Factors associated with late HIV diagnosis among adults living with HIV in Oman: a 32-year ambidirectional cohort study at a tertiary referral hospital Authors: Zainab M. Al-Zadjali, Faryal Khamis, Sanjay Jaju, Sulaiman Dawood Al Sabei, Ruhina Aimaq, Jalila Al-Naamani, Ibtisam K. Al-Maskari, Zakariya Al-Balushi, Saif Al-Abri, Muna Ba’Omar, Laila S. Al-Saadi, Yahya M. Al-Farsi Published: 2026 ABSTRACT BackgroundDespite advances in the management of Human Immunodeficiency Virus (HIV), delayed diagnosis remains a major public health challenge. Evidence on late HIV diagnosis in Oman and the wider Gulf region remains limited. This study aimed to identify factors associated with late HIV diagnosis among adult patients of people living with HIV (PLWH) in Oman.MethodsA 32-year ambidirectional cohort study was conducted at the Royal Hospital, a tertiary referral hospital and the national HIV referral centre in Muscat, Oman, including adults aged ≥ 18 years diagnosed with HIV between 1992 and 2024. Participants with available baseline Cluster of Differentiation 4 (CD4) count, viral load, and haemoglobin measurements were included. Late diagnosis was defined as a CD4 count ≤ 350 cells/mm³ at diagnosis. Data were extracted from electronic medical records and analysed using multivariable Poisson regression with robust variance. Adjusted relative risks (aRRs) with 95% confidence intervals (CIs) were reported.ResultsAmong 549 patients, 30.1% were diagnosed late. Of those diagnosed late, 72.1% acquired HIV through sexual transmission, 64.8% were asymptomatic at presentation, and 61.8% had no comorbidities. Younger age was independently associated with higher risk of late diagnosis among individuals aged 18–27 years (aRR 5.66; 95% CI: 2.26–14.21), 28–37 years (aRR 3.25; 95% CI: 1.33–7.94), and 38–47 years (aRR 3.51; 95% CI: 1.40–8.77), compared with those aged > 47 years. Low haemoglobin (≤ 10 g/dL) was associated with increased risk (aRR 3.03; 95% CI: 1.42–6.67). Heart disease (aRR 5.38; 95% CI: 1.36–21.31) and hypertension (aRR 3.34; 95% CI: 1.13–9.91) were also significant predictors. Sex, mode of HIV transmission, and reason for HIV testing were not significantly associated with late HIV diagnosis in the adjusted analysis. WHO clinical stage 2 was also not significantly associated.ConclusionLate HIV diagnosis remains common in Oman and is associated with younger age, anaemia, and comorbidities. These findings highlight missed opportunities for earlier HIV testing, as a substantial proportion of late-diagnosed individuals were asymptomatic and many were young adults. Strengthening targeted testing strategies, reducing stigma, and improving early linkage to care are essential to support timely diagnosis and progress toward the UNAIDS 95–95–95 targets. Background Human Immunodeficiency Virus (HIV) has remained a major global health challenge since the 1980s, with no effective cure or vaccine despite medical advances [1]. The global HIV/AIDS response is progressing slowly. Immediate action is needed to prevent millions of new infections over the next decade [2]. The Joint United Nations Programme on HIV/AIDS (UNAIDS) has set a target of no more than 370,000 new infections annually by 2025, yet 1.5 million people were newly infected in 2020 [3]. In 2023, the rate of new HIV infections remained high, 1.3 million [2]. To make progress, 95% of people living with HIV (PLWH) must be identified, 95% of them should be on comprehensive antiretroviral therapy (cART), and 95% of those must achieve viral suppression [4, 5]. These 95:95:95 targets aim to contribute to the eradication of AIDS by 2030 [6]. While some Gulf Cooperation Council (GCC) countries struggle to meet these targets, Bahrain, Oman, and the UAE are on track to achieve the first 95 criteria by 2025 [7]. Late diagnosis (LD) remains a major barrier to HIV/AIDS control, leading to worse clinical outcomes and increased healthcare costs [1, 8]. Older individuals (≥ 50 years) are especially prone to non-communicable comorbidities and complex multimorbidity at diagnosis [8]. It also adversely impacts HIV epidemic control by increasing the likelihood of ongoing transmission among undiagnosed individuals; additionally, delayed engagement in care is a primary factor contributing to virological failure due to postponed initiation of antiretroviral therapy [1, 8]. It may facilitate onward transmission during the early, highly infectious phase of infection, thereby contributing to new HIV cases and hindering reductions in disease prevalence [1]. LD is defined as presenting for HIV care with a CD4 count ≤ 350 cells/mm³ [1, 9]. In Europe, the prevalence of LD ranges from 15% to 66% [10], while in Asia, it is between 72% and 83% [11], and in Africa, it varies from 35% to 89% [12]. Specifically, in Iran, the prevalence of LD is 58.2% [13]. In Saudi Arabia, 50% of patients present with AIDS and 20% with LD [14, 15], whereas in Yemen, 83% of patients are classified as late with diagnosis, with 20% in advanced