|Home | About | Journals | Submit | Contact Us | Français|
The prevalence of hepatitis C virus (HCV) infection is high among adult incarcerated populations, but HCV-related mortality data are lacking. The study purpose was to assess HCV-related mortality over time and across racial/ethnic categories from 1994 through 2003 among male prisoners in the Texas Department of Criminal Justice (TDCJ).
TDCJ decedent data were linked with Texas Vital Statistics multiple-cause-of-death data. Crude annual HCV death rates, age- and race-adjusted summary rates, and average annual percent changes were estimated. The proportion of deaths due to chronic liver disease/cirrhosis, liver cancer, hepatitis B, and HIV for which HCV was identified as an intervening or contributing cause of death was calculated.
Among Texas male prisoners, HCV death rates were high and increased over the 10-year study period by an average 21% annually, with the largest increase occurring among Hispanic prisoners. HCV was identified as an intervening or contributing cause of death in 15% of chronic liver disease/cirrhosis deaths, 33% of liver cancer deaths, 81% of hepatitis B deaths, and 7% of HIV deaths.
Because HCV-related deaths among Texas male prisoners are high and increasing, particularly among Hispanics, targeted prevention, screening, and treatment of HCV infections should be among the priorities of U.S. correctional healthcare systems.
Chronic hepatitis C infection and its sequelae represent a significant public health problem in the United States. An estimated 3.2 million Americans are chronically infected with hepatitis C virus (HCV) (1, 2). These chronic infections translate into 8,000–10,000 estimated annual deaths due to chronic liver disease/cirrhosis deaths (CLD) (3). Between 1% and 5% of individuals with chronic HCV infection will develop primary liver cancer (PLC) (4, 5). Annual costs for HCV-associated end stage liver disease in the United States have been estimated at more than $600 million (3). The highest prevalence of HCV infection is currently among persons between 30–49 years of age (1, 3). As this cohort ages over the next 2 decades, HCV-related morbidity, mortality, and associated costs are expected to increase (4).
Chronic HCV infection and related conditions have a disproportionate impact on U.S. correctional healthcare systems. Studies over the past decade have consistently shown high prevalence of HCV serologic markers in incarcerated populations, ranging from ~20%–40% (6–16). Recent studies have indicated high prevalence of end stage liver disease and liver cancer in prison populations, with elevated risk for these conditions among prisoners with HCV infection (17, 18). High prevalence of HCV infection and related conditions must be considered in the context of a prison population that is both growing and aging (19). The U.S. prison population has more than quadrupled since 1980, exceeding 2 million in 2002 (20), and inmates over 45 years of age comprised 21.3% of all inmates in 2007, up from 13.6% in 1997 (21, 22). As a result, correctional healthcare systems are expected to experience substantial increases in HCV-related conditions and associated costs in the coming years (16).
Despite the high prevalence of HCV and related conditions in incarcerated populations (6–16), there are no published studies describing the extent to which HCV contributes to mortality in a U.S. correctional population. Because HCV-related mortality data may serve as a useful starting point for correctional healthcare planners and policy-makers in assessing, predicting, and addressing the burden of these conditions, we conducted a study to assess the extent of HCV-related mortality in the nation's largest state prison system (23). Mortality data obtained from the Texas Department of Criminal Justice (TDCJ) were linked with Texas Vital Statistics multiple-cause-of-death data. Linked data were used to examine trends in HCV death rates from 1994 through 2003 among male prisoners in TDCJ custody, to compare these death rates across racial/ethnic categories, and to quantify the proportion of CLD and PLC deaths that were related to HCV. Additionally, because prison populations have a relatively high prevalence of hepatitis B virus (HBV) and HIV infections (6, 7, 10, 11, 14), which are often comorbid with HCV, the proportion of HBV and HIV deaths that were HCV-related was also calculated.
