Pre-Exposure Prophylaxis Reduces HIV Test Positivity Among Men Who Have Sex with Men in Kazakhstan: An Interrupted Time Series Analysis

Pre-Exposure Prophylaxis Reduces HIV Test Positivity Among Men Who Have Sex with Men in Kazakhstan: An Interrupted Time Series Analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Pre-Exposure Prophylaxis Reduces HIV Test Positivity Among Men Who Have Sex with Men in Kazakhstan: An Interrupted Time Series Analysis Indira Karibayeva, Botagoz Turdaliyeva, Gulzar Shah, Manshuk Ramazanova, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9163384/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 11 You are reading this latest preprint version Abstract Background Eastern Europe and Central Asia faces an accelerating HIV epidemic among gay, bisexual, and other men who have sex with men (GBMSM), yet population-level effectiveness data for pre-exposure prophylaxis (PrEP) remain limited. In May 2021, Kazakhstan implemented a national PrEP programme for GBMSM. Methods We conducted interrupted time series (ITS) analysis of national HIV surveillance data from January 2020 through December 2024 (n = 60 months). The primary outcome was monthly HIV test positivity rate among GBMSM at substantial risk. Standard segmented regression was used to test for immediate and slope changes. We also performed counterfactual analysis to test the cumulative effects against projected pre-intervention trends. Sensitivity analyses controlled for testing volume and examined alternative outcomes. Results By December 2024, the programme achieved 20.1% coverage (1,936 of 9,630 GBMSM at substantial risk). Counterfactual analysis demonstrated significant cumulative reduction in monthly HIV test positivity (mean monthly divergence from projected trends: 0.537 percentage points, 95% CI: 0.393–0.681, p < 0.001). Descriptive comparison showed a 25.7% relative reduction from 3.38% to 2.51%. Effects accumulated gradually over the implementation period rather than as an immediate step-change. Findings remained robust after controlling for a 96% increase in testing volume and across alternative outcomes. Assuming 86% real-world effectiveness (from meta-analyses), the programme currently prevents an estimated 146 HIV infections annually, with a number needed to treat of 13. Conclusions Kazakhstan's PrEP programme demonstrates significant, robust population-level impact on HIV test positivity among GBMSM (25.7% reduction, p < 0.001). The programme currently prevents an estimated 146 infections annually (NNT = 13), with potential to prevent 510 infections annually at 70% coverage. This study provides quasi-experimental evidence for PrEP effectiveness in Kazakhstan, demonstrating that biomedical prevention can achieve measurable population-level impact in concentrated epidemics. HIV prevention Pre-exposure prophylaxis GBMSM Interrupted time series Kazakhstan Central Asia Figures Figure 1 INTRODUCTION Eastern Europe and Central Asia (EECA) faces one of the world's fastest-accelerating HIV epidemics. New infections among gay, bisexual, and other men who have sex with men (GBMSM) have risen by 144% between 2010 and 2022, accounting for 94% of new infections among key populations in the region ( 1 ). During this period, Kazakhstan experienced the steepest increase in Central Asia, with new diagnoses rising by 73% between 2010 and 2020, representing the 12th highest percent increase globally ( 2 , 3 ). The most recent integrated bio-behavioural surveillance (IBBS) from 2023 reported an annual HIV incidence rate of 8.8% among GBMSM, nearly 30 times higher than the general population rate of 0.3%, indicating a concentrated and escalating epidemic ( 2 – 4 ). Three interconnected structural barriers impede HIV prevention among GBMSM in Kazakhstan. First, legislative restrictions on information about sexual minorities, combined with documented discrimination in healthcare settings, have created an environment that discourages healthcare-seeking among GBMSM ( 5 – 7 ). Second, insufficient provider training and pervasive societal stigma result in discrimination within clinical settings, undermining trust in HIV services ( 8 – 10 ). Third, elevated rates of substance use, including chemsex (the intentional use of drugs to enhance sexual activity), increase HIV transmission risk through both biological and behavioural mechanisms ( 4 , 6 , 11 , 12 ). Addressing these structural determinants is the ideal long-term strategy. However, legal and social reforms require sustained political will and may take years to decades to achieve meaningful change. Biomedical prevention strategies such as pre-exposure prophylaxis (PrEP) offer a complementary approach, providing immediate protection that operates independently of structural barriers while broader reforms are pursued. Pre-exposure prophylaxis with oral tenofovir-based regimens offers 99% efficacy among adherent users and 70–90% effectiveness in real-world settings ( 13 – 16 ). The WHO recommends PrEP for key populations at substantial risk ( 17 , 18 ). However, achieving population-level impact from individual-level efficacy requires adequate coverage, sustained adherence and retention in care ( 19 – 21 ). Emerging evidence from high-income countries and research demonstration projects shows that PrEP programmes can reduce HIV diagnoses by 19–38% when coverage reaches key populations ( 22 – 25 ). These evaluations used ecological or cohort designs in settings with an established healthcare infrastructure or intensive external support. It remains unexamined whether population-level effectiveness translates to routine national programme implementation in Central Asia. In May 2021, Kazakhstan introduced oral PrEP through the Guaranteed Amount of Free Medical Care National Program ( 26 ). State AIDS Centres, regional public health facilities providing HIV prevention, testing, and treatment services, led a phased expansion through site activation, clinician training, and community outreach targeting GBMSM and other key populations. The national programme defines 'substantial risk' as having multiple sexual partners, inconsistent condom use during anal intercourse, or a recent sexually transmitted infection. By December 2024, 1,936 individuals were receiving PrEP, achieving 20.1% coverage of an estimated 9,630 GBMSM at substantial risk ( 27 ). Despite robust evidence of PrEP's individual-level efficacy from clinical trials, population-level effectiveness data specifically from Central Asia remain limited. However, no published studies have examined PrEP's population-level impact on HIV testing volume, test positivity rates, or disease severity at diagnosis in this region. This study evaluated Kazakhstan's national PrEP programme using interrupted time series analysis of HIV surveillance data from January 2020 through December 2024. We pursued three specific aims: ( i ) to quantify the programme's impact on HIV test positivity among GBMSM; ( ii ) to estimate the programme's current coverage and the number of infections prevented annually; and ( iii ) to project infections preventable under strategic coverage scenarios (30%, 50%, 70%, and 90%) to inform scale-up planning. METHODS Study Design and Data Sources We conducted an interrupted time series analysis to evaluate Kazakhstan's national PrEP programme for GBMSM. ITS is a quasi-experimental design that evaluates intervention effects by comparing post-implementation trends with pre-intervention trajectories, while controlling for secular trends ( 28 , 29 ). Nationwide implementation precluded randomised or non-randomised control group designs ( 30 ). We obtained monthly surveillance data from Kazakhstan's national HIV testing and PrEP monitoring system, covering January 2020 to December 2024 (60 months). The PrEP programme began in May 2021. The pre-implementation period (January 2020–April 2021) coincided with pandemic-related restrictions; we discuss implications in the Limitations section. The surveillance system captures all HIV testing conducted through public sector AIDS Centres, which perform approximately 95% of GBMSM testing nationwide. We extracted monthly data on: ( i ) current and new PrEP users among GBMSM; ( ii ) HIV tests conducted among GBMSM; ( iii ) new HIV diagnoses among GBMSM; ( iv ) late diagnoses among GBMSM, defined as CD4 count < 350 cells/µL or AIDS-defining illness at initial presentation; and ( v ) HIV-negative GBMSM population estimates derived from national size estimation studies (representing the total HIV-negative GBMSM population in Kazakhstan, independent of testing coverage). Monthly aggregate statistics were ≥ 95% complete throughout the study period, with no systematic missingness patterns. Outcome Measures and Definitions The primary outcome was HIV test positivity rate, calculated as (monthly diagnoses ÷ monthly tests) × 100. Test positivity serves as a proxy for population-level incidence, accounting for fluctuations in testing volume ( 31 , 32 ). Despite being influenced by testing patterns and programme characteristics, including repeat testing and coverage of previously diagnosed individuals ( 32 – 34 ), this metric enables standardised monitoring where cohort-based incidence measurement is infeasible due to population mobility and incomplete registration. We address these influences through sensitivity analyses. Secondary outcomes included: ( 1 ) diagnosis rate per 1,000 HIV-negative GBMSM, calculated as (monthly diagnoses ÷ HIV-negative GBMSM) × 1,000, providing an incidence measure less influenced by testing volume changes; and ( 2 ) monthly late diagnosis rate, calculated as (late diagnoses ÷ total diagnoses) × 100. Late diagnosis was defined as CD4 < 350 cells/µL or AIDS-defining illness at first presentation, with CD4 counts measured routinely at diagnosis per national protocols. Statistical Analysis Descriptive Statistics. We calculated means and standard deviations for all outcomes during pre-implementation (January 2020–April 2021, n = 16 months) and post-implementation (May 2021–December 2024, n = 44 months) periods. Percentage change was calculated as ((post-mean – pre-mean) / pre-mean) × 100. Primary Analysis. We specified the ITS model as a segmented linear regression, allowing for both immediate level changes and slope changes at the intervention point: Y ₜ = β ₀ + β ₁ ×Time + β ₂ ×Intervention + β ₃ ×Time since intervention + ε ₜ where: Yₜ = test positivity rate at month t Time = months from study start (1–60) Intervention = binary indicator (0 = pre-PrEP, 1 = post-PrEP) Time since intervention = months post-intervention (0 before May 2021) β₀ = baseline intercept β₁ = baseline trend (pre-intervention slope) β₂ = immediate level change at intervention β₃ = slope change post-intervention ε t = random error We included both immediate level change ( β₂ ) and slope change ( β₃ ) parameters following best practices for ITS analysis ( 29 , 35 ), which recommends specifying the full segmented regression model a priori. This approach enables detection of effects manifesting as immediate shifts ( β₂ ), gradual accumulation ( β₃ ), or both, without constraining the model based on theoretical expectations. While we hypothesized gradual effect accumulation based on incremental scale-up from 0 to 1,936 users over 44 months, specifying the complete model allows the data to reveal the actual nature of intervention effects. We estimated the model using ordinary least squares with Newey-West heteroscedasticity and autocorrelation consistent (HAC) robust standard errors ( 36 ). These account for potential autocorrelation between adjacent time points and heteroscedasticity in time series data. Counterfactual Analysis. Kazakhstan's PrEP programme scaled gradually from 0 to 1,936 users over 44 months rather than achieving immediate full implementation. For gradually scaled interventions, standard ITS decomposition into immediate level changes ( β₂ ) and slope changes ( β₃ ) may lack sensitivity ( 35 ). Effects accumulate incrementally over time rather than occurring as discrete shifts at programme launch. With gradual scale-up, month-to-month incremental changes remain small even when cumulative programme effects are substantial. To address this limitation, we conducted counterfactual analysis, which tests cumulative effects across the entire post-intervention period without decomposing them into discrete parameters, thereby providing greater statistical power. We estimated the counterfactual trajectory - what would have occurred had the pre-intervention trend ( β₀ + β₁×Time ) continued unchanged into the post-implementation period (May 2021–December 2024, n = 44 months). For each post-intervention month t, the intervention effect was calculated as: Effect = Counterfactual − Observed The mean intervention effect across all 44 post-intervention months was tested using a one-sample t-test against the null hypothesis of zero effect. This approach examines total cumulative divergence from projected trends rather than incremental month-to-month changes, increasing sensitivity for detecting gradually accumulating programme effects. Model Diagnostics. We conducted model diagnostics to verify regression assumptions, following established standards for time series analysis ( 30 ). These included the Durbin-Watson test for autocorrelation, visual inspection of residual plots for heteroscedasticity, Q-Q plots for normality assessment, and autocorrelation/partial autocorrelation functions (Figure S1 ). Sensitivity Analyses. We assessed robustness through alternative model specifications and outcome measures. First, a quadratic time trend model included a Time² term to test for non-linear temporal patterns. Second, a lag-adjusted model excluded the first three months post-intervention to account for potential delayed effects. Third, we included log(testing volume) as a covariate to assess whether positivity changes were independent of testing volume, controlling for potential selection bias. We compared models using Akaike Information Criterion (AIC) and R² values. To ensure findings were not artefacts of outcome selection, we repeated the ITS analysis using alternative outcomes. These included diagnosis rate per 1,000 HIV-negative GBMSM, which reflects absolute incidence less influenced by testing volume changes, and late diagnosis rate, which indicates disease severity at diagnosis. Power Analysis. We conducted post-hoc power analysis to evaluate the precision of our ITS estimates. Effect size was calculated as Cohen's f² = R² / (1–R²) , classified per standard thresholds: small (0.02), medium (0.15), and large (0.35). Achieved power was calculated using standard formulas for multiple regression with α = 0.05. Infections Prevented Estimation. We estimated infections prevented using current programme coverage, IBBS-derived incidence, and meta-analytic effectiveness estimates. The national PrEP programme defines substantial risk using behavioural screening criteria: multiple sexual partners, inconsistent condom use during anal intercourse, or recent sexually transmitted infection within six months. Based on 2023 IBBS data, 15% of HIV-negative GBMSM self-reported these risk indicators ( 2 ). We applied this 15% proportion to the December 2024 HIV-negative GBMSM population of 64,202 individuals ( 37 ), yielding an estimate of 9,630 GBMSM at substantial risk. Current coverage was calculated as (current PrEP users / at-risk population) × 100. We used the IBBS-derived incidence estimate of 8.8% annually from the 2023 study among GBMSM with behavioural risk factors ( 2 ), representing the most recent national estimate that accounts for concentrated epidemic dynamics through targeted sampling. For comparison, surveillance-based incidence calculated from monthly HIV diagnoses among GBMSM yielded a substantially lower estimate of 0.55% annually. The surveillance-based estimate likely underestimates true incidence due to: ( i ) limited testing coverage, ( ii ) window period infections, and ( iii ) selection bias among those