and the lack of clear data upon which to base interventions to improve cancer care. A scoping review is a methodology to systematically evaluate knowledge gaps in the literature 8 ; to inform future cancer control efforts, this study conducts a scoping review of the existing literature on cancer prevalence, incidence, mortality, and racial and ethnic disparities in these measures for individuals who were incarcerated or formerly incarcerated at the time of their cancer diagnosis.

| Study design and data sources
We conducted a scoping literature review using PubMed, an online biomedical literature database consisting of over 32 million citations from MEDLINE, life science journals, and online books. This study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis--Scoping Review (PRISMA-ScR) guidelines (Table S1). As a review of published studies, this study is exempt from review and was not submitted for approval from an Institutional Review Board.

| Article selection
In May 2021, we searched PubMed for all English language articles published between 1 January 1990 and 30 April 2021, and included the search terms: cancer AND (prevalence OR incidence OR mortality) AND (jail OR prison OR incarcerat* OR "criminal justice" NOT hernia). PubMed was last searched on 18 June 2021 to capture articles that were published by 30 April 2021 but had not been indexed in PubMed at the time of the initial search in May 2021. Articles on incarcerated hernias were excluded. Abstracts were independently reviewed by two investigators (CM and VO) for initial screening, as were full text articles reviewed after the initial screening. Articles were excluded if they did not report data on the study outcomes of cancer prevalence, incidence, or mortality for incarcerated or formerly incarcerated patients, or if the study population was not based in the United States. Inconsistent scoring after independent selection was resolved by discussion and consensus of study authors.
In addition to articles included through the PubMed search, two study investigators (CM and VO) reviewed references in included articles for additional studies that met inclusion criteria. Finally, study investigators reviewed and included the most recent reports from the US Bureau of Justice Statistics, which publishes annual reports describing deaths that occur in US jails and prisons. This review was not registered; the review protocol, templates used for data collection, and original data collection spreadsheets are available upon request.

| Data synthesis and analysis
For each included article, we abstracted whether the study involved incarcerated patients, formerly incarcerated patients, or both; whether the study population was in jail or prison; whether the correctional facilities were local, state, or federal; study population characteristics including number of patients, sex, years of inclusion, and location; source of data; and primary findings regarding (1) cancer prevalence, incidence, or mortality, as absolute numbers or rates and subdivided by cancer type when available, for incarcerated or formerly incarcerated patients, (2) comparison of findings to non-incarcerated populations, if reported; and (3) reported racial or ethnic disparities in any of the findings. Two investigators (CM and VO) conducted the abstraction independently, results were merged by consensus, and reviewed by the third investigator (DS). The qualitative synthesis of major findings regarding cancer prevalence, incidence, and mortality in the included studies was summarized in a table.

| RESULTS
A total of 252 studies were screened and assessed for eligibility, and 20 were included in this review ( Figure 1). Of note, one study 9 was excluded that contained a subset of data reported in another, parent publication that was included in this review. 10 Three other studies fit this criteria but were nonetheless included in this review because the studies reported additional data or analyses that were not included in the parent papers (see Table 1). [11][12][13] Two additional studies were excluded as they provided data only on cancer deaths in a hospital 14 and cancer prevalence in a small study of individuals in prison who smoke 15 as these studies do not reflect cancer prevalence, incidence, or mortality in jail or prison populations as a whole.
A summary of the characteristics of included studies is presented in Table 1 and a summary of study design and findings for the included studies are presented in Table 2.
Of the 20 studies included in this review, 16 studies included prisons, two studies included jails, and two studies included both prisons and jails. Two, 18, and four studies examined federal, state, and local correctional facilities, respectively (sum is greater than 20 as some studies included multiple levels of government). Only five of the 20 studies used data that was less than 10 years old as of 2021. 7,[16][17][18][19] Except in studies sharing datasets (e.g., Spaulding et al 10 and Zlotorynzska et al 12 ), studies used incompatible methodologies, data sources, time periods, and patient populations that prevent direct quantitative comparisons of study findings. A qualitative review of major findings on cancer prevalence, incidence, mortality, and racial and ethnic disparities in those measures follows.

