Abstract
Background
Viral load is the most reliable indicator of antiretroviral therapy response and treatment failure. In Ethiopia, virologic failure remains a major challenge despite expanded ART coverage.
Objective
To identify determinants of virologic failure among adults on first-line antiretroviral therapy in health facilities of Awi Zone, Northwest Ethiopia.
Methods
An institution-based unmatched case–control study with a 1:2 ratio was conducted among 369 participants (123 cases and 246 controls). Cases were HIV-infected clients aged ≥ 15 years with virologic failure (viral load > 1000 copies/mL on two consecutive tests after adherence support) between 30 October 2020 and 30 December 2024. Controls were HIV-infected clients aged ≥ 15 years on first-line ART with viral load ≤ 1000 copies/mL on two consecutive tests after ≥ 6 months of treatment during the study period. Participants were selected using computer-generated simple random sampling from selected facilities. Bivariable and multivariable binary logistic regression analyses were performed to identify independent determinants of virologic failure.
Results
Multivariable analysis identified non-disclosure of HIV status (AOR = 2.24, 95% CI: 1.11–5.22), viral load after ART initiation > 1000 copies/mL (AOR = 5.32, 95% CI: 3.32–7.15), poor treatment adherence (AOR = 5.64, 95% CI: 2.01–7.32), treatment interruption (AOR = 6.45, 95% CI: 3.95–14.01), being single (AOR = 2.34, 95% CI: 1.54–5.32), and CD4 count < 200 cells/mm³ (AOR = 4.90, 95% CI: 2.40–8.25) as significant determinants of virologic failure.
Conclusion
Strengthening adherence support, retention in care, and disclosure counseling is essential to reduce virologic failure.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12981-026-00904-2.
Keywords: Virologic failure, Determinants, Adult HIV patients, First-line ART, Awi Zone
Introduction
The introduction of modern antiretroviral therapy has markedly improved HIV treatment outcomes by achieving viral suppression and reducing disease progression. In Ethiopia, expanded access to free ART has increased national coverage to 73% among people living with HIV [1].
Virologic failure, defined as a viral load exceeding 1000 copies per mL on two consecutive measurements after adherence support, remains a major challenge to HIV care [2]. Early identification of virologic failure is essential to enable timely regimen optimization and prevent drug resistance [3].
Despite ART scale-up, viral suppression remains suboptimal. In Ethiopia, one quarter of ART patients experienced virologic failure in 2016, and only 70.1% of urban adults achieved viral suppression in 2018 [4]. In the Amhara region, HIV continues to contribute substantially to morbidity and mortality, with over 208,000 people living with HIV and an average of 2,658 HIV-related deaths annually [5]. Although routine viral load monitoring and enhanced adherence counseling are recommended, only 66.4% of patients achieve viral suppression following adherence support [6].
Understanding these determinants is essential for improving ART outcomes, guiding clinical decisions, and achieving the UNAIDS 95-95-95 targets by 2030, particularly the third target, which aims for 95% of individuals receiving antiretroviral therapy to achieve viral suppression, as part of efforts to end AIDS as a public health threat [7]. However, evidence remains limited in the Awi Zone. The zone is predominantly rural, with the majority of the population engaged in subsistence agriculture, which may limit consistent health facility attendance due to seasonal farming activities and time constraints. In addition, the geographically dispersed settlement pattern and variable terrain may create physical access barriers to health facilities. Although ART services are available in multiple health facilities, specialized laboratory services, including routine viral load monitoring, are limited and often depend on sample referral systems, which may delay result turnaround and clinical decision-making. These combined demographic and geographic constraints may influence adherence patterns, continuity of care, and timely detection of treatment failure. Therefore, determinants of virologic failure identified in other settings may not be fully generalizable to this context, necessitating local evidence.
Methods and materials
Study area and period
Data for this study were collected retrospectively from medical records of highly active antiretroviral therapy (HAART) clients between 14 April and 08 June 2025. The study included clients who had follow-up visits between 30 October 2020 and 30 December 2024 in five public hospitals and seven health centers in the Awi Zone, Amhara Regional State, Northwest Ethiopia. Awi Zone is located approximately 118 km southwest of Bahir Dar and 448 km northwest of Addis Ababa. Administratively, it is divided into twelve woredas, with nine rural and three urban towns. The estimated population is about 1.03 million, predominantly rural, with Amharic and Awugni commonly spoken and Orthodox Christianity the main religion. Most residents rely on subsistence agriculture, with limited formal employment, low household income, and low literacy levels. Access to health services is constrained by geography, transportation, and economic barriers, which may influence adherence to long-term HIV care. ART services in the zone are provided at five public hospitals and seventeen health centers, offering free antiretroviral therapy, routine viral load monitoring, and adherence counseling within a resource-limited public health system supported by regional and federal health authorities.
Study design
An institution-based unmatched case-control study design was conducted.
Population
The source population comprised adult HIV-infected clients aged 15 years and above receiving first-line ART in public health facilities of Awi Zone. The study population included adult first-line ART clients with follow-up between 30 October 2020 and 30 December 2024.
Variables of the study
The dependent variable was virologic failure, categorized as cases and controls (see operational definitions for details). The independent variables in this study included socio-demographic, clinical, treatment-related, and behavioral factors. Socio-demographic factors comprised age, sex, residence, religion, marital status, employment, and educational status. Clinical factors included viral load after ART initiation, body mass index (BMI), cluster of differentiation 4 (CD4) count, opportunistic infections, duration on antiretroviral therapy (ART), WHO clinical stage (I-IV), disclosure status, anemia, functional status, and comorbidities. Treatment-related factors consisted of type of first-line ART regimen, ART duration, adherence to treatment, opportunistic infection (OI) prophylaxis, and treatment interruptions. Behavioral factors included alcohol consumption and khat chewing.
