Study design and data source
This study employed a cross-sectional design to analyse the age-standardized prevalence of HIV/AIDS and STIs in Sierra Leone from 2000 to 2019. Data were sourced from the WHO HEAT, an online interactive platform designed to facilitate the examination, analysis, and reporting of health inequality data. The HEAT platform draws from the Sierra Leone Demographic Health Survey Health Inequality Data Repository, which includes disaggregated health indicators from various global sources. The toolkit is publicly accessible at (https://www.who.int/data/inequality-monitor/assessment_toolkit). For this study, we focused on the sex-standardized prevalence of HIV/AIDS and STIs (per 100,000 population) as reported in the HEAT database. The data provided by the WHO HEAT platform are pre-processed and nationally representative, ensuring consistency and comparability across time points. All available data points were included in the analysis. However, it is important to note that no additional weighting was applied beyond this standardization process. This approach ensures the comparability of prevalence rates across different age groups while accurately reflecting the underlying data. The data adheres to stringent data processing and quality assurance protocols. Missing data were assessed, and appropriate methods were applied, such as imputation or exclusion, based on the extent and nature of the missing data. Quality checks include validation processes that assess the completeness and accuracy of the data, such as cross-referencing with national health statistics and demographic surveys. Additionally, the WHO employs statistical techniques to identify and correct anomalies or inconsistencies in the data. Before analysis, the data undergoes age-standardization to account for variations in age distribution across populations. These comprehensive quality assurance measures ensure that the data used in our analysis are reliable and accurately reflect the prevalence of HIV/AIDS and STIs in Sierra Leone.
The WHO HEAT accounted for co-infections by counting each individual infection separately. For instance, individuals diagnosed with both HIV and another STI were included in the prevalence counts for both conditions. This approach enables us to present a detailed overview of the burden of HIV/AIDS and STIs in Sierra Leone, reflecting the true extent of these health issues in the population. We recognize that this method may lead to individuals being counted multiple times; however, it is critical for understanding the epidemiology of co-infections in the context of public health interventions. The analysis included five time points: 2000, 2005, 2010, 2015, and 2019. This longitudinal approach enabled a comprehensive evaluation of trends in prevalence and disparities over two decades.
Outcome measure and dimension of inequality
The primary outcome measure was the age-standardized prevalence of HIV/AIDS and STIs (per 100,000 population). This measure was chosen because it accounts for differences in population age structures, allowing for meaningful comparisons across time points and between subgroups. The dimension of inequality examined in this study was sex, with the population disaggregated into two subgroups: male and female.
Although the WHO HEAT platform supports the analysis of additional dimensions of inequality such as socio-economic status, geographic region, and urban-rural residence, this study focused exclusively on sex-based disparities due to the scope of the available data and the study’s primary objective. The analysis aimed to provide insights into sex-based differences in the prevalence of HIV/AIDS and STIs, a critical public health concern in Sierra Leone given the disproportionate burden of these infections among females and the socio-cultural factors influencing health outcomes. Combining HIV/AIDS and STI is grounded in the recognition that HIV and STIs share common transmission pathways and risk factors, making their combined analysis relevant for understanding the overall sexual health landscape in Sierra Leone. By examining these populations together, we aim to highlight the interconnectedness of these health issues, which can inform more effective public health strategies and interventions. While HIV and STIs are distinct, their co-occurrence necessitates an integrated approach to address the comprehensive sexual health needs of the population.
Data analysis
The data analysis for this study was conducted in the WHO HEAT using R version 4.1.0, a widely used statistical computing environment known for its robust data analysis capabilities and extensive package ecosystem. To facilitate the data cleaning and analysis process, custom scripts in R was developed. These scripts were designed to automate the following key tasks:
Data importation
Scripts were created to import data from the WHO HEAT database efficiently, ensuring that all relevant variables were captured.
Data cleaning
Custom functions were implemented to perform checks for missing values, identify outliers, and standardize variable formats. This automation enhanced the efficiency and reproducibility of the data cleaning process.
Age standardization
Specific scripts were utilized to perform age-standardization of prevalence rates, allowing for accurate comparisons across different age groups.
Statistical analysis
Custom scripts were also employed to conduct the statistical analyses necessary for the study, including prevalence calculations and the generation of visualizations.
