Introduction

Human immunodeficiency virus (HIV) and hepatitis C virus (HCV) infections remain major public health problems worldwide [1]. According to World Health Organization (WHO) 2024 data, approximately 40.8 million people are living with HIV and 50 million have chronic HCV infection [2]. Although effective treatments and prevention strategies exist, undiagnosed infections persist and continue to drive transmission [3].

HIV and HCV share common transmission routes: both are bloodborne and can be transmitted through contact with infected blood, sexual contact, and vertically from mother to child [4]. Both infections have a prolonged clinical course and are often asymptomatic or only mildly symptomatic, which makes early diagnosis difficult. Early HIV diagnosis and initiation of antiretroviral therapy (ART) reduces mortality and prevents sexual transmission; individuals with a sustained undetectable viral load have zero risk of sexually transmitting HIV (U = U principle). Early HCV diagnosis in the era of direct-acting antivirals (DAA) enables viral eradication, typically within 8-12 weeks of treatment, with sustained virologic response rates of 95-99% [5].

The clinical consequences of late diagnosis are substantial. For HIV, late presentation is associated with acquired immune deficiency syndrome (AIDS)-defining illnesses including Pneumocystis jirovecii pneumonia, Kaposi sarcoma, AIDS-defining cancers, and increased all-cause mortality, alongside a reduced response to ART [6, 7]. For HCV, untreated chronic infection progresses in 5-30% of cases over 20-30 years to compensated and decompensated cirrhosis, hepatocellular carcinoma, and the need for liver transplantation [8]. Both infections therefore impose major individual and population-level harms when diagnosis is delayed.

In Europe, late diagnosis remains widespread. In 2024, 105,922 new HIV diagnoses were reported in the WHO European Region and 24,164 in the European Union/European Economic Area (EU/EEA), corresponding to crude rates of 11.8 per 100,000 in the wider Region and 5.3 per 100,000 in the EU/EEA. Forty-eight percent of new EU/EEA diagnoses met the criterion for late presentation (CD4 below 350 cells/mm³) [9]. An estimated 1 in 14 people living with HIV in the EU/EEA remains undiagnosed [10]. In the EU/EEA, approximately 1.8 million people are living with HCV, with a combined chronic HBV/HCV burden of 5.4 million; a substantial proportion of those infected remain undiagnosed [11].

Polish data show comparable challenges. The cumulative number of HIV diagnoses by the end of 2024 reached 35,175, with 2,257 new diagnoses in that year [12, 13]. Late presentation rates of 53-63% have been reported from Polish HIV cohorts [14]. The active HCV infection rate (HCV-RNA positive) is estimated at 0.5% of the population, corresponding to 140,000-165,000 individuals, with the majority unaware of their infection [15]. Although the present study was conducted during the active study period 2013-2018 (with retrospective chart review extending back to 2008 for the post-exposure prophylaxis cohort in Study 5), the underlying issues of underdiagnosis and late presentation remain relevant today.

Several models have been proposed for testing in acute care settings. Universal opt-out testing (an automatic offer of testing with the option to decline) has been implemented in UK emergency departments serving high-prevalence areas, yielding over 750 new HIV diagnoses by April 2024 alongside high patient acceptance (> 90%) [16, 17]. Universal opt-in testing (test offered, requires explicit consent) has been used in Spain in selected ED programmes [18]. Indicator condition-guided testing, validated by the European HIDES I and HIDES II studies, focuses on patients presenting with conditions where HIV prevalence exceeds 0.1% [19, 20]. Risk factor-based testing targets individuals with defined behavioural or epidemiological risk factors. Each model has trade-offs: universal approaches maximize coverage but require operational integration and may face acceptability challenges; targeted approaches improve yield but depend on accurate risk assessment in time-pressured environments [21, 22]. Evidence on which models work best in low-prevalence Central European countries remains limited.

The aim of this study was to evaluate HIV and HCV testing strategies in a specialist infectious diseases emergency department, with a view to reducing the number of undiagnosed or late-diagnosed infections. A cycle of five sub-studies was conducted at the Hospital for Infectious Diseases in Warsaw, addressing the following questions: 1) What proportion of patients hospitalized for mononucleosis-like syndrome are tested for HIV, and what is the HIV detection rate? 2) Can indicator condition-guided HIV testing serve as a useful diagnostic tool in the ED? 3) What is the HCV seroprevalence among unselected ED patients? 4) What is the HCV seroprevalence among ED patients with defined risk factors? 5) Which population groups should be prioritized for routine HCV testing?