stages (stage 3) of the disease [16]. Oman has made substantial progress in HIV control and, in 2022, became the first country in the Eastern Mediterranean Region validated by WHO for eliminating mother-to-child transmission of HIV and syphilis [17]. This reflects strong national investment in maternal health, surveillance, and equitable care [17]. Continued efforts are essential to prevent new infections, combat stigma, and ensure inclusive healthcare for all [18]. One of the major challenges confronting the HIV program in Oman is the high rate of late HIV diagnosis. In 2024, baseline CD4 testing was performed for 64% (150 out of 235) of newly diagnosed PLWH, and among those tested, 61% had a CD4 count ≤ 350 cells/mm³, indicating late-stage diagnosis [17]. Studying the determinants of late HIV diagnosis is vital for reducing HIV-related morbidity, mortality, and transmission[19]. It supports evidence-based policymaking, aligns with UNAIDS 95–95–95 targets [20], and advances the Ministry of Health’s Vision 2050 by translating research into improved care [21]. Although factors associated with late HIV diagnosis have been reported in several settings, evidence from Oman and the wider Gulf region remains limited. In addition, long-term national cohort data examining determinants of late HIV diagnosis in this region are scarce. Therefore, this study aimed to identify factors associated with late HIV diagnosis using data from a 32-year national ambidirectional cohort of people living with HIV in Oman. Methods Study design This study was conducted using a cohort study design on HIV/AIDS adult patients seeking treatment at the tertiary Royal Hospital, a National HIV Centre (Ministry of Health, Oman). The study included retrospective data collected from 1 January 1992 to 26 August 2024, along with prospective data gathered from 27 August to 31 December 2024, encompassing an overall study period from January 1992 to December 2024 at the Royal Hospital. Study setting This study was conducted at the Royal Hospital in Muscat, Oman, which serves as the national tertiary referral center for HIV care. The hospital offers specialized management of complex HIV cases under the supervision of infectious disease specialists and maintains a national registry for clinical follow-up and treatment monitoring. As a major referral facility with substantial inpatient capacity, it provides HIV care for patients referred from across the country [22]. Characteristics of participants: Inclusion, exclusion criteria The study included all adults aged 18 and older who had confirmed HIV diagnosis by laboratory testing for CD4 cell count, viral load, and hemoglobin, and were treated with cART in Royal Hospital during the selected period (1992–2024). Patients with incomplete demographics, clinical, or laboratory data were excluded from this study. Sample size calculation This study utilized registry data from the national HIV referral centre, and all eligible adults meeting the inclusion criteria during the study period (1992–2024) were included in the analysis. Although an initial minimum sample size estimation was performed using OpenEpi version 3.01 with a 95% confidence level, 80% statistical power, and a 5% significance level, based on estimates from a previous study by [23], the final analysis included all available eligible registry records. The calculated minimum required sample size was 224 participants (112 per group), which increased to 246 after adding a 10% contingency for missing data. However, the final sample size (n = 549) exceeded the minimum requirement, reflecting a near-complete census of available national registry data and providing adequate statistical power and precision for the analyses performed [8, 24]. Sampling method A total of 549 adult patients were included from the Royal Hospital HIV registry using a convenience sampling method. As Oman’s national referral centre for HIV care, the hospital receives patients from across the country, which improves the representativeness of the sample. Although convenience sampling may limit generalisability, the national coverage of the registry supports the external validity of the findings [25]. Convenience sampling was used because it involves selecting available patients from the target population, and therefore, the results are interpreted within the accessible study population from which the sample was drawn [26]. Data source and data collection tools Data were extracted from electronic patient records using a data extraction sheet developed by the primary researcher, reviewed by two HIV clinical experts, and approved by the academic supervisor. Data were collected in accordance with ethical standards for human subjects’ research. Eligible patients’ clinical records, covering the period from 1st January 1992 to 31st December 2024, were extracted from the Royal Hospital’s Shifaa3 + electronic medical system and documented using a customized Excel-based checklist. Study variables The study