A database containing information on all prisoners who died in TDCJ's custody during 1994–2003 was obtained from TDCJ. From this database, demographic and other identifying information on decedents were extracted and linked to a multiple-cause-of-death database obtained from the Texas Bureau of Vital Statistics. Vital statistics data had been drawn from death certificates, and then corrected for common data entry errors and for certifier errors in delineating multiple causes of death according to current selection rules (24). Causes of death were recorded on the death certificate by TDCJ healthcare providers after being ascertained by autopsy, chart reviews, narrative death summaries, and/ or findings of the Mortality Review Committee. As in noncorrectional medical settings, autopsy rates declined over the study period from greater than or equal to 90% of prisoner deaths for 1989–1997 to an average of 50% for 1998–2003 (J. Pulvino, personal communication, TDCJ December 2008); however, these autopsy rates were higher than autopsy rates reported in most noncorrectional medical settings for the entire study period (25). Use of TDCJ and vital statistics data was approved by two institutional review boards with prisoner advocate members (Texas Department of State Health Services and the University of Texas Houston Health Science Center) and was subsequently approved by TDCJ.
Average male prisoner census was used as the denominator for calculating annual death rates. “Snapshot” census data of the in-custody population is taken on the last day of each month by TDCJ. For each year, the average of the 12 monthly snapshot censuses for males within each stratum (race/ethnicity by 10-year age categories) was calculated.
Six personal identifiers (first, middle, and last names, race/ethnicity, dates of birth and death) representing 10 separate codes were used to link the TDCJ decedent data set with the State decedent data set. Linkage was accomplished using the select query function in Microsoft Access in an iterative fashion (i.e., linkage attempted with all 10 codes, then with 9 codes, then with a different set of 9 codes, etc.). Individual links were checked for quality. A successful link required a 90% match (9 of 10 codes) with at least seven identical codes and no more than two codes reflecting an apparent match, defined as non-identical codes that subjectively appeared to be due to coding errors (e.g., inversion of first and last names, inversion of numbers in birth or death dates).
Prisoner deaths with HCV, CLD, PLC, HBV, and HIV indicated as the underlying cause of death were identified. ICD-9 and ICD-10 codes for these causes were as follows: HCV = 070.4, 70.5, B17.1, B18.2; CLD = 571.0–571.3, 571.8, 571.9, K70.09, K74.6; PLC = 155.0, 155.2, C22.0, C22.2, C22.9; HBV = 070.2–070.3, B16.0–16.9, B18.0–B18.1; HIV = 042–044, B20–B23. Similar to the methods used by Vong and Bell (26), deaths due to CLD, PLC, HBV, and HIV were considered HCV-related if HCV infection was recorded as an intervening or contributing cause of death. Intervening causes (or “antecedent” causes) are those listed in Part I of the death certificate but are not selected as the underlying cause of death. Contributing causes are listed in Part II of the death certificate.
The total number of recorded prisoner deaths was 3993 for the study period, 98% of which (n = 3916) were linked successfully to the state multiple-cause-of-death data. Of these linked deaths, 168 were among women and 118 were among males under the age of 25 or over the age of 84. Because very fewprisonerdeathsweredue to the causesof interestamong females or those outside the age range of 25–84 years, analyses were limited to male prisoners 25–84 years of age (n = 3630).
For deaths with HCV indicated as the underlying cause, crude annual death rates per 100,000 population along with 95% exact Poisson confidence intervals (CI) (27) were estimated for male decedents aged 25–84 years and stratified by race/ethnicity. Poisson regression on a log normal scale was used to model trends in crude annual HCV death rates. Regression coefficients were estimated using weighted least-squares, where the weights were equal to the death count for each year or an adjusted count of 0.5 if the observed death count was zero. Trends in crude HCV death rates were described by calculating the average annual percent change (AAPC) with 95% CI. The AAPC over any fixed interval is a weighted average of the slope coefficients of the underlying regression line with the weights equal to the number of years covered by the interval (28). If the CI for the AAPC did not contain zero, this supported the hypothesis that the true AAPC was different from zero at the 0.05 α level. The AAPC can be advantageous over the more traditional annual percent change, particularly when data are sparse or highly variable over a specified time interval, because the AAPC does not assume linearity over time and produces more stable estimates with narrower confidence intervals (28).
Crude average annual HCV death rates for 5-year periods per 100, 000 population along with 95% exact Poisson confidence intervals were also estimated. These summary rates were used to improve stability and precision in examining differences over time and across race/ethnicity categories and in age- and race-adjusting rates, given the small number of HCV deaths per year within each stratum (10-year age by race/ethnicity categories). Average annual rates were ageand race-adjusted using the direct method based on the composition of the male populations of the U.S. prison system, 1999, (29) and also the state of Texas, 2000 (30), which, respectively, are similar to the Texas prison population in their large proportions of younger persons and Hispanic persons. Because significant differences between death rates may exist even when the CI of the rates overlap (31), we calculated rate differences (RD) with 95% CI in cases where summary rates overlapped (32) and noted RD values that were significant at the 0.05 level.