Material and methods

Setting

All sub-studies were conducted at the Emergency Department and the Outpatient Prophylactic-Treatment Clinic of the Hospital for Infectious Diseases in Warsaw, the largest specialist infectious diseases referral centre in central Poland. As a specialist referral centre, this hospital serves a self-selected population in which bloodborne infections are overrepresented compared with general emergency departments. This setting is reflected in the interpretation of all results.

Sub-study designs

Figure 1 presents the timelines of the five sub-studies. Each sub-study is described below.

Study 1 (retrospective HIV chart review). Records of all patients hospitalized for mononucleosis-like syndrome between 1 May 2014 and 30 April 2015 were reviewed. Mononucleosis-like syndrome was defined as an acute febrile illness with at least one of the following: pharyngitis, lymphadenopathy, or hepatosplenomegaly, diagnosed by an infectious diseases specialist on initial assessment. The primary outcome was documented HIV testing during admission and HIV positivity based on enzyme immunoassay with confirmation by Western blot.

Study 2 (prospective HIV cohort). Conducted from 11 May 2015 to 10 March 2018 as part of the
European OptTEST by the EuroTEST consortium. Inclusion criteria: ED presentation with mononucleosis-like syndrome (defined as in Study 1) or community-acquired pneumonia (CAP, diagnosed according to the 2019 American Thoracic Society/Infectious Diseases Society of America guideline on the basis of clinical presentation, physical examination, and chest radiography [23]); age 18-65 years; written informed consent. HIV DUO testing (Geenius HIV 1/2 Confirmatory Assay) was offered to all eligible patients; reactive results were confirmed by Western blot (MP Diagnostics HIV Blot 2.2) and HIV RNA quantification (Abbott Real Time HIV-1). CD4 cell count and HIV viral load were determined at diagnosis. Detailed clinical, virological and linkage-to-care data for this sub-study are reported separately [24].

Study 3 (cross-sectional HCV screening of unselected patients). Cross-sectional study with prospective consecutive enrolment over a 2-month window from 1 July to 2 September 2013. Routine anti-HCV testing was offered to all consecutive ED and outpatient clinic patients during this period; the only exclusion criterion was patients presenting with already-diagnosed acute or chronic viral hepatitis. The term “unselected” in this study denotes the absence of any prior risk-factor selection. Anti-HCV antibodies were detected using the VITROS anti-HCV immunoassay (specificity 99.76%, sensitivity 100%).

Study 4 (prospective risk factor-based HCV testing). Cross-sectional study with prospective enrolment over 22 months from 15 September 2016 to 23 July 2018. Inclusion criteria followed the Polish Group of Experts Recommendations for HCV: at least one of the following risk factors: blood transfusion before 1992, current or past injecting drug use, more than three lifetime hospitalizations, history of incarceration, attendance at a voluntary counselling and testing (VCT) centre for HIV/HCV testing, elevated aminotransferase activity, or suspected liver disease of any cause. A rapid anti-HCV immunochromatographic assay (sensitivity 100%) was used; reactive results were confirmed by HCV RNA quantification at the outpatient clinic.

Study 5 (retrospective post-exposure prophylaxis cohort). Retrospective chart review of all patients receiving occupational and non-occupational post-exposure prophylaxis (PEP) between 2008 and 2016. Occupational exposure was defined as workplace contact with potentially infectious material, including needlestick injuries. Non-occupational exposure was defined as sexual contact (heterosexual or same-sex), accidental needlestick injury outside the workplace, or injecting drug use. The primary outcome was anti-HCV positivity at the first PEP visit; HBsAg and HIV positivity were also documented.

Statistical analysis

Categorical variables were compared using the χ² test or Fisher’s exact test where expected cell counts were below 5. Continuous variables were compared using Student’s t-test for normally distributed data and the Mann-Whitney U test otherwise; normality was assessed by the Shapiro-Wilk test. Prevalence estimates are presented with Clopper-Pearson exact binomial 95% confidence intervals (CI). For Study 5, both univariate and multivariate logistic regression models were fitted to identify factors associated with anti-HCV positivity; odds ratios are reported alongside 95% profile-likelihood confidence intervals. Variables included in the multivariate model were route of exposure, sex, age (years), and type of exposure (occupational vs. non-occupational). HIV cases (n = 4) were not included in regression modelling due to small numbers. Statistical significance was set at p < 0.05. Analyses were performed using SPSS Statistics version 25 (IBM Corp., Armonk, NY, USA).

