Five Years of HIV Hospital Admissions Reveal Late Diagnosis and a Stark Mortality Signal

More than four decades after the first cases of AIDS were described, hospital wards in many countries still admit people living with HIV who arrive profoundly immunocompromised, often without ever having been tested for the virus. A new five-year retrospective study from a tertiary referral hospital in Türkiye offers a detailed window into what that reality looks like on the ground, documenting why hospitalized patients with HIV/AIDS needed inpatient care, how advanced their immune suppression was at the moment of admission, and which factors were associated with dying in hospital. The research, published in BMC Infectious Diseases, analyzed 128 adults with laboratory-confirmed HIV infection admitted between 2021 and 2025, and its findings carry uncomfortable implications for diagnosis and continuity of care even in the era of combined antiretroviral therapy.

The study team, led by Alper Tahmaz and colleagues in the Department of Infectious Diseases and Clinical Microbiology at Antalya Training and Research Hospital, designed the analysis around a single admission per patient. When a patient had been hospitalized more than once during the study window, the researchers selected the chronologically latest admission, irrespective of whether the patient survived. This approach ensured that each individual contributed only one observation to the dataset, avoiding the statistical distortion that repeated admissions can introduce when the same severely ill patients dominate hospital records. The investigators then combed retrospectively through demographic characteristics, immune status, viral load measurements, comorbidities, and the recorded indications for hospitalization, which were treated as non-mutually exclusive categories, meaning a single admission could carry several simultaneous reasons for care.

The demographic and clinical profile of the cohort was striking. The median age of admitted patients was 45 years, and nearly nine in ten participants, 89.8 percent, were male. More revealing were the immunological numbers: at admission, 56.3 percent of patients had CD4 cell counts below 200 cells per cubic millimeter, the threshold that defines advanced immunosuppression and the level at which opportunistic infections become a dominant threat. Only 37.5 percent of the hospitalized patients had a previously established HIV diagnosis, and just 27.3 percent were receiving combined antiretroviral therapy at the time of admission. In other words, the majority of people ending up in hospital beds were either newly diagnosed in the course of their acute illness or had known HIV but were not on suppressive treatment, a pattern that echoes the persistent challenge of late diagnosis documented across many health systems.

The reasons these patients required hospitalization mapped closely onto the classical spectrum of HIV-related disease. Bacterial infections were the most frequently recorded indication, appearing in 58.6 percent of admissions, a reminder that even in the antiretroviral era, bacterial pneumonia, sepsis, and other common bacterial pathogens remain leading causes of serious illness among immunocompromised hosts. Pneumocystis jirovecii pneumonia, the fungal opportunistic infection that once defined the AIDS epidemic, was recorded in 28.9 percent of patients, while cytomegalovirus infection, another hallmark of severe cellular immune deficiency, appeared in 21.9 percent. Because the indications were not mutually exclusive, many patients carried multiple overlapping diagnoses at once, reflecting the compounded vulnerability of bodies whose CD4 defenses had collapsed.

Beyond the classic opportunistic infections, the study also captured the substantial burden of other medical conditions, which were recorded in 67 patients, or 52.3 percent of the cohort. Rather than forcing these into narrow diagnostic boxes, the researchers described them by organ system category, acknowledging that hospitalized people living with HIV frequently present with a tangle of concurrent problems affecting the respiratory tract, the gastrointestinal system, the nervous system, and beyond. This overlap of infectious and non-infectious conditions is one of the study’s central messages: the modern hospitalized HIV patient is rarely admitted for a single clean diagnosis, and clinical management must contend with simultaneous, interacting pathologies.

The severity of these admissions was underscored by intensive care utilization. Nearly a quarter of the cohort, 24.2 percent, required admission to the intensive care unit at some point during their hospital stay. That figure speaks to how far disease had progressed before these patients reached effective care, and it aligns with the broader clinical understanding that late presenters with CD4 counts below 200 cells per cubic millimeter face dramatically elevated risks of respiratory failure, septic shock, and central nervous system complications that demand critical care support.

The primary outcome of the study was documented in-hospital death, and the mortality figure was sobering: 20 of the 128 patients died, an in-hospital mortality rate of 15.6 percent, with a 95 percent confidence interval of 10.3 to 22.9 percent. To identify factors associated with this outcome, the researchers employed an exploratory Firth penalized logistic regression model, a statistical technique specifically designed for situations where the number of outcome events is small relative to the number of predictor variables. Ordinary logistic regression can fail or produce wildly unstable estimates in such sparse-data settings, and the Firth penalty shrinks coefficient estimates toward zero to reduce this small-sample bias. The model included three predictors: age, a CD4 count below 200 cells per cubic millimeter at admission, and a history of malignancy.

Of these three candidates, one emerged with a statistically clear signal. A history of malignancy was associated with roughly fourfold higher odds of in-hospital death, with an adjusted odds ratio of 4.00 and a 95 percent confidence interval of 1.25 to 12.37, yielding a p-value of 0.020. The authors are careful to frame this as an exploratory finding, and for good reason: with only 20 deaths in the entire cohort, the confidence interval is wide, spanning from just over one to more than twelve, which means the true strength of the association could plausibly range from modest to very large. The direction of the result, however, is biologically coherent. Cancer history in people living with HIV can reflect prior AIDS-defining malignancies such as lymphomas, the immunosuppressive effects of chemotherapy, and generally diminished physiological reserve, all of which plausibly compound the danger of an acute infectious admission.

The authors themselves are explicit about the limitations that should temper interpretation. The cohort was drawn from a single tertiary referral hospital, which selects for the most complex and severe cases in a region and limits generalizability to primary care settings or to the broader population of people living with HIV who never require admission. The retrospective design depends on the completeness and accuracy of medical records, and the regression model adjusted for only a small number of covariates, leaving open the possibility of confounding by unmeasured factors such as treatment adherence, nutritional status, or the specific pathogens involved. The selection of the latest admission per patient, while methodologically defensible, also means the analysis captures a particular snapshot of each patient’s trajectory rather than the full arc of their hospitalizations.

Even with those caveats, the study lands on a set of conclusions that resonate far beyond one Turkish hospital. The finding that more than half of hospitalized patients had CD4 counts below 200 cells per cubic millimeter, and that fewer than three in ten were on antiretroviral therapy, points to failures upstream of the hospital door: missed opportunities for earlier testing, gaps in linkage to care, and interruptions in treatment continuity. The authors argue that their findings support greater attention to timely HIV diagnosis, sustained continuity of care, and systematic comorbidity assessment in this population. As antiretroviral therapy transforms HIV into a manageable chronic condition for people who are diagnosed early and treated consistently, studies like this one document what happens in the gap between that promise and its delivery, and they quantify, in hard numbers, the cost of arriving at the hospital too late.

Subject of Research: Hospitalization indications, clinical course, and in-hospital mortality among people living with HIV/AIDS

Article Title: Indications for hospitalization, clinical course, and mortality analysis among patients with HIV/AIDS in a tertiary care hospital: a 5-year retrospective study (2021–2025)

Article References: Tahmaz, A., Yildiz Dikmen, M., Ersari, S. S., Dinç, E., & Seyman, D. (2026). Indications for hospitalization, clinical course, and mortality analysis among patients with HIV/AIDS in a tertiary care hospital: a 5-year retrospective study (2021–2025). BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14560-4

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14560-4

Keywords: HIV, AIDS, hospitalization, in-hospital mortality, CD4 count, opportunistic infections, Pneumocystis jirovecii pneumonia, cytomegalovirus, antiretroviral therapy, late diagnosis, malignancy, retrospective study