Estimating and projecting the number of new HIV diagnoses and incidence in Spectrum's case surveillance and vital registration tool.

File Description SizeFormat 
application-pdf.pdfFile embargoed until 02 August 2020531.14 kBAdobe PDF    Request a copy
Title: Estimating and projecting the number of new HIV diagnoses and incidence in Spectrum's case surveillance and vital registration tool.
Authors: Mahiane, SG
Marsh, K
Glaubius, R
Eaton, JW
Item Type: Journal Article
Abstract: OBJECTIVE: The United Nations Program on HIV/AIDS-supported Spectrum software package is used by most countries worldwide to monitor the HIV epidemic. In Spectrum, HIV incidence trends among adults (aged 15-49 years) are derived by either fitting to seroprevalence surveillance and survey data or generating curves consistent with case surveillance and vital registration data, such as historical trends in the number of newly diagnosed infections or AIDS-related deaths. This article describes development and application of the case surveillance and vital registration (CSAVR) tool for United Nations Program on HIV/AIDS' 2019 estimate round. METHODS: Incidence in CSAVR is either estimated directly using single logistic, double logistic, or spline functions, or indirectly via the 'r-logistic' model, which represents the (log-transformed) per-capita transmission rate using a logistic function. The propensity to get diagnosed is assumed to be monotonic, following a Gamma cumulative distribution function and proportional to mortality as a function of time since infection. Model parameters are estimated from a combination of historical surveillance data on newly reported HIV cases, mean CD4 at HIV diagnosis and estimates of AIDS-related deaths from vital registration systems. Bayesian calibration is used to identify the best fitting incidence trend and uncertainty bounds. RESULTS: We used CSAVR to estimate HIV incidence, number of new diagnoses, mean CD4 at diagnosis and the proportion undiagnosed in 31 European, Latin American, Middle Eastern, and Asian-Pacific countries. The spline model appeared to provide the best fit in most countries (45%), followed by the r-logistic (25%), double logistic (25%), and single logistic models. The proportion of HIV-positive people who knew their status increased from about 0.31 [interquartile range (IQR): 0.10-0.45] in 1990 to about 0.77 (IQR: 0.50-0.89) in 2017. The mean CD4 at diagnosis appeared to be stable, decreasing from 410 cells/μl (IQR: 224-567) in 1990 to 373 cells/μl (IQR: 174-475) by 2017. CONCLUSION: Robust case surveillance and vital registration data are routinely available in many middle-income and high-income countries while HIV seroprevalence surveillance and survey data may be scarce. In these countries, the CSAVR offers a simpler, improved approach to estimating and projecting trends in both HIV incidence and knowledge of HIV status.
Issue Date: 2-Aug-2019
Date of Acceptance: 1-Aug-2019
ISSN: 0269-9370
Publisher: Lippincott, Williams & Wilkins
Journal / Book Title: AIDS
Copyright Statement: © 2019 Wolters Kluwer Health, Inc. This is a non-final version of an article published in final form in
Sponsor/Funder: National Institutes of Health
Medical Research Council (MRC)
Funder's Grant Number: 5776-ICS-DHHS-6664
Keywords: 06 Biological Sciences
11 Medical and Health Sciences
17 Psychology and Cognitive Sciences
Publication Status: Published online
Conference Place: England
Embargo Date: 2020-08-02
Open Access location:
Online Publication Date: 2019-08-02
Appears in Collections:Epidemiology, Public Health and Primary Care

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Creative Commons