Eighteen percent of patients who came to a large referral hospital in Bahir Dar, Northwest Ethiopia, with symptoms suggesting viral hepatitis tested positive for hepatitis B surface antigen, according to a cross-sectional study conducted at Felege Hiwot Comprehensive Specialized Hospital between October and December 2025. The figure, 36 positive results among 204 participants, carries a 95 percent confidence interval of 12.4 to 22.9 percent, and places the tested group in what the World Health Organization would classify as a high-endemicity setting, where surface antigen prevalence exceeds 8 percent.
The study, published in the Springer Nature journal Discover Viruses, enrolled adults who presented to the hospital’s outpatient department with at least one clinical feature suggestive of viral hepatitis, such as jaundice, dark urine, pale stools, right upper quadrant pain, or hepatomegaly. All 204 eligible participants consented and completed the study, with no exclusions after enrollment. Because the investigators used consecutive sampling, every clinically suspected patient who arrived during the study window was recruited until the required sample size was met, a design the authors acknowledge may introduce selection bias and limit how far the findings generalize beyond this single referral center.
Laboratory testing relied on the SNIBE Maglumi chemiluminescence immunoassay analyzer, which captures hepatitis B surface antigen from serum using magnetic microbeads coated with specific antibodies. After unbound components are removed by magnetic separation, an enzyme-labeled detection antibody is added, and the instrument measures light emitted when the enzyme reacts with a chemiluminescent substrate. The signal is proportional to the antigen concentration in the sample. Results were classified as positive when a cutoff index reached 1.0 or above, and borderline or equivocal readings flagged by the instrument were retested, with repeatedly reactive samples counted as positive. Daily internal and external quality controls were run, reagents were checked for expiration and storage conditions, and the analyzer was calibrated regularly.
Sample size was calculated in advance using a single population proportion formula, assuming a 95 percent confidence level, a 5 percent margin of error, and an expected prevalence of 14 percent drawn from a 2024 study at Guhalla Primary Hospital in Northwest Ethiopia. The initial calculation of 185 participants was inflated by 10 percent for anticipated non-response, yielding a final target of 204. Data collectors administered a structured questionnaire covering sociodemographic characteristics, behaviors such as alcohol use and sharing of sharp instruments, and medical exposures including blood transfusion, surgery, dental procedures, tattooing, and vaccination status. Venous blood samples of 3 to 5 milliliters were drawn aseptically with single-use needles and tubes.
The participant group was evenly split by sex, with 101 men and 103 women. Nearly half, 98 people or 48 percent, were aged 25 to 34 years, while 58 participants, 28 percent, were 35 or older and 48, or 24 percent, were under 25. Clinically, 40 percent reported jaundice, 45 percent reported dark urine or pale stools, 48 percent reported fatigue or weakness, 43 percent reported right upper quadrant abdominal pain, and 32 percent reported nausea or vomiting. Behaviorally, 62 percent reported alcohol consumption within the past year, 45 percent reported sharing sharp materials such as razors, blades, or nail clippers, and 59 percent reported a previous sexually transmitted infection diagnosis or treatment.
In bivariable analysis, three factors stood out. Participants aged over 35 had roughly three times the odds of surface antigen positivity compared with those under 25, with a crude odds ratio of 3.15 and a 95 percent confidence interval of 1.14 to 8.74. Jaundice showed the strongest crude association, with an odds ratio of 5.2 and a confidence interval of 2.34 to 11.54. Sharing sharp materials doubled the odds, with a crude odds ratio of 2.2 and a confidence interval of 1.05 to 4.60.
