397. Long-Term Mortality of HIV Patients Following Cryptococcal Infection

BACKGROUND: Prevalence of cryptococcosis in HIV-positive patients in the developed world has decreased considerably in the modern antiretroviral (ART) era. While early mortality of cryptococcal infection is lower than in non-HIV-infected patients, late mortality in HIV+ patients has not been previously evaluated. Here, we describe the presentation and outcomes of HIV+ patients with cryptococcosis. METHODS: We conducted a retrospective cohort study of patients with HIV infection and cryptococcosis from January 2002 to June 2017 at our institution. Data included demographics, clinical features, diagnostics, and outcomes. Death date was obtained from the hospital system’s Medical Informatics database and the Social Security Death Index. RESULTS: We reviewed 105 HIV+ patients with cryptococcosis. At time of analysis: 55 were living (52.4%), 17 died within 90 days of cryptococcal diagnosis (early mortality, 16.2%), and 33 died after 90 days (late mortality, 31.4%) (Figure 1). Late mortality patients were more likely to have known HIV+ status at the time of cryptococcal diagnosis (97%) than living (70.9%) or early mortality (70.6%) (P = 0.03); less likely to be ART adherent (15.2%) than living (43.6%) or early mortality (35.3%) (P = 0.02); less likely to have private insurance (6.1%) than living (34.5%) or early mortality (17.6%) (P = 0.007); and more likely to have Medicaid (51.5%) than living (29.1%) or early mortality (17.6%) (P = 0.03). Presenting symptoms and diagnostics were similar between groups. Prevalence of substance abuse (48.6%) and psychiatric history (31.4%) were high in all groups but not significantly different. CONCLUSION: Despite improvements in ART, HIV+ patients have high mortality following cryptococcal infection which persists beyond their initial hospitalization. Identifying patients at higher risk for mortality is critical for successful treatment and outcomes. In our study, nonadherence to ART was associated with a higher risk of dying. Follow-up studies of late mortality in other opportunistic infections would be beneficial. DISCLOSURES: All authors: No reported disclosures.