was purely observational and did not involve any intervention, interviews, or direct physical contact with patients. This data included demographics, such as age at diagnosis, sex, marital status, occupation, mode of HIV transmission, clinical data such as WHO clinical staging [27], comorbidities, combination antiretroviral therapy (cART) status, first cART date, cART side effects and laboratory data such as haemoglobin levels (Hb) baseline CD4 (Cell count/mm3), viral load (copies/ml) at diagnosis and after cART. Outcome definition LD was defined as presentation for medical care with a CD4 cell count below 350 cells/mm³ [1, 9, 28]. The outcome variable was categorized as a binary variable, classified as late diagnosis or early diagnosis based on these criteria. Statistical analysis Data analysis was performed using Statistical Package for the Social Sciences (SPSS) software (Version 28). Descriptive statistics were used to summarize categorical variables using frequencies and percentages. Associations between categorical variables were examined using chi-square tests to assess relationships between groups [29]. A Poisson regression model with robust variance was applied to examine factors associated with LD, including age, sex, marital status, occupation, injection drug use, mode of transmission, co-infection, disease stage, and advanced HIV stage, while adjusting for potential confounding factors. Controlling for confounders helps to reduce bias and provides more accurate estimates of the independent association between each variable and the outcome [30]. Adjusted relative risks (aRR) were reported instead of odds ratios (OR), as ORs may overestimate associations when outcomes are common [31, 32]. Relative risk (RR) is generally preferred in cohort studies as it provides a more direct and interpretable measure of association. Although logistic regression commonly produces ORs, these may approximate RR only when the outcome is rare (incidence < 10%); otherwise, ORs can substantially overestimate effect sizes even after adjustment for confounders [33]. Therefore, a modified Poisson regression approach (Poisson regression with robust error variance) is often used to estimate this effect directly [34]. Therefore, RR is generally considered more interpretable and appropriate in such contexts[31]. All covariates were included simultaneously in the multivariable Poisson regression model using a forced-entry approach. This approach was adopted to ensure adjustment for all clinically and epidemiologically relevant variables based on prior evidence and to minimize residual confounding in the estimation of associations. Potential multicollinearity among variables was considered during model development. Statistical significance was set at p ≤ 0.05. Statistical significance was set at p ≤ 0.05. Results Socio-demographic characteristics of HIV/AIDS patients Among 549 participants, 379 (69.0%) were male, and 170 (31.0%) were female. Of the males, 278 (72.4%) were in the early diagnosis group, and 101 (61.2%) were in the late diagnosis group, while among females, 106 (27.6%) were in the early diagnosis group and 64 (38.8%) were in the late diagnosis group (p = 0.009) (Fig. 1).In terms of age, participants aged 28–37 years accounted for 38.8% of late diagnoses compared with 30.0% in the early diagnosis group, while those aged ≥ 47 years represented 5.5% of the late diagnosis group (p = 0.001) (Fig. 2). No statistically significant differences were observed between early and late diagnosis according to year of diagnosis in the quarterly analysis (p = 0.105). Fig. 1Sex distribution in early versus late HIV diagnosis among HIV/AIDS adult patients, Oman (1992–2024) Sex distribution in early versus late HIV diagnosis among HIV/AIDS adult patients, Oman (1992–2024) Age distribution in late HIV diagnosis, Oman (1992–2024) The distribution of early versus late diagnosis by age at diagnosis is presented in Fig. 2. Fig. 2Age distribution in early versus late HIV diagnosis among HIV/AIDS adult patients, Oman (1992–2024) Age distribution in early versus late HIV diagnosis among HIV/AIDS adult patients, Oman (1992–2024) Clinical characteristics of HIV/AIDS patients Table 1 presents the clinical characteristics of participants stratified by diagnosis timing. Overall, 359 (65.4%) participants acquired HIV through sexual transmission, comprising 240 (62.5%) of the early diagnosis group and 119 (72.1%) of the late diagnosis group (p = 0.019) (Fig. 3). HIV-related symptoms were the reason for testing in 304 (55.4%) participants, including 247 (64.3%) of the early diagnosis group and 57 (34.5%) of the late diagnosis group (p = 0.001) (Fig. 4). According to WHO clinical staging, 213 (38.8%) participants were classified as stage 1 (asymptomatic), comprising 106 (27.6%) of the early diagnosis group and 107 (64.8%) of the late diagnosis group (p = 0.001) (Fig. 5). In the unadjusted descriptive analysis, a greater proportion of late-diagnosed individuals had