seeking testing. We therefore used the IBBS-derived 8.8% estimate for all infections prevented calculations. PrEP Effectiveness and Calculations. Primary real-world effectiveness of 86% was derived from meta-analyses of PrEP programmes accounting for real-world adherence patterns ( 13 , 14 ). This reflects a conservative estimate midway between clinical trial efficacy of 99% with perfect adherence and typical field effectiveness of 70–90%. Sensitivity analyses used 99% optimal effectiveness. Annual infections prevented were calculated assuming constant incidence and effectiveness: Infections prevented = current users × annual incidence × effectiveness We calculated 95% confidence intervals using normal approximation to the binomial distribution. Number needed to treat (NNT) was calculated as NNT = total PrEP users / infections prevented. Coverage Projections. Projections were generated for strategic coverage targets: 30%, 50%, 70%, and 90%. For each scenario: users needed = coverage target × at-risk population (9,630 GBMSM), users needed × annual incidence (8.8%) × effectiveness (86%), and additional users = users needed − current users (1,936). Five-year cumulative impact was calculated as annual infections prevented multiplied by five, assuming constant incidence and coverage throughout the projection period. This assumption may underestimate true benefit if PrEP reduces population-level incidence. All analyses were performed using R version 4.5.0 ( 38 ) with packages: lmtest, sandwich, broom, and pwr. Ethics Approval This study used de-identified, aggregated surveillance data collected through Kazakhstan's national HIV monitoring system as part of routine public health activities. All data were aggregated at the programme level with no individual identifiers, exempting the study from individual informed consent requirements under Kazakhstan research regulations. The study was conducted in accordance with the principles of the Declaration of Helsinki. The Local Ethics Committee of the Kazakh Scientific Centre of Dermatology and Infectious Diseases granted ethical approval (Protocol No. 9, Approval No. 4-2024, dated 27 September 2024). RESULTS Descriptive Statistics and Program Scale-Up By December 2024, 1,936 GBMSM were receiving PrEP, representing 20.1% coverage of the 9,630 GBMSM at substantial risk (Table 1 , Figure S2). During the post-implementation period, HIV testing volume increased by 96.0% (p < 0.001), while absolute HIV diagnoses increased by 54.4% (p < 0.001). The primary outcome measure, HIV test positivity rate, decreased from 3.38% (95% CI: 2.58–4.17%) before implementation to 2.51% (95% CI: 2.15–2.87%) after implementation, representing a 25.7% relative reduction. Table 1 Descriptive statistics comparing pre- and post-PrEP implementation periods (January 2020–December 2024) Indicator Pre-implementation (n = 16 months) Post-implementation (n = 44 months) PrEP users (current) 0.0 ± 0.0 672.4 ± 566.9 PrEP users (new) 0.0 ± 0.0 85.9 ± 66.9 HIV tests conducted 657.4 ± 280.8 1288.7 ± 473.3 New HIV diagnoses 19.1 ± 4.8 29.5 ± 6.7 Test positivity rate, % 3.4 ± 1.6 2.5 ± 0.9 Late HIV diagnosis, n 3.1 ± 2.7 4.7 ± 2.7 Late HIV diagnosis rate, % 15.6 ± 11.6 15.8 ± 7.8 Values are presented as mean ± standard deviation. Pre-implementation period: January 2020–April 2021. Post-implementation period: May 2021–December 2024. PrEP, pre-exposure prophylaxis. The diagnosis rate per 1,000 HIV-negative GBMSM increased by 50.0% (p < 0.001), rising from 0.34 (95% CI: 0.26–0.42) to 0.51 (95% CI: 0.45–0.57) per 1,000, reflecting the 96% increase in testing volume. The late diagnosis rate remained stable throughout the study period at approximately 16% (pre-implementation: 15.7%; post-implementation: 15.8%; p = 0.87). Primary Interrupted Time Series Model Standard segmented regression analysis examined immediate and gradual effects of the intervention. Parameter estimates were: baseline level β₀ = 3.734 (95% CI: 2.534–4.935, p < 0.001); baseline trend β₁ = -0.047 percentage points per month (95% CI: -0.175 to 0.081, p = 0.50); immediate level change β₂ = -0.254 (95% CI: -1.728 to 1.220, p = 0.76); and slope change β₃ = +0.037 (95% CI: -0.095 to 0.169, p = 0.60). Neither the immediate level change nor the slope change reached statistical significance, with confidence intervals spanning both beneficial and harmful directions. These non-significant parameters are consistent with expected low statistical power (66%, below the conventional 80% threshold) for detecting gradually accumulating effects through parameter decomposition. Model fit was modest ( R² =0.127, adjusted R² =0.080), though diagnostics indicated acceptable model validity (Durbin-Watson = 1.80, p = 0.41). Model Diagnostics Model diagnostics indicated acceptable validity (Figure S1 ). The Durbin-Watson statistic (1.80, p = 0.41) indicated no significant autocorrelation. Visual inspection of residuals revealed random scatter with no evident patterns or heteroscedasticity. The Q-Q plot showed acceptable normality with only minor tail deviations. Autocorrelation and partial autocorrelation functions showed no significant serial correlation beyond lag 1, and no influential outliers were identified. Counterfactual Analysis Counterfactual analysis demonstrated a significant reduction in test positivity: mean effect of 0.537 percentage points per month (95% CI: 0.393–0.681, p < 0.001) (Fig. 1 ). This represents the divergence between observed outcomes, and the projected trajectory had pre-intervention patterns continued unchanged. At the mean post-implementation testing volume (1,289 tests per month), this reduction translates to approximately seven fewer diagnoses per month, or 84 annually. Sensitivity Analyses Sensitivity analyses examining multiple model specifications confirmed the robustness of primary findings (Table S1 ). Models including log(testing volume) as a covariate showed improved fit ( R² =0.663 vs 0.127; AIC = 135.2 vs 190.3), while intervention parameters remained in the beneficial direction (reduced positivity). Quadratic time trend and lag-adjusted models (excluding the first three post-intervention months) produced similar parameter estimates to the primary linear model ( R² =0.130, AIC = 192.1 and R² =0.115, AIC = 175.4, respectively). The primary linear model demonstrated the best balance between fit and parsimony (AIC = 190.3), with all alternative specifications showing consistent trends towards reduced positivity. Analysis using alternative outcomes demonstrated consistency across indicators (Table S2). The diagnosis rate per 1,000 HIV-negative GBMSM increased from 0.34 (95% CI: 0.26–0.42) to 0.51 (95% CI: 0.45–0.57) representing a 48.8% increase (p < 0.001), with non-significant ITS parameters. The late diagnosis rate remained stable throughout the study period at 15.7% (95% CI: 9.8–21.6%) pre-implementation and 15.8% (95% CI: 12.3–19.3%) post-implementation, with no significant change (p = 0.87). Power Analysis Post-hoc power analysis showed that the standard ITS model achieved 66% power (n = 60 observations, k = 3 predictors), below the conventional 80% threshold. Cohen's f² was 0.145, indicating a small effect size (Table S3). For the counterfactual analysis, we calculated Cohen's d as the mean monthly effect (0.537 percentage points) divided by the standard deviation of monthly effects (SD = 0.98), yielding d = 0.55, which represents a medium effect size. The one-sample t-test achieved 95% power (α = 0.05, n = 44 post-intervention months, two-tailed test). Current Program Impact and Projections At current coverage (20.1%), the programme prevents an estimated 146 infections annually (95% CI: 123–168) based on the IBBS-derived incidence rate of 8.8% and real-world effectiveness of 86%, yielding a number needed to treat of 13. With optimal effectiveness (99%), these figures rise to 168 infections prevented annually (95% CI: 148–188) and NNT of 12. Projections for strategic coverage targets demonstrate a direct correlation between coverage and infections prevented (see Table 2 ). Scaling up to 70% coverage would prevent 510 infections per year (95% CI: 467–552) and would require an additional 4,805 users. At the maximum feasible coverage level of 90%, 655 infections would be prevented each year (95% CI: 607–704). Assuming a constant incidence and coverage rate, the five-year cumulative impact would range from 1,090 infections prevented at a 30% coverage rate to 3,275 at a 90% coverage rate. Table 2 Projected annual infections prevented under varying PrEP coverage scenarios (86% real-world effectiveness) Coverage (%) Users Needed Additional Users a Infections Prevented b 95% CI 5-Year Cumulative Current (20.1%) 1,936 NA 146 (123–168) 730 30 2,889 953 218 (190–246) 1,090 50 4,815 2,879 364 (328–400) 1,820 70 6,741 4,805 510 (467–552) 2,550 90 8,667 6,731 655 (607–704) 3,275 Note : All estimates assume 86% real-world effectiveness and 8.8% annual incidence among GBMSM at substantial risk (n = 9,630). Five-year cumulative projections assume constant incidence and coverage throughout the period. a Additional users required beyond current programme (December 2024). b Annual infections prevented = current users × annual incidence × effectiveness. DISCUSSION This study provides the first quasi-experimental evidence for population-level PrEP impact in Central Asia, a region experiencing one of the world's fastest-accelerating HIV epidemics among GBMSM. Using interrupted time series analysis of 44 months of surveillance data from Kazakhstan's national programme, we demonstrate a significant association between programme scale-up to 20.1% coverage and HIV test positivity reduction (0.537 percentage points monthly, 95% CI: 0.393–0.681, p < 0.001). At current scale, the programme prevents approximately 146 infections annually with a number needed to treat of 13. This indicates high efficiency in a concentrated epidemic setting where GBMSM face HIV incidence 30 times higher than the general population. Evidence for Programme Effectiveness Three independent analytical approaches converge on consistent reductions in HIV test positivity, strengthening confidence in programme effectiveness. Counterfactual analysis demonstrated highly significant effects (p < 0.001) that remained robust across all sensitivity specifications, including models controlling for testing volume. The consistency between empirical estimates (84 diagnoses averted annually among tested individuals) and theoretical calculations (146 total infections prevented across all at-risk GBMSM) supports construct validity. This difference reflects measurement scope: empirical estimates capture only partial surveillance coverage, whilst theoretical calculations account for the complete at-risk population. Simple pre-post comparison showed a 25.7% relative reduction (from 3.38% to 2.51%). While quasi-experimental designs cannot establish definitive causation, the convergence across independent analytical approaches, combined with the biological plausibility of PrEP's preventive mechanism, provides strong evidence for population-level programme effectiveness. This 25.7% reduction closely aligns with population-level effects in high-income settings. Ecological analysis across all US states showed a 38% reduction in HIV diagnoses where PrEP coverage was highest ( 22 ), while Scotland's national program demonstrated a 19.7% reduction at 19.5% coverage ( 23 ). Kazakhstan achieved a similarly effective result with almost identical coverage (20.1%), despite operating in a middle-income setting with legislative restrictions on key populations and healthcare-related stigma. This consistency across different health systems and geographic contexts reinforces the reproducible population-level impact of PrEP when it is targeted appropriately. The robustness of intervention effects when controlling for testing volume indicates genuine prevention impact rather than testing artefacts. Including log(testing volume) as a covariate substantially improved model fit ( R² increased from 0.127 to 0.663), yet intervention parameters maintained their beneficial direction. This R² improvement demonstrates that testing volume explains substantial variance, but crucially, PrEP effects persist independently. The observed epidemiological pattern is characteristic of successful PrEP implementation combined with expanded testing infrastructure, as documented in other concentrated epidemic settings ( 13 , 15 ). This pattern includes increased case detection (+ 48.8% in diagnosis rate), decreased test positivity (− 25.7%), and stable late diagnosis rates (~ 16%). This combination suggests the programme both identifies previously undiagnosed infections through expanded testing and prevents new infections through PrEP, with no detectable change in diagnosis timing. Public Health Significance The programme demonstrates substantial public health impact. The standard segmented regression model showed modest statistical effect size (Cohen's f²=0.145) and low power (66%), consistent with the gradually accumulating nature of programme effects. However, the counterfactual analysis, which achieved 95% power by testing cumulative effects, demonstrated a medium effect (Cohen's d = 0.55) that translated to clinically and epidemiologically significant reductions in HIV test positivity: 0.537 percentage points monthly (p < 0.001), representing approximately 7 fewer diagnoses per month among tested individuals. In concentrated epidemics where GBMSM experience 30-fold higher HIV incidence than the general population, each PrEP user generates outsized prevention benefits. The number needed to treat of 13 indicates high intervention efficiency: approximately 13 person-years of PrEP prevent one HIV infection annually. This aligns with NNT estimates from clinical trials conducted among GBMSM at substantial risk for HIV, demonstrating that Kazakhstan's national programme replicates trial-level efficiency in real-world implementation ( 16 , 39 , 40 ). For broader context, NNTs for many established preventive interventions in chronic disease management range from 20 to over 100 ( 41 – 43 ). The programme's current impact (146 infections prevented annually) translates to substantial individual and societal benefits, given lifetime HIV treatment costs and morbidity. This efficiency arises from precise targeting of biomedical prevention to populations at highest risk, demonstrating that PrEP achieves meaningful population-level impact even at modest coverage levels in concentrated epidemic settings. Strategic Considerations for Programme Expansion Scaling up to 70% coverage would increase the programme's impact three and a half times, preventing 510 infections annually compared to the current 146. Although PrEP has been approved in 38 out of 53 countries in the WHO European Region, only 344,596 individuals received it in 2024, which is below the regional target of 500,000. Uptake remains particularly constrained in EECA due to legislative restrictions on key populations, inadequate primary care integration, and urban-only service delivery ( 44 ). Kazakhstan's experience shows that national programmes can have a significant impact on the population in resource-constrained EECA settings, providing a model for scaling up across the region. However, achieving this expansion requires addressing three interconnected barriers documented in Kazakhstan: legislative restrictions discouraging healthcare-seeking among sexual minorities, limited integration of PrEP services with primary care, and geographic concentration in major cities ( 5 – 7 ). Evidence from other concentrated epidemic settings demonstrates that service delivery diversification, including community-based distribution, telehealth, and pharmacy dispensing, substantially improves PrEP accessibility while reducing reliance on specialised AIDS Centres ( 45 – 47 ). The forthcoming availability of twice-yearly injectable lenacapavir presents additional opportunities for adherence improvement and effectiveness