Cardiovascular Death and Access to Health Care Among Individuals Incarcerated in US State Prisons From 2001 to 2019

Abstract

Background

Incarcerated individuals have a disproportionate burden of cardiovascular risk factors. However, there is a paucity of data focusing on cardiovascular death and access to adequate health care among incarcerated individuals.

Methods

We used the Mortality in Correctional Institutions database from the US Bureau of Justice Statistics to examine cardiovascular deaths in all state prisons from 2001 to 2019, health care provision, as well as differences in these measures between racial and ethnic groups.

Results

From 2001 to 2019, there were a total of 18 227 (28.0% of total) cardiovascular deaths among those incarcerated in state prisons. After declining from 2001, age‐standardized cardiovascular mortality rates increased since 2012, and were 128.2 per 100 000 in 2019. Most cardiovascular deaths (40.8%) occurred after ≥120 months of incarceration, with 24.4% developing the condition that caused death after incarceration. Approximately 19% of individuals were not medically evaluated, 31% did not receive any diagnostic testing, and 27% did not receive medications during their incarceration for the cardiovascular condition causing death. Compared with White counterparts, Black individuals who died while incarcerated had a higher proportion of death related to a cardiovascular cause (30.9% versus 27.8%, P<0.001) and were less likely to be medically evaluated (78.6% versus 81.9%, P<0.0001) or receive treatment during incarceration for the cardiovascular condition determined to be the cause of death (68.1% versus 72.6%, P<0.0001).

Conclusions

Cardiovascular disease is an important cause of death among individuals incarcerated in state prisons. A large number of individuals who died from cardiovascular disease, did not receive medical treatment during their incarceration, with racial disparities in carceral care provision. Strategies to ensure quality and equitable cardiovascular care for this population are needed.

Nonstandard Abbreviations and Acronyms

BJS

Bureau of Justice Statistics

DCRP

Deaths in Custody Reporting Program

Clinical Perspective.

What Is New?

• Nearly one‐third of deaths among those incarcerated were due to cardiovascular disease, and Black incarcerated individuals were more likely to die from a cardiovascular cause than White individuals.

• Despite an increasing age‐adjusted cardiovascular mortality rate among those incarcerated in US state prisons, a large proportion of incarcerated individuals did not receive any medical care before their death, with racial disparities in care provision.

What Are the Clinical Implications?

• There is a high and increasing burden of cardiovascular disease contributing to death in the incarcerated population.

With nearly 2 million people incarcerated at any given time, the United States incarcerates more people than any other country in the world. ¹ , ² Given the large proportion of the population that is incarcerated at any given time in the United States, the health care provided in prison can have significant public health implications. ² Although incarcerated individuals have a constitutional right to health care, they face several barriers to accessing timely and appropriate high‐quality care. ³ , ⁴ , ⁵ , ⁶ Cardiovascular death remains the leading cause of death in the United States. ⁷ While heart disease has been shown to be the second most common cause of death among incarcerated individuals, accounting for 26% of all deaths, there remains a paucity of data regarding the cardiovascular health among the justice‐involved population, particularly the rates of treatment for cardiovascular disease. ⁸

Incarcerated individuals have a disproportionate burden of cardiovascular risk factors and are more likely to belong to racially minoritized populations. ⁹ , ¹⁰ , ¹¹ , ¹² Incarceration exposure causes accelerated aging, provoking chronic disease development. ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ Additionally, mass incarceration both reflects and perpetuates structural racism. Despite efforts at decarceration, Black men remain 4 times more likely to be imprisoned compared with White counterparts. ¹⁹ , ²⁰ , ²¹ Given known racial disparities in cardiovascular disease outcomes, ²² it is possible that mass incarceration serves as a sociopolitical driver of cardiovascular health inequities. ²³ However, little is known about how the racialization of mass incarceration or disparities in care receipt while incarcerated contribute to cardiovascular health inequities.

Understanding the burden of cardiovascular disease death among incarcerated individuals is critical for increasing awareness of the cardiovascular health status of this population and to inform future equity efforts, not only in the United States but globally as well. Additionally, characterizing rates of care provision for cardiovascular disease among the incarcerated population and identifying potential gaps in care access is crucial. Therefore, we used data from the Bureau of Justice Statistics (BJS) to evaluate the burden and trends of cardiovascular death and use of health care before cardiovascular death as well as whether these measures varied on the basis of race.

METHODS

The University of Pennsylvania Institutional Review Board determined that this research was exempt from regulatory requirements. We followed the Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines. Requests to access the data set from qualified researchers trained in human subject confidentiality protocols may be sent to BJS. ²⁴

Data Sources

For this study, we used the database “Mortality in Correctional Institutions: ICD‐10 Diagnosis Codes for Natural Deaths Occurring in State Prison or Local Jail Custody, 2001–2019” from the US BJS. ²⁵ The Mortality in Correctional Institutions database collects quarterly inmate death records from each of the nation's 50 state prison systems, 50 state juvenile correctional facilities, and 3095 local jails.