haemoglobin > 10 g/dL (116, 70.3%) than ≤ 10 g/dL (p = 0.001) (Fig. 6); Fig. 6 presents unadjusted comparisons, whereas the adjusted association is shown in Table 2. Other clinical characteristics, including tuberculosis, HBsAg, syphilis, and HCV IgG, did not differ significantly between the two groups (p > 0.05). Table 1Clinical characteristics of HIV/AIDS adult patients, stratified by early (CD4 > 350 cells/mm3) and late (CD4 ≤ 350 cells/mm3) diagnosis of CD4 counts, in Oman from 1992 to 2024CharacteristicsTotalEarly HIV DiagnosisLate HIV diagnosisP-value(n = 549)(n = 384)(n = 165)n (%)n (%)n (%) Mode of Transmission Not Disclosed131 (23.9)105 (27.3)26 (15.8)0.019*Sexual359 (65.4)240 (62.5)119 (72.1)Vertical7 (1.3)6 (1.6)1 (0.6)Parenteral52 (9.5)33 (8.6)19 (11.5) Reason for HIV Testing 0.001*Not Disclosed6 (1.1)2 (0.5)4 (2.4)Antenatal Care42 (7.7)17 (4.4)25 (15.2)Hospital Checkup18 (3.3)10 (2.6)8 (4.8)HIV Contact103 (18.8)72 (18.8)31 (18.8)Test Request15 (2.7)5 (1.3)10 (6.1)HIV Symptoms304 (55.4)247 (64.3)57 (34.5)Official Document61 (11.1)31 (8.1)30 (18.2)WHOa Clinical Staging0.001*Stage 1: Asymptomatic213 (38.8)106 (27.6)107 (64.8)Stage 2: Mild symptoms25 (4.6)21 (5.5)4 (2.4)Stage 3: Advanced Symptoms146 (26.6)110 (28.6)36 (21.8)Stage 4: Severe symptoms165 (30.1)147 (38.3)18 (10.9) Comorbidity 0.037*No Comorbidity288 (52.5)186 (48.4)102 (61.8)Anemia66 (12.0)56 (14.6)10 (6.1)Diabetes Mellitus34 (6.2)24 (6.3)10 (6.1)Hypertension22 (4.0)14 (3.6)8 (4.8)Heart Disease15 (2.7)10 (2.6)5 (3.0)Cancer51 (9.3)42 (10.9)9 (5.5)Thyroid Disease9 (1.6)5 (1.3)4 (2.4)Renal Disease20 (3.6)12 (3.1)8 (4.8)Psychiatric Disorder22 (4.0)17 (4.4)5 (3.0)Dyslipidemia8 (1.5)6 (1.6)2 (1.2)Respiratory Disease14 (2.6)12 (3.1)2 (1.2) Haemoglobin 0.001*≤ 10 g/dL271 (49.4)222 (57.8)49 (29.7)> 10 g/dL278 (50.6)162 (42.2)116 (70.3) Tuberculosis 0.120*Not done91 (16.6)56 (14.6)35 (21.2)Positive58 (10.6)44 (11.5)14 (8.5)Negative400 (72.9)284 (74.0)116 (70.3) HBsAg b 0.346Not done18 (3.3)12 (3.1)6 (3.6)Positive32 (5.8)26 (6.8)6 (3.6)Negative499 (90.9)346 (90.1)153 (92.7) Syphilis 0.285Not done50 (9.1)36 (9.4)14 (8.5)Positive36 (6.6)21 (5.5)15 (9.1)Negative463 (84.3)327 (85.2)136 (82.4) HCV IgG c 0.487Not done25 (4.6)20 (5.2)5 (3.0)Positive64 (11.7)43 (11.2)21 (12.7)Negative460 (83.8)321 (83.6)139 (84.2)aWHO: World Health Organization, b HBsAg: Hepatitis B surface antigen, c HCV IgG: Hepatitis C IgG antibody; * Significant p-value ≤ 0.05 Table 1Clinical characteristics of HIV/AIDS adult patients, stratified by early (CD4 > 350 cells/mm3) and late (CD4 ≤ 350 cells/mm3) diagnosis of CD4 counts, in Oman from 1992 to 2024CharacteristicsTotalEarly HIV DiagnosisLate HIV diagnosisP-value(n = 549)(n = 384)(n = 165)n (%)n (%)n (%) Mode of Transmission Not Disclosed131 (23.9)105 (27.3)26 (15.8)0.019*Sexual359 (65.4)240 (62.5)119 (72.1)Vertical7 (1.3)6 (1.6)1 (0.6)Parenteral52 (9.5)33 (8.6)19 (11.5) Reason for HIV Testing 0.001*Not Disclosed6 (1.1)2 (0.5)4 (2.4)Antenatal Care42 (7.7)17 (4.4)25 (15.2)Hospital Checkup18 (3.3)10 (2.6)8 (4.8)HIV Contact103 (18.8)72 (18.8)31 (18.8)Test Request15 (2.7)5 (1.3)10 (6.1)HIV Symptoms304 (55.4)247 (64.3)57 (34.5)Official Document61 (11.1)31 (8.1)30 (18.2)WHOa Clinical Staging0.001*Stage 1: Asymptomatic213 (38.8)106 (27.6)107 (64.8)Stage 2: Mild symptoms25 (4.6)21 (5.5)4 (2.4)Stage 3: Advanced Symptoms146 (26.6)110 (28.6)36 (21.8)Stage 4: Severe symptoms165 (30.1)147 (38.3)18 (10.9) Comorbidity 0.037*No Comorbidity288 (52.5)186 (48.4)102 (61.8)Anemia66 (12.0)56 (14.6)10 (6.1)Diabetes Mellitus34 (6.2)24 (6.3)10 (6.1)Hypertension22 (4.0)14 (3.6)8 (4.8)Heart Disease15 (2.7)10 (2.6)5 (3.0)Cancer51 (9.3)42 (10.9)9 (5.5)Thyroid Disease9 (1.6)5 (1.3)4 (2.4)Renal Disease20 (3.6)12 (3.1)8 (4.8)Psychiatric Disorder22 (4.0)17 (4.4)5 (3.0)Dyslipidemia8 (1.5)6 (1.6)2 (1.2)Respiratory Disease14 (2.6)12 (3.1)2 (1.2) Haemoglobin 0.001*≤ 10 g/dL271 (49.4)222 (57.8)49 (29.7)> 10 g/dL278 (50.6)162 (42.2)116 (70.3) Tuberculosis 0.120*Not done91 (16.6)56 (14.6)35 (21.2)Positive58 (10.6)44 (11.5)14 (8.5)Negative400 (72.9)284 (74.0)116 (70.3) HBsAg b 0.346Not done18 (3.3)12 (3.1)6 (3.6)Positive32 (5.8)26 (6.8)6 (3.6)Negative499 (90.9)346 (90.1)153 (92.7) Syphilis 0.285Not done50 (9.1)36 (9.4)14 (8.5)Positive36 (6.6)21 (5.5)15 (9.1)Negative463 (84.3)327 (85.2)136 (82.4) HCV IgG c 0.487Not done25 (4.6)20 (5.2)5 (3.0)Positive64 (11.7)43 (11.2)21 (12.7)Negative460 (83.8)321 (83.6)139 (84.2)aWHO: World Health Organization, b HBsAg: Hepatitis B surface antigen, c HCV IgG: Hepatitis C IgG antibody; * Significant p-value ≤ 0.05 Clinical characteristics of HIV/AIDS adult patients, stratified by early (CD4 > 350 cells/mm3) and late (CD4 ≤ 350 cells/mm3) diagnosis of CD4 counts, in Oman from 1992 to 2024 aWHO: World Health Organization, b HBsAg: Hepatitis B surface antigen, c HCV IgG: Hepatitis C IgG antibody; * Significant p-value ≤ 0.05 Table 2Multivariable poisson regression analysis of factors associated with late HIV diagnosis (CD4 ≤ 350 cells/mm³) among adult patients