gains ( 48 , 49 ). Our findings suggest that even modest coverage expansion yields substantial prevention returns in concentrated epidemics, making targeted outreach to underserved subpopulations, particularly chemsex participants with elevated HIV risk, a high-priority strategy. The 13-percentage-point gap between trial-based (99%) and real-world (86%) PrEP effectiveness represents substantial unrealised prevention potential. At current programme scale (1,936 users), closing this adherence gap would prevent an additional 22 infections annually (168 versus 146). Across the full at-risk population (9,630 GBMSM), perfect adherence at 70% coverage would prevent 580 infections annually versus 510 with 86% effectiveness, a 14% gain achievable through strengthened adherence support rather than recruitment expansion. Evidence-based adherence interventions, including peer counselling, proactive side-effect management, and event-driven dosing protocols, offer cost-effective pathways to maximise prevention returns from existing users. The efficient NNT of 13 and substantial lifetime costs of HIV treatment suggest that both coverage expansion and adherence optimisation would likely be cost-effective, although formal economic evaluation is needed. Continued surveillance investment, particularly repeat IBBS to monitor population-level incidence trends and assess programme reach across risk subgroups, remains essential for evaluating whether these improvements translate to measurable epidemic impact. Study Limitations Several methodological limitations require consideration. While the ITS design is robust to confounding from time-invariant characteristics such as age, socioeconomic status, or baseline geography ( 29 ), the use of aggregate surveillance data prevents assessment of PrEP effectiveness heterogeneity across subgroups. Individual-level data would enable more granular understanding of which GBMSM subpopulations (defined by age, geography, or specific risk profiles) benefit most from programme participation, though the robust overall trend provides strong evidence of population-level effectiveness. Pre-implementation period (January 2020–April 2021) coincided with pandemic-related healthcare disruptions and altered sexual behaviour patterns. While temporally unavoidable for programmes launched during COVID-19, this timing means our counterfactual projections use pandemic-affected baselines rather than stable pre-pandemic epidemic dynamics. However, this limitation likely biases our estimates conservatively: pandemic restrictions reduced both HIV transmission opportunities and testing access, potentially underestimating true programme impact as conditions normalised. The introduction of PrEP scale-up, along with a 96% surge in testing volume, has led to a more complex attribution landscape. While models controlling for log(testing volume) substantially improved fit (R²: 0.127→0.663) and maintained beneficial intervention effects, complete separation of PrEP's direct preventive impact from testing pattern changes remains infeasible with aggregate data. Expanded testing could reduce observed positivity through two mechanisms: recruiting lower-risk individuals (denominator effects) or identifying previously undiagnosed prevalent infections (numerator effects that would naturally decline over time). However, three findings suggest these mechanisms do not fully explain observed reductions: ( 1 ) intervention effects persisted across all testing volume-adjusted models, ( 2 ) late diagnosis rates remained stable rather than increasing (as expected if testing recruited healthier individuals), and ( 3 ) the magnitude of positivity reduction (25.7%) exceeded what denominator effects alone could plausibly generate given testing expansion magnitude. Finally, infections prevented estimates assume homogeneous incidence across PrEP users, constant 86% effectiveness, and no risk compensation or network effects. These simplifying assumptions likely produce conservative estimates for three reasons: ( 1 ) PrEP programmes typically achieve higher coverage among highest-risk individuals, amplifying prevention impact; ( 2 ) network effects (reduced transmission from protected individuals to their partners) compound direct prevention benefits; and ( 3 ) our 86% effectiveness estimate is conservative relative to newer long-acting formulations. While agent-based transmission models could better capture transmission dynamics, our linear calculations provide reasonable lower-bound estimates suitable for programme planning. The 8.8% annual incidence estimate from 2023 IBBS, while subject to sampling uncertainty, represents the most robust available measure through purposive sampling of GBMSM at substantial risk. The 16-fold discrepancy from surveillance-based estimates (0.55%) reflects known limitations of passive surveillance, incomplete testing coverage, window period infections, and selection bias, reinforcing our reliance on IBBS-derived parameters for impact calculations. These limitations are offset by methodological strengths that enhance confidence in our conclusions. Counterfactual analysis achieved 95% statistical power, compared to 66% for standard segmented regression, making it more sensitive to gradually scaling interventions. Our findings remained consistent across multiple analytical approaches, sensitivity specifications, and alternative outcomes. Alongside biological plausibility and concordance between empirical and theoretical estimates, this provides substantial evidence of the effectiveness at a population level. CONCLUSIONS Kazakhstan's national PrEP programme has had a measurable impact on the population in an area with a concentrated HIV epidemic. Counterfactual analysis of 44 months of surveillance data revealed a significant association between scaling up the programme to achieve 20.1% coverage and a reduction in HIV test positivity. This association remained robust across sensitivity specifications. At its current scale, the programme prevents around 146 infections each year scaling up to 70% coverage could prevent around 510 infections each year. This provides the first quasi-experimental evidence of PrEP's population-level effectiveness in Central Asia, where 94% of new HIV infections occur among GBMSM. From a methodological perspective, counterfactual analysis achieved 95% statistical power compared to 66% for standard segmented regression, offering a valuable framework for evaluating the gradual scaling of interventions. These findings demonstrate that biomedical prevention can have a meaningful population-level impact in concentrated epidemics, even at modest coverage, with implications for similar contexts across Eastern Europe and Central Asia. Declarations Conflict of Interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Ethics statement The Local Ethics Committee of the Kazakh Scientific Centre of Dermatology and Infectious Diseases granted ethical approval (Protocol No. 9, Approval No. 4-2024, dated 27 September 2024). Clinical Trial number Not applicable. Funding This article is the result of work carried out as part of a scientific project funded by the Ministry of Science and Higher Education of the Republic of Kazakhstan, IRN AR25794027 “Improving HIV prevention in MSM by examining epidemiological, socio-economic, psychological, and biobehavioral factors influencing pre-exposure prophylaxis use.” Author Contribution N.L. and S.K. contributed equally to this work and share senior authorship. I.K. contributed to conceptualization, methodology, data collection, study design, funding acquisition, writing – original draft, and writing – review and editing. B.T. contributed to methodology, software, validation, funding acquisition, and writing – review and editing. G.S. contributed to data curation, supervision, and writing – review and editing. M.R. and G.A. contributed to validation, software, and visualization. E.A., E.G., and A.U. contributed to investigation, supervision, and project administration. S.K. contributed to conceptualization, investigation, supervision, project administration, and writing – review and editing. N.L. contributed to conceptualization, methodology, formal analysis, statistical analysis, data curation, visualization, writing – original draft, and writing – review and editing. All authors reviewed the manuscript. Acknowledgement The authors wish to acknowledge the healthcare professionals at State AIDS Centres across Kazakhstan who deliver PrEP services, often in challenging conditions shaped by stigma and resource constraints. Their commitment to providing non-judgmental, evidence-based care to GBMSM has been essential to the programme's success. We are equally grateful to the community-based organizations whose outreach, advocacy, and trust-building efforts have made PrEP uptake possible. In particular, we thank Public Fund "Community Friends" and Public Fund "Human Health Institute" for their sustained contributions to HIV prevention among key populations in Kazakhstan. We also acknowledge the broader network of civil society organizations and peer educators working across the country whose engagement with communities at substantial risk underpins the programme's reach. We thank Arailym Abilbay for her contribution to data and reference management. Finally, we express our gratitude to the GBMSM community members who engage with prevention services and whose participation makes this research possible. Data Availability The dataset analyzed during the current study is available from the corresponding author (N.K. [ [email protected] ] (mailto: [email protected] ) ) and the first author (I.K. [ [email protected] ] (mailto: [email protected] ) ) upon reasonable request. References UNAIDS. What the data tell us: Projections for the HIV epidemic in Eastern Europe and Central Asia in 2030. Geneva: Joint United Nations Programme on HIV/AIDS; 2025. March 2025. Republican Center for the Prevention and Control of AIDS MoHotRoK. Monitoring and Evaluation of the HIV Response in the Republic of Kazakhstan, 2023. National HIV Monitoring & Evaluation Report (Annual report). Almaty, Republic of Kazakhstan. 2024. Contract No.: № 05-0186. UNAIDS. Kazakhstan: Progress towards 90-90-90 https://aidsinfo.unaids.org/?did=554c9b9c05b5b281750af202&r=world&t=null&tb=g&bt=gnli&ts=0,0&gid=5977091c3ba08432558d099c&gr=KAZ.2023[ Bilibayeva G, Ospanova D, Nurkerimova A, Kussainova F, Tukeev M, Shokybaeva M, et al. Epidemiological Analysis of HIV/AIDS in Kazakhstan During 2018–2020. J Res Health Sci. 2023;23(2):e00580. Reuters. Kazakhstan lawmakers give preliminary approval to ban on LGBT 'propaganda' 2025 [updated October 29, 2025. Available from: https://www.reuters.com/world/asia-pacific/kazakhstan-lawmakers-give-preliminary-approval-ban-lgbt-propaganda-2025-10-29/ Paine EA, Lee YG, Mergenova G, Vinogradov V, Laughney CI, Davis A et al. Compounding vulnerabilities: victimization and discrimination is associated with COVID-19 disruptions to HIV-related care among gay, bisexual, and other men and transgender and nonbinary people who have sex with men in Kazakhstan. AIDS Care. 2023:1–7. Eurasian Coalition on Health R, Gender and Sexual Diversity,. National report on violations of the rights of gay men, other MSM and trans* people in Kazakhstan. 2022. Tallinn; 2023. Stringer KL, Mukherjee T, McCrimmon T, Terlikbayeva A, Primbetovac S, Darisheva M, et al. Attitudes towards people living with HIV and people who inject drugs: A mixed method study of stigmas within harm reduction programs in Kazakhstan. Int J Drug Policy. 2019;68:27–36. Iskakova B, King EJ, Yucel RM, DeHovitz J, Nugmanova Z. I think they are infected because of their ignorance and lack of responsibility: A mixed-methods study on HIV-related stigma in the healthcare system in Kazakhstan. PLoS ONE. 2025;20(9):e0331201. King EJ, Kussainova AZ, Erlon-Baurjan J, Nugmanova ZS. Healthcare Providers’ Perspectives and Role in Improving Patient Engagement in HIV Treatment and Care in Almaty, Kazakhstan. Int J Environ Res Public Health. 2021;18(22):12138. Lunchenkov N, Cherchenko N, Altynbekov K, Terlikbayeva A, Primbetova S, Gryazev D, et al. A way to liberate myself: A qualitative study of perceived benefits and risks of chemsex among gay, bisexual, and other men who have sex with men in Almaty, Kazakhstan. Drug Alcohol Depend. 2024;264:112464. Lee YG, Zhakupova G, Vinogradov V, Paine EA, Laughney CI, Reeder K, et al. Polydrug use, sexual risk, and HIV testing among cisgender gay, bisexual, and other men and transgender and nonbinary individuals who have sex with men in Kazakhstan. AIDS Educ Prev. 2022;34(5):413–26. Murchu EO, Marshall L, Teljeur C, Harrington P, Hayes C, Moran P, et al. Oral pre-exposure prophylaxis (PrEP) to prevent HIV: a systematic review and meta-analysis of clinical effectiveness, safety, adherence and risk compensation in all populations. BMJ open. 2022;12(5):e048478. Sun Z, Gu Q, Dai Y, Zou H, Agins B, Chen Q, et al. Increasing awareness of HIV pre-exposure prophylaxis (PrEP) and willingness to use HIV PrEP among men who have sex with men: a systematic review and meta‐analysis of global data. J Int AIDS Soc. 2022;25(3):e25883. Khurana N, Yaylali E, Farnham PG, Hicks KA, Allaire BT, Jacobson E, et al. Impact of improved HIV care and treatment on PrEP effectiveness in the United States, 2016–2020. JAIDS J Acquir Immune Defic Syndr. 2018;78(4):399–405. McCormack S, Dunn DT, Desai M, Dolling DI, Gafos M, Gilson R, et al. Pre-exposure prophylaxis to prevent the acquisition of HIV-1 infection (PROUD): effectiveness results from the pilot phase of a pragmatic open-label randomised trial. Lancet. 2016;387(10013):53–60. WHO. WHO implementation tool for pre-exposure prophylaxis (PrEP) of HIV infection: Module 1: Clinical. World Health Organization; 2017. WHO. WHO expands the recommendation on oral pre-exposure prophylaxis of HIV infection (PrEP). 2015. Nunn AS, Brinkley-Rubinstein L, Oldenburg CE, Mayer KH, Mimiaga M, Patel R, et al. Defining the HIV pre-exposure prophylaxis care continuum. Aids. 2017;31(5):731–4. Kelley CF, Kahle E, Siegler A, Sanchez T, Del Rio C, Sullivan PS, et al. Applying a PrEP continuum of care for men who have sex with men in Atlanta, Georgia. Clin Infect Dis. 2015;61(10):1590–7. Sullivan PS, Siegler AJ. Getting pre-exposure prophylaxis (PrEP) to the people: opportunities, challenges and emerging models of PrEP implementation. Sex Health. 2018;15(6):522–7. Sullivan PS, Juhasz M, DuBose SN, Le G, Brisco K, Islek D, et al. Association of state-level PrEP coverage and new HIV diagnoses in the USA from 2012 to 2022: an ecological analysis of the population impact of PrEP. Lancet HIV. 2025;12(6):e440–8. Estcourt C, Yeung A, Nandwani R, Goldberg D, Cullen B, Steedman N, et al. Population-level effectiveness of a national HIV preexposure prophylaxis programme in MSM. Aids. 2021;35(4):665–73. Grulich AE, Guy R, Amin J, Jin F, Selvey C, Holden J, et al. Population-level effectiveness of rapid, targeted, high-coverage roll-out of HIV pre-exposure prophylaxis in men who have sex with men: the EPIC-NSW prospective cohort study. lancet HIV. 2018;5(11):e629–37. Koss CA, Havlir DV, Ayieko J, Kwarisiima D, Kabami J, Chamie G, et al. HIV incidence after pre-exposure prophylaxis initiation among women and men at elevated HIV risk: a population-based study in rural Kenya and Uganda. PLoS Med. 2021;18(2):e1003492. Ministry of Health of the Republic of Kazakhstan. On approval of the list of medicinal products and medical devices for free and/or concessional outpatient provision to certain categories of citizens of the Republic of Kazakhstan with specified diseases (conditions). Nur-Sultan (now Astana), Kazakhstan2021. Kazakh Scientific Center for Dermatology and Infectious Diseases KDI. Стратегический план РГП на ПХВ «Казахский научный центр дерматологии и инфекционных заболеваний» на 2025–2029 годы. Almaty, Republic of Kazakhstan: Казахский научный центр дерматологии и инфекционных заболеваний; 2025. p. 48 pp. Anderson-Cook CM. Experimental and quasi-experimental designs for generalized causal inference. Taylor & Francis; 2005. Bernal JL, Cummins S, Gasparrini A. Interrupted time series regression for the evaluation of public health interventions: a tutorial. Int J Epidemiol. 