Under the BJS Deaths in Custody Reporting Program (DCRP), respondents from jails and state departments of corrections are asked to report causes of deaths, identifiers, and characteristics of each incarcerated individual who died. ²⁶ Under the Death in Custody Report Act, all state prison facilities are required to complete a standardized DCRP reporting form for each inmate who dies while incarcerated. This form is completed by prison administrators and correctional medical staff and includes individual‐level information on the deceased's sociodemographic characteristics including self‐identified race and ethnicity; correctional information (ie, length of stay); the circumstances of the death and information regarding the medical condition causing death, including if individuals received evaluation by a medical provider/staff, diagnostic test performed, treatment with medications, nonmedication treatments (or surgery provided) for the condition causing death other than emergency care (binary variables as yes/no); details of the death itself; and the cause of death. ²⁶ Respondents report the final cause of death as recorded by a medical examiner or through another official medical investigation. The DCRP uses the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD‐10) codes for cause of death. This standardized DCRP reporting form allows respondents to select “yes” or “no” for questions regarding the medical condition determined to be the cause of death such as whether it was preexisting before incarceration and whether the incarcerated individual received evaluation, diagnostic testing, medication, or nonmedication treatments including surgery for this condition, not including emergency care. Therefore, DCRP data on correctional care and treatment are reported as binary without further details on specific testing and medications provided. The DCRP had a 100% response rate among 50 state correctional departments during the study period. ²⁶ The deceased individual's race and ethnicity was acquired from the DCRP, which is collected by BJS self‐reported prison surveys upon prison admission and uses the Office of Management and Budget definitions of race and ethnicity (ethnicity [Hispanic] as a separate category, and race categories are defined exclusive of ethnicity). ²⁷

For this analysis, we included only data from the state prison system, excluding deaths that occurred in jails and juvenile correctional facilities (n=20 321 total deaths, of which n=2017 were cardiovascular deaths). The overall number of people incarcerated annually in state prisons was obtained from the BJS for state prisons, which provides demographic information on all individuals in custody annually. ²⁸

Study Population

The study population consisted of all individuals who had a death in state prisons while incarcerated from 2001 to 2019.

Cardiovascular Death

We identified individuals who died due to cardiovascular disease on the basis of ICD codes as defined in Tables S1. We examined death rates of all combined cardiovascular causes and each separate cardiovascular disease subset.

Rate and Trends in Cardiovascular Death

We determined the annual number of cardiovascular deaths from 2001 to 2019, including by racial and ethnic group as well as the proportion of all deaths that were due to cardiovascular causes. We determined national mortality rates using the BJS annual prison population at the end of the year as the denominator for the overall population and for each racial and ethnic group separately using annual numbers of individuals incarcerated by race and ethnicity. ²⁸

Statistical Analysis

We calculated summary measures for demographic and clinical variables for each individual in the database. Annual mortality rates for the overall population were calculated, and rates were compared among the following subgroups of race and ethnicity (using Pearson's χ² or Fisher's exact test where appropriate [Fisher's exact test used when ≥1 cells had values <5]): American Indian/Alaska Native, Asian/Native Hawaiian or Other Pacific Islander, Black, Hispanic or Latino, and White. We also compared the proportion of deaths that were due to cardiovascular diseases among individuals from different race and ethnicity subgroups. Age‐standardized annual mortality rates were determined for the overall population using the 2000 US census population. ²⁹ To evaluate for deviations in trends, the annual percentage change of the age‐adjusted mortality rate was evaluated using a Joinpoint model. BJS does not provide custody data on age groups for each racial and ethnic group precluding age standardization for mortality rates by race and ethnicity.

To determine the association of sociodemographic factors and care delivery, we performed a multivariable, mixed‐effects logistic regression analysis, with the outcome of interest being a composite treatment variable of patients having received any of the following: evaluation by medical staff, had diagnostic test performed, treatment with medications, treatment other than medications, or surgery. The following covariates were included in the model: age group, race and ethnicity, sex, amount of time imprisoned when death occurred, state (as random intercept), geographic region, and timing of development of medical condition causing death (preexisting versus after prison admission). We also fit a fixed‐effect model including state fixed effects to characterize the association of each state with the composite treatment variable.

Individuals with missing sex or race were excluded. Multiple imputations were performed for the timing of development of medical condition causing death variable (whether the condition was preexisting or not before incarceration) given the frequency of missingness of this variable (30.6%). Imputation was conducted using the Multiple Imputation by Chained Equations algorithm with the use of PROC MI in SAS (SAS Institute, Cary, NC). A logistic regression model was used as the imputation model and included the following variables: age category, race and ethnicity, evaluation by medical staff, year, time served, state, and ICD‐10 cause of death. We created 31 imputed data sets because 31% of the subjects had missing data. ³⁰

All analyses were performed using SAS version 9.4. We separately used the National Cancer Institute Joinpoint Regression Program 5.3.0.0 to analyze change in trends of the age‐adjusted mortality rate. Adjustment for multiple comparisons was not made, and subgroup analyses should be considered exploratory.

