Impact of delayed diagnoses bias on the estimation of AIDS incubation / 中华流行病学杂志

ABSTRACT

Objective Using the HIV positive paid blood donors (PBDs) as example, to investigate the impact of the delayed diagnosed HIV/AIDS patients on the calculated median incubation of AIDS. Methods Information was retrospectively collected on HIV/AIDS patients who were infected through paid blood donation. Data was from the national HIV/AIDS case reporting system. The time of diagnosed HIV/AIDS after the screening was defined as delayed diagnoses. All the cases were diagnosed between Jan. 1st 1995 and the end of the screening were classified as group 1, while the cases diagnosed between Jan. 1st 1995 and Dec. 31st 2010 were classified as group 2.Kaplan-Meier method was used to calculate the median AIDS incubation of cases from the two groups. Results from the two groups on theft incubation periods were compared. Results From the data of group 1, we noticed that the median AIDS incubation was 9.62 years but the 95％ CI were not provided in the results of SAS software. When using the data from group 2,we found that the median AIDS incubation was 9.83 years, with the 95％CI as 9.73 to 9.83 years. Conclusion Our data showed that, if excluding the delayed diagnosed cases, the AIDS incubation periods were shorter and biased.