The rationale for using R and developing custom scripts was to ensure flexibility and precision in the analysis. R’s extensive libraries and community support provided us with the tools necessary to perform complex statistical operations, while custom scripts allowed to tailor the data processing steps to the specific requirements of the study. This approach not only streamlined the workflow but also enhanced the reproducibility of the findings.
Data analysis was conducted for individuals aged 15–49 years, using the WHO HEAT software, which facilitates the computation of estimates, confidence intervals (CIs), and summary measures of inequality. The HEAT platform provides tools for generating both absolute and relative measures of inequality, enabling a detailed exploration of disparities. Four measures were used to assess sex-based disparities in the prevalence of HIV/AIDS and STIs: Difference (D), Ratio (R), Population Attributable Fraction (PAF), and Population Attributable Risk (PAR).
Difference (D)
This metric quantifies the absolute disparity in prevalence rates between females and males by directly comparing their values, providing a clear indication of the magnitude of inequality in absolute terms.
Ratio (R)
This metric calculates the proportional difference in prevalence rates by dividing the prevalence among females by that of males, offering a relative measure of the disparity.
Population attributable fraction (PAF)
This metric represents the percentage of the overall prevalence attributable to sex-based inequality, reflecting the potential reduction in prevalence if the disparity were eliminated.
Population attributable risk (PAR)
This metric indicates the excess prevalence of HIV/AIDS and STIs attributable to sex-based inequality in the population, providing an absolute measure of the burden.
These metrics collectively provided a comprehensive understanding of both the magnitude and context of sex-based disparities in HIV/AIDS and STI prevalence. It important to note that negative PAF and PAR values indicate that addressing health disparities could lead to a reduction in HIV/AIDS and STI prevalence. Specifically, these negative values suggest that certain population groups experienced a lower prevalence of the health outcome than would be expected if disparities were eliminated. In the context of public health interventions, negative PAF and PAR values imply an opportunity for significant health improvements by focusing on reducing disparities; and emphasize the need for equitable health policies that prioritize vulnerable populations. It also guides policymakers in allocating resources effectively, ensuring that interventions are directed toward populations that would benefit the most from reduced disparities.
Estimates and 95% confidence intervals were computed for each subgroup (male and female) and for each year of analysis. The results were presented in tabular and graphical formats to facilitate interpretation and to highlight trends over time. The WHO HEAT platform’s standardized computations ensured robust comparisons across time points and subgroups, enhancing the reliability and validity of the findings. For further details on the inequality measures used, readers are referred to the relevant literature [23,24,25].
Population growth and demographics
We utilized age-standardization techniques to adjust for changes in population structure and growth. This approach allowed us to compare prevalence rates across different years while controlling for variations in age distribution. Additionally, we referenced demographic data from national censuses to ensure our estimates reflected the evolving population characteristics in Sierra Leone.
Changes in testing practices
We acknowledged that advancements in testing methodologies and increased access to testing services may have impacted reported prevalence rates. To address this, we examined trends in testing coverage and accessibility over the study period. While we did not have specific adjustment factors for testing practices, we discussed these changes in the context of our findings, noting that improved testing could lead to higher reported prevalence rates due to increased detection rather than an actual increase in incidence.
Age standardization
To account for variations in age distribution across different years and ensure comparability of prevalence estimates over time, we employed age standardization in our analysis. The following steps outline the process:
Selection of standard population
We used the WHO’s World Standard Population as the reference for age standardization. This population is widely recognized and facilitates comparisons across different studies and time periods.
Calculation of age-specific rates
We calculated age-specific prevalence rates for HIV/AIDS and sexually transmitted infections for each year of the study. This involved determining the number of cases within specific age groups and dividing by the total population in those age groups.
Application of weights
The age-specific rates were then multiplied by the corresponding weights from the World Standard Population. This step ensured that each age group contributed proportionally to the overall prevalence estimate based on the standard population structure.
Summation of standardized rates
Finally, we summed the weighted age-specific rates to obtain the age-standardized prevalence estimate for each year. This method allowed us to compare prevalence rates across the two decades while controlling for changes in age distribution, thus providing a clearer understanding of trends in HIV/AIDS and STIs over time.