living with HIV in Oman, 1992–2024VariableCrude ModelAdjusted ModelRR95%CI(Lower, Upper)P-valueaRRa95%CI(Lower, Upper)P-value Sex Male1.41(0.96, 2.08)0.0691.51(0.85,2.70)0.163Female1.00**- Age at Diagnosis 18–27 years3.46(1.61, 7.47)0.0025.66(2.26, 14.21)0.001*28–37 years2.80(1.32, 5.95)0.0073.25(1.33, 7.94)0.009*38–47 years2.79(1.26, 6.18)0.0123.51(1.40, 8.77)0.007*Above 47 years1.00--1.00-- Mode of Transmission Not Disclosed1.85(0.92, 3.85)0.0811.61(0.67, 4)0.295Sexual1.09(0.60,2.04)0.7642.08(0.95,4.55)0.670Vertical2.56(0.29,25)0.3486.67(0.57,7.14)0.101Parenteral1.001.00- Reason for HIV Testing Antenatal care1.21(0.64, 2.29)0.5591.48(0.52, 4.22)0.455Hospital checkup0.90(0.39, 2.08)0.8010.95(0.29, 3.13)0.943HIV Contact0.61(0.34, 1.10)0.1030.45(0.20, 1.01)0.053Test Request1.36(0.61, 3.04)0.4522.54(0.69, 9.30)0.157HIV Symptoms0.38(0.21, 0.68)0.0010.72(0.34, 1.54)0.407Official Document1.00--1.00-- Hemoglobin ≤ 10 g/dL3.33(1.82, 5.88)0.0013.03(1.42,6.67)0.004*> 10 g/dL1.00--1.00--WHOb clinical stagingStage 1: Asymptomatic4.61(2.92, 7.28)0.0017.34(3.12, 14.29)0.001*Stage 2: Mild symptoms1.47(0.54, 3.99)0.461.73(0.44, 5.57)0.404Stage 3: Advanced symptoms2.26(1.34, 3.81)0.0012.93(1.40, 5.68)0.003*Stage 4: Severe symptoms1.00--1.00-- Comorbidity Anemia0.43(0.21, 0.88)0.0200.81(0.33, 2.02)0.661Diabetes mellitus0.83(0.35, 1.97)0.6811.54(0.60, 3.96)0.362Hypertension1.03(0.42, 2.53)0.9483.34(1.13, 9.91)0.029*Heart disease0.94(0.30, 2.91)0.1965.38(1.36, 21.31)0.016*Cancer0.50(0.23, 1.07)0.0731.27(0.51, 3.15)0.603Thyroid disease1.25(0.37, 4.18)0.7531.92(0.40, 9.17)0.410Renal disease1.13(0.47, 2.72)0.8092.78(0.91, 8.49)0.072Psychiatric disorder0.64(0.25, 1.64)0.3430.46(0.14, 1.49)0.196Dyslipidemia0.71(0.12, 4.25)0.6480.75(0.11, 4.68)0.753Respiratory disease0.40(0.08, 2.02)0.1720.25(0.04, 1.55)0.137No comorbidity1.00--1.00--aaRR: Adjusted Relative Risk; b WHO: World Health Organization.: * Significant (p-value ≤ 0.05), **Reference category Table 2Multivariable poisson regression analysis of factors associated with late HIV diagnosis (CD4 ≤ 350 cells/mm³) among adult patients living with HIV in Oman, 1992–2024VariableCrude ModelAdjusted ModelRR95%CI(Lower, Upper)P-valueaRRa95%CI(Lower, Upper)P-value Sex Male1.41(0.96, 2.08)0.0691.51(0.85,2.70)0.163Female1.00**- Age at Diagnosis 18–27 years3.46(1.61, 7.47)0.0025.66(2.26, 14.21)0.001*28–37 years2.80(1.32, 5.95)0.0073.25(1.33, 7.94)0.009*38–47 years2.79(1.26, 6.18)0.0123.51(1.40, 8.77)0.007*Above 47 years1.00--1.00-- Mode of Transmission Not Disclosed1.85(0.92, 3.85)0.0811.61(0.67, 4)0.295Sexual1.09(0.60,2.04)0.7642.08(0.95,4.55)0.670Vertical2.56(0.29,25)0.3486.67(0.57,7.14)0.101Parenteral1.001.00- Reason for HIV Testing Antenatal care1.21(0.64, 2.29)0.5591.48(0.52, 4.22)0.455Hospital checkup0.90(0.39, 2.08)0.8010.95(0.29, 3.13)0.943HIV Contact0.61(0.34, 1.10)0.1030.45(0.20, 1.01)0.053Test Request1.36(0.61, 3.04)0.4522.54(0.69, 9.30)0.157HIV Symptoms0.38(0.21, 0.68)0.0010.72(0.34, 1.54)0.407Official Document1.00--1.00-- Hemoglobin ≤ 10 g/dL3.33(1.82, 5.88)0.0013.03(1.42,6.67)0.004*> 10 g/dL1.00--1.00--WHOb clinical stagingStage 1: Asymptomatic4.61(2.92, 7.28)0.0017.34(3.12, 14.29)0.001*Stage 2: Mild symptoms1.47(0.54, 3.99)0.461.73(0.44, 5.57)0.404Stage 3: Advanced symptoms2.26(1.34, 3.81)0.0012.93(1.40, 5.68)0.003*Stage 4: Severe symptoms1.00--1.00-- Comorbidity Anemia0.43(0.21, 0.88)0.0200.81(0.33, 2.02)0.661Diabetes mellitus0.83(0.35, 1.97)0.6811.54(0.60, 3.96)0.362Hypertension1.03(0.42, 2.53)0.9483.34(1.13, 9.91)0.029*Heart disease0.94(0.30, 2.91)0.1965.38(1.36, 21.31)0.016*Cancer0.50(0.23, 1.07)0.0731.27(0.51, 3.15)0.603Thyroid disease1.25(0.37, 4.18)0.7531.92(0.40, 9.17)0.410Renal disease1.13(0.47, 2.72)0.8092.78(0.91, 8.49)0.072Psychiatric disorder0.64(0.25, 1.64)0.3430.46(0.14, 1.49)0.196Dyslipidemia0.71(0.12, 4.25)0.6480.75(0.11, 4.68)0.753Respiratory disease0.40(0.08, 2.02)0.1720.25(0.04, 1.55)0.137No comorbidity1.00--1.00--aaRR: Adjusted Relative Risk; b WHO: World Health Organization.: * Significant (p-value ≤ 0.05), **Reference category Multivariable poisson regression analysis of factors associated with late HIV diagnosis (CD4 ≤ 350 cells/mm³) among adult patients living with HIV in Oman, 1992–2024 aaRR: Adjusted Relative Risk; b WHO: World Health Organization.: * Significant (p-value ≤ 0.05), **Reference category Other clinical characteristics, including tuberculosis, HBsAg, syphilis, and HCV IgG, did not differ significantly between the two groups (p > 0.05). Fig. 3Mode of transmission and distribution in early versus late HIV diagnosis among HIV/AIDS adult patients, Oman (1992–2024) Mode of transmission and distribution in early versus late HIV diagnosis among HIV/AIDS adult patients, Oman (1992–2024) Fig. 4Reasons for testing and distribution in early versus late HIV diagnosis among HIV/AIDS adult patients in Oman (1992–2024) Reasons for testing and distribution in early versus late HIV diagnosis among HIV/AIDS adult patients in Oman (1992–2024) Fig. 5WHO clinical staging status