RESULTS

From 2001 to 2019, there were a total of 65 021 deaths among individuals incarcerated at US state prisons. Of those, 18 227 (28.0%) were due to a cardiovascular cause.

Among individuals with a cardiovascular cause of death, 30.4% (n=5532) were aged >65 years and 3.4% (n=627) were women (Table 1). Of this population, 52.3% (n=9535) were White individuals, 37.8% (n=6887) Black individuals, and 8.2% (n=1497) Hispanic individuals. Of all cardiovascular deaths, the majority occurred in the South (52.8%), followed by the Midwest (18.2%) and the West (15.7%), with the lowest proportion in the Northeast (13.4%). The largest proportion of deaths due to cardiovascular death (40%) occurred after 120 months of incarceration. Among all deaths due to cardiovascular disease, 77% were due to ischemic heart disease, 10.8% due to hypertensive disorders, and 6.2% due to heart failure (Table 2). Cardiovascular death rates for each state overall and by racial and ethnic group are summarized in Table S2.

Table 1.

Baseline Characteristics of Individuals With Cardiovascular Death While Incarcerated at State Prisons in the United States From 2000 to 2019 by Racial Group^*

	Overall (18 227)†	White (n=9535)	Black (n=6887)	Hispanic or Latino (n=1497)	American Indian/Alaska Native (n=159)	Asian/Native Hawaiian or Other Pacific Islander (n=85)	P value
Age, y							<0.0001
≤34	1013 (5.6)	307 (3.2)	593 (8.6)	92 (6.2)	12 (7.6)	5 (5.9)
35–44	2264 (12.4)	861 (9.0)	1134 (16.5)	222 (14.8)	22 (13.8)	16 (18.8)
45–54	4422 (24.3)	1908 (20.0)	2006 (29.1)	428 (28.6)	43 (27.0)	22 (25.9)
55–64	4996 (27.4)	2675 (28.1)	1872 (27.2)	375 (25.1)	41 (25.8)	15 (17.7)
≥65	5532 (30.4)	3784 (39.7)	1282 (18.6)	380 (25.4)	41 (25.8)	27 (31.8)
Sex							<0.0001
Male	17 598 (96.6)	9230 (96.8)	6615 (96.1)	1461 (97.6)	150 (94.3)	81 (95.3)
Female	627 (3.4)	305 (3.2)	272 (4.0)	36 (2.4)	9 (5.7)	4 (4.7)
Geographic region of state prison‡							<0.0001
Midwest	3313 (18.2)	1963 (59.3)	1224 (37.0)	84 (2.5)	31 (0.9)	4 (0.1)
Northeast	2439 (13.4)	1154 (47.3)	951 (39.0)	308 (12.6)	6 (0.3)	6 (0.3)
South	9623 (52.8)	4888 (50.8)	4057 (42.2)	608 (6.3)	46 (0.5)	15 (0.2)
West	2852 (15.7)	1530 (53.7)	655 (23.0)	497 (17.4)	76 (2.7)	60 (2.1)
Timing of development of medical condition causing death							<0.0001
Medical condition causing death WAS Preexisting	17 598 (96.6)	4373 (45.9)	2877 (41.8)	795 (53.1)	78 (49.1)	42 (49.4)
Developed after admission to prison	627 (3.4)	2235 (23.4)	1778 (25.8)	372 (24.9)	34 (21.4)	16 (18.8)
Missing	5584 (30.6)	2927 (30.7)	2232 (32.4)	330 (22.0)	47 (29.6)	27 (31.8)
Evaluated by a medical physician or medical staff for the condition causing death (not including emergency care)	14 702 (80.7)	7810 (81.9)	5413 (78.6)	1234 (82.4)	128 (80.5)	70 (82.4)	<0.0001
Had diagnostic test performed for the medical condition causing death (not including emergency care)	12 356 (67.8)	6580 (69.0)	4496 (65.3)	1080 (72.1)	101 (63.5)	58 (68.2)	<0.0001
Treated with medications for medical condition causing death (not including emergency care)	12 923 (70.9)	6921 (72.6)	4693 (68.1)	1083 (72.3)	119 (74.8)	64 (75.3)	<0.0001
Treatment other than medications administered for condition causing death (not including emergency care)	9453 (51.9)	5034 (52.8)	3381 (49.1)	882 (58.9)	74 (46.5)	51 (60.0)	<0.0001
Treatment with surgery for condition causing death (not including emergency care)	2681 (14.7)	1550 (16.3)	884 (12.8)	192 (12.8)	28 (17.6)	16 (18.8)
Time served when death occurred							<0.0001
<6 mo	1843 (10.1)	854 (9.0)	745 (10.8)	194 (13.0)	25 (15.7)	17 (20.0)
6–11.9 mo	1154 (6.3)	534 (5.6)	477 (6.9)	122 (8.2)	13 (8.2)	1 (1.2)
12–23.9 mo	1660 (9.1)	868 (9.1)	610 (8.9)	152 (10.2)	14 (8.8)	9 (10.6)
24–35.9 mo	1145 (6.3)	614 (6.4)	391 (5.7)	121 (8.1)	10 (6.3)	7 (8.2)
36–59.9 mo	1900 (10.4)	1039 (10.9)	663 (9.6)	170 (11.4)	16 (10.1)	7 (8.2)
60–119.9 mo	3089 (17.0)	1752 (18.4)	1036 (15.0)	252 (16.8)	21 (13.2)	15 (17.7)
≥120 mo	7436 (40.8)	3874 (40.6)	2965 (43.1)	486 (32.5)	60 (37.7)	29 (34.1)
Medical examiner evaluation							<0.0001
Yes	16 731 (91.8)	8717 (91.4)	6401 (92.9)	1342 (89.7)	142 (91.8)	70 (82.4)
No	1270 (7.0)	698 (7.3)	404 (5.9)	136 (9.1)	11 (6.9)	15 (17.7)
Missing	226 (1.2)	120 (1.3)	82 (1.2)	19 (1.3)	2 (1.3)	0 (0)