Operational definitions
Confirmed virologic failure: Two consecutive viral load measurements > 1000 copies/mL, following national guidelines.
Antiretroviral therapy (ART): Use of three or more antiretroviral drugs for HIV treatment [8].
Baseline: Refers to the first available clinical or laboratory measurement recorded at or shortly after initiation of antiretroviral therapy (ART) during follow-up, before the occurrence of virologic failure.
Viral load after ART initiation: The first HIV RNA viral load measurement recorded after initiation of antiretroviral therapy during follow-up, extracted from medical records and categorized for analysis.
Adherence to ART: Adherence to antiretroviral therapy was extracted from patients’ medical records, where it was routinely assessed during ART follow-up visits by clinicians using standardized programmatic tools. The assessment was based on patient self-report of missed doses over a one-month recall period documented in ART follow-up forms. Adherence was categorized as good (≥ 95% adherence or ≤ 2 missed doses in 30 days or < 3 in 60 days), fair (85–94% adherence or 3–5 missed doses in 30 days or 4–9 in 60 days), and poor (< 85% adherence or ≥ 6 missed doses in 30 days or > 9 in 60 days) [9].
Virological categories were defined as suppressed viral load (≤ 1000 copies/mL) and virologic failure (> 1000 copies/mL confirmed on two consecutive measurements ≥ 3 months apart with adherence support).
Cases: Adult first-line ART clients with virologic failure, defined as a viral load greater than 1000 copies per mL during follow-up [10].
Controls: Adult first-line ART clients with viral suppression, defined as a viral load ≤ 1000 copies per mL, including undetectable levels, sustained for more than six months [10].
Interruption after treatment: Missing ART for ≥ 28 days from last appointment.
Opportunistic infection (OI): HIV-infected individuals with conditions such as tuberculosis, herpes zoster, candidiasis, pneumonia, chronic diarrhea, or cryptococcal meningitis [11].
Comorbidities: Presence of ≥ 1 chronic medical condition other than HIV (e.g., diabetes, hypertension, TB) at ART initiation.
ART Duration: Total months on first-line ART from initiation to data collection.
Eligibility criteria
Inclusion criteria
Cases: HIV-infected clients aged ≥ 15 years with virologic failure, defined as a viral load > 1000 copies/mL on two consecutive measurements after 3–6 months of enhanced adherence support, during the period 30 October 2020 to 30 December 2024.
Controls: HIV-infected clients aged ≥ 15 years on first-line ART with viral load ≤ 1000 copies/mL on two consecutive measurements after ≥ 6 months of treatment during the study period.
Exclusion criteria: Clients with incomplete sociodemographic, clinical, or treatment-related records were excluded. A total of 4 patient charts were excluded due to incomplete data.
Sample size determination
The sample size was calculated using EPI Info version 4.6, considering different predictors based on previous studies [12, 13]. A 95% confidence level, 80% power, and a 2:1 control-to-case ratio were used. The largest sample size of 224 was obtained and then adjusted for a design effect of 1.5 to account for potential variability between health facilities, as participants were drawn from multiple hospitals and health centers rather than a single facility. Finally, a 10% allowance for incomplete records was added, resulting in a final sample size of 369 participants.
Sampling procedure
Among ART-providing facilities in the Awi Zone, five public hospitals and seven health centers were selected using simple random sampling. Medical record numbers of adult HIV patients with viral load testing between 30 October 2020 and 30 December 2024 were extracted and reviewed. Patients were classified strictly based on repeated laboratory viral load measurements into virologically suppressed and virologic failure groups. Among the total ART cohort, 2,187 were on first-line ART, and 321 had switched to second-line ART; however, this information was reported descriptively and was not used for case selection or classification. Case identification was based exclusively on laboratory-confirmed virologic failure (viral load > 1000 copies/mL on two consecutive measurements after enhanced adherence support). Using these criteria, 123 cases of virologic failure with complete records were identified. A total of 369 participants were selected using proportional allocation, including 123 cases and 246 controls, through computer-generated simple random sampling using Microsoft Excel. The hospitals included Injibara General Hospital (32/65), Dangila Primary Hospital (24/49), Jawi Primary Hospital (16/33), Agew Gimjabet Primary Hospital (14/27), and Chagni Primary Hospital (22/43). The health centers included Injibara Health Center (8/16), Dangila Health Center (7/13), Chagni Health Center (5/11), Gimjabet Health Center (4/8), Azena Health Center (3/5), Tillilli Health Center (3/5), and Addis Kidam Health Center (5/11) (cases/controls).
Data collection tools and procedures
Data were collected using KoboCollect through record review of ART clients, identifying cases and controls based on pre-determined criteria. A checklist was prepared following the routine ART registration protocol and standardized ART entry and follow-up forms. Information was extracted from patient cards, including ART registers, charts, follow-up cards, and intake forms. Data collectors, who were ART nurses and clerks, received training on chart review and confidentiality. Collected data were securely stored to ensure accuracy and confidentiality.
Data quality control
Data quality was ensured before, during, and after collection. A checklist based on the standardized ART entry and follow-up form was used to extract information from electronic and paper-based records. Data collectors received two days of training on sampling and data collection procedures. The principal investigator monitored daily data completeness and consistency and supervised data collectors, providing feedback to correct any issues, ensuring accuracy and reliability of the collected data.