by early versus late HIV diagnosis among HIV/AIDS adult patients, Oman (1992–2024) WHO clinical staging status by early versus late HIV diagnosis among HIV/AIDS adult patients, Oman (1992–2024) Fig. 6Haemoglobin distribution by early and late HIV diagnosis among HIV/AIDS adult patients, Oman (1992–2024) Haemoglobin distribution by early and late HIV diagnosis among HIV/AIDS adult patients, Oman (1992–2024) Multivariable poisson regression models for late HIV diagnosis and sociodemographic/clinical factors among HIV/AIDS adult patients in Oman, 1992–2024 Table 2 presents the results of a multivariable Poisson regression analysis examining factors associated with late HIV diagnosis, defined as CD4 ≤ 350 cells/mm³ [1, 9], among HIV/AIDS patients in Oman from 1992 to 2024. Age at diagnosis emerged as a significant predictor, with younger individuals more likely to be diagnosed late. Specifically, those aged 18–27 years had the highest adjusted risk (aRR: 5.66; 95% CI: 2.26–14.21; p = 0.001) compared to those over 47 years. Low haemoglobin (≤ 10 g/dL) was independently associated with late HIV diagnosis (aRR: 3.03; 95% CI: 1.42–6.67; p = 0.004). WHO clinical staging at diagnosis showed strong associations with late HIV diagnosis. Using stage 4 as the reference category, individuals in stage 1 (asymptomatic) and stage 3 (advanced symptoms) had higher risks of late diagnosis (stage 1: aRR = 7.34, p = 0.001; stage 3: aRR = 2.93, p = 0.003), compared with those in stage 4. Among comorbidities, hypertension (aRR: 3.34; p = 0.029) and heart disease (aRR: 5.38; p = 0.016) were significantly associated with late diagnosis. Due to small numbers in certain exposure categories, particularly vertical transmission and some comorbidities, effect estimates were unstable and should be interpreted with caution. The associations reported in Table 2 are adjusted estimates and may differ from the unadjusted descriptive patterns shown in Table 1 due to confounding. In the multivariable Poisson regression model, the independent predictors of late HIV diagnosis were younger age, low haemoglobin, WHO clinical stage 1 and 3, hypertension, and heart disease. Conversely, sex, mode of HIV transmission, and reason for HIV testing (including HIV-related symptoms) were not significantly associated with late HIV diagnosis after adjustment. Discussion HIV/AIDS remains a major global public health concern, and national guidelines emphasize the importance of translating evidence into practice to strengthen healthcare systems and improve patient outcomes [17, 21]. In this context, this study provides important evidence on determinants of late HIV diagnosis in Oman, based on a 32-year national cohort spanning from 1992 to 2024. The findings contribute to the limited literature from the Gulf region and offer insights into factors associated with delayed presentation to care among people living with HIV. Late diagnosis based on age The study demonstrated that younger individuals were significantly more likely to experience late HIV diagnosis compared with those aged over 47 years. Participants aged 18–27 years exhibited nearly a fourfold increased risk of late diagnosis (aRR = 5.66; 95% CI: 2.26, 14.21), followed by those aged 28–37 years (aRR = 3.25; 95% CI: 1.33, 7.94) and 38–47 years (aRR = 3.51; 95% CI: 1.40, 8.77). Although male sex was associated with late HIV diagnosis in the unadjusted analysis (Table 3), this association was not statistically significant in the adjusted model. This suggests that the crude association was likely confounded by other factors such as age distribution and clinical characteristics, including WHO stage and comorbidities, which were accounted for in the multivariable analysis. Similarly, the apparently protective association of presenting with HIV-related symptoms in the unadjusted analysis did not persist after adjustment, indicating that this crude association was largely explained by WHO clinical stage and other clinical factors. Individuals are diagnosed years after contracting the virus, often unknowingly transmitting it before treatment [35]. A previous local study noted that missing CD4 data for one-third of patients, particularly those from high-risk groups, resulted in an underestimation of late diagnoses and advanced disease rates [18]. Excluding CD4 counts over 350 cells/mm³ further compromised accuracy [19]. Additionally, the stigma surrounding HIV remains a barrier to timely testing and treatment [20]. Consistent with our findings, a Canadian study also reported that older age correlates with lower odds of late diagnosis [36]. In contrast, a previous study in Oman indicated that older individuals had a higher rate of late diagnosis. Earlier research also showed that those who were diagnosed late generally had a higher median age than early presenters [37–39]. Factors influencing this may include better access to care for younger patients and misconceptions among older individuals about their risk for HIV [11, 40, 41]. This