Differences in Clinical Outcomes of Adults Referred to a Homeless Transitional Care Program Based on Multimorbid Health Profiles: A Latent Class Analysis

Abstract

Background: People experiencing homelessness face significant medical and psychiatric illness, yet few studies have characterized the effects of multimorbidity within this population. This study aimed to (a) delineate unique groups of individuals based on medical, psychiatric, and substance use disorder profiles, and (b) compare clinical outcomes across groups.Methods: We extracted administrative data from a health system electronic health record for adults referred to the Durham Homeless Care Transitions program from July 2016 to June 2020. We used latent class analysis to estimate classes in this cohort based on clinically important medical, psychiatric and substance use disorder diagnoses and compared health care utilization, overdose, and mortality at 12 months after referral.Results: We included 497 patients in the study and found 5 distinct groups: “low morbidity” (referent), “high comorbidity,” “high tri-morbidity,” “high alcohol use,” and “high medical illness.” All groups had greater number of admissions, longer mean duration of admissions, and more ED visits in the 12 months after referral compared to the “low morbidity” group. The “high medical illness” group had greater mortality 12 months after referral compared to the “low morbidity” group (OR, 2.53, 1.03–6.16; 95% CI, 1.03–6.16; p = 0.04). The “high comorbidity” group (OR, 5.23; 95% CI, 1.57–17.39; p < 0.007) and “high tri-morbidity” group (OR, 4.20; 95% CI, 1.26–14.01; p < 0.02) had greater 12-month drug overdose risk after referral compared to the referent group.Conclusions: These data suggest that distinct groups of people experiencing homelessness are affected differently by comorbidities, thus health care programs for this population should address their risk factors accordingly.

Social determinants of multimorbidity patterns: A systematic review

“…Similarly, with a few exceptions ( 51 , 65 ), most studies did not consider ethnic or racial differences. In fact, most reviewed works were oriented to Western countries' populations and mainly applied to white Caucasians groups, even though specific multimorbidity patterns have been found to be more prevalent in African Americans ( 147 ).…”

Risk profiles of adults with heavy alcohol use: Drinking patterns, behavioral and metabolic factors, health problems, and racial and ethnic disparities

“…To identify risk profiles, we employed latent class analysis (LCA), a data-driven method that can identify subgroups of people based on frequently co-occurring patterns in the data. For example, LCA has been applied to characterizing subgroups of people based on patterns of alcohol use (Robins et al, 2021), other behavioral factors such as physical inactivity and poor diet (Cook et al, 2020), and multimorbidity (Smith et al, 2021).…”

“…To identify risk profiles, we employed latent class analysis (LCA), a data‐driven method that can identify subgroups of people based on frequently co‐occurring patterns in the data. For example, LCA has been applied to characterizing subgroups of people based on patterns of alcohol use (Robins et al., 2021), other behavioral factors such as physical inactivity and poor diet (Cook et al., 2020), and multimorbidity (Smith et al., 2021). This study aimed to: (1) identify and describe risk profiles of adults with heavy alcohol use by heavy drinking patterns, other behavioral and metabolic factors, and morbidity, using LCA; and (2) examine whether there were racial and ethnic disparities in the risk profiles identified.…”