Data processing and analysis
Data were exported to Stata version 17 for analysis. Descriptive statistics, including frequencies, percentages, mean, and standard deviation, were used to summarize participant characteristics. Bivariable binary logistic regression was performed to identify candidate variables for multivariable analysis using a p-value < 0.25 as the screening criterion. Multivariable logistic regression was then fitted to identify independent determinants of virologic failure. Model adequacy was assessed using the Hosmer–Lemeshow goodness-of-fit test, and multicollinearity was evaluated using the variance inflation factor (VIF < 10). Statistical significance was declared at a p-value < 0.05. An intraclass correlation coefficient (ICC) was calculated to assess clustering at the facility level. The ICC was 0.002, indicating negligible between-facility variation. According to multilevel modeling literature, very low ICC values suggest minimal clustering effects and limited impact on standard error estimation; therefore, standard logistic regression was considered appropriate for analysis [14, 15].
Results
Socio-demographic, clinical and substance related characteristics
A total of 369 participants (123 cases and 246 controls) on ART follow-up were included in the study. The mean age of participants was 35.96 years, with a standard deviation of ± 10.36. Among the study participants, 84 (68.3%) cases and 224 (91.1%) controls had disclosed their HIV status. A viral load above 1000 copies/mL after ART initiation was observed in 62 (50.4%) cases and 32 (13.0%) controls. A baseline CD4 count below 200 cells/mm³ was observed in 53 (43.1%) cases and 36 (14.6%) controls. Regarding adherence, 79 (64.2%) cases and 229 (93.1%) controls had good adherence. Opportunistic infection (OI) prophylaxis was more commonly recorded among cases 101 (82.1%) compared to controls 160 (65.0%) (Table 1).
Table 1.
Socio-demographic, clinical, and substance-related characteristics of adult first-line ART clients in Awi Zone, Northwest Ethiopia (n = 369)
| Socio-demographic characteristics | ||||
|---|---|---|---|---|
| Variable | Category | Cases n (%) | Controls n (%) | Total (%) |
| Sex | Male | 46 (37.4) | 95 (38.6) | 141 (38.2) |
| Female | 77 (62.6) | 151 (61.4) | 228 (61.8) | |
| Age at ART initiation | < 35 years | 68 (55.3) | 134 (54.5) | 202 (54.7) |
| ≥ 35 years | 55 (44.7) | 112 (45.5) | 167 (45.3) | |
| Marital status | Married | 48 (39.0) | 129 (52.4) | 177 (48) |
| Single | 28 (22.8) | 29 (11.8) | 57 (15.4) | |
| widowed | 11 (8.9) | 18 (7.3) | 29 (7.9) | |
| Divorced | 36 (29.3) | 70 (28.5) | 106 (28.7) | |
| Residence | Urban | 92 (74.8) | 188 (76.4) | 280 (75.9) |
| Rural | 31 (25.2) | 58 (23.6) | 89 (24.1) | |
| Religion | Orthodox | 114 (92.7) | 230 (93.5) | 344 (93.2) |
| Others* | 9 (7.3) | 16 (6.5) | 25 (6.8) | |
| Education | No education | 41 (33.3) | 70 (28.4) | 111 (30.1) |
| primary | 51 (41.5) | 100 (40.7) | 151 (40.9) | |
| secondary | 23 (18.7) | 56 (22.8) | 79 (21.4) | |
| Above secondary | 8 (6.5) | 20 (8.1) | 28 (7.6) | |
| Employment | Formal employee | 5 (4.0) | 13 (5.3) | 18 (4.9) |
| Informal employee | 68 (55.3) | 141 (57.3) | 209 (56.6) | |
| Unemployed | 50 (40.7) | 92 (37.4) | 142 (38.5) | |
| Occupation | Civil servant | 5 (4.1) | 13 (5.3) | 18 (4.9) |
| House wife | 35 (28.4) | 79 (32.1) | 114 (30.9) | |
| Merchant | 17 (13.8) | 32 (13.0) | 49 (13.3) | |
| Farmer | 28 (22.8) | 57 (23.2) | 85 (23.1) | |
| Daily laborer | 15 (12.2) | 32 (13.0) | 47 (12.7) | |
| Female Sex Worker | 5 (4.1) | 14 (5.7) | 19 (5.1) | |
| Student | 15 (12.2) | 13 (5.3) | 28 (7.6) | |
| Others** | 3 (2.4) | 6 (2.4) | 9 (2.4) | |
| Clinical characteristics | ||||
| Disclosure status | Disclosed | 84 (68.3) | 224 (91.1) | 308 (83.5) |
| Not disclose | 39 (31.7) | 22 (8.9) | 61 (16.5) | |
| Viral load after ART initiation (copies/mL) | > 1000 copies/mL | 62 (50.4) | 32 (13.0) | 94 (25.5) |
| ≤ 1000 copies/mL | 61 (49.6) | 214 (87.0) | 275 (74.5) | |
| Baseline WHO clinical staging | Stage I | 60 (48.8) | 110 (44.7) | 170 (46.1) |
| Stage II | 49 (39.8) | 95 (38.6) | 144 (39.0) | |
| Stage III | 10 (8.1) | 30 (12.2) | 40 (10.8) | |
| Stage IV | 4 (3.3) | 11 (4.5) | 15 (4.1) | |