Finally, to assess the burden of CLD, liver cancer, HBV, and HIV deaths that were related to HCV, we calculated the proportion of CLD, liver cancer, HBV, and HIV deaths (underlying cause) for which HCV was recorded as an intervening or contributing cause of death.
Stata 8.2 was used for initial data preparation and for production of frequencies and cross tabulations (33). Stats-Direct 2.7.2 was used to calculate crude annual and average annual HCV death rates, rate differences, and confidence intervals and to produce line graphs of regression models described below (34). Poisson regression models and AAPCs were produced using Joinpoint 3.3 (35), which has been used widely for analysis of cancer mortality rates (35–37).
From 1994 to 2003, the average daily census of TDCJ male prisoners, 25–84 years of age, increased more than 200%, from 64,864 to 116,640 prisoners. The proportion of White prisoners remained fairly steady (~30%); the proportion of Black prisoners declined from 47% to 40%; and the proportion of Hispanic prisoners increased monotonically from 24% to 28%. Prisoners aged 25–34 years accounted for 52% of the prison population in 1994 and monotonically declined to 38% in 2003. Conversely, prisoners aged 35–54 years accounted for 44% of the prison population in 1994 and increased to 56% in 2003. Prisoners aged 55–84 years accounted for 3% of the population in 1994 and 6% in 2003. In the 1999–2003 period, the proportion of middle-aged and elderly inmates (ages 45–84) was slightly higher in White prisoners (27%) than in Hispanic prisoners (19%) and Black prisoners (21%) (data not shown). Thus, over the study period, the size ofthe prison population grew considerably, and the proportions of older prisoners and Hispanic prisoners increased, whereas the proportion of Black prisoners decreased.
The number of deaths overall grew steadily from 287 in 1994 to 452 in 2000, then declined to 360 in 2003, whereas the proportion of prisoners who died each year remained fairly steady (0.4% in 1994 to 0.3% in 2003). HCV accounted for an increasing proportion of deaths over time, from an average annual 1% for the period from 1994 through 1997 to nearly 8% (28/360) in 2003 [data not shown]. HCV was recorded as the underlying cause of death for more than 3% of prisoner deaths during the study period (125/3630). The largest proportion of HCV deaths was among Hispanic prisoners (43%), followed by White prisoners (30%) and Black prisoners (28%).
From 1994 to 2003, HCV death rates increased in the study population overall and within each category of race/ethnicity (Table 1, Fig. 1). Confidence intervals for crude annual HCV death rates were wide, suggesting a lack of precision, and were overlapping across categories of race/ethnicity in most cases. However, the direction of these point estimates and their confidence intervals suggested increasing HCV death rates, particularly toward the end of the study period (~1999–2003) and among Hispanic prisoners. Confidence intervals for AAPCs were also somewhat wide and overlapping across categories of race/ethnicity. However, AAPC point estimates and confidence intervals suggested that HCV death rates increased at a slower rate among Black prisoners compared to White and Hispanic prisoners over the study period. HCV death rates seem to have increased at a somewhat faster rate among Hispanic prisoners compared to White prisoners.
These trends in HCV death rates were also shown by crude and adjusted average annual HCV death rates for the 1994–1998 and 1999–2003 periods (Table 2). Over time, Hispanic and White prisoners showed a significant increase from 1994–1998 to 1999–2003 in crude average annual HCV death rates, and all race/ethnic groups showed an increase over time in adjusted average annual HCV death rates. Across categories of race/ethnicity, no significant differences in crude rates were indicated for 1994–1998. However, for 1999–2003, Hispanic and White prisoners showed higher crude rates than Black prisoners [White to Black, RD = 7.6 (0.7–14.5)], and Hispanic prisoners showed higher crude rates than White prisoners [RD = 11.4 (1.2–21.6)]. Hispanic-White differences in rates adjusted to the 1999 U.S. prison population were significant for the 1994–1998 period [RD = 2.5 (0.3–4.7)].