| Prevalence
Patient-reported cancer prevalence in jails or prisons was 0.4%-9.0% in three studies 6,16,24 and in one study was comparable to that of the general public. 6 In one survey study, White men reported higher cancer prevalence than Black men (1.3% vs. 0.4%). 24 The most common cancers were cervical, lung, colorectal, non-Hodgkin's lymphoma, and hepatocellular carcinoma, 20,21,23 and these cancers were more common in incarcerated than non-incarcerated patient populations in the two studies that compared findings to SEER cancer prevalence. 21,23 One study found that women in jails and prison were at higher risk of cancer due to high rates of gender-specific cancers, 22 especially cervical cancer for which prevalence was high. 20,22,23

| Incidence
Zlotorzynska et al. is the only study to provide standardized cancer incidence ratios, but this registry-based, 20-year follow-up of a cohort of individuals incarcerated in 1991 does not differentiate cancers diagnosed during or after incarceration. 12 Nonetheless, it found that cancer incidence for HIVnegative patients was slightly lower (standardized incidence ratio 0.9, 95% CI 0.8-0.9) than the general population in Georgia. The 4% of the cohort with HIV infection had higher cancer incidence than the general population (standardized incidence ratio 2.0, 95% CI 1.5-2.6). Consequently, the true cancer incidence for the total incarcerated population (which was not reported) may be slightly higher than the estimate for HIV-negative patients and roughly equivalent to that of the general population. This study found that 36%, 23%, and 32% of patients were diagnosed with localized, regional, and metastatic disease, respectively, but no comparison to the non-incarcerated population was reported. The most common cancers were lung, prostate, colorectal, kidney, and non-Hodgkin lymphoma, consistent with the two other studies that provided data on cancer types. 20,23

| Mortality
Data from all state prisons found that cancer accounted for 16,277 deaths (27% of all deaths) in state prisons from 2000 to 2018, growing from 58 to 92 per 100,000 incarcerated individuals over that period, 7 but only 698 deaths (4% of all deaths) in local jails from 2000 to 2016. 17 Cancer deaths in all state prisons were more common in White than Black and Hispanic patients (154 vs. 89 vs. 38 per 100,000 incarcerated individuals), 7 findings echoed in data from Texas. 26 Notably, these estimates are not adjusted for age, and do not include cancer deaths in federal prisons, which have a prison population approximately 1/10 of the combined state prison populations.
Among individuals incarcerated in Georgia prisons in 1991, Zlotorzynska et al. found that this cohort had higher cancer mortality (standardized mortality ratio (SMR) 1.2, 95% CI 1.1-1.3) than the general population in Georgia over the following 20 years. 12 However, only 22% of these cancer deaths occurred during incarceration, while the remainder occurred after release. 10 In North Carolina, White incarcerated men had higher age-adjusted SMR for all cancers (1.5, 95% CI 1.3-1.9) compared to non-incarcerated White men, while Black incarcerated men had lower mortality for all cancers (0.7, 95% CI 0.6-0.9) compared to non-incarcerated Black men. 27 In this study, 24% of incarcerated patient deaths were attributed to cancer versus 28% for the general population, but this finding was not standardized or statistically tested.
Finally, a study of patients with cancer incarcerated in Texas prisons found that these patients had a median overall survival of 21 months compared with 54 months for age, gender, and race-matched SEER patients, though this analysis did not match on stage. 23

| Formerly incarcerated patients
Seven studies, including three studies with mixed populations of incarcerated and formerly incarcerated patients, -Thirty-six percent presented with localized, 23% with regional, and 32% with distant cancer, with 9% unknown.
-All-cancer incidence was lower than the Georgia general population (Standardized incidence ratio 0.9, 95% CI 0.8-0.9) Mortality: -There were 481 cancer deaths, 12 in patients with HIV. The mortality rate was 181 per 100,000.
-All-cancer mortality was higher than the Georgia general population (Standardized mortality ratio 1. reported data on formerly incarcerated patients: one study reported data on prevalence, 19 one study reported data on incidence, 12 and six studies evaluated cancer mortality in this population. 10,12,13,18,28,29

