BackgroundDespite increased antiretroviral therapy (ART) coverage and the raised CD4 threshold for starting ART, opportunistic infections (OIs) are still one of the leading causes of death in sub-Saharan Africa. There are few studies from resource-limited settings on long-term reporting of OIs other than tuberculosis.MethodsPatients starting ART between April 2004 and April 2005 were enrolled and followed-up for 10 years in Kampala, Uganda. We report incidences, patterns and risk factors using Cox proportional hazards models of OIs among all patients and among patients with CD4 cell counts >200 cells/μL.ResultsOf the 559 patients starting ART, 164 patients developed a total of 241 OIs during 10 years of follow-up. The overall incidence was highest for oral candidiasis (25.4, 95% confidence interval (CI): 20.5–31.6 per 1000 person-years of follow-up), followed by tuberculosis (15.3, 95% CI: 11.7–20.1), herpes zoster (12.3, 95% CI: 9.1–16.6) and cryptococcal meningitis (3.0, 95% CI: 1.7–5.5). Incidence rates for all OIs were highest in the first year after ART initiation and decreased with the increase of the current CD4 cell count. Factors independently associated with development of OIs were baseline nevirapine-based regimens, time-varying higher viral load, time-varying lower CD4 cell count and time-varying lower hemoglobin. In patients developing OIs at a current CD4 cell count >200 cells/μL, factors independently associated with OI development were time-varying increase in viral load and time-varying decrease in hemoglobin, whereas a baseline CD4 cell count <50 cells/μL was protective.ConclusionWe report high early incidences of OIs, decreasing with increasing CD4 cell count and time spent on ART. Ongoing HIV replication and anemia were strong predictors for OI development independent of the CD4 cell count. Our findings support the recommendation for early initiation of ART and suggest close monitoring for OIs among patients recently started on ART, with low CD4 cell count, high viral load and anemia.

WHAT THIS PAPER ADDSThe study analyses the diagnostic accuracy of two imaging modalities in the detection of infected aortic aneurysms. The accuracy of positron emission tomography/computed tomography (PET/CT) is higher than contrast enhanced CT. The high metabolic activity of infected aortic aneurysms, measurable and quantifiable only by PET/ CT by means of maximum standardised uptake values (SUV max ), may account for the excellent sensitivity of PET/ CT. However, its specificity is hampered, owing to false positive findings in inflammatory aneurysms and in arteritis. Findings may influence future clinical practice, as PET/CT may become the imaging modality of choice in infected aortic aneurysms.Objective: Infected aortic aneurysms are highly lethal, and management is very demanding, requiring an early diagnosis. The aim of this study was to evaluate the diagnostic accuracy of positron emission tomography/ computed tomography with 18 F-fluorodeoxyglucose (PET/CT) and contrast enhanced CT (CE-CT) in patients with suspected infected aortic aneurysms. Methods: PET/CT was performed in patients with clinically suspected infected aortic aneurysms, and additional CE-CT was performed if feasible. Diagnostic accuracy was assessed by two independent readers using a four point grading score for both imaging modalities. Maximum standardised uptake values (SUV max ) were calculated for quantitative measurements of metabolic activity in PET/CT. The reference standard was a combination of clinical presentation, laboratory findings, and imaging. Results: Ten patients were included prospectively in the study, 24 retrospectively; 16 patients (47%) prior to the start of antimicrobial treatment and all 34 patients prior to any vascular intervention. Thirteen of the 34 patients had an infected aortic aneurysm (38%). Proven infected aortic aneurysms were all metabolically active on PET/CT with a median SUV max of 6.6 (interquartile range 4.7e21.8). Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of PET/CT for the diagnosis of infected aortic aneurysm was 100%, 71%, 68%, 100%, and 82%, for reader 1 and 85%, 71%, 65%, 88%, and 77%, for reader 2. Respective values for CE-CT, performed in 20 patients (59%), were 63%, 75%, 63%, 75%, and 70%, for reader 1 and 88%, 50%, 54%, 86%, and 65%, for reader 2. Conclusion:The diagnostic accuracy of PET/CT in the detection of infected aortic aneurysms (n ¼ 13) is high, and higher than CE-CT. While PET/CT demonstrates an excellent sensitivity, its specificity is hampered because of false positive findings.