Ethics

Studies were approved by the Bioethical Committee of the Medical University of Warsaw (Study 2: approval no. AKBE/35/15; Study 3: approval no. KB/187/2013; Study 4: approval no. KB/123/2017). Studies 1 and 5, based on retrospective chart review of routinely collected clinical data, were conducted in accordance with institutional research governance procedures. Written informed consent was obtained from all prospectively enrolled participants in Studies 2, 3 and 4.

Results

Study 1 – Retrospective HIV testing in mononucleosis-like syndrome

During the study period, 72 patients were hospitalized for mononucleosis-like syndrome. HIV testing was performed in 54 (75%) during admission; in the remaining 18 records, no documented reason for omission of HIV testing was identified. Four patients (5.5%; 95% CI: 1.5-13.6%) were newly diagnosed with HIV – all male, with male-to-male sex as the route of exposure, median age 24 years (range 20-41), and median CD4 count 564.7 cells/µl (range 347-855). All four patients were linked to care and initiated on ART within 1-6 months. Detection rates and study population sizes for both HIV studies are summarized in Table 1.

Study 2 – Prospective HIV testing in MONO and CAP

Eight hundred and eighty-five patients met inclusion criteria during the study period: 794 (89.7%) with MONO and 91 (10.3%) with CAP. HIV testing was accepted by 636 patients (71.9% overall uptake; MONO 573/794, 72.2%; CAP 63/91, 69.2%). Among 636 tested patients, 22 (3.5%; 95% CI: 2.2-5.2%) were confirmed HIV-positive: 15 of 573 MONO patients (2.6%; 95% CI: 1.5-4.3%) and 7 of 63 CAP patients (11.1%; 95% CI: 4.6-21.6%; p = 0.003 between groups). None of the HIV-positive patients had been previously diagnosed; all 22 were linked to specialist care, with a median of 8 days from diagnosis to antiretroviral therapy initiation. Detection rates and population sizes for both HIV studies are summarized in Table 1. Detailed clinical, immunological, virological and longitudinal linkage-to-care characteristics of the HIV-positive cohort in Study 2 are reported separately [24].

Study 3 – Cross-sectional HCV screening in unselected ED patients

Anti-HCV testing was offered to 411 consecutive patients; 393 consented and were tested (18 declined). The study population comprised 125 men (31.8%) and 268 women (68.2%) with a median age of 46 years (range 18-79). Anti-HCV antibodies were detected in 2 patients (0.5%), median age 59.5 years. Both seropositive patients were linked to care; one was confirmed HCV RNA-positive and subsequently cured with DAA therapy.

Study 4 – Prospective risk factor-based HCV testing

Of 1,502 ED patients consulted with HCV risk factors during the study period, 1,487 were tested (15 declined). The study population comprised 712 women (47.9%) with a median age of 38 years; 288 (19.4%) were hospitalized. Twenty-five patients tested anti-HCV positive, yielding an overall seroprevalence of 1.68% (95% CI: 1.09-2.47%). Seropositive patients were significantly older than seronegative patients (median 49 vs. 38 years; p = 0.025); the female proportion did not differ significantly between groups (36.0% vs. 47.3%; ns). Seroprevalence by individual risk factor group is presented in Table 2; the highest yields were in PWID (15.8%), patients who had received a transfusion before 1992 (7.0%), and former inmates (6.7%). Risk factor accumulation was associated with higher probability of seropositivity (OR = 2.2 per additional risk factor; p < 0.001): seroprevalence rose from 1.0% in patients with one risk factor to 6.3% in those with three or four, and 50% in those with five or seven risk factors. HCV RNA was confirmed in 21 of 25 seropositive patients (84.0%). Median time from diagnosis to first specialist visit was 168 days (range 21-288); seven patients (28.0%) required urgent hospital admission for decompensated liver disease (ascites, hepatic encephalopathy, severe thrombocytopenia).

