Predictors of rehospitalization of military veterans who abuse substances

Abstract

The study reported in this article is of a systematic random sample of 600 homeless Vietnam veterans, ages 46 to 65, who abuse substances, many of whom are comorbid with psychological afflictions. All of these veterans were in a Midwestern residential program for homeless substance abusers at the time of the interview. Cox's proportional-hazards model was used to estimate the relative rate of rehospitalization (hazard function) across the follow-up interval of two years by the predictors. The ecological predictors include, but are not limited to, demographic characteristics, history of drug and psychiatric treatment, psychological afflictions, abuse before 18 years of age, inner strengths, social support, religiosity, and direct combat experience in Vietnam. The range of ecological factors investigated presents more comprehensive findings for future conceptual models and for more thorough social work intervention. Key words: comorbidity; drug abuse; ecological predictors; homeless veterans; survival analysis ********** Large numbers of homeless military veterans are being treated for substance abuse at Veterans' Affairs medical centers (VAMCs). For example, in 1996, 36 percent of the more than 250,000 discharges of Vietnam veterans nationwide were related to substance abuse (Virgo, Price, Spitznagel, & Ji, 1999). In addition, more than 25 percent of all veterans treated at VAMCs in an inpatient setting in 1996 had a substance abuse diagnosis. Forty percent had a diagnosis of comorbidity, or the coexistence of chemical abuse and other psychological problems. In 1996 one-third of the 10.9 million days of inpatient care provided by VAMCs were for veterans with substance use disorders. Veterans who suffer comorbidity have longer and more frequent hospitalizations than veterans treated in inpatient settings for other reasons (Piette, Baisden, & Moos, 1997). Comorbidity, in fact, appears to be one of the strongest predictors of rehospitalization among veterans (Dixon, McNary, & Lehman, 1997; Marshall, Jorm, Grayson, & O'Toole, 1998; Murray, Anthenelli, & Maxwell, 2000; Rosenheck & Cicchetti, 1998). At the same time, substance abuse historically has been treated in a different system than other psychological or mental health problems. Typically, these systems have had dissimilar, even contradictory, treatment orientations and approaches. Given these dissimilarities, substance abuse frequently is not properly assessed or is overlooked in mental health centers, whereas mental health problems generally are not appropriately diagnosed or treated in drug rehabilitation programs. Often, chemical abusers who suffer other psychological afflictions are shuffled between drug rehabilitation and mental health systems, using scarce resources, without receiving adequate treatment in either system to resolve their problems (Dixon et al., 1997; Nuttbrock, Rahav, Rivera, Ng-Mak, & Link, 1998; Rosenheck, & Cicchetti, 1998). Whereas integrative programs have been designed recently to treat substance abuse and a variety of psychological problems, there are still few of these programs for homeless veterans (Nuttbrock et al., 1998; Rosenheck & Cicchetti, 1998). Moreover, little information exists about what factors predict rehospitalization of homeless veterans who abuse substances (Helzer & Przybeck, 1988; Hoff, Rosenheck, Meterko, & Wilson, 1999; Murray et al., 2000; Olfson, Mechanic, Hansell, Boyer, & Walkup, 1999; Rosenheck & Cicchetti, 1998). Yet, a major assumption underlying the present study is that programs are efficacious to the extent that they target the factors that predict rehospitalization (Andrews & Bonta, 1998). Factors that predict rehospitalization can be categorized into two useful classes for intervention purposes: (1) static factors, which are not alterable by social work intervention, such as age or history of substance abuse, and (2) dynamic factors, which are amenable to social work change, such as self-esteem or depression (Andrews & Bonta). …