Over the study period, HCV was recorded as an intervening or contributing cause for 9% of all deaths in this prisoner sample (332/3,603), nearly 15% of CLD deaths (27/186), 33% of liver cancer deaths (41/125), 81% of HBV deaths (20/24), and 7% of HIV deaths (41/575).
These data indicate a relatively high and growing burden of HCV-related deaths among male prisoners in Texas. Point estimates for crude annual HCV death rates in this sample of male prisoners ranged from 11.5 per 100,000 in 1999 to 24.0/100,000 in 2003. These rates far exceeded state and national rates, which respectively averaged only ~4/100,000 and ~3/100,000 for the same years (38). HCV death rates increased in this prisoner sample by an average of 21% annually over the study period, with the largest average annual percent increases occurring among Hispanic prisoners. Moreover, substantial proportions of HBV and liver cancer deaths (81% and 33%, respectively) and nontrivial proportions of CLD and HIV deaths (15% and 7%, respectively) were HCV-related.
Elevated HCV death rates among male prisoners may be explained in part by the higher prevalence in male prisoners of injection drug use, sexual risk behaviors, and alcohol use (6–14, 39–40), which are known risk factors for HCV infections and their sequelae. High volume, extended use of alcohol among prisoners may be particularly important. Indeed, as much as 30% of newly admitted male inmates in Texas meet diagnostic criteria for alcohol abuse and dependence (40). It should also be noted, although the evidence is somewhat limited, that tattooing may also contribute to elevated HCV prevalence and mortality in the prison population. Receipt of tattoos in prison has been associated with HCV prevalence in both prison and community samples (41–44).
The Hispanic-White difference in HCV death rates among male prisoners was similar to that difference seen in the general U.S. population over the study period (30). The disparity in the general population is commonly attributed to earlier initiation of risk behaviors for HCV among Hispanic males (45–47) and, thus, earlier age of infection (47), and to more frequent or heavier volume alcohol use among Hispanic males, particularly Mexican-American males, compared to White males (48–50). More recently, it has been suggested that Mexican-American males may be at increased risk for progression to liver disease because of a genetic predisposition, gene-environment interactions, or interaction of primary risk factors with comorbid conditions, such as obesity or diabetes (51–55). In Texas, nearly all Hispanic prisoners are Mexican-American, so any or all of these factors could potentially have contributed to elevated risk of HCV-related deaths among Hispanic prisoners observed in our study.
The study's limitations should be considered in interpreting these findings. Because of the heterogeneity of prison populations with respect to prevalence of HCV infection (6–16), our findings may not generalize to other male state prison populations. The Texas prison population was somewhat small for observing this relatively uncommon cause of death. However, the effects of random fluctuations in HCV-related deaths or their coding were diminished by using regression modeling and summary rates for comparing mortality over time and across race/ethnicity and by considering HCV as an intervening and/or contributing cause of death.
Increases in HCV mortality must be interpreted in light of common problems with mortality data based on death certificates. Misclassification bias may be introduced in coding and certifying cause of death because of changes in medical opinion or general change in emphasis regarding certain conditions over time (56, 57). In this case, increased awareness of the problem of HCV among U.S. prisoners incited changes in TDCJ screening policies in 1998, which resulted in increased HCV screening at that time, and may also have increased identification or selection of HCV as a cause of death for prisoners. Thus, apparent increases in HCV-related prisoner mortality around this time may reflect changes in screening and/or certification practices along with real trends. HCV screening protocols have remained consistent since 1998, however, so it seems unlikely that changes in screening procedures would account entirely for the continued and substantial increases in HCV-related mortality since that time. Furthermore, because HCV screening and testing protocols in TDCJ reflect community standards of care, it seems unlikely that differential screening or testing practices affected comparisons between TDCJ HCV death rates and general population rates.
Any study of HCV-related mortality among male prisoners in custody captures only a small fraction of the total burden of HCV-related morbidity on a large correctional healthcare system. It is highly probable that far fewer prisoners die from HCV-related conditions while in custody than are treated for these conditions and ultimately released to their communities. In a recent study of TDCJ prisoners, the number of inmates diagnosed with end-stage liver disease (n = 484) was approximately twice that of those who died from that condition (n = 213) over the same time period (2003–2006) (17). Additionally, an average of 63 ill prisoners was released annually on “special needs parole” or “medically related intensive supervision” over the study period, but the proportion of these releases that was HCV-related could not be determined.
