Using a retrospective cohort study of all inmates (30,237 people, after exclusions) released from the Washington State Department of Corrections from July 1999 to December 2003, the authors studied the risk of death in this cohort and compared it to the risk of death among other state residents. Indirect adjustments were made for age, sex and race. Death rate for this cohort was assessed using the NDI, a computerized database abstracted from death certificates from all 50 states, Washington D.C., Puerto Rico and the Virgin Islands. Prisoners' reentry to society, particularly during the first few weeks, is stressful due to factors such as rejoining friends and family, and finding employment, housing, and health care. In an attempt to provide data for future targeted interventions for this high-risk group, the authors also described the major causes of death, and examined mortality rates in the first two weeks post-release.
- The adjusted relative rate of mortality for released prisoners was 3.5 times that of other Washington residents.
- Within the first two weeks, the adjusted relative risk was 12.7 times that of other state residents, with the most deaths occurring in the first week.
- The adjusted relative death risk among women was higher than among men.
- Drug overdose represented nearly 25 percent of all deaths; cardiovascular diseases were the second highest cause of death, and homicide the third; other major causes were suicide, cancer and motor vehicle accidents.
- Rates of homicide and drug overdose were considerably higher post-release than during incarceration.
The authors report various limitations, such as that prison populations tend to be less educated, which may contribute to their higher risk of mortality. Tobacco use is also high among this population. Classification errors may have occurred, such as suicides being classified as drug overdoses and the reverse. Lastly, the study may not be applicable to other states, although there is no reason to assume so. U.S. prisons released more than 600,000 prisoners in 2002, making the post-release population of people potentially vulnerable to the deaths described here substantial.