The black shade was to censor cells with less than 5 people to protect the identify of these people. Any cells with 5 or less people should be shaded black.

^*Any number <10 is censored to protect individual privacy.

^† A total of 64 (0.35%) had missing race information.

^‡ Regions of the United States are defined as South: Alabama; Arkansas; Delaware; Washington, DC; Florida; Georgia; Kentucky; Louisiana; Maryland; Mississippi; North Carolina; Oklahoma; South Carolina; Tennessee; Texas; Virginia; West Virginia; Northeast: Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont; Midwest: Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, Wisconsin; West: Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, Wyoming.

Table 2.

Causes of Cardiovascular Death Based on ICD‐10 Codes Among Incarcerated Individuals in State Prisons From 2000 to 2019^*

	White (n=34 262)	Black (n=22 274)	Hispanic or Latino (n=7122)	American Indian/Alaska Native (n=690)	Asian/Native Hawaiian or Other Pacific Islander (n=407)	P value
Total deaths due to cardiovascular disease	9535 (27.8)	6887 (30.9)	1497 (21.0)	159 (23.0)	85 (20.9)	<0.0001
Ischemic heart disease†	7507 (78.7)	5168 (75.0)	1126 (75.2)	126 (78.6)	55 (64.7)	0.0036
Acute myocardial infarction	1088 (11.4)	657 (9.5)	169 (11.3)	21 (13.2)	0 (0)	0.0057
Other acute ischemic heart diseases	126 (1.3)	90 (1.3)	4 (0.3)	4 (2.5)	0 (0)	<0.0001
Other forms of chronic ischemic heart disease	2067 (21.7)	1065 (15.5)	265 (17.7)	26 (16.4)	11 (12.9)	<0.0001
Atherosclerotic cardiovascular disease	572 (6.0)	281 (4.1)	61 (4.07)	13 (8.2)	4 (4.7)	<0.0001
All other forms of chronic ischemic heart disease	1154 (12.1)	594 (8.6)	172 (11.5)	11 (6.9)	7 (8.2)	0.0044
Atherosclerosis	41 (0.43)	18 (0.3)	5 (0.3)	1 (0.6)	0 (0)	<0.0001
Heart failure	661 (6.9)	370 (5.4)	82 (5.5)	17 (10.7)	5 (5.9)	0.0008
Hypertensive cardiovascular death causes	1013 (10.6)	815 (11.8)	125 (8.4)	13 (8.2)	8 (9.4)	0.0082
Hypertensive heart disease	918 (9.6)	707 (10.3)	116 (7.8)	8 (5.0)	6 (7.1)	0.873
Hypertensive heart and renal disease	2 (0.02)	0 (0)	0 (0)	0 (0)	0 (0)	0.0003
Essential (primary) hypertension and hypertensive renal disease	93 (1.0)	108 (1.6)	9 (0.6)	5 (3.1)	2 (2.4)	0.0852
Acute and subacute endocarditis	49 (0.5)	22 (0.3)	13 (0.9)	1 (0.6)	0 (0)	0.3497
Diseases of pericardium and acute myocarditis	67 (0.7)	67 (1.0)	15 (1)	2 (1.3)	0 (0)	0.9705
Acute rheumatic fever and chronic rheumatic heart diseases	7 (0.07)	5 (0.07)	2 (0.1)	0 (0)	0 (0)	<0.0001
Cerebrovascular disease	807 (8.5)	784 (11.4)	215 (14.4)	17 (10.7)	17 (20.0)	<0.0001
Pulmonary hypertension and pulmonary heart disease‡	288 (0.8)	398 (1.8)	66 (0.9)	5 (0.7)	1 (0.3)	<0.0001
Nonrheumatic valvular disease§	73 (0.2)	29 (0.1)	14 (0.2)	2 (0.29)	1 (0.3)	0.29
Cardiomyopathy‖	131 (0.4)	209 (0.9)	37 (0.5)	2 (0.3)	1 (0.3)	<0.0001
Arrhythmia and conduction disease#	2217 (6.5)	1826 (8.2)	358 (5.0)	32 (4.6)	18 (4.4)	<0.0001
Cardiac arrest	1769 (5.2)	1381 (6.2)	276 (3.9)	22 (3.2)	12 (3.0)	<0.0001
Atrial fibrillation and flutter	7 (0.02)	5 (0.02)	1 (0.01)	0 (0)	0 (0)	0.99
Paroxysmal tachycardia	7 (0.02)	2 (0.01)	0 (0)	0 (0)	0 (0)	0.76
Atrioventricular and left bundle‐branch block	3 (0.01)	1 (0)	1 (0)	0 (0)	0 (0)	0.98
Other conduction disorders	10 (0.03)	5 (0.02)	0 (0)	0 (0)	0 (0)	0.77
Other cardiac arrhythmias	421 (1.23)	432 (1.94)	80 (1.12)	10 (1.45)	6 (1.47)	<0.0001
Other heart disease complications and ill‐defined descriptions of heart disease	699 (2.0)	416 (1.9)	98 (1.4)	13 (1.9)	6 (1.5)	0.01