Nonetheless, our study is the first to use existing multiple-cause-of-death data to assess HCV-related mortality among prisoners and represents an important step toward assessing the burden of these conditions on a correctional healthcare system. If these mortality data are viewed as suggestive of the extent of severe morbidity in the prison population and as indicative of future trends, these findings point to the great and growing burden of HCV and HCV-related conditions in prison populations. Costs of treating HCV-related conditions in prisons are considerable and will likely increase. Although annual per case costs of asymptomatic chronic hepatitis infection or mild cirrhosis have been estimated at just $145, annual per case treatment costs for compensated cirrhosis have been estimated at $1053 (U.S. dollars, 2007) (58). For the first year alone, annual per case costs related to severe cirrhosis complications such as ascites, variceal hemorrhage or hepatic encephalopathy have been estimated at $3765, $20,822, and $13,365, respectively (U.S. dollars, 2001) (59), and annual per case costs of treating hepatocellular carcinoma in prison have been estimated at $42,255 (U.S. dollars, 2007) (58).
Targeted prevention, screening, and treatment of HCV infection in prison populations have been recommended to reduce HCV-related morbidity, mortality, and costs (15, 60). Substance abuse treatment and behavioral risk reduction education for HCV-infected inmates and those with HCV risk factors may reduce morbidity among HCV-infected inmates by lessening alcohol and drug use after release and may also reduce HCV transmission by decreasing unsafe injection drug use after release (60). Targeted screening of inmates with HCV risk factors has been recommended to maximize identification of HCV-infected inmates while simultaneously avoiding the costs of universal screening (15, 60). Treatment of HCV-infected prisoners with pegylated interferon and ribavirin according to published guidelines has produced sustained virologic response at rates comparable to the general population (60). Finally, collaboration of correctional healthcare systems with community providers has been recommended to ensure continuity of medical care for HCV-infected inmates and to provide other services as well for the purposes of preventing re-infection and relapse (15, 60).
Despite these recommendations, current policies and practices regarding screening and treatment for alcohol and substance abuse and for HCV infection vary widely across state prison systems (61, 62). Cost is often cited as a primary barrier to the implementation of recommended strategies (60). It is widely thought, because a vast majority of prisoners return to their communities, that correctional healthcare systems would bear the initial financial brunt of implementing recommended strategies while directly reaping only a small fraction of the long-term benefits (60). In general, however, the cost-effectiveness of these recommended strategies has not been well-studied, so their relative costs and benefits to correctional healthcare systems and to the broader U.S. healthcare system are not well-understood. Notably, a recent study showed that treatment of eligible HCV-infected inmates with pegylated interferon and ribavirin would be cost-effective for the U.S. healthcare system (58). Further studies are needed to determine the cost-effectiveness of recommended strategies for reducing HCV-related morbidity and mortality and to guide correctional healthcare administrators regarding which strategy or combination of strategies should be implemented.
To determine correctional healthcare priorities and policy directions, of course, the costs of HCV-related morbidity and mortality must be considered alongside costs associated with other causes of morbidity and mortality in the prison population. Because state correctional healthcare systems are often underfunded, other public funding will likely be required to provide HCV-related preventive and therapeutic services that meet community standards. As prison populations continue to age and increase in size, and as the number of annual releasees increases as well (19–21), the growing role of correctional healthcare systems in the broader U.S. healthcare system must be addressed.
Dr. Harzke was funded by Pre-doctoral Cancer Prevention and Control Fellowship during the early phases of this research (R25 CA 57712 to Patricia Dolan Mullen, University of Texas, School of Public Health). The research described herein was coordinated in part by the Texas Department of Criminal Justice (TDCJ), research agreement 470-RM05. The contents of this article reflect the views of the authors and do not necessarily reflect the views or policies of the TDCJ. Dr. David Ramsey at the University of Texas, School of Public Health in Houston provided technical assistance in data linkage. Leonard Pechacek, from Correctional Managed Care at the University of Texas Medical Branch, provided editorial assistance.