Meta-analysis of risk for relapse to substance use after transplantation of the liver or other solid organs

Abstract

For patients receiving liver or other organ transplants for diseases associated with substance use, risk for relapse posttransplantation is a prominent clinical concern. However, there is little consensus regarding either the prevalence or risk factors for relapse to alcohol or illicit drug use in these patients. Moreover, the evidence is inconsistent as to whether patients with pretransplantation substance use histories show poorer posttransplantation medical adherence. We conducted a meta-analysis of studies published between 1983 and 2005 to estimate relapse rates, rates of nonadherence to the medical regimen, and the association of potential risk factors with these rates. The analysis included 54 studies (50 liver, 3 kidney, and 1 heart). Average alcohol relapse rates (examined only in liver studies) were 5.6 cases per 100 patients per year (PPY) for relapse to any alcohol use and 2.5 cases per 100 PPY for relapse with heavy alcohol use. Illicit drug relapse averaged 3.7 cases per 100 PPY, with a significantly lower rate in liver vs. other recipients (1.9 vs. 6.1 cases). Average rates in other areas (tobacco use, immunosuppressant and clinic appointment nonadherence) were 2 to 10 cases per 100 PPY. Risk factors could be examined only for relapse to any alcohol use. Demographics and most pretransplantation characteristics showed little correlation with relapse. Poorer social support, family alcohol history, and pretransplantation abstinence of < or =6 months showed small but significant associations with relapse (r = 0.17-0.21). Future research should focus on improving the prediction of risk for substance use relapse, and on testing interventions to promote continued abstinence posttransplantation.

Figure 1

Identification of independent studies for current meta-analysis.

Figure 2

Pooled estimates of post-transplant rates of substance use and other nonadherence outcomes. ^aNumber of studies examining each nonadherence outcome. ^bCannot be estimated for < 3 studies. *p<.01 **p<.001

Figure 3

Associations between post-transplant alcohol relapse (any use) and psychosocial variables.^{a a}Positive effect sizes indicate associations of alcohol recidivism with male gender, greater age, greater education, unmarried status, pre-transplant unemployment, poorer social support, pre-transplant history of psychiatric disorder, pre-transplant history of illicit drug use, family history of alcohol abuse/dependence, pre-transplant alcohol abstinence ≤ 6 mos, shorter pre-transplant abstinence time, and lack of pre-transplant participation in an alcohol rehabilitation program. Abbreviations: ES, effect size; CI, confidence interval; tx, transplant; dep, dependence; rehab, rehabilitation

Figure 4. Time to alcohol relapse (any use) after liver transplant

For each of 11 studies reporting time to relapse after adjusting for differences in patients' follow-up duration, the range of time to relapse is plotted (earliest point, midpoint [mean or median], latest point), as well as length of total follow-up in the sample. *see Appendix A for full listing of studies and authors (Garhardt, 1996 is listed with Goldstein, 1993 in Appendix A; Campbell, 1998 is listed with Beresford, 1992; Cuadrado, 2005 is listed with Fábrega,1998)

Figure 5. Rate of alcohol relapse as a function of time since transplant

Solid line shows estimate from meta-analysis (alcohol relapse, any use: 5.6% increment annually, i.e., 5.6 cases per 100 patients per year; alcohol relapse, heavy use: 2.5% increment annually). This line becomes dashed after 9 years due to the limited number of studies available. Points show individual studies' rates. Individual studies' rates were calculated at the end of followup for studies that followed all patients for the same duration or for studies that adjusted for differences in followup duration through survival analysis; otherwise, studies' rates represent the midpoint of each study's follow-up period.