| Prevalence
In a national survey of the general population, individuals who indicated criminal justice involvement in the prior year reported higher rates of lung (4.6% vs. 2.4%), cervical (13.8% vs. 9.1%), and alcohol-related cancers (3.5% vs. 1.7%) compared to individuals without recent criminal justice involvement, but not for colon, breast, prostate, and smoking-related cancers. 19

| Incidence
The cancer incidence estimates of individuals incarcerated in Georgia prisons discussed above included individuals who were currently and formerly incarcerated at the time of cancer diagnosis. 12 A companion study that looked at causes of death over the subsequent 20 years for this cohort found that 78% of cancer deaths occurred outside of prison, 10 providing indirect evidence that, for individuals with a history of incarceration, cancer diagnoses are likely more common after release into the community. No studies provide precise estimates on cancer estimates solely for formerly incarcerated patients.

| Mortality
Cancer was the second leading cause of death among formerly incarcerated patients in Georgia at 15% of all deaths. 10 In Washington State, cancer was the third leading cause of death after release from prison at 71 per 100,000 person-years despite a median follow-up of only 2 years, with released patients have a SMR of 1.9 compared to age, sex-, and race-matched US general population. 28 A study of a subset of the same patients found that formerly incarcerated patients had higher cancer mortality than incarcerated patients in Washington state prisons (68 vs. 42 per 100,000 person-years). 13 In North Carolina, individuals formerly incarcerated in prisons had higher cancer mortality than the general population (SMR 3.9, 95% CI 3.3-4.6), 18

