Blood transfusion is life-saving but sometimes also associated with morbidity and mortality. There is limited data on mortality in patients transfused with whole blood in sub-Saharan Africa. We described the 30-day all-cause mortality and its associated factors in patients transfused with whole blood to inform appropriate clinical intervention and research priorities to mitigate potential risks. A retrospective study was performed on purposively sampled patients transfused with whole blood at the Uganda Cancer Institute (UCI) and Mulago hospital in the year 2018. Two thousand twelve patients with a median (IQR) age of 39 (28–54) years were enrolled over a four month period. There were 1,107 (55%) females. Isolated HIV related anaemia (228, 11.3%), gynaecological cancers (208, 10.3%), unexplained anaemia (186, 9.2%), gastrointestinal cancers (148, 7.4%), and kidney disease (141, 7.0%) were the commonest diagnoses. Most patients were transfused with only one unit of blood (n = 1232, 61.2%). The 30 day all-cause mortality rate was 25.2%. Factors associated with mortality were isolated HIV related anaemia (HR 3.2, 95% CI, 2.3–4.4), liver disease (HR 3.0, 95% CI, 2.0–4.5), kidney disease (HR 2.2, 95% CI, 1.5–3.3; p<0.01), cardiovascular disease (HR 2.9, 95% CI, 1.6–5.4; p<0.01), respiratory disease (HR 3.0, 95% CI 1.8–4.9; p<0.01), diabetes mellitus (HR 4.1, 95% CI, 2.3–7.4; p<0.01) and sepsis (HR 6.2, 95% CI 3.7–10.4; p<0.01). Transfusion with additional blood was associated with survival (HR 0.8, 95% CI 0.7–0.9, p<0.01). In conclusion, the 30-day all-cause mortality was higher than in the general inpatients. Factors associated with mortality were isolated HIV related anaemia, kidney disease, liver disease, respiratory disease, cardiovascular disease, diabetes mellitus and sepsis. Transfusion with additional blood was associated with survival. These findings require further prospective evaluation.

Background: Most recent cancer registry data suggests a change in cancer occurrence in Uganda with a decrease in incidence of Kaposi sarcoma (KS) but an increase in cervical, prostate and breast cancer. Anecdotal data suggest that KS, non-Hodgkin's lymphoma and breast cancer were the most common cancers among patients at Uganda Cancer Institute (UCI) by 2006. Aim: To describe the spectrum of cancer diagnoses among new patients that presented for care at UCI over the past 2-years. Methods: We conducted a cross sectional study of patients admitted into care at UCI with a histologic or clinical diagnosis of cancer from January 2015 to December 2016. Cancer diagnoses were reported as proportions by gender and age - children (0-14 years) and adults (above 14 years). Results: Overall, 8279 new patients were registered during the study period but only 7588 (92%) were recorded in the electronic database and had information on cancer diagnosis. Of these, 53% were admitted in 2015, and 55% were females. Median age was 48 years (IQR: 34-62). Among 2997 female adults, 30% had cervical, 17% breast, 5% Kaposi sarcoma (KS), 4% leukemia and 3.9% esophageal cancer. Among 2136 male adults, 17% had KS, 12% prostate, 10% esophageal, 6% leukemia and 4% colorectal cancer. Among the 486 children, 17% had leukemia, 16.7% nephroblastoma (Wilms tumor), 15.9% Burkitt lymphoma (BL), 8% rhabdomyosarcoma, and 6% Kaposi sarcoma. Conclusion: The distribution of cancer diagnoses among patients seen at UCI reflects the population level cancer incidence with cervical, breast, KS, prostate, esophageal, and colorectal cancer in adults, and nephroblastoma in children as the leading cause of cancer related morbidity. The overrepresentation of leukemia may be due to referral bias but warrants further study. The correlation of our findings with incidence data suggests that missing information did not significantly skew our findings. However more investments are needed to improve the quality of data captured electronically.