Study 5 – Retrospective post-exposure prophylaxis cohort

Of 5,187 PEP records screened, 1,594 (30.7%) were excluded for incomplete first-visit data, leaving 3,593 individuals for analysis. Anti-HCV antibodies were positive in 60 individuals (1.7%); HBsAg was positive in 34 (1.0%); HIV antibodies/p24 antigen were positive in 4 (0.1%). The HCV-positive group differed from the HCV-negative group in three respects: a lower proportion of women (41.7% vs. 57.8%; p = 0.018), older mean age (40.0 vs. 35.7 years; p = 0.013), and a lower proportion of occupational exposures (31.7% vs. 63.9%; p < 0.001). HIV cases were not included in regression modelling due to small numbers. Multivariate logistic regression results are presented in Table 3. Independent predictors of anti-HCV positivity were injecting drug use (OR = 162.60, 95% CI: 38.55-1139.85; p < 0.001), male-to-male sexual contact (OR = 3.59, 95% CI: 1.01-22.87; p = 0.048), and age in years (OR = 1.05, 95% CI: 1.03-1.07; p < 0.001).

Discussion

This series of five sub-studies provides the first prospective and retrospective Polish data on HIV and HCV testing in a specialist infectious diseases ED setting between 2013 and 2018. The data demonstrate three findings of practical relevance: 1) indicator condition-guided HIV testing detects new infections in 3.5-5.5% of patients with mononucleosis-like syndrome or CAP – substantially higher than the general population detection rate of 0.05 per 100 tests in Poland; 2) HCV testing of unselected ED patients yields only 0.5% positivity, whereas risk factor-based testing reaches 15.8% in PWID and 7.0% in patients with a history of transfusion before 1992; 3) the median delay from HCV diagnosis to first specialist visit (168 days) contrasts sharply with the 8-day median time from HIV diagnosis to ART initiation in Study 2, identifying a potential gap in the HCV care continuum.

Indicator condition-guided HIV testing in the ED

The detection rates observed in Studies 1 and 2 (5.5% and 3.5%, respectively) exceed the 0.1% threshold above which IC-guided testing is considered cost-effective according to EU guidelines [19] by more than thirtyfold. The particularly high yield in CAP patients (11.1%) may reflect a high prevalence of severe immunosuppression in this subgroup, consistent with the broader European pattern in which a substantial proportion of new HIV diagnoses are made at advanced disease stages [6]. The high yield in MONO patients (2.6%) is consistent with detection of acute or recent HIV infection; patients in this group had high viral loads and relatively preserved CD4 counts, consistent with primary HIV syndrome. Identifying these patients during the acute retroviral phase is doubly valuable: rapid ART initiation provides individual benefit, and detection during the period of highest infectiousness reduces onward transmission [25].

These findings align with the European HIDES I and HIDES II studies, which validated IC-guided testing across multiple European countries [19, 20], and with implementation experience from the United Kingdom [22], Spain [18, 26], the Netherlands [27], and Italy [28]. Despite this evidence base, IC-guided HIV testing has not been routinely implemented in Polish ED settings, and the barrier to testing in our prospective study was overwhelmingly failure to offer (87.1% of non-tested cases) rather than patient refusal (12.9%) – a pattern consistent with reports from other countries [21].

Risk factor-based HCV testing

Study 3 demonstrates that universal HCV screening of unselected ED patients yields too few cases (0.5%) to justify resource allocation in a low-prevalence setting. This finding does not contradict the case for population-level HCV elimination strategies, which require approximately 3 million tests annually in Poland to meet WHO 2030 targets [29, 30]. Rather, it indicates that the ED is not the most efficient channel for unselected screening: the same population is more efficiently reached through primary care, antenatal screening, and birth-cohort programmes. Within the ED, however, risk-based selection generates yields that justify routine testing: 15.8% in PWID and 7.0% in patients who received transfusions before 1992, both far above reported thresholds for cost-effective testing. ED-based risk-stratified testing should be viewed as complementary to population-level screening, not as an alternative.

The independent risk factors identified in Study 5 – injecting drug use, male-to-male sexual contact, and older age – are consistent with established predictors of HCV positivity in European cohorts [11]. The very high OR for injecting drug use (162.60) indicates a strong association with HCV positivity in Poland, consistent with epidemiological data from across the WHO European Region [29].

Linkage to care: the gap between diagnosis and treatment

The 168-day median delay from HCV diagnosis to first specialist visit observed in Study 4 highlights a major gap in the care continuum. Twenty-eight percent of newly diagnosed HCV patients required urgent hospitalization for decompensated liver disease before being seen, indicating that the wait itself contributes to morbidity. By contrast, HIV-positive patients in Study 2 reached ART initiation at a median of 8 days from diagnosis, demonstrating that rapid linkage is achievable when the system is designed for it. The discrepancy between these two pathways, in the same hospital and during overlapping periods, suggests that HCV linkage delays are organizational rather than clinical – reflecting limited outpatient hepatology capacity rather than diagnostic complexity. European civil society monitoring across 25 countries during 2020-2023 found persistent gaps in the HCV care continuum, particularly among PWID [31], confirming that this is not a uniquely Polish problem.