| DISCUSSION
This scoping review found 20 studies evaluating cancer prevalence, incidence, and mortality in US jails and prisons. There are numerous gaps in the literature, but these studies demonstrate that cancer is a leading and increasing cause of death for patients incarcerated in prison and for formerly incarcerated patients, and these patient populations are likely at higher risk of cancer mortality than the general population. Cancer mortality in jails is low, likely due to the short period of incarceration and younger age distribution. Overall cancer incidence and prevalence may be relatively similar for incarcerated patients compared to the general population but elevated for specific diseases like cervical cancer, but the data are very limited and less clear. Overall, methodologically robust comparisons in cancer incidence and mortality between incarcerated, formerly incarcerated, and general populations are scarce. Despite limitations of the included studies, there are four important points that stand out. First, many of the most common cancers in patients who have experienced incarceration--lung, cervical, colorectal, and hepatocellular cancers--are preventable or have effective screening. Thus, incarceration or release from prison may present a public health opportunity to provide cancer control interventions that can reduce risk and cancer disparities by facilitating access to care. Indeed, as 95% of incarcerated individuals are ultimately released and 78% of cancer deaths in the Georgia cohort occurred after release from prison, 10 the public health benefits of improved cancer prevention and screening for incarcerated and formerly incarcerated patients accrue primarily to the communities that otherwise bear the costs of cancer diagnosis and treatment.
Second, incarcerated patients may receive delayed or substandard cancer screening and treatment, contributing to worse cancer mortality. Compared to the general population, the Georgia cohort study showed lower cancer incidence but higher cancer mortality, 12 and a Texas study showed higher mortality for incarcerated patients with cancer. 23 Both studies have methodologic limitations (a mixed incarcerated/formerly incarcerated patient population and insufficient risk adjustment, respectively) that prevent drawing of strong conclusions, and further investigation into access to and quality of cancer care for incarcerated and formerly incarcerated patients is needed.
Third, the relationship of incarceration and cancer disparities may vary by race and by state. Incarcerated Black men have lower rates of cancer death than incarcerated White and Hispanic men, a reversal of disparities typically seen in the general population (though not all studies adjusted for age). Furthermore, the North Carolina studies showed that both incarcerated and formerly incarcerated White men had worse age-adjusted cancer mortality than never-incarcerated White men, while the reverse was true for Black men. 27,29 It is possible that cancer care in prison is of lower quality than care received by White men in the community but better quality than care received by Black men, particularly in states such as North Carolina that have large racial disparities in cancer mortality in the general population, 3 but additional research in this area is needed.
Formerly incarcerated individuals appear to be at higher risk of cancer mortality than both incarcerated and never-incarcerated patients, but this disparity may also vary by state. 13,28 While many factors may mediate this disparity, formerly incarcerated individuals are often uninsured, due in part to suspension or cancelation of Medicaid enrollment upon incarceration, barriers to reenrollment upon release, and strict eligibility criteria, particularly in states that have not expanded Medicaid. 30,31 As studies have demonstrated that improved access to Medicaid reduces cancer mortality and reduces racial disparities in time to treatment in the community, 32,33 research is needed to determine how state polices on Medicaid enrollment and post-release healthcare transitions may influence disparities in cancer mortality for formerly incarcerated individuals compared to incarcerated and never-incarcerated individuals.
Fourth, possible incarceration-related disparities in cancer remain invisible as much data are simply unavailable. Data comparing age-adjusted cancer incidence for the general population to that of incarcerated and, separately, formerly incarcerated patients are absent. Similarly, there are no comparisons of stage of presentation or cancer care quality between incarcerated, formerly incarcerated, and general populations to assess whether these patient populations have delays in diagnosis or inferior care that may contribute to cancer disparities. An ongoing study will link state cancer registry and incarceration data to address some of these questions in Connecticut. 34 As many of the studies in this review came from just a few states, similar studies are needed with a diverse array of states to provide a comprehensive view on the relationship between incarceration and cancer disparities. Such studies are painstaking to assemble given the fragmented nature of the US criminal justice system and appropriate safeguards for research involving incarcerated individuals that nonetheless increase barriers to studying cancer in this population. When layered upon incarcerated patients' exclusion from Medicare and commercial payer databases, these barriers currently render possible disparities mostly invisible and additional research is necessary to help prioritize disease-specific interventions for improved cancer control. A cancer registry variable that indicates incarceration status at the time of diagnosis would increase transparency and enable states to evaluate the care quality for which they are paying. Finally, there were only four studies that included jails, likely because data are especially fragmented among the more than 3000 county jails in the United States. The comprehensive jail mortality data indicates that cancer mortality is uncommon, but the transitory nature of jail populations may mean that the consequences of potentially inferior cancer care in jails occur after release. This review has several limitations. Of studies outside of PubMed, the review only included studies from the Bureau of Justice Statistics; though the Bureau of Justice Statistics compiles data reported from the states, individual states may publish additional statistics on cancer in their incarcerated populations. Only 25% of the included studies reported data from within the past 10 years; US prison populations have declined slightly over that time and changes to insurance and Medicaid expansion during that time may change the relationship between incarceration and cancer disparities. However, the US prison population older than 55 continues to grow. Most studies focused on prisons in single states and there was considerable heterogeneity in study design; particularly given the potential importance of state demographics and healthcare policies, these study characteristics limit the ability to draw robust conclusions from individual study findings. Finally, accurate data on cancer diagnoses and outcomes are only a starting point for identifying and ameliorating disparities, and data on access to cancer prevention, screening, and quality of care for incarcerated patients are even harder to find and sorely needed.

| CONCLUSION
There is a growing spotlight on the US criminal justice system and its long-lasting effects on health for incarcerated individuals. While the inequities of the criminal justice system itself must be addressed, the current reality is that the US prison population is aging and a substantial proportion of individuals, particularly Black men, spend at least some time in jail or prison. This scoping review highlights that cancer mortality is common among incarcerated and formerly incarcerated patients and may be higher than the general population. A more nuanced understanding of the relationship between incarceration and disparities in cancer incidence and mortality is critical to inform efforts to remediate cancer disparities for patients who have been incarcerated and the communities in which they reside.