Documented Obstetric Haemorrhage in a General Intensive Care Unit in Ghana: Outcomes and the Unadjusted Association Between Recorded GCS and Mortality

Background: Obstetric haemorrhage remains an important cause of maternal death and severe morbidity. Outcomes depend on timely recognition, resuscitation, referral, access to blood and essential treatment, and critical care. This study described the clinical characteristics, documented haemorrhage diagnoses, and ICU outcomes of women admitted with obstetric haemorrhage to the main intensive care unit (ICU) of Komfo Anokye Teaching Hospital (KATH), Kumasi, Ghana. Methods: This retrospective cohort analysis included 79 women with an obstetric haemorrhage-related diagnosis documented in ICU records, drawn from 344 obstetric ICU admissions between 17 July 2022 and 16 July 2023. Demographic and clinical characteristics, ICU treatments, and outcomes were summarized. Survivors and non-survivors were compared using the Wilcoxon rank-sum test and the chi-square or Fisher exact test, as appropriate. Only unadjusted comparisons are reported because the patient-level dataset required for a prespecified admission-only model was not available for analysis. Results: Among 344 obstetric admissions to KATH’s main ICU, 79 women (23.0%) had documented obstetric haemorrhage and were included. Postpartum haemorrhage was the most common diagnosis (32/79, 40.5%), followed by abruptio placentae (18/79, 22.8%) and uterine rupture (12/79, 15.2%). Forty-four women died in the ICU (55.7%). The GCS variable, labelled “admission GCS” in the source dataset, was lower among non-survivors than survivors (median 8 vs. 15; p < 0.001), as was diastolic blood pressure (71.5 vs. 86 mmHg; p = 0.010). Mechanical ventilation was more frequent among non-survivors (70.5% vs. 40.0%; p = 0.013), while hospital stay was shorter (4 vs. 9 days; p < 0.001). GCS measurement timing could not be verified; treatment and length-of-stay differences describe post-admission care and course. Conclusion: Among women admitted to a general mixed ICU with documented obstetric haemorrhage, ICU mortality was high and postpartum haemorrhage was the most common diagnosis. Lower recorded GCS among non-survivors was an unadjusted descriptive finding; the source dataset’s GCS timing could not be confirmed, and no adjusted prognostic effect can be inferred. Prospective studies with verified admission measures, explicit case criteria, and treatment timing are needed.