2017;46(1):348–55. Kontopantelis E, Doran T, Springate DA, Buchan I, Reeves D. Regression based quasi-experimental approach when randomisation is not an option: interrupted time series analysis. BMJ. 2015;350. Brookmeyer R. Measuring the HIV/AIDS epidemic: approaches and challenges. Epidemiol Rev. 2010;32(1):26–37. Godin A, Eaton JW, Giguère K, Marsh K, Johnson LF, Jahn A, et al. Inferring population HIV incidence trends from surveillance data of recent HIV infection among HIV testing clients. Aids. 2021;35(14):2383–8. Martin EG, MacDonald RH, Smith LC, Gordon DE, Lu T, O'Connell DA. Modeling the declining positivity rates for human immunodeficiency virus testing in New York state. J Public Health Manage Pract. 2015;21(6):556–63. Jones HS, Hensen B, Musemburi S, Chinyanganya L, Takaruza A, Chabata ST, et al. Interpreting declines in HIV test positivity: an analysis of routine data from Zimbabwe's national sex work programme, 2009–2019. J Int AIDS Soc. 2022;25(7):e25943. Lopez Bernal J, Cummins S, Gasparrini A. The use of controls in interrupted time series studies of public health interventions. Int J Epidemiol. 2018;47(6):2082–93. Newey WK, West KD. A simple, positive semi-definite, heteroskedasticity and autocorrelationconsistent covariance matrix. 1986. Казахский научный центр дерматологии и инфекционных заболеваний. Страновой Отчет, 2024 год. Almaty, Republic of Kazakhstan; 2025 17.04.2025. Report No.: № 05-0567. RStudio Team. RStudio: Integrated Development Environment for R,. 4.1.3. ed. Boston, MA: RStudio, PBC; 2021. Molina J-M, Capitant C, Spire B, Pialoux G, Cotte L, Charreau I, et al. On-demand preexposure prophylaxis in men at high risk for HIV-1 infection. N Engl J Med. 2015;373(23):2237–46. Buchbinder SP, Glidden DV, Liu AY, McMahan V, Guanira JV, Mayer KH, et al. HIV pre-exposure prophylaxis in men who have sex with men and transgender women: a secondary analysis of a phase 3 randomised controlled efficacy trial. Lancet Infect Dis. 2014;14(6):468–75. Byrne P, Cullinan J, Smith A, Smith SM. Statins for the primary prevention of cardiovascular disease: an overview of systematic reviews. BMJ open. 2019;9(4):e023085. Zheng SL, Roddick AJ. Association of aspirin use for primary prevention with cardiovascular events and bleeding events: a systematic review and meta-analysis. JAMA. 2019;321(3):277–87. Wright JM, Musini VM, Gill R. First-line drugs for hypertension. Cochrane Database Syst reviews. 2018(4). European Centre for Disease Prevention and Control. Pre-exposure prophylaxis for HIV prevention in Europe and Central Asia. Monitoring implementation of the Dublin Declaration on partnership to fight HIV/AIDS in Europe and Central Asia: 2024 progress report. Stockholm; 2025. Kamitani E, Higa DH, Crepaz N, Wichser M, Mullins MM, Control UCD, et al. Identifying best practices for increasing HIV pre-exposure prophylaxis (PrEP) use and persistence in the United States: a systematic review. AIDS Behav. 2024;28(7):2340–9. Bonett S, Li Q, Sweeney A, Gaither-Hardy D, Safa H. Telehealth models for PrEP delivery: a systematic review of acceptability, implementation, and impact on the PrEP care continuum in the United States. AIDS Behav. 2024;28(9):2875–86. Hillis A, Germain J, Hope V, McVeigh J, Van Hout MC. Pre-exposure prophylaxis (PrEP) for HIV prevention among men who have sex with men (MSM): a scoping review on PrEP service delivery and programming. AIDS Behav. 2020;24(11):3056–70. Blackwell CW, Armstrong F, Castillo HL. Lenacapavir for HIV PrEP: Interim Phase III Clinical Data Evaluation. J Nurse Practitioners. 2025:105360. Kelley CF, Acevedo-Quiñones M, Agwu AL, Avihingsanon A, Benson P, Blumenthal J, et al. Twice-yearly lenacapavir for HIV prevention in men and gender-diverse persons. N Engl J Med. 2025;392(13):1261–76. Additional Declarations No competing interests reported. Supplementary Files Supplementarymaterials.docx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 29 Apr, 2026 Reviews received at journal 28 Apr, 2026 Reviews received at journal 16 Apr, 2026 Reviewers agreed at journal 03 Apr, 2026 Reviewers agreed at journal 31 Mar, 2026 Reviewers agreed at journal 31 Mar, 2026 Reviewers invited by journal 26 Mar, 2026 Editor invited by journal 25 Mar, 2026 Editor assigned by journal 24 Mar, 2026 Submission checks completed at journal 24 Mar, 2026 First submitted to journal 18 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9163384","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":613910981,"identity":"96babdcb-ee60-4a17-b712-541162f07be2","order_by":0,"name":"Indira Karibayeva","email":"","orcid":"","institution":"Georgia Southern University","correspondingAuthor":false,"prefix":"","firstName":"Indira","middleName":"","lastName":"Karibayeva","suffix":""},{"id":613910982,"identity":"2ea37087-281c-406e-8bf9-4a20e547f1a9","order_by":1,"name":"Botagoz Turdaliyeva","email":"","orcid":"","institution":"Kazakh Scientific Centre of Dermatology and Infectious Diseases","correspondingAuthor":false,"prefix":"","firstName":"Botagoz","middleName":"","lastName":"Turdaliyeva","suffix":""},{"id":613910983,"identity":"3b14918b-e478-4636-8b03-2b93630002c9","order_by":2,"name":"Gulzar Shah","email":"","orcid":"","institution":"Georgia Southern University","correspondingAuthor":false,"prefix":"","firstName":"Gulzar","middleName":"","lastName":"Shah","suffix":""},{"id":613910984,"identity":"41792af2-02aa-4c26-81f6-14ee0e586bd6","order_by":3,"name":"Manshuk Ramazanova","email":"","orcid":"","institution":"Kazakhstan Medical University","correspondingAuthor":false,"prefix":"","firstName":"Manshuk","middleName":"","lastName":"Ramazanova","suffix":""},{"id":613910985,"identity":"72e10ef1-9af2-4f84-ac94-d51e9c91b322","order_by":4,"name":"Gulzhahan Akhmetova","email":"","orcid":"","institution":"Kazakh Scientific Centre of Dermatology and Infectious Diseases","correspondingAuthor":false,"prefix":"","firstName":"Gulzhahan","middleName":"","lastName":"Akhmetova","suffix":""},{"id":613910986,"identity":"07408580-5ddc-4d12-a227-1a53fd761076","order_by":5,"name":"Evgenia Artamonova","email":"","orcid":"","institution":"Technical University of Munich","correspondingAuthor":false,"prefix":"","firstName":"Evgenia","middleName":"","lastName":"Artamonova","suffix":""},{"id":613910987,"identity":"fefe6fa3-5c46-438c-8503-5bd7c35b6d60","order_by":6,"name":"Elena German","email":"","orcid":"","institution":"ECOM - Eurasian Coalition on Health, Rights, Gender and Sexual Diversity","correspondingAuthor":false,"prefix":"","firstName":"Elena","middleName":"","lastName":"German","suffix":""},{"id":613910988,"identity":"50f2e2f1-8e60-4d53-a307-c4a49b1c430b","order_by":7,"name":"Alya Utegenova","email":"","orcid":"","institution":"Kazakh Scientific Centre of Dermatology and Infectious Diseases","correspondingAuthor":false,"prefix":"","firstName":"Alya","middleName":"","lastName":"Utegenova","suffix":""},{"id":613910989,"identity":"8df5c483-6768-4172-b757-3d88dc0abf05","order_by":8,"name":"Sairankul Kassymbekova","email":"","orcid":"","institution":"Kazakh Scientific Centre of Dermatology and Infectious Diseases","correspondingAuthor":false,"prefix":"","firstName":"Sairankul","middleName":"","lastName":"Kassymbekova","suffix":""},{"id":613910990,"identity":"34babfaa-beec-4173-ba01-1f7e7939f9f4","order_by":9,"name":"Nikolay Lunchenkov","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABAklEQVRIiWNgGAWjYDACZjYGhgdAmg/EBmIDIHUAJMHYgE9LApBmQ2hhS2zAq4UBUwuPIV4t5uxsiQ8Sag4zsEnkHvxcUHHHmF8i5/tj3hwG2X4cWiyb2Q4bJBwDaclLlp5x5pmZ5Izcjc282xiMZ+KwxuAwe5tEAhtIS44ZM2/bYRuDM2fBWhI3HMCn5R9Myz+QljMPwVr249TCdkwisQ2mpeGwmcHxHkaILTj8AtSSbJDYl87DxvPGWJrn2GFjyfY2w5lzt0kYz8Bly/ljhg8+fLOW42fPMfzMU3PYsJ+Z+cGHt9tsZPtxeB8KmnnQRSTwqgeCOkIKRsEoGAWjYCQDALSGVyAKyp+6AAAAAElFTkSuQmCC","orcid":"","institution":"Technical University of Munich","correspondingAuthor":true,"prefix":"","firstName":"Nikolay","middleName":"","lastName":"Lunchenkov","suffix":""}],"badges":[],"createdAt":"2026-03-18 23:08:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9163384/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9163384/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105904114,"identity":"795d5e69-9616-40c3-8e16-fe947b2f5963","added_by":"auto","created_at":"2026-04-01 10:04:26","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":176601,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eInterrupted time series analysis of HIV test positivity rate among GBMSM in Kazakhstan, January 2020–December 2024. The vertical dashed line (yellow) indicates PrEP programme initiation (May 2021). Blue points represent observed monthly test positivity rates; blue solid line shows the fitted segmented regression. Red dashed line represents the counterfactual scenario (projected trend had pre-intervention patterns continued). Green shaded area illustrates cumulative intervention effect (difference between observed and counterfactual trajectories)\u003c/em\u003e\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9163384/v1/2f8e53693d4e1f00d55a3619.png"},{"id":105906348,"identity":"9e7cc22e-f16d-4548-930a-bef298501614","added_by":"auto","created_at":"2026-04-01 10:20:59","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1048198,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9163384/v1/6dd3818d-20c4-4b6b-a558-d19bf28604ed.pdf"},{"id":105764237,"identity":"c57cc97c-8cad-4166-ba63-bfc25bf106d3","added_by":"auto","created_at":"2026-03-30 19:27:09","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":496946,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-9163384/v1/bb69e01d062836095645a423.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Pre-Exposure Prophylaxis Reduces HIV Test Positivity Among Men Who Have Sex with Men in Kazakhstan: An Interrupted Time Series Analysis","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eEastern Europe and Central Asia (EECA) faces one of the world's fastest-accelerating HIV epidemics. New infections among gay, bisexual, and other men who have sex with men (GBMSM) have risen by 144% between 2010 and 2022, accounting for 94% of new infections among key populations in the region (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). During this period, Kazakhstan experienced the steepest increase in Central Asia, with new diagnoses rising by 73% between 2010 and 2020, representing the 12th highest percent increase globally (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). The most recent integrated bio-behavioural surveillance (IBBS) from 2023 reported an annual HIV incidence rate of 8.8% among GBMSM, nearly 30 times higher than the general population rate of 0.3%, indicating a concentrated and escalating epidemic (\u003cspan additionalcitationids=\"CR3\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThree interconnected structural barriers impede HIV prevention among GBMSM in Kazakhstan. First, legislative restrictions on information about sexual minorities, combined with documented discrimination in healthcare settings, have created an environment that discourages healthcare-seeking among GBMSM (\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). Second, insufficient provider training and pervasive societal stigma result in discrimination within clinical settings, undermining trust in HIV services (\u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). Third, elevated rates of substance use, including chemsex (the intentional use of drugs to enhance sexual activity), increase HIV transmission risk through both biological and behavioural mechanisms (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). Addressing these structural determinants is the ideal long-term strategy. However, legal and social reforms require sustained political will and may take years to decades to achieve meaningful change. Biomedical prevention strategies such as pre-exposure prophylaxis (PrEP) offer a complementary approach, providing immediate protection that operates independently of structural barriers while broader reforms are pursued.\u003c/p\u003e \u003cp\u003ePre-exposure prophylaxis with oral tenofovir-based regimens offers 99% efficacy among adherent users and 70\u0026ndash;90% effectiveness in real-world settings (\u003cspan additionalcitationids=\"CR14 CR15\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e). The WHO recommends PrEP for key populations at substantial risk (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e). However, achieving population-level impact from individual-level efficacy requires adequate coverage, sustained adherence and retention in care (\u003cspan additionalcitationids=\"CR20\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e). Emerging evidence from high-income countries and research demonstration projects shows that PrEP programmes can reduce HIV diagnoses by 19\u0026ndash;38% when coverage reaches key populations (\u003cspan additionalcitationids=\"CR23 CR24\" citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e). These evaluations used ecological or cohort designs in settings with an established healthcare infrastructure or intensive external support. It remains unexamined whether population-level effectiveness translates to routine national programme implementation in Central Asia.\u003c/p\u003e \u003cp\u003eIn May 2021, Kazakhstan introduced oral PrEP through the Guaranteed Amount of Free Medical Care National Program (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e). State AIDS Centres, regional public health facilities providing HIV prevention, testing, and treatment services, led a phased expansion through site activation, clinician training, and community outreach targeting GBMSM and other key populations. The national programme defines 'substantial risk' as having multiple sexual partners, inconsistent condom use during anal intercourse, or a recent sexually transmitted infection. By December 2024, 1,936 individuals were receiving PrEP, achieving 20.1% coverage of an estimated 9,630 GBMSM at substantial risk (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e). Despite robust evidence of PrEP's individual-level efficacy from clinical trials, population-level effectiveness data specifically from Central Asia remain limited. However, no published studies have examined PrEP's population-level impact on HIV testing volume, test positivity rates, or disease severity at diagnosis in this region.\u003c/p\u003e \u003cp\u003eThis study evaluated Kazakhstan's national PrEP programme using interrupted time series analysis of HIV surveillance data from January 2020 through December 2024. We pursued three specific aims: (\u003cem\u003ei\u003c/em\u003e) to quantify the programme's impact on HIV test positivity among GBMSM; (\u003cem\u003eii\u003c/em\u003e) to estimate the programme's current coverage and the number of infections prevented annually; and (\u003cem\u003eiii\u003c/em\u003e) to project infections preventable under strategic coverage scenarios (30%, 50%, 70%, and 90%) to inform scale-up planning.