Survival and prognostic factors of adult human immunodeficiency virus (HIV) patients in Hospital Universiti Sains Malaysia (HUSM) from 2002 to 2006 / Mohd Sapuan Baharuddin, , Wan Mohd Zahiruddin Wan Mohammad and

The estimated number of persons living with HIV around the world in 2007 was 33.2 million. It was estimated
that 80, 938 people have been infected with HIV in Malaysia and 10, 334 died from AIDS from 1986 till
December 2007. Drug abuse through injection was the dominant mode of HIV transmission in Malaysia.The
objectives of this study were to identify the prognostic factors of HIV patients in HUSM. The design for this
study was a retrospective record review. All the patients diagnosed as HIV positive in HUSM during the study
period of 1st January 2002 until 31st December 2006 (5 years) will be recruited as respondents in this study.
Additional follow up for this study was about 12 months, from 1st January 2007 till 31st December 2007. All HIV
cases who fulfill the selection criteria in this study were 107 patients. Data of HIV patients were collected from
medical record unit using data collection sheet. The patient’s survival status was obtained from medical record
or contact by phone call. The process of data entry for this study was done using SPSS version 12.0.1. Data was
transferred into STATA version 9.0 program using STAT Transfer software version 6.0. The statistical analysis
were used in this study was Cox Proportional Hazard Regression analysis. Prognostic factors that were found
to be significant during simple cox regression analysis were co-infections, opportunistic infections, clinical
stage and clinic visit frequency. During multiple cox regression analysis, prognostic factors such as clinical
stage and clinic visit frequency were significant in this study. Final model in this study only had clinical stage as
prognostic factor (Adjusted HR=2.40; 95%CI: 1.42, 4.01, p=0.001). Overall median survival time in this study
and survivorship among HIV/AIDS patients in HUSM was low. The five years survival in this study was affected
by late diagnosis as HIV positive and the unavailability of HAART in HUSM. The prognostic factor that influenced the risk of death among HIV patients was clinical stage. Prognostic factors such as CD4 count, viral load and HAART will be included in this final model in future in order to increase the strength of the model to predict survival of HIV patients in HUSM.