HIV, tuberculosis, diabetes mellitus and hypertension admissions and premature mortality among adults in Uganda from 2011 to 2019: is the tide turning?

Abstract

Background The growing burden of diabetes mellitus (DM) and hypertension (HTN) on the background of endemic Human Immuno-deficiency Virus (HIV) and tuberculosis (TB) is a concern in low- and middle-income countries. We aimed to describe annual trends in admissions, mortality rates and premature mortality (years of potential life lost—YPLLs) due to HIV, tuberculosis (TB), diabetes mellitus (DM) and hypertension (HTN) in Uganda. Methods We conducted a retrospective cohort study, retrieving electronic records of adults admitted to Mulago and Kiruddu national referral hospitals medical wards between 1st January 2011 and 31st December 2019. We used STATA BE 17.0 and GraphPad Prism 8.0.2 to compute total admissions, inpatient crude mortality rates, and YPLLs; and demonstrate trends using Mann–Kendall test. Results Of 108,357 admissions, 55,620 (51.3%) were female, 15,300 (14.1%) were recorded in 2012, and 22,997 (21.2%) were aged 21–30 years. HIV, TB, DM and HTN accounted for 26,021 (24.0%); 9537 (8.8%); 13,708 (12.7) and 13,252 (12.2%) of all admissions, respectively. Overall inpatient mortality was 16.7% (18,099/108,357), 53.5% (9674/18,099) were male, 21.5% (3898) were aged 31–40 years and 2597 (14.4%) were registered in 2013. HIV, TB, DM and HTN accounted for 35.6% (6444), 14.6% (2646), 9.1% (1648) and 11.8% (2142) of all deaths, respectively. Total admissions (Kendall’s tau-B = − 0.833, p < 0.001) and deaths declined (Kendall’s tau-B = − 0.611, p = 0.029). A total of 355,514 (mean = 20.8 years, SD 30.0) YPLLs were recorded, of which 54.6% (191,869) were in males; 36.2% (128,755) were among those aged 21–30 years and were recorded in 2012 (54,717; 15.4%). HIV, TB, DM and HTN accounted for 46.5% (165,352); 19.5% (69,347); 4.8% (16,991) and 4.5% (16,167) of YPLLs, respectively. Proportionate contribution of HIV to deaths and YPLLs declined, remained stagnant for TB; and increased for both DM and HTN. Conclusion TB and HIV account for higher though declining, while DM and HTN account for lower albeit rising morbidity and premature mortality among adult medical patients in Uganda. TB prevention and treatment; and DM/HTN service integration in HIV care should be optimized and scaled up.

Un miembro del personal de la Building Foundation for Development ayuda a preparar suministros para su entrega a instalaciones sanitarias y campos de desplazados en Al Jazirah (Sudán), que actualmente acoge a desplazados internos que huyen del conflicto en Jartum (Sudán).

“…The pooled prevalence of diabetes in medical settings was 11.9% (95% CI: 9.9–14.3; 15 studies; Table 2 ). 39 , 49 , 50 , 52 , 53 , 62 , 64 , 66 , 69 – 71 , 79 , 84 , 85 , 92 In emergency departments, we found an overlapping prevalence estimate of 9.3% (95% CI: 5.1–16.2; three studies). 58 , 75 , 80 A high degree of heterogeneity of estimates in both settings is noted (medical wards 95% PI: 5.4–24.3; I 2 : 97.6% and emergency departments I 2 : 98.5%; < 5 studies).…”

“…Fourteen studies reported comorbid chronic diseases alongside a primary condition (online repository). 20 Hypertension and diabetes co-prevalence was reported in three studies: 3790/108 357 (3.5%); 64 33/353 (9.3%); 66 and 273/3961 (6.9%). 58 HIV and hypertension co-prevalence was reported in three studies: 665/108 357 (0.6%); 64 6/972 (0.6%); 48 and 140/2316 (6.0%).…”

“…1 ). These articles included 17 prospective cohort studies; 37 – 53 11 retrospective cohorts; 54 – 64 and 33 cross-sectional studies. 65 – 97 The pooled sample size was 376 676 participants, including 97 737 participants admitted to the emergency department, and 278 939 admitted to medical wards.…”

“…HIV prevalence was reported in 32 studies ( Table 2 ). The pooled prevalence of HIV in medical wards was 36.4% (95% CI: 31.3–41.8; 25 studies) 39 , 40 , 43 , 45 , 47 – 49 , 51 , 52 , 55 , 56 , 60 , 61 , 64 , 69 – 71 , 79 , 81 , 84 – 87 , 89 higher than the prevalence in emergency departments (21.9%; 95% CI: 14.5–31.7; seven studies). 37 , 44 , 65 , 72 , 80 , 82 , 96 HIV infection was reported using laboratory or point-of-care diagnostics in 18 studies, 37 , 40 , 43 – 45 , 47 – 51 , 56 , 65 , 68 , 72 , 82 , 84 , 89 , 96 and by medical records or clinical history in the remaining studies ( Table 1 ).…”

“…We estimated the prevalence of hypertension from 16 studies. Pooled prevalence was 24.4% (95% CI: 16.7–34.2; 14 studies) 39 , 49 , 50 , 52 , 55 , 64 , 66 , 67 , 69 – 71 , 79 , 84 , 85 in medical wards, and 31.5% (95% CI: 26.4–37.1; two studies) 58 , 80 in emergency departments. A high degree of heterogeneity in prevalence estimates was observed in the medical setting (95% PI: 4.1–71.1; I 2 : 99.7%).…”