Increased antimicrobial resistance among Human Immunodeficiency Virus (HIV)-infected individuals to commonly used antibiotics in the treatment of gastroenteritis is a public health concern, especially in resource-limited settings. We set out to compare the antimicrobial susceptibility pattern of Escherichia coli (E. coli) isolates from HIV-infected and HIV-uninfected individuals at a tertiary hospital in Lusaka, Zambia. An analytical cross-sectional study was conducted at the University Teaching Hospital from May 2019 to August 2019. Stool samples were screened, and 79 HIV-infected individuals matched by age and sex with 84 HIV-uninfected individuals that presented with E. coli associated gastroenteritis were studied. Demographics were collected from the Laboratory Information System (LIS) and stool samples were collected in a sterile leak-proof container. Samples were cultured and only those where E. coli was isolated were included in the study and tested for antimicrobial susceptibility by the Kirby–Bauer disk diffusion technique. HIV-positive individuals were 3 times (adjusted odds ratio (AOR) = 3.17; 95% CI (1.51, 6.66); p < 0.001) more likely to be resistant to quinolones compared with their HIV-negative counterparts. Similarly, HIV-positive individuals were almost 4 times (AOR = 3.97, 95% CI (1.37, 11.46); p = 0.011) more likely to have multidrug-resistant E. coli compared with those who were HIV-negative. HIV infection was associated with reduced E. coli susceptibility to commonly used antibiotics, and most cases showed resistance.

The coronavirus disease 2019 (COVID-19) has wreaked havoc globally, resulting in millions of cases and deaths. The objective of this study was to predict mortality in hospitalized COVID-19 patients in Zambia using machine learning (ML) methods based on factors that have been shown to be predictive of mortality and thereby improve pandemic preparedness. This research employed seven powerful ML models that included decision tree (DT), random forest (RF), support vector machines (SVM), logistic regression (LR), Naïve Bayes (NB), gradient boosting (GB), and XGBoost (XGB). These classifiers were trained on 1,433 hospitalized COVID-19 patients from various health facilities in Zambia. The performances achieved by these models were checked using accuracy, recall, F1-Score, area under the receiver operating characteristic curve (ROC_AUC), area under the precision-recall curve (PRC_AUC), and other metrics. The best-performing model was the XGB which had an accuracy of 92.3%, recall of 94.2%, F1-Score of 92.4%, and ROC_AUC of 97.5%. The pairwise Mann–Whitney U-test analysis showed that the second-best model (GB) and the third-best model (RF) did not perform significantly worse than the best model (XGB) and had the following: GB had an accuracy of 91.7%, recall of 94.2%, F1-Score of 91.9%, and ROC_AUC of 97.1%. RF had an accuracy of 90.8%, recall of 93.6%, F1-Score of 91.0%, and ROC_AUC of 96.8%. Other models showed similar results for the same metrics checked. The study successfully derived and validated the selected ML models and predicted mortality effectively with reasonably high performance in the stated metrics. The feature importance analysis found that knowledge of underlying health conditions about patients’ hospital length of stay (LOS), white blood cell count, age, and other factors can help healthcare providers offer lifesaving services on time, improve pandemic preparedness, and decongest health facilities in Zambia and other countries with similar settings.

Persons living with HIV (PLWH) receiving tenofovir disoproxil fumarate (TDF)-based antiretroviral therapy (ART) risk suffering TDF-associated nephrotoxicity (TDFAN). TDFAN can result in short- and long-term morbidity, including permanent loss of kidney function, chronic kidney disease (CKD), and end-stage kidney disease (ESKD) requiring dialysis. Currently, there is no model to predict this risk or discern which patients to initiate TDF-based therapy. Consequently, some patients suffer TDFAN within the first few months of initiating therapy before switching to another suitable antiretroviral or a lower dose of TDF. In a prospective observational cohort study of adult Zambian PLWH, we modelled the risk for TDFAN before initiating therapy to identify individuals at high risk for experiencing AKI after initiating TDF-based therapy. We enrolled 205 HIV-positive, ART-naïve adults initiating TDF-based therapy followed for a median of 3.4 months for TDFAN at the Adult Infectious Disease Research Centre (AIDC) in Lusaka, Zambia. We defined TDFAN as meeting any of these acute kidney disease (AKD) criteria: 1) An episode of estimated glomerular filtration rate (eGFR)< 60ml/ min/1.73m2 within 3 months, 2) reduced eGFR by> 35% within 3 months or 3) increased serum creatinine by> 50% within 3 months. A total of 45 participants (22%) developed acute kidney disease (AKD) after TDF-based therapy. The development of AKD within the first 3 months of commencing TDF-based therapy was associated with an increase in baseline serum creatinine, age, baseline eGFR and female sex. We concluded that baseline characteristics and baseline renal function biomarkers predicted the risk for AKD within the first 3-months of TDF-based therapy.

Medication administration omission errors (MAOE) are very common and often affect patient outcomes and length of stay in the hospitals. This study was a cross-sectional study in which the frequency and causes of MAOE over four weeks at Ndola Teaching Hospital (NTH) was assessed. It involved reviewing patients' drug charts and observation of nurses during the administration of medications to inpatients to detect the MAOE. A total of 259 drug charts were reviewed using a checklist and administered semistructured questionnaires to 50 nurses who were involved in medication administration to solicit the cause of MAOE. To assess factors associated with MAOE, multivariate logistic regression was used. In this study, 259 drug charts were reviewed of which 220 (84.9%) had one or more MAOE. Of the 1100 doses prescribed to 259 inpatients, 317 doses were omitted resulting in an overall MAOE frequency of 28.8%. In multivariate regression analysis, increased number of medications that the patient used (AOR: 2.18, CI: 1.62-2.94; p=0.0001), being male (AOR: 2.42, Cl: 1.05-5.53: p=0.036) and surgical wards (AOR: 8.56, CI: 3.04 -24.1; p=0.0001) were significant predictors of MAOE. The most common causes of MAOE were the unavailability of medication on the ward followed by work overload. The most omitted class of medication was anti-infective. Medication omission errors are common and affect adult inpatients at Ndola Teaching. There is a need to highlight the magnitude of this problem to promote awareness so that specific interventions are put in place to address the identified causes.