ICD indicates International Statistical Classification of Diseases and Related Health Problems. ICD codes for other diagnoses are summarized in Table S1. The black shade was to censor cells with less than 5 people to protect the identify of these people. Any cells with 5 or less people should be shaded black.

^*Any number <10 is censored to protect individual privacy. A total of 64 (0.35% had missing race information).

^† Ischemic heart disease includes Ischemic heart diseases (I20–I25), Acute myocardial infarction (I21–I22), Other acute ischemic heart diseases (I24), Other forms of chronic ischemic heart disease (I20–I25), Atherosclerotic cardiovascular disease, so described (I25.0) All other forms of chronic ischemic heart disease (I20, I25.1–I25.9), and atherosclerosis (I70).

^‡ ICD‐10 codes I26–I28.

^§ ICD‐10 codes I34–I38.

^‖ ICD‐10 codes I42 and I43.

^# ICD‐10 codes I44–49 (I44, Atrioventricular and left bundle‐branch block; I45, Other conduction disorders; I46, Cardiac arrest, I47, Paroxysmal tachycardia; I48, Atrial fibrillation and flutter; I49, Other cardiac arrhythmias).

From 2001 to 2019, the crude all‐cause mortality rate among all incarcerated individuals increased from 242.0 to 330.4 per 100 000 persons. The crude cardiovascular disease–related mortality rate increased from 71.3 to 94.0 per 100 000 persons (Figure 1A). Age‐standardized mortality rates decreased from 200.9 to 117.3 per 100 000 from 2001 to 2012 but then increased to 130.1 per 100 000 in 2019 (Figure 1A). A statistically significant Joinpoint was detected in 2012 with an annual percentage change of −4.94% (95% CI, −5.68% to −4.35%) between 2001 and 2012 and 2.09% (95% CI, 0.687%–3.92%) between 2012 and 2019. From 2001 to 2019, the cardiovascular mortality rate increased from 108.0 to 164.6 per 100 000 among incarcerated White individuals, 65.5 to 97.5 per 100 000 people among Black incarcerated individuals, and 31.2 to 36.2 per 100 000 people among Hispanic individuals (Figure 1B).

Figure 1. Trends in death rates (crude and age‐adjusted) from cardiovascular death among state prison population from 2001 to 2019 (A) and crude death rates from cardiovascular death among state prison population by racial groups from 2001 to 2019 (B).

Among those who died while incarcerated, the proportion of all deaths that were due to cardiovascular causes decreased from 29.5% in 2001 to 25.9% in 2012 and then increased annually back up to 29.0% in 2019. Trends in the proportion of deaths overall and those due to a cardiovascular cause by race and ethnicity from 2001 to 2019 are summarized in Figure S1. The age distribution of the incarcerated population used for age standardization is shown in Figure S2.

Of incarcerated individuals who died from a cardiovascular cause (n=18 227), 19.9% (n=3525) were not evaluated by medical staff before death, 32.2% (n=5871) did not have a diagnostic test performed, and 29.1% (n=5304) did not receive any medications for the cardiovascular condition causing death. Trends in health care receipt are shown in Figure 2 (with annual rates summarized in Table S3). There was significant geographic variation in care delivery; individuals incarcerated in US state prisons in the South had the highest rates of not being evaluated by medical staff (21.3%), of not receiving any diagnostic testing (35.9%), of not receiving treatment with medications (32.8%), and not receiving any medical treatment other than medications (48.9%) for the condition that caused death (Table S4). Variability in rates of care provision by state are shown in Table S5. Forest plots of the odds by state of receiving care or treatment for the cardiovascular condition causing death on multivariable mixed‐effect analysis by state are summarized in Figure S3 (with Alabama having the lowest odds and New Hampshire the highest odds of care provision).