| Baseline ART regimen | TDF + 3TC + EFV | 58 (47.2) | 123 (50.0) | 181 (49.1) |
| AZT + 3TC + EFV | 30 (24.4) | 55 (22.3) | 85 (23.0) | |
| TDF + 3TC + NVP | 20 (16.3) | 40 (16.3) | 60 (16.2) | |
| AZT + 3TC + NVP | 15 (12.1) | 28 (11.4) | 43 (11.7) | |
| Baseline BMI | < 18.5 kg/m2 | 51 (41.5) | 58 (23.6) | 109 (29.5) |
| 18.5–24 kg/m2 | 70 (56.9) | 182 (74.0) | 252 (68.3) | |
| > 25 kg/m2 | 2 (1.6) | 6 (2.4) | 8 (2.2) | |
| Baseline CD4 count (cells/mm³) | < 200 | 53 (43.1) | 36 (14.6) | 89 (24.1) |
| 200–349 | 40 (32.5) | 90 (36.6) | 130 (35.2) | |
| ≥ 350 | 30 (24.4) | 120 (48.8) | 150 (40.7) | |
| ART Duration (months) | 6–24 | 25 (20.3) | 50 (20.3) | 75 (20.3) |
| 25–48 | 30 (24.4) | 61 (24.8) | 91 (24.7) | |
| 49–72 | 35 (28.5) | 62 (25.2) | 97 (26.3) | |
| ≥ 73 | 33 (26.8) | 73 (29.7) | 106 (28.7) | |
| Functional status at baseline | Working | 45 (36.6) | 140 (56.9) | 185 (50.1) |
| Ambulatory | 55 (44.7) | 85 (34.6) | 140 (38.0) | |
| Bedridden | 23 (18.7) | 21 (8.5) | 44 (11.9) | |
| History of opportunistic infections | Yes | 40 (32.5) | 44 (17.9) | 84 (22.8) |
| No | 83 (67.5) | 202 (82.1) | 285 (77.2) | |
| Comorbidity | Yes | 41 (33.3) | 48 (19.5) | 89 (24.1) |
| No | 82 (66.7) | 198 (80.5) | 280 (75.9) | |
| Opportunistic infections prophylaxis | Yes | 101 (82.1) | 160 (65.0) | 261 (70.7) |
| No | 22 (17.9) | 86 (35.0) | 108 (29.3) | |
| Interruption after treatment | Yes | 43 (35.0) | 13 (5.3) | 56 (15.2) |
| No | 80 (65.0) | 233 (94.7) | 313 (84.8) | |
| Adherence to treatment | Good | 79 (64.2) | 229 (93.1) | 308 (83.5) |
| Fair | 21 (17.1) | 10 (4.1) | 31 (8.4) | |
| Poor | 23 (18.7) | 7 (2.8) | 30 (8.1) | |
| Baseline Hgb (g/dL) | ≥ 12 | 87 (70.7) | 211 (85.8) | 298 (80.8) |
| < 12 | 36 (29.3) | 35 (14.2) | 71 (19.2) | |
| Substance-related characteristics | ||||
| Alcohol use | Yes | 6 (4.9) | 5 (2.0) | 11 (3.0) |
| No | 117 (95.1) | 241 (98.0) | 358 (97.0) | |
| Khat chewing | Yes | 7 (5.7) | 4 (1.6) | 11 (3.0) |
| No | 116 (94.3) | 242 (98.4) | 358 (97.0) | |
* Muslim, protestant, catholic. ** Teacher, Driver, Tailor/Artisan
Determinants of virologic failure
In the bi-variable logistic regression analysis, variables with a p-value < 0.25 were included as candidate variables in the multivariable logistic regression model. In multivariable logistic regression analysis, six factors were independently associated with virologic failure among adult first-line ART clients in health facilities of Awi Zone. These included non-disclosure of HIV status, viral load after ART initiation > 1000 copies/mL, being single, treatment interruption, baseline CD4 count < 200 cells/mm³, and poor adherence to ART (Table 2). The model showed a good fit to the data (Hosmer–Lemeshow p-value = 0.53), and multicollinearity was not observed (VIF < 10 for all variables).
Table 2.
Bivariable and multivariable logistic regression analysis of determinants of virologic failure among adult first-line ART clients in Awi Zone, Northwest Ethiopia (n = 369)
| Variables | β-Crude | SE | COR (95% CI) | β-Adjusted | SE | AOR (95% CI) | p-value |
|---|---|---|---|---|---|---|---|
| Disclosure status | |||||||
| Disclosed | 0 | – | 1 | 0 | – | 1 | – |
| Not disclosed | 1.55 | 0.30 | 4.72 (2.64–8.44) | 0.81 | 0.40 | 2.24 (1.11–5.22)* | 0.01 |
| Viral load after ART initiation (copies/mL) | |||||||
| > 1000 copies/mL | 1.92 | 0.26 | 6.79 (4.11–11.23) | 1.67 | 0.20 | 5.32 (3.32–7.15)* | < 0.001 |
| ≤ 1000 copies/mL | 0 | – | 1 | 0 | – | 1 | – |
| Baseline BMI | |||||||
| < 18.5 kg/m² | 0.67 | 0.25 | 1.95 (1.21–3.16) | 0.47 | 0.28 | 1.60 (0.92–2.80) | 0.09 |
| 18.5–24 kg/m² | 0 | – | 1 | 0 | – | 1 | – |
| > 25 kg/m² | − 0.14 | 0.83 | 0.87 (0.17–4.41) | − 0.05 | 0.84 | 0.95 (0.18–4.90) | 0.94 |
| Marital status | |||||||
| Single | 0.95 | 0.31 | 2.59 (1.40–4.80) | 0.85 | 0.32 | 2.34 (1.54–5.32)* | 0.02 |
| Divorced | 0.32 | 0.27 | 1.38 (0.82–2.32) | 0.19 | 0.34 | 1.21 (0.56–2.11) | 0.54 |
| Widowed | 0.49 | 0.42 | 1.64 (0.72–3.72) | 0.20 | 0.47 | 1.22 (0.48–3.04) | 0.43 |
| Married | 0 | – | 1 | 0 | – | 1 | – |
| Baseline WHO clinical staging | |||||||
| Stage I | 0 | – | 1 | 0 | – | 1 | – |
| Stage II | − 0.05 | 0.25 | 0.95 (0.59–1.52) | − 0.08 | 0.24 | 0.92 (0.57–1.48) | 0.72 |
| Stage III | − 0.50 | 0.39 | 0.61 (0.28–1.32) | − 0.24 | 0.44 | 0.79 (0.33–1.88) | 0.59 |
