Estimate of undiagnosed HIV cases in the Cajamarca region of Peru: probabilistic linkage between databases

Abstract

Undiagnosed HIV cases contribute to the increase in new infections, therefore estimating this figure is important in order to assess strategies in HIV control programs. The objective of this study was to estimate the number of undiagnosed HIV cases in the Cajamarca region in Peru between 2015 and 2021. HIV cases were obtained from three sources of information: The Regional Health Strategy for HIV Prevention and Control (ESPC-HIV); the HIV epidemiological notification application (Noti-HIV), and the laboratory information system (Netlab). The three databases were linked; a capture-recapture analysis using a log-linear model provided estimates of the number of undiagnosed HIV cases, taking into account interactions and the Akaike information criterion. After linkage, 991 registered HIV cases were obtained. An estimated 1388 cases (95%CI: 1265.6-1542.8) of people living with HIV were estimated, of which 393 (28.4%) were not diagnosed. The underreporting of each source was: 51.9% in the ESPC-HIV; 63.6% in Netlab; and 88% in Noti-HIV. The conclusion drawn was that a high number of HIV cases went undiagnosed, and strategies need to be reconsidered to increase the systematic detection of HIV cases.

Key words:
HIV; Health Information Systems; Disease Notification