Limited data exist on temporal changes in antibiotic use in low and middle-income countries. We evaluated trends in antibiotic use at tertiary hospitals in Uganda.

Retrospective trend analysis of a repeated point prevalence survey (PPS).

This study utilised antibiotic use data from quarterly PPS conducted among inpatients at nine regional referral hospitals in Uganda between October 2020 and December 2023.

We determined the proportions of antibiotic use, prescriptions guided by culture and sensitivity tests (CST), WHO AWaRe (Access, Watch and Reserve) categories, and prescriptions without documented indication. Linear regression was used to derive slope coefficients and 95% confidence interval (CI).

Of 15,154 patients surveyed, 8,892 (58.7%) received systemic antibiotics. The median age was 23 years (IQR: 11–38), 5,338 (60.5%) were female, and 4,583 (51.5%) were on treatment for infectious syndromes, including sepsis (1,400, 15.7%) and pneumonia (867, 9.8%). The drug utilisation index (DU75) consisted of ceftriaxone, metronidazole, gentamicin and ampicillin, which accounted for 76.9% (12,291/15,989) of total antibiotic use. The distribution of prescribed antibiotics was 46.6% Access, 45.5% Watch, 0.1% Reserve and 7.7% unrecommended combinations. Overall, 5,402 (60.8%) prescriptions were aligned with national guidelines, 2,147 (24.1%) prescriptions were issued without an indication, and CST guided 271 (3%) prescriptions. Over time, there was no significant change in antibiotic prescription prevalence (slope=0.09, CI –0.93 to 1.10) and prescriptions without indication (slope=–0.70, CI –1.79 to 3.98). However, adherence to treatment guidelines (slope=2.06, CI 0.14 to 3.98) and prescriptions based on CST results (slope=0.62, CI 0.12 to 1.13) significantly increased, while ‘Watch’ antibiotics prescriptions decreased (slope=–0.40, CI –0.63 to –0.17).

The antibiotic prescription rate remained high, with no significant change over time. Improvements were seen in adherence to treatment guidelines, use of CST and reduced use of ‘Watch’ antibiotics. Strengthening antibiotic stewardship is recommended to further improve practices.

Intrapartum-related complications are a leading cause of adverse perinatal outcomes, including stillbirths, neonatal deaths and intrapartum-related neonatal encephalopathy (IP-NE). We assessed the prevalence of adverse intrapartum-related outcomes, evaluated the association between IP-NE and obstetric and fetal risk factors, and examined whether emergency referral and emergency caesarean section (CS) modified this association through interaction effects.

Cross-sectional with a nested case–control study.

Two hospitals in rural Eastern Uganda.

Women giving birth to a live or stillborn baby weighing >2000 g between June and December 2022.

We used prospectively collected perinatal e-registry data to assess the prevalence of adverse perinatal outcomes. Logistic regression with interaction with postregression margins analysis was used to determine the association between IP-NE and emergency referral and emergency CS across risk groups of hypertensive disorders, antepartum haemorrhage, prolonged/obstructed labour and birth weight.

Adverse perinatal outcomes were stillbirths, 24-hour neonatal deaths and IP-NE (defined as Apgar score

Of 6550 births, 10.2% had an adverse perinatal outcome: 3.8% stillbirths, 0.6% neonatal deaths and 5.7% IP-NE. Adverse outcomes were higher among neonates whose mothers had antepartum haemorrhage (31.3%) or prolonged/obstructed labour (27.2%) compared with those whose mothers had no complications. Emergency referral and CS did not change the association between IP-NE and obstetric risk, except in prolonged/obstructed labour. Without emergency CS, the predicted probability of IP-NE was 0.73 (95% CI 0.51 to 0.95); with CS, it decreased to 0.45 (95% CI 0.39 to 0.50).

Neonates born to mothers with obstetric complications had low healthy survival rates. Emergency referral and CS did not alter the risks of IP-NE in women with obstetric complications except for obstructed or prolonged labour, highlighting that these interventions may not be implemented with sufficient timeliness or quality, and/or that additional, more targeted strategies beyond referral and CS are needed to address IP-NE.