Figure 2. Rates of care receipt for the cardiovascular condition causing death among incarcerated individuals who died from a cardiovascular cause from 2001 to 2019.

Black individuals who died from a cardiovascular cause were younger on average as compared with other racial and ethnic groups (Table 1). Black individuals were more likely than White individuals to have a cardiovascular death occur within 6 months of time served (10.8% versus 9.0%) but also had higher rates of death occurring after ≥120 months of time served (43.1% versus 40.6%, P<0.001). Black individuals were less likely to have the medical condition causing death to be considered preexisting before incarceration than White individuals (41.8% versus 45.9%, P<0.001). Of all deaths occurring while incarcerated, Black individuals were more likely to have a death related to a cardiovascular cause than White individuals (30.9% versus 27.8%, P<0.001) (Table 2). Black individuals had higher rates of hypertensive‐related cardiovascular death (11.8% versus 10.6%, P<0.001) and death from cerebrovascular disease (11.4% versus 8.5%, P<0.001) as compared with White individuals. Cardiovascular deaths due to heart failure were the highest among American Indian/Alaska Native individuals. The racial composition of those with cardiovascular death annually are summarized in Figure S4.

In univariable analysis, Black individuals with cardiovascular death while incarcerated were less likely to be evaluated by a physician or medical staff at the prison than White individuals (78.6% versus 81.9%, P<0.0001) and were less likely to have a diagnostic test performed (65.3% versus 69%, P<0.0001) for the medical condition causing death (Table 1). Black individuals were also less likely to receive medications (68.1% versus 72.6%, P<0.0001) and were less likely to receive treatment other than medications (49.1% versus 52.8%, P<0.0001) for the condition determined to be the cause of death. Rates of diagnostic testing and treatment other than medicine for the condition determined to be the cause of death were also significantly lower among American Indian/Alaska Native individuals as compared with White counterparts.

In the multivariable analysis, the odds of receiving any care or treatment before death did not differ significantly between White and Black individuals (adjusted odds ratio, 1.09 [95% CI, 0.97–1.21]) (Table 3). Odds differed significantly between regions, with individuals incarcerated in prisons in the South (adjusted odds ratio, 0.51 [95% CI, 0.28–0.91]) and West regions (adjusted odds ratio, 0.52 [95% CI, 0.30–0.90]) having significantly lower odds of treatment than those in the Northeast. Younger age and the condition causing death being preexisting (versus developed after admission) were strongly associated with care/treatment receipt.

Table 3.

Factors Associated With Receipt of Medical Evaluation or Treatment for the Condition Causing Death Among Those Incarcerated in State Prisons in the United States on Multivariable Mixed‐Effects Logistic Regression Analysis

Variable	Odds ratio	95% CI	P value
Race and ethnicity (White as reference*)
Black	1.09	0.97–1.21	0.151
Hispanic	1.21	1.02–1.45	0.033
American Indian/Alaska Native	0.86	0.55–1.36	0.527
Asian	1.46	0.67–3.17	0.341
Age category, y (<34 y as reference)
35–44	1.42	1.21–1.66	<0.0001
45–54	1.84	1.53–2.21	<0.0001
55–64	2.69	2.02–3.57	<0.0001
≥65	3.87	2.88–5.21	<0.0001
Sex (male as reference)
Female	1.73	1.29–2.32	0.0003
Time served, mo (<6 mo as reference)
6–11.9	1.17	0.88–1.56	0.281
12–23.9	1.16	0.95–1.42	0.141
24–35.9	1.06	0.83–1.36	0.624
36–59.9	1.24	1.00–1.54	0.048
60–119.9	1.21	0.95–1.53	0.127
≥120	1.45	1.15–1.83	0.002
Timing of medical condition development (developed after admission to prison as reference)
Preexisting	2.90	2.45–3.44	<0.0001
Region (Northeast as reference)†
Midwest	0.65	0.37–1.12	0.122
South	0.51	0.28–0.91	0.022
West	0.52	0.30–0.90	0.019
Year (2001 as reference)
2002	2.36	1.16–4.80	0.017
2003	2.53	1.11–5.73	0.027
2004	2.70	1.14–6.4	0.024
2005	4.13	1.84–9.28	0.001
2006	5.38	2.29–12.62	0.000
2007	7.63	3.13–18.62	<0.0001
2008	6.39	1.94–21.05	0.002
2009	2.63	0.6–11.61	0.202
2010	6.24	2.37–16.46	0.000
2011	4.78	1.74–13.14	0.002
2012	4.81	1.9–12.19	0.001
2013	4.11	1.5–11.24	0.006
2014	3.54	1.6–7.81	0.002
2014	4.31	1.76–10.55	0.001
2015	3.53	1.62–7.7	0.002
2016	3.04	1.2–7.71	0.019
2017	3.00	1.17–7.7	0.022
2018	2.51	1.02–6.18	0.045
2019	2.36	1.16–4.8	0.017

^*White selected as reference given highest mortality rate at baseline.

