STRATIFIED RANDOMIZED RESPONSE TECHNIQUE FOR MODELLING HIV SEROPREVALENCE RATES IN NIGERIA

Aliyu Usman; Mamuda Ibrahim Kukasheka

Aliyu Usman Department of Mathematic and Statistics, Kaduna Polytechnic, Nigeria
Mamuda Ibrahim Kukasheka Department of Mathematic and Statistics, Kaduna Polytechnic, Nigeria

Keywords: Randomized response techniques, seroprevalence, design parameter, stratified random sampling

Abstract

Warner (1965) proposed the pioneering RRT for estimating the proportion of persons bearing a socially disapproved character. The RRT are used to avoid the concealment of sensitive information from respondents. Similarly, Su (2021) proposed a randomized response techniques (RRT) for tracking drug usage. The RRT guarantees the anonymity of respondents in surveys aimed at determining the frequency of stigmatic, embarrassing or criminal behaviour where direct techniques for data collection may induce respondents to refuse to answer or give false responses. Different randomized response techniques have been devised in the past decades. Most of these RRTs have been proposed without some specific applications to HIV seroprevalence surveys. The objective here is to apply the RRT to estimate HIV seroprevalence rates. Quatember (2009) produced unified criteria for all RRTs, Kim and Warde (2005) proposed a stratified randomized response model (RRM) and so many others. The proposed RRM for HIV seroprevalence surveys was relatively more efficient than the Kim and Warde (2005) stratified estimator for a fixed sample size. Using the criteria of Quatember (2009) who derived the statistical properties of the standardized estimator for general probability sampling and privacy protection, the chosen design parameter was . The procedure of the field work and sampling design were well coordinated for the target population using a sample size of 400. Furthermore, the model was used to estimate the HIV seroprevalence rate of adults attending a clinic in Kaduna, Nigeria. Using the survey data, the model estimated the HIV seroprevalence rate is 1.1% with a standard error of 0.0024 and 95% confidence bands of [0.6%, 1.6%]. These estimates are for adults who are 18 years and above who attend a hospital. These results are consistent with that of Nigerian sentinel survey (2018) conducted by Nigeria HIV/AIDS Indicator and Impact survey (NAIIS) and United Nations Programme on HIV/AIDS (UNAIDS) which estimated the HIV seroprevalence in Nigeria as 1.4%. Accordingly, this is within the 95% confidence interval. Hence, the RRT can serve as cheaper and faster viable methods for HIV seroprevalence surveys.

, , , .