Evaluation of Mortality among HIV patients on HAART in Lusaka, Zambia : An observational study

Thumbnail Image
Mweemba, Aggrey
Journal Title
Journal ISSN
Volume Title
To evaluate mortality and determine the causes of death in HIV patients initiating HAART in Lusaka District clinics . Mortality was assessed in first patients enrolled between May, 2004 and May, 2005 in the Lusaka District Management clinics. Medical charts and interviews with medical staff, who were attending to the patients before they died, were done to ascertain the probable cause of death. For the predictors of death analysis, we estimated Kaplan-Meier probability of survival for those who remained alive on antiretroviral therapy compared to those who died while on ART. To investigate the relationship between those who remained alive on ART and those who died while on ART, we used univariate and multivariate Cox's proportional hazard models, confidence interval (CI) and log-rank tests. We evaluated 13,672 individuals who were seen in the eight clinics. Of these 8,497 were females (62%) and the median age was 34 years. Patients with CD4 < 50 cells/mm'' and those with CD4 between 51 and 200 cells/mm^ accounted for the majority of patients in all the groups except for "surviving" patients who had "not initiated HAART". A preponderance of patients that died had a BMI < 18.5 kg/m^ and 20kg/m^ for males and females respectively (P-value < 0.0001). Probable causes of death were identified for 294 (44%) of the 670 deaths. In the majority of patients, cause of death could not be determined. Dying at a health centre did not improve the probability of determining the cause of death. For instance, 59 percent of patients that died at a health facility had no diagnosis at the time of death compared to 51 percent for patients that died at home. The cause of death in most cases was attributed to Opportunistic infections. Mortality rate was found to be high in the initial stages of HAART. Ninety seven percent (391) of deaths occurred in the first six months (180 days) of ART initiation (p-value <0.0001). The mortality rate was 21 deaths per 100 person-years (95% CI) in the first month of ART. The rate declined five-fold (4 deaths per 100 person-years; 95%) CI) after one to six months of ART. The mortality rate continued to decline such that after 6 months the rate was 0.43 per 100 person years (95% CI).The overall mortality among patients who were on ART was 4.9 deaths per 100 person-years. Kaplan-Meier survival curves for all patients enrolled into the programme showed that WHO stage 3 & 4, baseline CD4<50 cells/mm^ BMI <1 8.5 kg/mm^ haemoglobin < 6.0g/dl were associated with a substantially lower survival probability compared with other patients. Overall WHO stage 3 & 4 diseases and CD4 < 50 cells/mm^ accounted for 86.4 percent (579) and 41.5 percent (278) respectively. In contrast, patients with WHO stage 1 & 2, CD4 >201 cell/mm-\ Haemoglobin > 6g/dl and BMI > 18.5 kg/m^ had a lower risk of death. Mortality was high in the early phase of initiation of HAART due to severe immunosuppression and late presentation of patients. Benefits of HAART were noted by declining mortality rates over a period of time. The lack of diagnostics and the critical shortage of staff were reflected by failure to identify cause of death in the majority of patients. There is need to increase staffing levels and equip laboratories in the clinics to improve diagnosis and management of opportunistic infections in HIV patients. Robust follow of HIV patients initiating HAART with severe disease, low BMI, Anaemia and Low CD4 is required to in order to reduce associated mortality.
AIDS (Disease) - patients , AIDS (Disease) -- Patients -- Zambia , Aids(disease) -- Treatment