\u003c/p\u003e"},{"header":"METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy Design and Data Sources\u003c/h2\u003e \u003cp\u003eWe conducted an interrupted time series analysis to evaluate Kazakhstan's national PrEP programme for GBMSM. ITS is a quasi-experimental design that evaluates intervention effects by comparing post-implementation trends with pre-intervention trajectories, while controlling for secular trends (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e). Nationwide implementation precluded randomised or non-randomised control group designs (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWe obtained monthly surveillance data from Kazakhstan's national HIV testing and PrEP monitoring system, covering January 2020 to December 2024 (60 months). The PrEP programme began in May 2021. The pre-implementation period (January 2020\u0026ndash;April 2021) coincided with pandemic-related restrictions; we discuss implications in the Limitations section. The surveillance system captures all HIV testing conducted through public sector AIDS Centres, which perform approximately 95% of GBMSM testing nationwide. We extracted monthly data on: (\u003cem\u003ei\u003c/em\u003e) current and new PrEP users among GBMSM; (\u003cem\u003eii\u003c/em\u003e) HIV tests conducted among GBMSM; (\u003cem\u003eiii\u003c/em\u003e) new HIV diagnoses among GBMSM; (\u003cem\u003eiv\u003c/em\u003e) late diagnoses among GBMSM, defined as CD4 count\u0026thinsp;\u0026lt;\u0026thinsp;350 cells/\u0026micro;L or AIDS-defining illness at initial presentation; and (\u003cem\u003ev\u003c/em\u003e) HIV-negative GBMSM population estimates derived from national size estimation studies (representing the total HIV-negative GBMSM population in Kazakhstan, independent of testing coverage). Monthly aggregate statistics were \u0026ge;\u0026thinsp;95% complete throughout the study period, with no systematic missingness patterns.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eOutcome Measures and Definitions\u003c/h3\u003e\n\u003cp\u003eThe primary outcome was HIV test positivity rate, calculated as (monthly diagnoses\u0026thinsp;\u0026divide;\u0026thinsp;monthly tests) \u0026times; 100. Test positivity serves as a proxy for population-level incidence, accounting for fluctuations in testing volume (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e). Despite being influenced by testing patterns and programme characteristics, including repeat testing and coverage of previously diagnosed individuals (\u003cspan additionalcitationids=\"CR33\" citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e), this metric enables standardised monitoring where cohort-based incidence measurement is infeasible due to population mobility and incomplete registration. We address these influences through sensitivity analyses. Secondary outcomes included: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) diagnosis rate per 1,000 HIV-negative GBMSM, calculated as (monthly diagnoses\u0026thinsp;\u0026divide;\u0026thinsp;HIV-negative GBMSM) \u0026times; 1,000, providing an incidence measure less influenced by testing volume changes; and (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) monthly late diagnosis rate, calculated as (late diagnoses\u0026thinsp;\u0026divide;\u0026thinsp;total diagnoses) \u0026times; 100. Late diagnosis was defined as CD4\u0026thinsp;\u0026lt;\u0026thinsp;350 cells/\u0026micro;L or AIDS-defining illness at first presentation, with CD4 counts measured routinely at diagnosis per national protocols.\u003c/p\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003e \u003cem\u003eDescriptive Statistics.\u003c/em\u003e We calculated means and standard deviations for all outcomes during pre-implementation (January 2020\u0026ndash;April 2021, n\u0026thinsp;=\u0026thinsp;16 months) and post-implementation (May 2021\u0026ndash;December 2024, n\u0026thinsp;=\u0026thinsp;44 months) periods. Percentage change was calculated as ((post-mean \u0026ndash; pre-mean) / pre-mean) \u0026times; 100.\u003c/p\u003e \u003cp\u003e \u003cem\u003ePrimary Analysis.\u003c/em\u003e We specified the ITS model as a segmented linear regression, allowing for both immediate level changes and slope changes at the intervention point:\u003c/p\u003e \u003cp\u003e \u003cem\u003eY\u003c/em\u003eₜ \u003cem\u003e= β\u003c/em\u003e₀ \u003cem\u003e+ β\u003c/em\u003e₁\u003cem\u003e\u0026times;Time\u0026thinsp;+\u0026thinsp;β\u003c/em\u003e₂\u003cem\u003e\u0026times;Intervention\u0026thinsp;+\u0026thinsp;β\u003c/em\u003e₃\u003cem\u003e\u0026times;Time since intervention\u0026thinsp;+\u0026thinsp;ε\u003c/em\u003eₜ\u003c/p\u003e \u003cp\u003ewhere:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eYₜ = test positivity rate at month t\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eTime\u0026thinsp;=\u0026thinsp;months from study start (1\u0026ndash;60)\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eIntervention\u0026thinsp;=\u0026thinsp;binary indicator (0\u0026thinsp;=\u0026thinsp;pre-PrEP, 1\u0026thinsp;=\u0026thinsp;post-PrEP)\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eTime since intervention\u0026thinsp;=\u0026thinsp;months post-intervention (0 before May 2021)\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eβ₀ = baseline intercept\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eβ₁ = baseline trend (pre-intervention slope)\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eβ₂ = immediate level change at intervention\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eβ₃ = slope change post-intervention\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eε\u003csub\u003et\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;random error\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eWe included both immediate level change (\u003cem\u003eβ₂\u003c/em\u003e) and slope change (\u003cem\u003eβ₃\u003c/em\u003e) parameters following best practices for ITS analysis (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e), which recommends specifying the full segmented regression model a priori. This approach enables detection of effects manifesting as immediate shifts (\u003cem\u003eβ₂\u003c/em\u003e), gradual accumulation (\u003cem\u003eβ₃\u003c/em\u003e), or both, without constraining the model based on theoretical expectations. While we hypothesized gradual effect accumulation based on incremental scale-up from 0 to 1,936 users over 44 months, specifying the complete model allows the data to reveal the actual nature of intervention effects.\u003c/p\u003e \u003cp\u003eWe estimated the model using ordinary least squares with Newey-West heteroscedasticity and autocorrelation consistent (HAC) robust standard errors (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e). These account for potential autocorrelation between adjacent time points and heteroscedasticity in time series data.\u003c/p\u003e \u003cp\u003e \u003cem\u003eCounterfactual Analysis.\u003c/em\u003e Kazakhstan's PrEP programme scaled gradually from 0 to 1,936 users over 44 months rather than achieving immediate full implementation. For gradually scaled interventions, standard ITS decomposition into immediate level changes (\u003cem\u003eβ₂\u003c/em\u003e) and slope changes (\u003cem\u003eβ₃\u003c/em\u003e) may lack sensitivity (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e). Effects accumulate incrementally over time rather than occurring as discrete shifts at programme launch. With gradual scale-up, month-to-month incremental changes remain small even when cumulative programme effects are substantial. To address this limitation, we conducted counterfactual analysis, which tests cumulative effects across the entire post-intervention period without decomposing them into discrete parameters, thereby providing greater statistical power. We estimated the counterfactual trajectory - what would have occurred had the pre-intervention trend (\u003cem\u003eβ₀ + β₁\u0026times;Time\u003c/em\u003e) continued unchanged into the post-implementation period (May 2021\u0026ndash;December 2024, n\u0026thinsp;=\u0026thinsp;44 months). For each post-intervention month t, the intervention effect was calculated as:\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eEffect = Counterfactual − Observed\u003c/h3\u003e\n\u003cp\u003eThe mean intervention effect across all 44 post-intervention months was tested using a one-sample t-test against the null hypothesis of zero effect. This approach examines total cumulative divergence from projected trends rather than incremental month-to-month changes, increasing sensitivity for detecting gradually accumulating programme effects.\u003c/p\u003e \u003cp\u003e \u003cem\u003eModel Diagnostics.\u003c/em\u003e We conducted model diagnostics to verify regression assumptions, following established standards for time series analysis (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e). These included the Durbin-Watson test for autocorrelation, visual inspection of residual plots for heteroscedasticity, Q-Q plots for normality assessment, and autocorrelation/partial autocorrelation functions (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cem\u003eSensitivity Analyses.\u003c/em\u003e We assessed robustness through alternative model specifications and outcome measures. First, a quadratic time trend model included a Time\u0026sup2; term to test for non-linear temporal patterns. Second, a lag-adjusted model excluded the first three months post-intervention to account for potential delayed effects. Third, we included log(testing volume) as a covariate to assess whether positivity changes were independent of testing volume, controlling for potential selection bias. We compared models using Akaike Information Criterion (AIC) and \u003cem\u003eR\u0026sup2;\u003c/em\u003e values. To ensure findings were not artefacts of outcome selection, we repeated the ITS analysis using alternative outcomes. These included diagnosis rate per 1,000 HIV-negative GBMSM, which reflects absolute incidence less influenced by testing volume changes, and late diagnosis rate, which indicates disease severity at diagnosis.\u003c/p\u003e \u003cp\u003e \u003cem\u003ePower Analysis.\u003c/em\u003e We conducted post-hoc power analysis to evaluate the precision of our ITS estimates. Effect size was calculated as Cohen's \u003cem\u003ef\u0026sup2; = R\u0026sup2; / (1\u0026ndash;R\u0026sup2;)\u003c/em\u003e, classified per standard thresholds: small (0.02), medium (0.15), and large (0.35). Achieved power was calculated using standard formulas for multiple regression with α\u0026thinsp;=\u0026thinsp;0.05.\u003c/p\u003e \u003cp\u003e \u003cem\u003eInfections Prevented Estimation.\u003c/em\u003e We estimated infections prevented using current programme coverage, IBBS-derived incidence, and meta-analytic effectiveness estimates. The national PrEP programme defines substantial risk using behavioural screening criteria: multiple sexual partners, inconsistent condom use during anal intercourse, or recent sexually transmitted infection within six months. Based on 2023 IBBS data, 15% of HIV-negative GBMSM self-reported these risk indicators (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). We applied this 15% proportion to the December 2024 HIV-negative GBMSM population of 64,202 individuals (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e), yielding an estimate of 9,630 GBMSM at substantial risk. Current coverage was calculated as (current PrEP users / at-risk population) \u0026times; 100.\u003c/p\u003e \u003cp\u003eWe used the IBBS-derived incidence estimate of 8.8% annually from the 2023 study among GBMSM with behavioural risk factors (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e), representing the most recent national estimate that accounts for concentrated epidemic dynamics through targeted sampling. For comparison, surveillance-based incidence calculated from monthly HIV diagnoses among GBMSM yielded a substantially lower estimate of 0.55% annually. The surveillance-based estimate likely underestimates true incidence due to: (\u003cem\u003ei\u003c/em\u003e) limited testing coverage, (\u003cem\u003eii\u003c/em\u003e) window period infections, and (\u003cem\u003eiii\u003c/em\u003e) selection bias among those seeking testing. We therefore used the IBBS-derived 8.8% estimate for all infections prevented calculations.\u003c/p\u003e \u003cp\u003e \u003cem\u003ePrEP Effectiveness and Calculations.\u003c/em\u003e Primary real-world effectiveness of 86% was derived from meta-analyses of PrEP programmes accounting for real-world adherence patterns (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e). This reflects a conservative estimate midway between clinical trial efficacy of 99% with perfect adherence and typical field effectiveness of 70\u0026ndash;90%. Sensitivity analyses used 99% optimal effectiveness. Annual infections prevented were calculated assuming constant incidence and effectiveness:\u003c/p\u003e\n\u003ch3\u003eInfections prevented = current users × annual incidence × effectiveness\u003c/h3\u003e\n\u003cp\u003eWe calculated 95% confidence intervals using normal approximation to the binomial distribution. Number needed to treat (NNT) was calculated as NNT\u0026thinsp;=\u0026thinsp;total PrEP users / infections prevented.\u003c/p\u003e \u003cp\u003e \u003cem\u003eCoverage Projections.\u003c/em\u003e Projections were generated for strategic coverage targets: 30%, 50%, 70%, and 90%. For each scenario: users needed\u0026thinsp;=\u0026thinsp;coverage target \u0026times; at-risk population (9,630 GBMSM), users needed \u0026times; annual incidence (8.8%) \u0026times; effectiveness (86%), and additional users\u0026thinsp;=\u0026thinsp;users needed\u0026thinsp;\u0026minus;\u0026thinsp;current users (1,936). Five-year cumulative impact was calculated as annual infections prevented multiplied by five, assuming constant incidence and coverage throughout the projection period. This assumption may underestimate true benefit if PrEP reduces population-level incidence. All analyses were performed using R version 4.5.0 (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e) with packages: lmtest, sandwich, broom, and pwr.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eEthics Approval\u003c/h2\u003e \u003cp\u003eThis study used de-identified, aggregated surveillance data collected through Kazakhstan's national HIV monitoring system as part of routine public health activities. All data were aggregated at the programme level with no individual identifiers, exempting the study from individual informed consent requirements under Kazakhstan research regulations. The study was conducted in accordance with the principles of the Declaration of Helsinki. The Local Ethics Committee of the Kazakh Scientific Centre of Dermatology and Infectious Diseases granted ethical approval (Protocol No. 9, Approval No. 4-2024, dated 27 September 2024).\u003c/p\u003e \u003c/div\u003e"},{"header":"RESULTS","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eDescriptive Statistics and Program Scale-Up\u003c/h2\u003e \u003cp\u003eBy December 2024, 1,936 GBMSM were receiving PrEP, representing 20.1% coverage of the 9,630 GBMSM at substantial risk (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Figure S2). During the post-implementation period, HIV testing volume increased by 96.0% (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), while absolute HIV diagnoses increased by 54.4% (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The primary outcome measure, HIV test positivity rate, decreased from 3.38% (95% CI: 2.58\u0026ndash;4.17%) before implementation to 2.51% (95% CI: 2.15\u0026ndash;2.87%) after implementation, representing a 25.7% relative reduction.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDescriptive statistics comparing pre- and post-PrEP implementation periods (January 2020\u0026ndash;December 2024)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIndicator\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePre-implementation (n\u0026thinsp;=\u0026thinsp;16 months)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePost-implementation (n\u0026thinsp;=\u0026thinsp;44 months)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePrEP users (current)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.0\u0026thinsp;\u0026plusmn;\u0026thinsp;0.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e672.4\u0026thinsp;\u0026plusmn;\u0026thinsp;566.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePrEP users (new)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.0\u0026thinsp;\u0026plusmn;\u0026thinsp;0.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e85.9\u0026thinsp;\u0026plusmn;\u0026thinsp;66.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHIV tests conducted\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e657.4\u0026thinsp;\u0026plusmn;\u0026thinsp;280.