OBJECTIVE: To determine factors associated with the interruption of outpatient care of HIV-positive patients. METHODS: Non-concurrent prospective study carried out in a public AIDS referral center in Belo Horizonte, Brazil. Medical records were reviewed in order to assess factors associated with the interruption of clinical care of HIV patients admitted between 1993 and 1995. Patients should have attended at least one follow-up visit within a period of 7 months. Statistical analysis was carried out using Chi-square and relative hazard (RH) with 95% confidence interval (CI) estimated by Cox Regression Model. RESULTS: Cumulative incidence of interruption was 54% among 517 patients included in the study (mean follow-up=24.6 months; 26.5/100 person-years). Multivariate analysis indicated that those individuals who had fewer (<2) CD4+ T lymphocyte cell counts (RH=1.94; 95% CI=1.32-2.84) did not have viral load measured (RH=14.94; IC 95%=5.44-41.04), attended <7 medical follow-up visits (RH =2.80; IC 95%=1.89-4.14), did not change clinical category (RH =1.40; IC 95% =1.00-1.93) and did not undergo any anti-retroviral therapy (RH =1.43; IC 95% =1.06-1.93) had independently an increased risk of interrupting clinical care. CONCLUSIONS: The rate of clinical interruption in this center is high. The results suggest that interruption may be a function of better clinical outcome, i.e. the service may give priority to those patients with more severe clinical condition, and interruption of clinical care may be a marker for future antiretroviral compliance.
OBJECTIVE: To determine factors associated with the interruption of outpatient care of HIV-positivepatients. METHODS: Non-concurrent prospective study carried out in a public AIDS referral center in Belo Horizonte, Brazil. Medical records were reviewed in order to assess factors associated with the interruption of clinical care of HIVpatients admitted between 1993 and 1995. Patients should have attended at least one follow-up visit within a period of 7 months. Statistical analysis was carried out using Chi-square and relative hazard (RH) with 95% confidence interval (CI) estimated by Cox Regression Model. RESULTS: Cumulative incidence of interruption was 54% among 517 patients included in the study (mean follow-up=24.6 months; 26.5/100 person-years). Multivariate analysis indicated that those individuals who had fewer (<2) CD4+ T lymphocyte cell counts (RH=1.94; 95% CI=1.32-2.84) did not have viral load measured (RH=14.94; IC 95%=5.44-41.04), attended <7 medical follow-up visits (RH =2.80; IC 95%=1.89-4.14), did not change clinical category (RH =1.40; IC 95% =1.00-1.93) and did not undergo any anti-retroviral therapy (RH =1.43; IC 95% =1.06-1.93) had independently an increased risk of interrupting clinical care. CONCLUSIONS: The rate of clinical interruption in this center is high. The results suggest that interruption may be a function of better clinical outcome, i.e. the service may give priority to those patients with more severe clinical condition, and interruption of clinical care may be a marker for future antiretroviral compliance.