International ED testing experience

Several European countries have implemented ED-based bloodborne virus testing programmes. In London, opt-out HIV testing in EDs serving high-prevalence boroughs had diagnosed over 750 HIV-positive individuals by April 2024, with patient acceptance exceeding 90% [16, 17]. In Catalonia, opt-in ED testing identified previously undiagnosed HIV in symptomatic and asymptomatic patients [18]. In the Netherlands, IC-guided testing in primary care reduced late presentation in a high-prevalence area [27]. Italian economic modelling suggested that universal opt-out ED testing for HIV is cost-effective compared with indicator-only approaches [28]. These models offer a range of options for Poland; the choice between them depends on local prevalence, ED workflow, and acceptability, and would benefit from pilot implementation studies in Polish settings.

Recent developments

Since this study concluded in 2018, several relevant policy developments have occurred in Poland, including the introduction of HIV testing into the primary care benefits package through the “Moje Zdrowie” programme in May 2025 and the launch of the ELIMINATE-HCV pilot in 2024. The effects of these initiatives on ED testing patterns are beyond the scope of the present analysis.

Limitations

This work has several limitations. First, all sub-studies were conducted at a single specialist infectious diseases hospital. As a referral centre, this hospital sees a self-selected population in which bloodborne infections are overrepresented compared with general emergency departments. Detection rates reported here therefore represent an upper bound; replication in non-specialist settings would be expected to yield lower absolute prevalence but the same relative pattern across risk groups. Second, data were collected during the active study period 2013-2018 (with retrospective records from 2008 onwards for the PEP cohort in Study 5), before the COVID-19 disruption of testing services and before the introduction of HIV testing into the Polish primary care benefits package through the “Moje Zdrowie” programme in 2025. The findings should be interpreted as a baseline against which post-pandemic and post-reform testing strategies can be compared. Third, the small CAP subgroup in Study 2 (n = 63) yields wide confidence intervals around the 11.1% prevalence estimate. Fourth, the retrospective designs of Studies 1 and 5 carry inherent limitations of chart-based data collection; in Study 5, 30.7% of records were excluded for incomplete first-visit data. Fifth, linkage-to-care data were limited to the first specialist visit, and longer-term retention in care was not analysed. Sixth, HCV genotype data were not available for the seropositive cases. Finally, indicator condition diagnoses in Studies 1 and 2 were assigned by clinicians on initial assessment, with potential for inter-rater variability.

Conclusions

Indicator condition-guided HIV testing in a specialist infectious diseases emergency department detects new infections in 3.5-5.5% of patients with mononucleosis-like syndrome or community-acquired pneumonia – a rate substantially higher than the general population detection rate.

Universal HCV screening of unselected ED patients in this setting yields 0.5% seroprevalence and appears to have low diagnostic yield; risk factor-based testing reaches up to 15.8% positivity in people who inject drugs.

Independent predictors of anti-HCV positivity identified by multivariate analysis are injecting drug use, male-to-male sexual contact, and older age; these groups warrant priority for routine HCV testing.

The 168-day median delay from HCV diagnosis to first specialist visit identifies a major gap in the care continuum, highlighting a need for improved linkage to specialist care, in contrast to the rapid HIV linkage observed in the same hospital (8-day median to ART).

Findings are most directly translatable to specialist infectious diseases emergency departments and require validation in general ED settings before broader implementation. Prospective evaluation of opt-out, opt-in, and risk-based models in Polish settings would inform national strategy.

Acknowledgements

We thank Prof. Justyna D. Kowalska for her supervision of the doctoral research on which this study is based, and Maksymilian Bielecki, for his support with the statistical analysis.

We thank the staff of the Emergency Department and HIV Out-Patient Clinic at the Hospital for Infectious Diseases in Warsaw for their contribution to patient recruitment and data collection.

We also acknowledge all members of the OptTEST by EuroTEST consortium for their collaboration.

Disclosures

This research received no external funding.

Approval of the Bioethics Committee – Approval number: KB/123/2017 KB/187/2013 AKBE/35/15.

The author declares no conflict of interest.

References