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e1288.7\u0026thinsp;\u0026plusmn;\u0026thinsp;473.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNew HIV diagnoses\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e19.1\u0026thinsp;\u0026plusmn;\u0026thinsp;4.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e29.5\u0026thinsp;\u0026plusmn;\u0026thinsp;6.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTest positivity rate, %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e3.4\u0026thinsp;\u0026plusmn;\u0026thinsp;1.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e2.5\u0026thinsp;\u0026plusmn;\u0026thinsp;0.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLate HIV diagnosis, n\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e3.1\u0026thinsp;\u0026plusmn;\u0026thinsp;2.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e4.7\u0026thinsp;\u0026plusmn;\u0026thinsp;2.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLate HIV diagnosis rate, %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e15.6\u0026thinsp;\u0026plusmn;\u0026thinsp;11.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e15.8\u0026thinsp;\u0026plusmn;\u0026thinsp;7.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eValues are presented as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation. Pre-implementation period: January 2020\u0026ndash;April 2021. Post-implementation period: May 2021\u0026ndash;December 2024. PrEP, pre-exposure prophylaxis.\u003c/em\u003e \u003c/p\u003e \u003cp\u003eThe diagnosis rate per 1,000 HIV-negative GBMSM increased by 50.0% (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), rising from 0.34 (95% CI: 0.26\u0026ndash;0.42) to 0.51 (95% CI: 0.45\u0026ndash;0.57) per 1,000, reflecting the 96% increase in testing volume. The late diagnosis rate remained stable throughout the study period at approximately 16% (pre-implementation: 15.7%; post-implementation: 15.8%; p\u0026thinsp;=\u0026thinsp;0.87).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003ePrimary Interrupted Time Series Model\u003c/h2\u003e \u003cp\u003eStandard segmented regression analysis examined immediate and gradual effects of the intervention. Parameter estimates were: baseline level \u003cem\u003eβ₀\u003c/em\u003e = 3.734 (95% CI: 2.534\u0026ndash;4.935, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001); baseline trend \u003cem\u003eβ₁\u003c/em\u003e = -0.047 percentage points per month (95% CI: -0.175 to 0.081, p\u0026thinsp;=\u0026thinsp;0.50); immediate level change \u003cem\u003eβ₂\u003c/em\u003e = -0.254 (95% CI: -1.728 to 1.220, p\u0026thinsp;=\u0026thinsp;0.76); and slope change \u003cem\u003eβ₃\u003c/em\u003e = +0.037 (95% CI: -0.095 to 0.169, p\u0026thinsp;=\u0026thinsp;0.60). Neither the immediate level change nor the slope change reached statistical significance, with confidence intervals spanning both beneficial and harmful directions. These non-significant parameters are consistent with expected low statistical power (66%, below the conventional 80% threshold) for detecting gradually accumulating effects through parameter decomposition. Model fit was modest (\u003cem\u003eR\u0026sup2;\u003c/em\u003e=0.127, adjusted \u003cem\u003eR\u0026sup2;\u003c/em\u003e=0.080), though diagnostics indicated acceptable model validity (Durbin-Watson\u0026thinsp;=\u0026thinsp;1.80, p\u0026thinsp;=\u0026thinsp;0.41).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eModel Diagnostics\u003c/h2\u003e \u003cp\u003eModel diagnostics indicated acceptable validity (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The Durbin-Watson statistic (1.80, p\u0026thinsp;=\u0026thinsp;0.41) indicated no significant autocorrelation. Visual inspection of residuals revealed random scatter with no evident patterns or heteroscedasticity. The Q-Q plot showed acceptable normality with only minor tail deviations. Autocorrelation and partial autocorrelation functions showed no significant serial correlation beyond lag 1, and no influential outliers were identified.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eCounterfactual Analysis\u003c/h2\u003e \u003cp\u003eCounterfactual analysis demonstrated a significant reduction in test positivity: mean effect of 0.537 percentage points per month (95% CI: 0.393\u0026ndash;0.681, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). This represents the divergence between observed outcomes, and the projected trajectory had pre-intervention patterns continued unchanged. At the mean post-implementation testing volume (1,289 tests per month), this reduction translates to approximately seven fewer diagnoses per month, or 84 annually.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eSensitivity Analyses\u003c/h2\u003e \u003cp\u003eSensitivity analyses examining multiple model specifications confirmed the robustness of primary findings (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Models including log(testing volume) as a covariate showed improved fit (\u003cem\u003eR\u0026sup2;\u003c/em\u003e=0.663 vs 0.127; AIC\u0026thinsp;=\u0026thinsp;135.2 vs 190.3), while intervention parameters remained in the beneficial direction (reduced positivity). Quadratic time trend and lag-adjusted models (excluding the first three post-intervention months) produced similar parameter estimates to the primary linear model (\u003cem\u003eR\u0026sup2;\u003c/em\u003e=0.130, AIC\u0026thinsp;=\u0026thinsp;192.1 and \u003cem\u003eR\u0026sup2;\u003c/em\u003e=0.115, AIC\u0026thinsp;=\u0026thinsp;175.4, respectively). The primary linear model demonstrated the best balance between fit and parsimony (AIC\u0026thinsp;=\u0026thinsp;190.3), with all alternative specifications showing consistent trends towards reduced positivity.\u003c/p\u003e \u003cp\u003eAnalysis using alternative outcomes demonstrated consistency across indicators (Table S2). The diagnosis rate per 1,000 HIV-negative GBMSM increased from 0.34 (95% CI: 0.26\u0026ndash;0.42) to 0.51 (95% CI: 0.45\u0026ndash;0.57) representing a 48.8% increase (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), with non-significant ITS parameters. The late diagnosis rate remained stable throughout the study period at 15.7% (95% CI: 9.8\u0026ndash;21.6%) pre-implementation and 15.8% (95% CI: 12.3\u0026ndash;19.3%) post-implementation, with no significant change (p\u0026thinsp;=\u0026thinsp;0.87).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003ePower Analysis\u003c/h2\u003e \u003cp\u003ePost-hoc power analysis showed that the standard ITS model achieved 66% power (n\u0026thinsp;=\u0026thinsp;60 observations, k\u0026thinsp;=\u0026thinsp;3 predictors), below the conventional 80% threshold. Cohen's f\u0026sup2; was 0.145, indicating a small effect size (Table S3). For the counterfactual analysis, we calculated Cohen's d as the mean monthly effect (0.537 percentage points) divided by the standard deviation of monthly effects (SD\u0026thinsp;=\u0026thinsp;0.98), yielding d\u0026thinsp;=\u0026thinsp;0.55, which represents a medium effect size. The one-sample t-test achieved 95% power (α\u0026thinsp;=\u0026thinsp;0.05, n\u0026thinsp;=\u0026thinsp;44 post-intervention months, two-tailed test).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eCurrent Program Impact and Projections\u003c/h2\u003e \u003cp\u003eAt current coverage (20.1%), the programme prevents an estimated 146 infections annually (95% CI: 123\u0026ndash;168) based on the IBBS-derived incidence rate of 8.8% and real-world effectiveness of 86%, yielding a number needed to treat of 13. With optimal effectiveness (99%), these figures rise to 168 infections prevented annually (95% CI: 148\u0026ndash;188) and NNT of 12.\u003c/p\u003e \u003cp\u003eProjections for strategic coverage targets demonstrate a direct correlation between coverage and infections prevented (see Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Scaling up to 70% coverage would prevent 510 infections per year (95% CI: 467\u0026ndash;552) and would require an additional 4,805 users. At the maximum feasible coverage level of 90%, 655 infections would be prevented each year (95% CI: 607\u0026ndash;704). Assuming a constant incidence and coverage rate, the five-year cumulative impact would range from 1,090 infections prevented at a 30% coverage rate to 3,275 at a 90% coverage rate.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eProjected annual infections prevented under varying PrEP coverage scenarios (86% real-world effectiveness)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCoverage (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUsers Needed\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAdditional Users\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eInfections Prevented\u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e95% CI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e5-Year Cumulative\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCurrent (20.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1,936\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e146\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(123\u0026ndash;168)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e730\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2,889\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e953\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e218\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(190\u0026ndash;246)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1,090\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4,815\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2,879\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e364\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(328\u0026ndash;400)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1,820\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6,741\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4,805\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e510\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(467\u0026ndash;552)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2,550\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e8,667\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6,731\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e655\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(607\u0026ndash;704)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3,275\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003e\u003cem\u003eNote\u003c/em\u003e: All estimates assume 86% real-world effectiveness and 8.8% annual incidence among GBMSM at substantial risk (n\u0026thinsp;=\u0026thinsp;9,630). Five-year cumulative projections assume constant incidence and coverage throughout the period. \u003csup\u003ea\u003c/sup\u003eAdditional users required beyond current programme (December 2024). \u003csup\u003eb\u003c/sup\u003eAnnual infections prevented\u0026thinsp;=\u0026thinsp;current users \u0026times; annual incidence \u0026times; effectiveness.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThis study provides the first quasi-experimental evidence for population-level PrEP impact in Central Asia, a region experiencing one of the world's fastest-accelerating HIV epidemics among GBMSM. Using interrupted time series analysis of 44 months of surveillance data from Kazakhstan's national programme, we demonstrate a significant association between programme scale-up to 20.1% coverage and HIV test positivity reduction (0.537 percentage points monthly, 95% CI: 0.393\u0026ndash;0.681, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). At current scale, the programme prevents approximately 146 infections annually with a number needed to treat of 13. This indicates high efficiency in a concentrated epidemic setting where GBMSM face HIV incidence 30 times higher than the general population.\u003c/p\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eEvidence for Programme Effectiveness\u003c/h2\u003e \u003cp\u003eThree independent analytical approaches converge on consistent reductions in HIV test positivity, strengthening confidence in programme effectiveness. Counterfactual analysis demonstrated highly significant effects (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) that remained robust across all sensitivity specifications, including models controlling for testing volume. The consistency between empirical estimates (84 diagnoses averted annually among tested individuals) and theoretical calculations (146 total infections prevented across all at-risk GBMSM) supports construct validity. This difference reflects measurement scope: empirical estimates capture only partial surveillance coverage, whilst theoretical calculations account for the complete at-risk population. Simple pre-post comparison showed a 25.7% relative reduction (from 3.38% to 2.51%). While quasi-experimental designs cannot establish definitive causation, the convergence across independent analytical approaches, combined with the biological plausibility of PrEP's preventive mechanism, provides strong evidence for population-level programme effectiveness.\u003c/p\u003e \u003cp\u003eThis 25.7% reduction closely aligns with population-level effects in high-income settings. Ecological analysis across all US states showed a 38% reduction in HIV diagnoses where PrEP coverage was highest (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e), while Scotland's national program demonstrated a 19.7% reduction at 19.5% coverage (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e). Kazakhstan achieved a similarly effective result with almost identical coverage (20.1%), despite operating in a middle-income setting with legislative restrictions on key populations and healthcare-related stigma. This consistency across different health systems and geographic contexts reinforces the reproducible population-level impact of PrEP when it is targeted appropriately.\u003c/p\u003e \u003cp\u003eThe robustness of intervention effects when controlling for testing volume indicates genuine prevention impact rather than testing artefacts. Including log(testing volume) as a covariate substantially improved model fit (\u003cem\u003eR\u0026sup2;\u003c/em\u003e increased from 0.127 to 0.663), yet intervention parameters maintained their beneficial direction. This \u003cem\u003eR\u0026sup2;\u003c/em\u003e improvement demonstrates that testing volume explains substantial variance, but crucially, PrEP effects persist independently. The observed epidemiological pattern is characteristic of successful PrEP implementation combined with expanded testing infrastructure, as documented in other concentrated epidemic settings (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e). This pattern includes increased case detection (+\u0026thinsp;48.8% in diagnosis rate), decreased test positivity (\u0026minus;\u0026thinsp;25.7%), and stable late diagnosis rates (~\u0026thinsp;16%). This combination suggests the programme both identifies previously undiagnosed infections through expanded testing and prevents new infections through PrEP, with no detectable change in diagnosis timing.