pattern may reflect increased health awareness among younger people and the impact of comorbidities on symptom recognition in older adults. Additionally, many elderly individuals were diagnosed only after hospitalization [42]. Table 3Sociodemographic characteristics of HIV/AIDS adult patients, stratified by early (CD4 > 350 cells/mm3) and late (CD4 ≤ 350 cells/mm3) diagnosis of CD4 counts, in Oman from 1992 to 2024CharacteristicsTotalEarly HIV DiagnosisLate HIV DiagnosisP-value(n = 549)(n = 384)(n = 165)n (%)n (%)n (%)Sex0.009*Male379 (69.0)278 (72.4)101 (61.2)Female170 (31.0)106 (27.6)64 (38.8) Age (years) 0.001*18 to 27146 (26.6)90 (23.4)56 (33.9)28 to 37206 (37.5)142 (37.0)64 (38.8)38 to 47116 (21.1)80 (20.8)36 (21.8)> 4781 (14.8)72 (18.8)9 (5.5) Marital status 0.070Not Disclosed16 (2.9)12 (3.1)4 (2.4)Single194 (35.3)147 (38.3)47 (28.5)Married287 (52.3)185 (48.2)102 (61.8)Divorced31 (5.6)24 (6.3)7 (4.2)Widow Employment 21 (3.8)16 (4.2)5 (3.0)0.667Not Disclosed44 (8.0)33 (8.6)11 (6.7)Employed251 (45.7)172 (44.8)79 (47.9)Unemployed254 (46.3)179 (46.6)75 (45.5) Year of diagnosis (quarterly) 0.105In 2008 or earlier142 (25.9)104 (27.1)38 (23.0)2009 to 2014147 (26.8)111 (28.9)36 (21.8)2015 to 2019142 (25.9)94 (24.5)48 (29.1)2020 to 2024118 (21.5)75 (19.5)43 (26.1)* Significant p-value ≤ 0.05 Table 3Sociodemographic characteristics of HIV/AIDS adult patients, stratified by early (CD4 > 350 cells/mm3) and late (CD4 ≤ 350 cells/mm3) diagnosis of CD4 counts, in Oman from 1992 to 2024CharacteristicsTotalEarly HIV DiagnosisLate HIV DiagnosisP-value(n = 549)(n = 384)(n = 165)n (%)n (%)n (%)Sex0.009*Male379 (69.0)278 (72.4)101 (61.2)Female170 (31.0)106 (27.6)64 (38.8) Age (years) 0.001*18 to 27146 (26.6)90 (23.4)56 (33.9)28 to 37206 (37.5)142 (37.0)64 (38.8)38 to 47116 (21.1)80 (20.8)36 (21.8)> 4781 (14.8)72 (18.8)9 (5.5) Marital status 0.070Not Disclosed16 (2.9)12 (3.1)4 (2.4)Single194 (35.3)147 (38.3)47 (28.5)Married287 (52.3)185 (48.2)102 (61.8)Divorced31 (5.6)24 (6.3)7 (4.2)Widow Employment 21 (3.8)16 (4.2)5 (3.0)0.667Not Disclosed44 (8.0)33 (8.6)11 (6.7)Employed251 (45.7)172 (44.8)79 (47.9)Unemployed254 (46.3)179 (46.6)75 (45.5) Year of diagnosis (quarterly) 0.105In 2008 or earlier142 (25.9)104 (27.1)38 (23.0)2009 to 2014147 (26.8)111 (28.9)36 (21.8)2015 to 2019142 (25.9)94 (24.5)48 (29.1)2020 to 2024118 (21.5)75 (19.5)43 (26.1)* Significant p-value ≤ 0.05 Sociodemographic characteristics of HIV/AIDS adult patients, stratified by early (CD4 > 350 cells/mm3) and late (CD4 ≤ 350 cells/mm3) diagnosis of CD4 counts, in Oman from 1992 to 2024 Widow Employment * Significant p-value ≤ 0.05 Late diagnosis based on haemoglobin level Low haemoglobin was significantly associated with late diagnosis (aRR = 3.03, 95% CI: 1.42–6.67), indicating that anemia is more likely a marker of advanced HIV disease and delayed diagnosis than a driver of earlier healthcare engagement. This observation aligns with evidence from previous research, which indicates that low haemoglobin can increase the risk of AIDS and death in affected individuals [43]. While anemia-related symptoms such as fatigue may prompt some patients to seek care, in this cohort, low haemoglobin more likely reflected advanced HIV disease at presentation [44]. Additionally, low baseline haemoglobin levels may increase the risk of opportunistic infections among those starting cART late in HIV progression [45]. Overall, anemia is frequently linked to advanced HIV infection and higher morbidity and mortality. Late diagnosis based on WHO clinical staging Our study found that asymptomatic and those with advanced symptoms (stage 1 and stage 3) individuals were more likely to be diagnosed late (aRR = 7.34, 95% CI: 3.12, 14.29) than those with AIDS stage or stage 4 disease. This finding indicates that many individuals classified as WHO stage 1 (asymptomatic) had already experienced significant immunological decline (CD4 ≤ 350 cells/mm³), despite the absence of symptoms, consistent with the definition of late HIV diagnosis. In contrast, WHO clinical stage 2 (mild symptoms) was not independently associated with late diagnosis (aRR = 1.73, p = 0.404), which may reflect the small number of patients in this category (n = 25) and, consequently, limited statistical power. Consistent with a previous study from South African study showed that 27% of individuals who were diagnosed late were classified as very late stage, with CD4 counts of 201–350 cells/mm³ or at an advanced WHO clinical stage [46]. A consistent study in Lebanon found that most of the PLWH who were diagnosed late are in the advanced stage (stage 3) [47]. Late diagnosis is often linked to barriers, such as societal stigma, confidentiality concerns, the criminalization of certain behaviors, and time constraints faced by healthcare providers [48, 49]. Late diagnosis based on comorbidities Regarding comorbidities, the analysis revealed that hypertension and heart disease were the only conditions significantly associated with LD (CD4 ≤ 350 