An analysis of the survival of HIV + patients

• Main Authors: Pacana, Jennifer P., Villaroman, Miriam Liza C.
• Format: text
• Language: English
• Published: Animo Repository 1997
• Subjects: Mathematical analysis; AIDS (Disease) > Patients; Mathematical models; Estimation theory
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/16451
• Institution: De La Salle University

This paper is an exposition of the article entitled, AIDS: Survival Analysis of Persons Testing HIV+, by Harry H. Panjer. It uses a study that observed a sample of 307 individuals who were grouped into six different classifications and observed in four different periods. From this study, the researchers have come up with a survival model that uses an exponential distribution. The maximum likelihood method is used to obtain the estimates of the parameters and an assumption of exposure using the midpoint method is applied to find the probabilities p s, the probability of not progressing for those in observation period i. The values of the estimates of the intensity function and the expected time in stage, and also their 95% confidence bounds the proportion of a person to progress to the next stage in a specified time the life expectancies of a person in any stage and the predicted and actual values of progression based on the model are all determined through extensive calculations. Finally, the exponential model is tested for its appropriateness on the study by applying the goodness-of-fit test that uses the chi-square, x2, distribution.