^† Regions of the United States are defined as South: Alabama; Arkansas; Delaware; Washington, DC; Florida; Georgia; Kentucky; Louisiana; Maryland; Mississippi; North Carolina; Oklahoma; South Carolina; Tennessee; Texas; Virginia; West Virginia; Northeast: Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont; Midwest: Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, Wisconsin; West: Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, Wyoming.

DISCUSSION

In a large study of all individuals who died while incarcerated in US state prisons from 2001 to 2019, our results demonstrate a considerable burden of death from cardiovascular causes. Additionally, a substantial portion of deceased individuals did not receive evaluation or treatment before death. These results provide, to our knowledge, the first evaluation of care provided by the state carceral health care system for cardiovascular conditions. As mass incarceration continues to be a pervasive crisis in the United States, these results highlight the critical need for strategies to ensure quality cardiovascular care for those incarcerated. We noted significant regional differences in access to health care. In addition, Black individuals who died in prison had higher rates of death from cardiovascular causes and lower rates of care receipt.

The United States has the highest incarceration rate of any country in the world. ¹ , ² Given the complex ways in which incarceration intersects with health, mass incarceration causes enormous public health harms. ² , ¹⁴ While efforts to dismantle the prison‐industrial complex are ongoing, population health initiatives must also focus on ensuring access to quality cardiovascular care for incarcerated individuals. While prisons are legally obligated to provide adequate medical care to those incarcerated, ³ , ⁴ , ⁵ this is often not the case. ⁶ While rates of provision of medical services increased throughout the study period, >20% of individuals who died from cardiovascular disease were not evaluated by a medical professional, >30% had no diagnostic testing, and >25% were not given medications for the condition that caused death in 2019.

Additionally, we found significant variation in care provision between geographic regions and states. There are no federal regulations on health care standards in prison systems. The National Commission on Correctional Health Care does publish standards, which provides best practices for constitutionally acceptable health services delivery, including for chronic diseases, and also offers accreditation for prison health care systems. ³¹ However, accreditation is voluntary with low uptake, and there remains a lack of comprehensive standardized regulations for cardiovascular screening, evaluation, and management. ³² As a result, there is substantial variability in basic cardiovascular health screening and assessments. ³³ Given the large proportion of deaths that occurred after 120 months, implementation of 10‐year cardiovascular risk prediction models may allow for early detection of those at highest cardiovascular risk. Additionally, Medicaid and Medicare are unable to pay for health care services for individuals while incarcerated. ³⁴ Enacting policies to allow for coverage continuation in prison may be 1 mechanism to allow for increased oversight and more standardized quality care provision.

Incarcerated individuals are much less likely to be diagnosed with chronic diseases after admission than on admission. ³⁵ In our cohort, most cardiovascular deaths (40.8%) occurred after 120 months of time served. Additionally, nearly a quarter of cardiovascular deaths were from a condition that developed after admission to prison. This may be due to accelerated aging, with every year of incarceration associated with a 2‐year reduction in life expectancy. ¹⁸ Incarcerated individuals have been shown to have higher rates of cardiovascular risk factors as compared with the general population, including hypertension, diabetes, and obesity. ⁹ , ¹⁰ , ¹¹ , ³⁶ , ³⁷ , ³⁸ , ³⁹ Incarceration produces and exacerbates cardiovascular risk factors by subjecting individuals to dehumanizing conditions, chronic stress, poor diet, variable access to physical activity, coupled with barriers to accessing quality health care. ¹² , ¹³ , ¹⁴ , ¹⁵ Incarceration has been shown to be a potential independent risk factor for cardiovascular disease development. ³⁶ Having a preexisting condition before incarceration was the factor most strongly associated with receipt of treatment for the condition causing death on multivariable analysis. Therefore, those without a known cardiovascular condition or who develop a condition while incarcerated are particularly vulnerable to not receiving needed care. In fact, cardiovascular risk has been shown to be almost a decade higher in those aged ≥40 years in the prison population versus the general community, which calls for more rigorous standardized screening assessments. ⁴⁰ Given that cardiovascular care often requires specialty and subspecialty care, cardiovascular health is particularly vulnerable to being suboptimal among the prison‐incarcerated population. A paucity of data on access to and quality of correctional health care delivery precludes a better understanding of potential targets for intervention to improve cardiovascular health in this population. ⁴¹ There exists a critical need for increased transparency and public reporting on health care delivery and outcomes as a requirement for all carceral systems. Carceral policies must be addressed to guarantee the human right to quality food and dignified health care. ¹² , ⁴²

After declining for more than a decade, the age‐adjusted cardiovascular mortality rate increased significantly between 2012 and 2019, signaling that cardiovascular disease is an increasingly important cause of death. Although race and ethnicity subgroup‐specific mortality rates could not be age adjusted, the crude mortality rate for both White and Black individuals also increased in recent years.