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003ePublic Health Significance\u003c/h2\u003e \u003cp\u003eThe programme demonstrates substantial public health impact. The standard segmented regression model showed modest statistical effect size (Cohen's f\u0026sup2;=0.145) and low power (66%), consistent with the gradually accumulating nature of programme effects. However, the counterfactual analysis, which achieved 95% power by testing cumulative effects, demonstrated a medium effect (Cohen's d\u0026thinsp;=\u0026thinsp;0.55) that translated to clinically and epidemiologically significant reductions in HIV test positivity: 0.537 percentage points monthly (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), representing approximately 7 fewer diagnoses per month among tested individuals. In concentrated epidemics where GBMSM experience 30-fold higher HIV incidence than the general population, each PrEP user generates outsized prevention benefits. The number needed to treat of 13 indicates high intervention efficiency: approximately 13 person-years of PrEP prevent one HIV infection annually. This aligns with NNT estimates from clinical trials conducted among GBMSM at substantial risk for HIV, demonstrating that Kazakhstan's national programme replicates trial-level efficiency in real-world implementation (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e). For broader context, NNTs for many established preventive interventions in chronic disease management range from 20 to over 100 (\u003cspan additionalcitationids=\"CR42\" citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e). The programme's current impact (146 infections prevented annually) translates to substantial individual and societal benefits, given lifetime HIV treatment costs and morbidity. This efficiency arises from precise targeting of biomedical prevention to populations at highest risk, demonstrating that PrEP achieves meaningful population-level impact even at modest coverage levels in concentrated epidemic settings.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eStrategic Considerations for Programme Expansion\u003c/h2\u003e \u003cp\u003eScaling up to 70% coverage would increase the programme's impact three and a half times, preventing 510 infections annually compared to the current 146. Although PrEP has been approved in 38 out of 53 countries in the WHO European Region, only 344,596 individuals received it in 2024, which is below the regional target of 500,000. Uptake remains particularly constrained in EECA due to legislative restrictions on key populations, inadequate primary care integration, and urban-only service delivery (\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e). Kazakhstan's experience shows that national programmes can have a significant impact on the population in resource-constrained EECA settings, providing a model for scaling up across the region. However, achieving this expansion requires addressing three interconnected barriers documented in Kazakhstan: legislative restrictions discouraging healthcare-seeking among sexual minorities, limited integration of PrEP services with primary care, and geographic concentration in major cities (\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). Evidence from other concentrated epidemic settings demonstrates that service delivery diversification, including community-based distribution, telehealth, and pharmacy dispensing, substantially improves PrEP accessibility while reducing reliance on specialised AIDS Centres (\u003cspan additionalcitationids=\"CR46\" citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e). The forthcoming availability of twice-yearly injectable lenacapavir presents additional opportunities for adherence improvement and effectiveness gains (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e). Our findings suggest that even modest coverage expansion yields substantial prevention returns in concentrated epidemics, making targeted outreach to underserved subpopulations, particularly chemsex participants with elevated HIV risk, a high-priority strategy.\u003c/p\u003e \u003cp\u003eThe 13-percentage-point gap between trial-based (99%) and real-world (86%) PrEP effectiveness represents substantial unrealised prevention potential. At current programme scale (1,936 users), closing this adherence gap would prevent an additional 22 infections annually (168 versus 146). Across the full at-risk population (9,630 GBMSM), perfect adherence at 70% coverage would prevent 580 infections annually versus 510 with 86% effectiveness, a 14% gain achievable through strengthened adherence support rather than recruitment expansion. Evidence-based adherence interventions, including peer counselling, proactive side-effect management, and event-driven dosing protocols, offer cost-effective pathways to maximise prevention returns from existing users. The efficient NNT of 13 and substantial lifetime costs of HIV treatment suggest that both coverage expansion and adherence optimisation would likely be cost-effective, although formal economic evaluation is needed. Continued surveillance investment, particularly repeat IBBS to monitor population-level incidence trends and assess programme reach across risk subgroups, remains essential for evaluating whether these improvements translate to measurable epidemic impact.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eStudy Limitations\u003c/h2\u003e \u003cp\u003eSeveral methodological limitations require consideration. While the ITS design is robust to confounding from time-invariant characteristics such as age, socioeconomic status, or baseline geography (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e), the use of aggregate surveillance data prevents assessment of PrEP effectiveness heterogeneity across subgroups. Individual-level data would enable more granular understanding of which GBMSM subpopulations (defined by age, geography, or specific risk profiles) benefit most from programme participation, though the robust overall trend provides strong evidence of population-level effectiveness. Pre-implementation period (January 2020\u0026ndash;April 2021) coincided with pandemic-related healthcare disruptions and altered sexual behaviour patterns. While temporally unavoidable for programmes launched during COVID-19, this timing means our counterfactual projections use pandemic-affected baselines rather than stable pre-pandemic epidemic dynamics. However, this limitation likely biases our estimates conservatively: pandemic restrictions reduced both HIV transmission opportunities and testing access, potentially underestimating true programme impact as conditions normalised.\u003c/p\u003e \u003cp\u003eThe introduction of PrEP scale-up, along with a 96% surge in testing volume, has led to a more complex attribution landscape. While models controlling for log(testing volume) substantially improved fit (R\u0026sup2;: 0.127\u0026rarr;0.663) and maintained beneficial intervention effects, complete separation of PrEP's direct preventive impact from testing pattern changes remains infeasible with aggregate data. Expanded testing could reduce observed positivity through two mechanisms: recruiting lower-risk individuals (denominator effects) or identifying previously undiagnosed prevalent infections (numerator effects that would naturally decline over time). However, three findings suggest these mechanisms do not fully explain observed reductions: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) intervention effects persisted across all testing volume-adjusted models, (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) late diagnosis rates remained stable rather than increasing (as expected if testing recruited healthier individuals), and (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) the magnitude of positivity reduction (25.7%) exceeded what denominator effects alone could plausibly generate given testing expansion magnitude.\u003c/p\u003e \u003cp\u003eFinally, infections prevented estimates assume homogeneous incidence across PrEP users, constant 86% effectiveness, and no risk compensation or network effects. These simplifying assumptions likely produce conservative estimates for three reasons: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) PrEP programmes typically achieve higher coverage among highest-risk individuals, amplifying prevention impact; (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) network effects (reduced transmission from protected individuals to their partners) compound direct prevention benefits; and (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) our 86% effectiveness estimate is conservative relative to newer long-acting formulations. While agent-based transmission models could better capture transmission dynamics, our linear calculations provide reasonable lower-bound estimates suitable for programme planning. The 8.8% annual incidence estimate from 2023 IBBS, while subject to sampling uncertainty, represents the most robust available measure through purposive sampling of GBMSM at substantial risk. The 16-fold discrepancy from surveillance-based estimates (0.55%) reflects known limitations of passive surveillance, incomplete testing coverage, window period infections, and selection bias, reinforcing our reliance on IBBS-derived parameters for impact calculations.\u003c/p\u003e \u003cp\u003eThese limitations are offset by methodological strengths that enhance confidence in our conclusions. Counterfactual analysis achieved 95% statistical power, compared to 66% for standard segmented regression, making it more sensitive to gradually scaling interventions. Our findings remained consistent across multiple analytical approaches, sensitivity specifications, and alternative outcomes. Alongside biological plausibility and concordance between empirical and theoretical estimates, this provides substantial evidence of the effectiveness at a population level.\u003c/p\u003e \u003c/div\u003e"},{"header":"CONCLUSIONS","content":"\u003cp\u003eKazakhstan's national PrEP programme has had a measurable impact on the population in an area with a concentrated HIV epidemic. Counterfactual analysis of 44 months of surveillance data revealed a significant association between scaling up the programme to achieve 20.1% coverage and a reduction in HIV test positivity. This association remained robust across sensitivity specifications. At its current scale, the programme prevents around 146 infections each year scaling up to 70% coverage could prevent around 510 infections each year. This provides the first quasi-experimental evidence of PrEP's population-level effectiveness in Central Asia, where 94% of new HIV infections occur among GBMSM. From a methodological perspective, counterfactual analysis achieved 95% statistical power compared to 66% for standard segmented regression, offering a valuable framework for evaluating the gradual scaling of interventions. These findings demonstrate that biomedical prevention can have a meaningful population-level impact in concentrated epidemics, even at modest coverage, with implications for similar contexts across Eastern Europe and Central Asia.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eConflict of Interest\u003c/h2\u003e \u003cp\u003eThe authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eEthics statement\u003c/h2\u003e \u003cp\u003eThe Local Ethics Committee of the Kazakh Scientific Centre of Dermatology and Infectious Diseases granted ethical approval (Protocol No. 9, Approval No. 4-2024, dated 27 September 2024).\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eClinical Trial number\u003c/h2\u003e \u003cp\u003eNot applicable.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis article is the result of work carried out as part of a scientific project funded by the Ministry of Science and Higher Education of the Republic of Kazakhstan, IRN AR25794027 \u0026ldquo;Improving HIV prevention in MSM by examining epidemiological, socio-economic, psychological, and biobehavioral factors influencing pre-exposure prophylaxis use.\u0026rdquo;\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eN.L. and S.K. contributed equally to this work and share senior authorship. I.K. contributed to conceptualization, methodology, data collection, study design, funding acquisition, writing \u0026ndash; original draft, and writing \u0026ndash; review and editing. B.T. contributed to methodology, software, validation, funding acquisition, and writing \u0026ndash; review and editing. G.S. contributed to data curation, supervision, and writing \u0026ndash; review and editing. M.R. and G.A. contributed to validation, software, and visualization. E.A., E.G., and A.U. contributed to investigation, supervision, and project administration. S.K. contributed to conceptualization, investigation, supervision, project administration, and writing \u0026ndash; review and editing. N.L. contributed to conceptualization, methodology, formal analysis, statistical analysis, data curation, visualization, writing \u0026ndash; original draft, and writing \u0026ndash; review and editing. All authors reviewed the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe authors wish to acknowledge the healthcare professionals at State AIDS Centres across Kazakhstan who deliver PrEP services, often in challenging conditions shaped by stigma and resource constraints. Their commitment to providing non-judgmental, evidence-based care to GBMSM has been essential to the programme's success. We are equally grateful to the community-based organizations whose outreach, advocacy, and trust-building efforts have made PrEP uptake possible. In particular, we thank Public Fund \"Community Friends\" and Public Fund \"Human Health Institute\" for their sustained contributions to HIV prevention among key populations in Kazakhstan. We also acknowledge the broader network of civil society organizations and peer educators working across the country whose engagement with communities at substantial risk underpins the programme's reach. We thank Arailym Abilbay for her contribution to data and reference management. Finally, we express our gratitude to the GBMSM community members who engage with prevention services and whose participation makes this research possible.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe dataset analyzed during the current study is available from the corresponding author (N.K. [[email protected]] (mailto:[email protected]) ) and the first author (I.K. [[email protected]] (mailto:[email protected]) ) upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eUNAIDS. What the data tell us: Projections for the HIV epidemic in Eastern Europe and Central Asia in 2030. Geneva: Joint United Nations Programme on HIV/AIDS; 2025. March 2025.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRepublican Center for the Prevention and Control of AIDS MoHotRoK. Monitoring and Evaluation of the HIV Response in the Republic of Kazakhstan, 2023. National HIV Monitoring \u0026amp; Evaluation Report (Annual report). Almaty, Republic of Kazakhstan. 2024. Contract No.: № 05-0186.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUNAIDS. Kazakhstan: Progress towards 90-90-90 https://aidsinfo.unaids.org/?did=554c9b9c05b5b281750af202\u0026amp;r=world\u0026amp;t=null\u0026amp;tb=g\u0026amp;bt=gnli\u0026amp;ts=0,0\u0026amp;gid=5977091c3ba08432558d099c\u0026amp;gr=KAZ.2023[\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBilibayeva G, Ospanova D, Nurkerimova A, Kussainova F, Tukeev M, Shokybaeva M, et al. Epidemiological Analysis of HIV/AIDS in Kazakhstan During 2018\u0026ndash;2020. J Res Health Sci. 2023;23(2):e00580.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReuters. Kazakhstan lawmakers give preliminary approval to ban on LGBT 'propaganda' 2025 [updated October 29, 2025. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.reuters.com/world/asia-pacific/kazakhstan-lawmakers-give-preliminary-approval-ban-lgbt-propaganda-2025-10-29/\u003c/span\u003e\u003cspan address=\"https://www.reuters.com/world/asia-pacific/kazakhstan-lawmakers-give-preliminary-approval-ban-lgbt-propaganda-2025-10-29/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePaine EA, Lee YG, Mergenova G, Vinogradov V, Laughney CI, Davis A et al. Compounding vulnerabilities: victimization and discrimination is associated with COVID-19 disruptions to HIV-related care among gay, bisexual, and other men and transgender and nonbinary people who have sex with men in Kazakhstan. AIDS Care. 2023:1\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEurasian Coalition on Health R, Gender and Sexual Diversity,. National report on violations of the rights of gay men, other MSM and trans* people in Kazakhstan. 2022. Tallinn; 2023.