| Stage IV | − 0.40 | 0.60 | 0.67 (0.21–2.17) | − 0.21 | 0.66 | 0.81 (0.22–2.97) | 0.75 |
| Baseline CD4 count (cells/mm³) | |||||||
| < 200 | 1.77 | 0.30 | 5.89 (3.25–10.68) | 1.59 | 0.32 | 4.90 (2.40–8.25)* | < 0.001 |
| 200–349 | 0.58 | 0.28 | 1.78 (1.03–3.09) | 0.42 | 0.27 | 1.52 (0.71–2.08) | 0.09 |
| ≥ 350 | 0 | – | 1 | 0 | – | 1 | – |
| Alcohol use | |||||||
| Yes | 0.90 | 0.64 | 2.47 (0.72–8.26) | 0.01 | 0.47 | 1.01 (0.34–2.11) | 0.23 |
| No | 0 | – | 1 | 0 | – | 1 | – |
| Khat use | |||||||
| Yes | 1.07 | 0.64 | 2.90 (0.82–10.23) | 0.19 | 0.50 | 1.21 (0.22–1.57) | 0.56 |
| No | 0 | – | 1 | 0 | – | 1 | – |
| Functional status | |||||||
| Working | 0 | – | 1 | 0 | – | 1 | – |
| Ambulatory | 0.70 | 0.24 | 2.01 (1.26–3.21) | 0.19 | 0.27 | 1.21 (0.72–2.05) | 0.46 |
| Bedridden | 1.23 | 0.34 | 3.41 (1.76–6.60) | 0.37 | 0.39 | 1.45 (0.68–3.10) | 0.33 |
| Baseline Hemoglobin (g/dL) | |||||||
| ≥12 | 0 | – | 1 | 0 | – | 1 | – |
| < 12 | 0.92 | 0.29 | 2.50 (1.45–4.32) | 0.50 | 0.34 | 1.65 (0.86–3.20) | 0.13 |
| Interruption after treatment | |||||||
| Yes | 2.26 | 0.34 | 9.63 (4.92–18.83) | 1.86 | 0.32 | 6.45 (3.95–14.01)* | 0.03 |
| No | 0 | – | 1 | 0 | – | 1 | – |
| History of opportunistic infections | |||||||
| Yes | 0.79 | 0.24 | 2.21 (1.35–3.63) | − 0.03 | 0.80 | 0.97 (0.20–4.65) | 0.24 |
| No | 0 | – | 1 | 0 | – | 1 | – |
| Adherence to treatment | |||||||
| Poor | 2.25 | 0.51 | 9.52 (3.54–25.62) | 1.73 | 0.32 | 5.64 (2.01–7.32)* | < 0.001 |
| Fair | 1.81 | 0.44 | 6.09 (2.57–14.44) | 0.22 | 0.26 | 1.24 (0.73–2.02) | 0.56 |
| Good | 0 | – | 1 | 0 | – | 1 | – |
| Comorbidity | |||||||
| Yes | 0.73 | 0.24 | 2.07 (1.28–3.34) | 0.08 | 0.77 | 1.09 (0.24–4.96) | 0.91 |
| No | 0 | – | 1 | 0 | – | 1 | – |
| Opportunistic infection prophylaxis | |||||||
| Yes | 0.90 | 0.27 | 2.46 (1.45–4.17) | 0.40 | 0.31 | 1.49 (0.81–2.73) | 0.19 |
| No | 0 | – | 1 | 0 | – | 1 | – |
*Statistically significant at p < 0.05; SE: Standard Error; COR: Crude Odds Ratio; AOR; Adjusted Odds Ratio
Discussion
This study identified disclosure status, viral load after ART initiation, treatment interruption, adherence to ART, baseline CD4 count, and marital status as significant determinants of virologic failure among adult clients on first-line ART in health facilities of Awi Zone, Northwest Ethiopia.
Clients who had not disclosed their HIV status had higher odds of virologic failure, consistent with findings from Kombolcha town [13], Debre Tabor Comprehensive Specialized Hospital [16], and West Hararghe [17]. This association may be explained through multiple interconnected pathways. Non-disclosure can limit access to social support from family or partners, reducing reminders and assistance with medication adherence and clinic attendance. It may also lead to concealment of medication-taking behavior, increasing the likelihood of missed doses, irregular intake, or delayed refills. In addition, the psychological burden of secrecy, including fear of stigma and discrimination, can contribute to stress and reduced engagement in care. Together, these factors negatively affect adherence and continuity of ART, thereby increasing the risk of virologic failure [18–20].
A viral load after ART initiation greater than 1000 copies/mL was strongly associated with virologic failure, in line with evidence from Tanzania [21]. Although one of these studies involved patients on second-line ART, where prior treatment exposure and resistance patterns may differ from first-line populations, both studies consistently highlight elevated post-treatment viral load as an early marker of subsequent treatment failure. Beyond reflecting inadequate early viral suppression, a persistently high viral load after ART initiation may indicate one or more underlying mechanisms, including poor adherence from the outset of therapy, suboptimal drug absorption or pharmacokinetics, or the presence of pre-existing (transmitted) or emerging acquired drug resistance. Early virologic non-suppression may also reflect high baseline viral burden and delayed immune recovery, which can further compromise treatment response. However, in the absence of baseline or follow-up resistance testing, the contribution of transmitted versus acquired resistance could not be determined in this study [22, 23].