Mass incarceration is a sociopolitical driver of health inequities. ² , ¹² Currently, 1 in 5 Black men born after 2001 will be incarcerated. ⁴³ Our results demonstrate a high proportion of death among Black incarcerated individuals from cardiovascular causes, along with concerning racial differences in carceral care provision. While Black race was not independently associated with treatment receipt, Black patients at baseline were less likely to receive treatment, which may be due to differences in baseline characteristics such as age and diagnosis of preexisting conditions which were strong predictors of care in multivariable analysis. Black individuals were less likely to have the medical condition causing death to be present before incarceration. While this may represent known racial inequities in care access before admission to prison, it is also possible that inequities in care delivery while incarcerated perpetuate disparities in disease development and outcomes. ⁴⁴ Given the racialization of mass incarceration, dismantling carceral structures that perpetuate health inequities is necessary. ²⁰ , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷

These results are rooted within the historic and ongoing impacts of structural racism, the oppressive systems and structures that disenfranchise and marginalize Black individuals in the United States. ⁴⁸ , ⁴⁹ Inequitable policies and racist practices have led to sustained and targeted divestment in communities of color, concentrated poverty, and economic distress, which increase both the risk of incarceration and development of cardiovascular disease. ⁵⁰ , ⁵¹ , ⁵² , ⁵³ , ⁵⁴ , ⁵⁵ , ⁵⁶

“Race‐neutral” policies that are inherently racist, such as the “war on drugs,” and a criminal‐legal system entrenched with racism and classism, have disproportionately institutionalized Black individuals in the United States. ⁵⁷ , ⁵⁸ The health implications of mass incarceration are profound, not only for the justice‐involved individual, but they have ripple effects among family and community members through not only direct health effects (ie, increased psychosocial stress) but also indirectly shape health through depleting community resources, societal disruption, and perpetuating neighborhood disadvantage. ² , ⁵⁹ , ⁶⁰ The mental, physical, and economic consequences of mass incarcerations converge to fuel poor health; the results of this study add to the complex ways in which incarceration harms health.

Our study has several limitations. This database included only state prison death data; the health of those in federal prison has been shown to differ from those in state prison. ⁶¹ However, close to 90% of the incarcerated population in the United States is incarcerated in state rather than federal prisons. ⁶² We were reliant on ICD coding for cause of death; however, 92% of the deaths were evaluated by a medical examiner. While reporting in the DCRP is standardized and there was 100% response rate of state correctional facilities during the study period, the accuracy of facility reporting may vary between facilities. We are unable to know the access to and quality of pre‐incarceration care, which may impact outcomes. This study evaluated only deaths occurring while the individual was incarcerated in prison. The majority of people who are incarcerated in prison return to the community. Therefore, this study does not fully capture the impact of incarceration on cardiovascular death more broadly for all individuals who have been incarcerated. Additionally, the racial and geographic differences in care receipt demonstrated were only among those who died while incarcerated and may not reflect the entire prison‐incarcerated population. While the BJS reports rates of evaluation, diagnostic testing, medical treatment for the condition causing death as a binary variable, we lack further granularity on the type or quality of care given (ie, which medications were given, which testing performed), or the timing of care in relation to the death other than it does not include emergency care. Additionally, given the amount of missingness for some variables (ie, whether the condition causing death was preexisting before incarceration), the reliability of these forms is unclear. This further supports the need for mandated reporting on quality care metrics among all prisons nationally. Finally, the BJS does not maintain custody data on racial and ethnic groups by age group, precluding age standardization for cardiovascular mortality rates for each racial and ethnic group. The lower rates of cardiovascular death for Black patients among those in custody is likely reflective of the younger age of Black incarcerated individuals. ²⁶ , ²⁷ Although adjustment for multiple comparisons was not made, based on the total number of univariable and multivariable hypothesis tests (n=20) and an α of 0.05, based on the family‐wise error rate, we estimate a 64.2% chance that at least 1 test would be statistically significant by chance. ⁶³ Finally, this study represents the US prison system only, but cardiovascular disease in the criminal justice population is a global public health issue. ³⁸ , ³⁹ , ⁶⁴ , ⁶⁵ , ⁶⁶ , ⁶⁷ , ⁶⁸

In conclusion, deaths due to cardiovascular causes are high among individuals incarcerated in US state prisons. Decarceral efforts to address mass incarceration must be coupled with strategies to ensure quality and equitable cardiovascular care for this population.

Sources of Funding

Dr Eberly is funded by grants from the National Heart, Lung, and Blood Institute (1K23HL166974‐01A1), the American Heart Association (23CDA1050650), and the Robert A. Winn Diversity in Clinical Trials Career Development Award. Dr Khatana is funded by grants from National Heart, Lung, and Blood Institute (K23HL153772 and R01HL171157).

Disclosures

None.

Supporting information

Tables S1–S5

Figures S1–S4