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStringer KL, Mukherjee T, McCrimmon T, Terlikbayeva A, Primbetovac S, Darisheva M, et al. Attitudes towards people living with HIV and people who inject drugs: A mixed method study of stigmas within harm reduction programs in Kazakhstan. Int J Drug Policy. 2019;68:27\u0026ndash;36.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIskakova B, King EJ, Yucel RM, DeHovitz J, Nugmanova Z. I think they are infected because of their ignorance and lack of responsibility: A mixed-methods study on HIV-related stigma in the healthcare system in Kazakhstan. PLoS ONE. 2025;20(9):e0331201.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKing EJ, Kussainova AZ, Erlon-Baurjan J, Nugmanova ZS. Healthcare Providers\u0026rsquo; Perspectives and Role in Improving Patient Engagement in HIV Treatment and Care in Almaty, Kazakhstan. Int J Environ Res Public Health. 2021;18(22):12138.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLunchenkov N, Cherchenko N, Altynbekov K, Terlikbayeva A, Primbetova S, Gryazev D, et al. A way to liberate myself: A qualitative study of perceived benefits and risks of chemsex among gay, bisexual, and other men who have sex with men in Almaty, Kazakhstan. Drug Alcohol Depend. 2024;264:112464.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee YG, Zhakupova G, Vinogradov V, Paine EA, Laughney CI, Reeder K, et al. Polydrug use, sexual risk, and HIV testing among cisgender gay, bisexual, and other men and transgender and nonbinary individuals who have sex with men in Kazakhstan. AIDS Educ Prev. 2022;34(5):413\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurchu EO, Marshall L, Teljeur C, Harrington P, Hayes C, Moran P, et al. Oral pre-exposure prophylaxis (PrEP) to prevent HIV: a systematic review and meta-analysis of clinical effectiveness, safety, adherence and risk compensation in all populations. BMJ open. 2022;12(5):e048478.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun Z, Gu Q, Dai Y, Zou H, Agins B, Chen Q, et al. Increasing awareness of HIV pre-exposure prophylaxis (PrEP) and willingness to use HIV PrEP among men who have sex with men: a systematic review and meta‐analysis of global data. J Int AIDS Soc. 2022;25(3):e25883.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhurana N, Yaylali E, Farnham PG, Hicks KA, Allaire BT, Jacobson E, et al. Impact of improved HIV care and treatment on PrEP effectiveness in the United States, 2016\u0026ndash;2020. JAIDS J Acquir Immune Defic Syndr. 2018;78(4):399\u0026ndash;405.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcCormack S, Dunn DT, Desai M, Dolling DI, Gafos M, Gilson R, et al. Pre-exposure prophylaxis to prevent the acquisition of HIV-1 infection (PROUD): effectiveness results from the pilot phase of a pragmatic open-label randomised trial. Lancet. 2016;387(10013):53\u0026ndash;60.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWHO. WHO implementation tool for pre-exposure prophylaxis (PrEP) of HIV infection: Module 1: Clinical. World Health Organization; 2017.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWHO. WHO expands the recommendation on oral pre-exposure prophylaxis of HIV infection (PrEP). 2015.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNunn AS, Brinkley-Rubinstein L, Oldenburg CE, Mayer KH, Mimiaga M, Patel R, et al. Defining the HIV pre-exposure prophylaxis care continuum. Aids. 2017;31(5):731\u0026ndash;4.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKelley CF, Kahle E, Siegler A, Sanchez T, Del Rio C, Sullivan PS, et al. Applying a PrEP continuum of care for men who have sex with men in Atlanta, Georgia. Clin Infect Dis. 2015;61(10):1590\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSullivan PS, Siegler AJ. Getting pre-exposure prophylaxis (PrEP) to the people: opportunities, challenges and emerging models of PrEP implementation. Sex Health. 2018;15(6):522\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSullivan PS, Juhasz M, DuBose SN, Le G, Brisco K, Islek D, et al. Association of state-level PrEP coverage and new HIV diagnoses in the USA from 2012 to 2022: an ecological analysis of the population impact of PrEP. Lancet HIV. 2025;12(6):e440\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEstcourt C, Yeung A, Nandwani R, Goldberg D, Cullen B, Steedman N, et al. Population-level effectiveness of a national HIV preexposure prophylaxis programme in MSM. Aids. 2021;35(4):665\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrulich AE, Guy R, Amin J, Jin F, Selvey C, Holden J, et al. Population-level effectiveness of rapid, targeted, high-coverage roll-out of HIV pre-exposure prophylaxis in men who have sex with men: the EPIC-NSW prospective cohort study. lancet HIV. 2018;5(11):e629\u0026ndash;37.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKoss CA, Havlir DV, Ayieko J, Kwarisiima D, Kabami J, Chamie G, et al. HIV incidence after pre-exposure prophylaxis initiation among women and men at elevated HIV risk: a population-based study in rural Kenya and Uganda. PLoS Med. 2021;18(2):e1003492.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMinistry of Health of the Republic of Kazakhstan. On approval of the list of medicinal products and medical devices for free and/or concessional outpatient provision to certain categories of citizens of the Republic of Kazakhstan with specified diseases (conditions). Nur-Sultan (now Astana), Kazakhstan2021.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKazakh Scientific Center for Dermatology and Infectious Diseases KDI. Стратегический план РГП на ПХВ \u0026laquo;Казахский научный центр дерматологии и инфекционных заболеваний\u0026raquo; на 2025\u0026ndash;2029 годы. Almaty, Republic of Kazakhstan: Казахский научный центр дерматологии и инфекционных заболеваний; 2025. p. 48 pp.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnderson-Cook CM. Experimental and quasi-experimental designs for generalized causal inference. Taylor \u0026amp; Francis; 2005.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBernal JL, Cummins S, Gasparrini A. Interrupted time series regression for the evaluation of public health interventions: a tutorial. Int J Epidemiol. 2017;46(1):348\u0026ndash;55.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKontopantelis E, Doran T, Springate DA, Buchan I, Reeves D. Regression based quasi-experimental approach when randomisation is not an option: interrupted time series analysis. BMJ. 2015;350.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrookmeyer R. Measuring the HIV/AIDS epidemic: approaches and challenges. Epidemiol Rev. 2010;32(1):26\u0026ndash;37.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGodin A, Eaton JW, Gigu\u0026egrave;re K, Marsh K, Johnson LF, Jahn A, et al. Inferring population HIV incidence trends from surveillance data of recent HIV infection among HIV testing clients. Aids. 2021;35(14):2383\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartin EG, MacDonald RH, Smith LC, Gordon DE, Lu T, O'Connell DA. Modeling the declining positivity rates for human immunodeficiency virus testing in New York state. J Public Health Manage Pract. 2015;21(6):556\u0026ndash;63.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJones HS, Hensen B, Musemburi S, Chinyanganya L, Takaruza A, Chabata ST, et al. Interpreting declines in HIV test positivity: an analysis of routine data from Zimbabwe's national sex work programme, 2009\u0026ndash;2019. J Int AIDS Soc. 2022;25(7):e25943.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLopez Bernal J, Cummins S, Gasparrini A. The use of controls in interrupted time series studies of public health interventions. Int J Epidemiol. 2018;47(6):2082\u0026ndash;93.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNewey WK, West KD. A simple, positive semi-definite, heteroskedasticity and autocorrelationconsistent covariance matrix. 1986.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eКазахский научный центр дерматологии и инфекционных заболеваний. Страновой Отчет, 2024 год. Almaty, Republic of Kazakhstan; 2025 17.04.2025. Report No.: № 05-0567.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRStudio Team. RStudio: Integrated Development Environment for R,. 4.1.3. ed. Boston, MA: RStudio, PBC; 2021.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMolina J-M, Capitant C, Spire B, Pialoux G, Cotte L, Charreau I, et al. On-demand preexposure prophylaxis in men at high risk for HIV-1 infection. N Engl J Med. 2015;373(23):2237\u0026ndash;46.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBuchbinder SP, Glidden DV, Liu AY, McMahan V, Guanira JV, Mayer KH, et al. HIV pre-exposure prophylaxis in men who have sex with men and transgender women: a secondary analysis of a phase 3 randomised controlled efficacy trial. Lancet Infect Dis. 2014;14(6):468\u0026ndash;75.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eByrne P, Cullinan J, Smith A, Smith SM. Statins for the primary prevention of cardiovascular disease: an overview of systematic reviews. BMJ open. 2019;9(4):e023085.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng SL, Roddick AJ. Association of aspirin use for primary prevention with cardiovascular events and bleeding events: a systematic review and meta-analysis. JAMA. 2019;321(3):277\u0026ndash;87.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWright JM, Musini VM, Gill R. First-line drugs for hypertension. Cochrane Database Syst reviews. 2018(4).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEuropean Centre for Disease Prevention and Control. Pre-exposure prophylaxis for HIV prevention in Europe and Central Asia. Monitoring implementation of the Dublin Declaration on partnership to fight HIV/AIDS in Europe and Central Asia: 2024 progress report. Stockholm; 2025.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKamitani E, Higa DH, Crepaz N, Wichser M, Mullins MM, Control UCD, et al. Identifying best practices for increasing HIV pre-exposure prophylaxis (PrEP) use and persistence in the United States: a systematic review. AIDS Behav. 2024;28(7):2340\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBonett S, Li Q, Sweeney A, Gaither-Hardy D, Safa H. Telehealth models for PrEP delivery: a systematic review of acceptability, implementation, and impact on the PrEP care continuum in the United States. AIDS Behav. 2024;28(9):2875\u0026ndash;86.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHillis A, Germain J, Hope V, McVeigh J, Van Hout MC. Pre-exposure prophylaxis (PrEP) for HIV prevention among men who have sex with men (MSM): a scoping review on PrEP service delivery and programming. AIDS Behav. 2020;24(11):3056\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlackwell CW, Armstrong F, Castillo HL. Lenacapavir for HIV PrEP: Interim Phase III Clinical Data Evaluation. J Nurse Practitioners. 2025:105360.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKelley CF, Acevedo-Qui\u0026ntilde;ones M, Agwu AL, Avihingsanon A, Benson P, Blumenthal J, et al. Twice-yearly lenacapavir for HIV prevention in men and gender-diverse persons. N Engl J Med. 2025;392(13):1261\u0026ndash;76.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-infectious-diseases","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"infd","sideBox":"Learn more about [BMC Infectious Diseases](http://bmcinfectdis.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/infd","title":"BMC Infectious Diseases","twitterHandle":"#bmcinfectdis","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"HIV prevention, Pre-exposure prophylaxis, GBMSM, Interrupted time series, Kazakhstan, Central Asia","lastPublishedDoi":"10.21203/rs.3.rs-9163384/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9163384/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eEastern Europe and Central Asia faces an accelerating HIV epidemic among gay, bisexual, and other men who have sex with men (GBMSM), yet population-level effectiveness data for pre-exposure prophylaxis (PrEP) remain limited. In May 2021, Kazakhstan implemented a national PrEP programme for GBMSM.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe conducted interrupted time series (ITS) analysis of national HIV surveillance data from January 2020 through December 2024 (n\u0026thinsp;=\u0026thinsp;60 months). The primary outcome was monthly HIV test positivity rate among GBMSM at substantial risk. Standard segmented regression was used to test for immediate and slope changes. We also performed counterfactual analysis to test the cumulative effects against projected pre-intervention trends. Sensitivity analyses controlled for testing volume and examined alternative outcomes.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eBy December 2024, the programme achieved 20.1% coverage (1,936 of 9,630 GBMSM at substantial risk). Counterfactual analysis demonstrated significant cumulative reduction in monthly HIV test positivity (mean monthly divergence from projected trends: 0.537 percentage points, 95% CI: 0.393\u0026ndash;0.681, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Descriptive comparison showed a 25.7% relative reduction from 3.38% to 2.51%. Effects accumulated gradually over the implementation period rather than as an immediate step-change. Findings remained robust after controlling for a 96% increase in testing volume and across alternative outcomes. Assuming 86% real-world effectiveness (from meta-analyses), the programme currently prevents an estimated 146 HIV infections annually, with a number needed to treat of 13.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eKazakhstan's PrEP programme demonstrates significant, robust population-level impact on HIV test positivity among GBMSM (25.7% reduction, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The programme currently prevents an estimated 146 infections annually (NNT\u0026thinsp;=\u0026thinsp;13), with potential to prevent 510 infections annually at 70% coverage. This study provides quasi-experimental evidence for PrEP effectiveness in Kazakhstan, demonstrating that biomedical prevention can achieve measurable population-level impact in concentrated epidemics.\u003c/p\u003e","manuscriptTitle":"Pre-Exposure Prophylaxis Reduces HIV Test Positivity Among Men Who Have Sex with Men in Kazakhstan: An Interrupted Time Series Analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-30 19:27:05","doi":"10.21203/rs.3.rs-9163384/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-29T06:47:32+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-28T20:31:57+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-16T18:35:35+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"92268711266409621456466421552995103019","date":"2026-04-03T12:51:48+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"159435768193281126533566047201505129504","date":"2026-03-31T14:36:51+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"215537113242223083653399274581471016037","date":"2026-03-31T04:05:14+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-26T09:43:17+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-25T05:59:42+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-24T11:15:17+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-24T11:14:37+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Infectious Diseases","date":"2026-03-18T22:59:02+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-infectious-diseases","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"infd","sideBox":"Learn more about [BMC Infectious Diseases](http://bmcinfectdis.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/infd","title":"BMC Infectious Diseases","twitterHandle":"#bmcinfectdis","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d2be4c58-c589-4525-b7c2-771639411d53","owner":[],"postedDate":"March 30th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-04-29T06:54:34+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-30 19:27:05","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9163384","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9163384","identity":"rs-9163384","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}