Treatment interruption was another strong predictor of virologic failure, consistent with studies from Debre Tabor Comprehensive Specialized Hospital [24] and a meta-analysis across Africa [25]. This indicates that patients with a history of treatment interruption had substantially higher odds of virologic failure, highlighting its clinical importance as a key behavioral determinant. Treatment interruption leads to suboptimal and inconsistent antiretroviral drug levels, which facilitates viral replication and increases the risk of drug resistance. Programmatically, this finding underscores the need for strengthened retention strategies, early tracing of missed appointments, differentiated service delivery models, and community-based adherence support to minimize interruptions and improve long-term viral suppression [26, 27]. However, the relatively low number of participants with treatment interruption among controls suggests potential sparse data in this category, which may affect the precision of the effect estimate and warrants cautious interpretation of the magnitude of the association.
Poor adherence to ART was also significantly associated with virologic failure, consistent with studies from Dessie City [4] and Kombolcha town [15]. This association is biologically plausible, as sustained antiretroviral drug exposure is essential to maintain viral suppression. Inadequate adherence leads to subtherapeutic drug concentrations, allowing ongoing viral replication and increasing the risk of viral rebound. Over time, intermittent drug exposure can also select for resistant viral strains, further reducing treatment efficacy and limiting future therapeutic options. From a behavioral perspective, poor adherence often reflects barriers such as forgetfulness, stigma, limited social support, or treatment fatigue, all of which disrupt continuity of care and compromise long-term virologic control [27, 28].
Single clients were more likely to experience virologic failure compared with married clients, consistent with studies from Kombolcha town [15]. This association may be explained by differences in social support structures. Married individuals often benefit from spousal support, including reminders for medication intake, shared responsibility for clinic attendance, and emotional encouragement, all of which enhance adherence. In contrast, single individuals may lack a consistent treatment partner, making them more vulnerable to missed doses, reduced motivation, and stigma-related barriers to care. From a social support theory perspective, reduced interpersonal support can negatively influence sustained engagement in HIV care and ultimately contribute to virologic failure [29, 30].
Clients with baseline CD4 counts below 200 cells per mm³ had higher odds of virologic failure, similar to findings from Woldia and Dessie hospitals [31]. This represents a clinically important effect size, indicating that patients with advanced immunosuppression are at markedly increased risk of poor virological outcomes. A low CD4 count reflects delayed ART initiation and advanced disease stage, which are strongly associated with impaired immune recovery and suboptimal treatment response. Clinically, this highlights the importance of early HIV diagnosis, prompt ART initiation, and intensified viral load monitoring and follow-up for late presenters to reduce the risk of treatment failure [32, 33].
The higher proportion of OI prophylaxis among cases likely reflects confounding by indication. In the bivariable analysis, OI prophylaxis was associated with higher odds of virologic failure (COR = 2.46, 95% CI: 1.45–4.17), but this association was not statistically significant in the multivariable model after adjusting for other clinical factors. Patients receiving OI prophylaxis are typically those with more advanced immunosuppression, lower CD4 counts, or higher baseline clinical risk, which predisposes them to virologic failure. Therefore, OI prophylaxis in this context should be interpreted as a marker of disease severity rather than an independent risk factor for virologic failure.
This study highlights key determinants of virologic failure that are largely modifiable in clinical practice, particularly poor adherence and treatment interruption. Strengthening routine adherence counseling, early identification of high-risk patients, and regular viral load monitoring are essential to improve treatment outcomes. From a public health perspective, improving continuity of ART services and implementing differentiated care models can reduce virologic failure and support progress toward the UNAIDS 95-95-95 targets. Overall, reinforcing adherence support and retention in care remains critical to achieving sustained viral suppression.
Strengths of the study
This study used a well-defined case–control design with a 1:2 case–control ratio and included participants selected from multiple public hospitals and health centers in Awi Zone, enhancing representativeness of the source population. Both cases and controls were drawn from the same facilities and study period using standardized operational definitions and computer-generated simple random sampling, which reduced selection bias and improved comparability between groups. All eligible ART-providing public health facilities in the zone were included using a systematic approach, further strengthening representativeness and minimizing facility selection bias. Restriction criteria were applied by including only adult clients (≥ 15 years) on first-line antiretroviral therapy with documented viral load measurements, which reduced baseline heterogeneity and improved internal validity. In addition, confounding was minimized through multivariable logistic regression and by selecting variables a priori based on prior literature and clinical relevance.
Limitations
This study was based on retrospectively collected secondary data from medical records, which may have resulted in incomplete or missing clinical information. In addition, not all health facilities in the study area were included, particularly those in more remote settings, which may limit the generalizability of the findings and may underrepresent patients with limited access to care. Consequently, the effect of distance-related and access-related factors on treatment outcomes may have been underestimated. Furthermore, dolutegravir (DTG)-based regimens were completely absent in the study population. This is likely because the study predominantly included patients who were initiated on antiretroviral therapy prior to or during the early phase of national transition to DTG-based first-line regimens in Ethiopia. Therefore, the cohort mainly reflects patients receiving non-nucleoside reverse transcriptase inhibitor (NNRTI)-based regimens, which were still widely in use during the study period in the selected facilities. This may limit the generalizability of the findings to the current ART context where DTG-based regimens are now the standard first-line therapy.