After adjustment in a multivariable logistic regression model, four factors remained independently associated with positivity. Age of 35 years or older carried an adjusted odds ratio of 2.85, with a 95 percent confidence interval of 1.08 to 7.55 and a p value of 0.035, meaning these participants were roughly 2.9 times more likely to test positive than the youngest age group. Those aged 25 to 34 did not differ significantly from the reference group, with an adjusted odds ratio of 0.90. Alcohol consumption was associated with an adjusted odds ratio of 2.52, a confidence interval of 1.13 to 5.63, and a p value of 0.037. A self-reported history of blood transfusion was associated with an adjusted odds ratio of 1.45, a confidence interval of 1.04 to 2.94, and a p value of 0.040, while sharing sharp materials yielded an adjusted odds ratio of 1.72, a confidence interval of 1.01 to 2.93, and a p value of 0.046.
The authors are explicit that these associations should not be read as causal. The cross-sectional design cannot establish whether the measured exposures preceded infection, and with only 36 positive outcomes relative to the number of candidate predictors, the risk of model overfitting, unstable coefficients, and inflated effect estimates cannot be excluded. The transfusion finding in particular is treated cautiously: the study did not record when transfusions occurred, how many units were given, or how donor blood was screened, so the association cannot demonstrate that hepatitis B was acquired through transfusion. Alcohol consumption may likewise serve as a marker for unmeasured behavioral or socioeconomic factors, and residual confounding remains possible.
Model diagnostics were nonetheless reassuring within their limits. Variance inflation factors peaked at 2.34 and tolerance values ranged from 0.43 to 0.89, indicating no substantial multicollinearity. The Hosmer-Lemeshow goodness-of-fit test produced a chi-square statistic of 8.14 with 8 degrees of freedom and a p value of 0.42, consistent with adequate calibration. No observation showed excessive influence, with Cook’s distance values below 1.0. The final model achieved an area under the receiver operating characteristic curve of 0.78, with a 95 percent confidence interval of 0.71 to 0.85, correctly classifying 75.0 percent of participants at the selected threshold, with 69.4 percent sensitivity and 76.2 percent specificity. A sensitivity analysis treating age as a continuous variable was reported in supplementary material.
The 18 percent prevalence sits within a broader landscape of hepatitis B burden in the region. It is comparable to the 14 percent reported among clinically suspected patients at Guhalla Primary Hospital, to a 21.2 percent figure from Northern Uganda, and to a 17.6 percent estimate from Ethiopia, but lower than rates of 22.3 percent among chronic hepatitis patients and 28 percent among patients with chronic liver disease in prior studies. It is higher than community-based estimates, including 5.4 percent in a general Ethiopian sample and 9.0 percent among adults in another study, a gap the authors attribute partly to the symptomatic, hospital-recruited population studied here. Globally, an estimated 254 million people were living with chronic hepatitis B surface antigen positivity in 2022, with about 1.2 million new infections and roughly 1.1 million deaths. In Ethiopia, where surface antigen prevalence ranges from 6 to 12 percent and the infant hepatitis B vaccine was added to the routine immunization schedule in 2007, the authors call for strengthened screening, vaccination coverage, blood safety, and infection-prevention measures, and note that longitudinal studies are needed to clarify the temporal and causal relationships suggested by their cross-sectional findings.
Subject of Research: Prevalence of and risk factors for hepatitis B surface antigen positivity among clinically suspected viral hepatitis patients in Northwest Ethiopia
Article Title: Prevalence and factors associated with HBsAg positivity among clinically suspected viral hepatitis patients in Northwest Ethiopia
Article References: Abate, A., Azage, M., & Munshea, A. (2026). Prevalence and factors associated with HBsAg positivity among clinically suspected viral hepatitis patients in Northwest Ethiopia. Discover Viruses, 3(1), Article 22. https://doi.org/10.1007/s44370-026-00058-9
Image Credits: AI Generated
DOI: 10.1007/s44370-026-00058-9
Keywords: hepatitis B, HBsAg, Ethiopia, seroprevalence, cross-sectional study, chemiluminescence immunoassay, blood transfusion, alcohol consumption, risk factors, public health, viral hepatitis, Bahir Dar
News Source: Kristina Jarvis. (October 28, 2026). Hepatitis B found in nearly one in five suspected hepatitis patients in Northwest Ethiopia. Scienmag.