cells/mm³) after adjustment. Patients with hypertension were more than three times more likely to present late compared to those without comorbidities (aRR: 3.34; 95% CI: 1.13–9.91; p = 0.029), while those with heart disease had an even higher risk (aRR: 5.38; 95% CI: 1.36–21.31; p = 0.016). Other comorbidities, including cancer, diabetes mellitus, renal disease, thyroid disease, psychiatric disorders, dyslipidemia, and respiratory disease, were not significantly associated with late HIV diagnosis in the adjusted model. These findings may suggest that individuals with chronic cardiovascular conditions are either diagnosed with HIV during unrelated medical evaluations or diagnosed late due to overlapping, non-specific symptoms that delay suspicion of HIV. However, the direction of this association cannot be determined due to the observational design of the study. It is also possible that late HIV diagnosis contributes to the development or progression of cardiovascular comorbidities through immune dysfunction and prolonged untreated infection. This observation aligns with evidence from a recent cross-sectional study conducted in Italy, which demonstrated that multimorbidity was more common in those diagnosed late compared to those who were diagnosed early across all age groups [50]. Additionally, a study conducted in Spain highlighted the association between LD and cardiovascular disease, reporting that PLWH who experienced a cardiovascular event were significantly higher among those who were diagnosed late (6.0%) compared to those who were diagnosed early (3.7%) [51]. Overall, the study objective to identify factors associated with late HIV diagnosis was achieved through multivariable analysis. Limitations This study has several limitations. First, although data were obtained from the national HIV referral centre in Oman, the use of a registry-based, non-random sampling approach may introduce referral bias, as only individuals who accessed care at the Royal Hospital were included. Therefore, the findings may not fully represent all people living with HIV in the general population, particularly those who remain undiagnosed or outside specialist care. Second, some important variables known to influence late HIV diagnosis, including stigma, mental health status, treatment literacy, education level, and income, were not available in the dataset. Their absence limits a more comprehensive understanding of social and structural determinants of late diagnosis. Third, the long study period (1992–2024) may have introduced information bias due to changes in diagnostic practices, treatment guidelines, and medical record documentation over time. Fourth, the lack of qualitative data limited insight into patient experiences and contextual reasons for delayed presentation to care. Future studies using mixed-methods designs are recommended to address these gaps. Finally, some estimates, particularly for vertical transmission and certain comorbidities, were imprecise due to small sample sizes and wide confidence intervals and should therefore be interpreted with caution. In addition, the inclusion of multiple covariates with sparse categories relative to the number of late-diagnosis events may have reduced model stability, as the forced-entry model did not incorporate formal variable-selection or overfitting diagnostics; these estimates should therefore be regarded as exploratory. Conclusion This ambidirectional cohort study identifies key factors associated with late HIV diagnosis in Oman from 1992 to 2024, providing important evidence to support data-driven public health decision-making. Late HIV diagnosis was associated with younger age, WHO stage 1 and stage 3, hypertension, heart disease, and low haemoglobin levels based on multivariable analysis. The study objective to identify factors associated with late HIV diagnosis was successfully achieved through multivariable analysis. These findings highlight the need for strengthened early HIV testing strategies, targeted interventions for younger populations, stigma reduction, and integrated clinical care to improve outcomes and support achievement of Oman’s 95–95–95 targets. Although convenience sampling was used, the findings are generalizable to the source population from which the sample was drawn, which may reduce selection bias. Future research should explore behavioral and psychosocial determinants of late diagnosis among younger populations in Oman, including social, cultural, and psychological barriers to early HIV testing and care engagement. In addition, public health strategies such as youth-focused awareness platforms, community-based testing services, and routine screening for comorbidities like cardiovascular disease and anaemia may help improve early diagnosis and linkage to care. Future analyses of these long-term registry data could also employ period-stratified, mixed-effects, or survival modelling to capture how determinants of late diagnosis have evolved across the epidemic.