Conclusion
Disclosure status, viral load after ART initiation, treatment interruption, poor adherence, marital status, and low CD4 count were significant determinants of virologic failure among adults on first-line ART. These findings highlight that virologic failure is largely driven by modifiable clinical and behavioral factors. From a policy perspective, strengthening viral load monitoring systems in Awi Zone is essential, including improving sample transport networks and considering expansion of near point-of-care viral load testing where feasible. Given the high impact of adherence and treatment interruption, differentiated service delivery models such as multi-month dispensing and community-based ART refill approaches should be prioritized to improve continuity of care. Strengthening patient tracking systems for missed appointments and integrating community health workers into ART follow-up could further reduce treatment interruption. Enhancing patient-centered counseling, disclosure support, and peer-led adherence groups within PLHIV associations can improve engagement in care and viral suppression outcomes. These context-specific interventions are critical for improving ART program effectiveness and achieving national and global HIV targets.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We would like to express our sincere gratitude to the ART clinic staff, data collectors, and colleagues who assisted and encouraged us throughout the study. We are grateful to the Awi Zone Health Department for providing background information and facilitating cooperation with the ART clinics.
Abbreviations
- ART
Antiretroviral Therapy
- HAART
Highly Active Antiretroviral Therapy
- TDF
Tenofovir Disoproxil Fumarate
- AZT
Zidovudine
- 3TC
Lamivudine
- EFV
Efavirenz
- NVP
Nevirapine
- PLHIV
People Living with HIV
- VL
Viral Load
- OI
Opportunistic Infection
- BMI
Body Mass Index
- CD4
Cluster of Differentiation 4
- HC
Health Center
- AOR
Adjusted Odds Ratio
- COR
Crude Odds Ratio
- Hgb
Hemoglobin
- Kg/m²
Kilogram per square meter
- g/dl
gram per deciliter
- mm³
Cubic millimeter
- mL
Milliliter
Author contributions
DME conceived and designed the study, coordinated data collection, conducted data analysis and interpretation, and drafted the manuscript. MGM contributed to study design, data collection, data analysis, and critically reviewed the manuscript. AY participated in data collection, field supervision, and manuscript review for intellectual content. HGM contributed to data analysis and interpretation of findings and critically revised the manuscript. MK contributed to study design, data interpretation, and manuscript review. All authors read and approved the final manuscript and agree to be accountable for all aspects of the work.
Funding
This research received no specific funding.
Data availability
The data are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
Ethical approval was obtained from the Institutional Research Ethics Review Committee of Injibara University (Ref. No.: IU/IRERC/125/25). Permission to access data was granted by the Amhara Public Health Institute and the respective health facilities. The study was conducted in accordance with the Declaration of Helsinki, and confidentiality was maintained through anonymous data extraction.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Barnabas G, Sibhatu MK, Berhane Y. Antiretroviral therapy program in Ethiopia benefits from virology treatment monitoring. Ethiop J health Sci. 2017;27(Suppl 1):1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Mengistu ST, Ghebremeskel GG, Ghebrat HB, Achila OO, Yohannes NA, Ghebrenegus AS, et al. Determinants of therapy failure among adults on first-line antiretroviral therapy in Asmara, Eritrea: a multicenter retrospective matched case–control study. BMC Infect Dis. 2022;22(1):834. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kanapathipillai R, McGuire M, Mogha R, Szumilin E, Heinzelmann A, Pujades-Rodríguez M. Benefit of viral load testing for confirmation of immunological failure in HIV patients treated in rural Malawi. Tropical Med Int Health. 2011;16(12):1495–500. [DOI] [PubMed] [Google Scholar]
- 4.Ethiopia Public Health Institute. Ethiopia Population based HIV Impact Assessment (EPHIA) 2017–2018: Final Report. Addis Ababa, Ethiopia: EPHI; 2020. [Google Scholar]
- 5.Worku ED, Asemahagn MA, Endalifer ML. Epidemiology of HIV infection in the Amhara region of Ethiopia, 2015 to 2018 surveillance data analysis. HIV/AIDS-Research and Palliative Care. 2020:307–14. [DOI] [PMC free article] [PubMed]
- 6.Diress G, Dagne S, Alemnew B, Adane S, Addisu A. Viral load suppression after enhanced adherence counseling and its predictors among high viral load HIV seropositive people in north wollo zone public hospitals, northeast Ethiopia, 2019: retrospective cohort study. AIDS Res Treat. 2020;2020(1):8909232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Frescura L, Godfrey-Faussett P, Feizzadeh A, El-Sadr W, Syarif O, Ghys PD, et al. Achieving the 95 95 95 targets for all: A pathway to ending AIDS. PLoS ONE. 2022;17(8):e0272405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Moreno S, Perno C, Mallon P, Behrens G, Corbeau P, Routy JP, et al. Two-drug vs. three‐drug combinations for HIV‐1: do we have enough data to make the switch? HIV Med. 2019;20:2–12. [DOI] [PubMed] [Google Scholar]
- 9.Desta AA, Kidane KM, Woldegebriel AG, Ajemu KF, Berhe AA, Zgita DN et al. Level of adherence and associated factors among HIV-infected patients on antiretroviral therapy in Northern Ethiopia: retrospective analysis. Patient Prefer Adherence. 2020:1585–94. [DOI] [PMC free article] [PubMed]
- 10.World Health Organization. What’s new in treatment monitoring: viral load and CD4 testing. Geneva: WHO. 2017 [Available from: https://www.who.int/publications/i/item/WHO-HIV-2017.22
- 11.Bogale B, Asefa A, Destaw A, Midaksa G, Asaye Z, Alemu Gebremichael M, et al. Determinants of virological failure among patients on first line highly active antiretroviral therapy (HAART) in Southwest Ethiopia: a case-control study. Front Public Health. 2022;10:916454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Siraj J, Feyissa D, Mamo Y, Zewudie A, Regesa T, Ejeta F, et al. Antiretroviral treatment failure and associated factors among HIV patients on the first-line antiretroviral therapy at Mizan-Tepi University teaching hospital, Southwest Ethiopia: a cross-sectional study. Medicine. 2021;100(51):e28357. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Meshesha HM, Nigussie ZM, Asrat A, Mulatu K. Determinants of virological failure among adults on first-line highly active antiretroviral therapy at public health facilities in Kombolcha town, Northeast, Ethiopia: a case–control study. BMJ open. 2020;10(7):e036223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Snijders TABBR. Multilevel analysis: an introduction to basic and advanced multilevel modeling. 2nd ed. London: Sage Publications Ltd; 2011. [Google Scholar]
- 15.Hox J, Moerbeek M, Van de Schoot R. Multilevel analysis: techniques and applications. 3rd ed. London: Routledge; 2017. [Google Scholar]
- 16.Mekonnen HA, Aregaw FM, Yesuf T, Derseh NM. Prevalence of virologic failure and determinants among adults on first-line antiretroviral therapy at debre tabor comprehensive specialized hospital in Northwest Ethiopia. Int J Infect Dis Ther. 2023;8:69–79. [Google Scholar]
- 17.Zerihun E, Tesema K, Abera F. Virological treatment failure and associated factors among adults on first-line antiretroviral therapy in West Hararghe, Ethiopia. Front Public Health. 2025;13:1440504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Deribe K, Woldemichael K, Wondafrash M, Haile A, Amberbir A. Disclosure experience and associated factors among HIV positive men and women clinical service users in southwest Ethiopia. BMC Public Health. 2008;8(1):81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Rintamaki LS, Davis TC, Skripkauskas S, Bennett CL, Wolf MS. Social stigma concerns and HIV medication adherence. AIDS Patient Care STDs. 2006;20(5):359–68. [DOI] [PubMed] [Google Scholar]
- 20.Earnshaw VA, Chaudoir SR. From conceptualizing to measuring HIV stigma: a review of HIV stigma mechanism measures. AIDS Behav. 2009;13(6):1160–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Mwavika ET, Kunambi PP, Masasi SJ, Lema N, Kamori D, Matee M. Prevalence, rate, and predictors of virologic failure among adult HIV-Infected clients on second-line antiretroviral therapy (ART) in Tanzania (2018–2020): a retrospective cohort study. Bull Natl Res Centre. 2024;48(1):96. [Google Scholar]
- 22.Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, Kumarasamy N, et al. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med. 2011;365(6):493–505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Mellors JW, Rinaldo CR Jr, Gupta P, White RM, Todd JA, Kingsley LA. Prognosis in HIV-1 infection predicted by the quantity of virus in plasma. Science. 1996;272(5265):1167–70. [DOI] [PubMed] [Google Scholar]
- 24.Asefa A, Asaye Z, Girma A, Hiko D. Predictors of clinical and immunological failure among patients on first-line antiretroviral therapy (ART) in Southwest Ethiopia. HIV/AIDS-Research and Palliative Care. 2019:377 – 86. [DOI] [PMC free article] [PubMed]
- 25.Abebe AD, Tassew WC, Ferede YA, Beyene JA, Assefa M. Prevalence and determinants of virologic failure among HIV patients on ART in Africa: a systematic review and meta-analysis. Virol J. 2025;22(1):395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Deeks SG, Lewin SR, Havlir DV. The end of AIDS: HIV infection as a chronic disease. lancet. 2013;382(9903):1525–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Bangsberg DR, Moss AR, Deeks SG. Paradoxes of adherence and drug resistance to HIV antiretroviral therapy. J Antimicrob Chemother. 2004;53(5):696–9. [DOI] [PubMed] [Google Scholar]
- 28.Paterson DL, Swindells S, Mohr J, Brester M, Vergis EN, Squier C, et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med. 2000;133(1):21–30. [DOI] [PubMed] [Google Scholar]
- 29.DiMatteo MR. Social support and patient adherence to medical treatment: a meta-analysis. Health Psychol. 2004;23(2):207. [DOI] [PubMed] [Google Scholar]
- 30.Bikaako-Kajura W, Luyirika E, Purcell DW, Downing J, Kaharuza F, Mermin J, et al. Disclosure of HIV status and adherence to daily drug regimens among HIV-infected children in Uganda. AIDS Behav. 2006;10(Suppl 1):85–93. [DOI] [PubMed] [Google Scholar]
- 31.Ahmed M, Merga H, Jarso H. Predictors of virological treatment failure among adult HIV patients on first-line antiretroviral therapy in Woldia and Dessie hospitals, Northeast Ethiopia: a case-control study. BMC Infect Dis. 2019;19(1):305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.INSIGHT START Study Group. Initiation of antiretroviral therapy in early asymptomatic HIV infection. N Engl J Med. 2015;373(9):795–807. 10.1056/NEJMoa1506816 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Mocroft A, Phillips AN, Gatell J, Ledergerber B, Fisher M, Clumeck N, et al. Normalisation of CD4 counts in patients with HIV-1 infection and maximum virological suppression who are taking combination antiretroviral therapy: an observational cohort study. Lancet. 2007;370(9585):407–13. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data are available from the corresponding author upon reasonable request.