Cohort study of access block trends in a public, regional hospital Emergency Centre in South Africa

Background: Access block is one of the most serious and frequently encountered system problems in the specialty of Emergency Medicine and, although well described internationally, within the South African setting limited research has been performed on the phenomenon. Objectives: To describe the total access block of adult patients and per specialty, the monthly variation of the degree of access block and its correlation to the monthly Emergency Centre head count, hospital bed occupancy rate and monthly patient days and the effect of any interventions made during the study period in an Emergency Centre (EC) at a regional hospital in Cape Town, South Africa. Methods: The study was a retrospective descriptive study of a pre-existing database of EC access block, monthly EC head count, hospital bed occupancy and hospital patient days from April 2015 to March 2017. Results: A total of 700 discrete data sets were analysed. The study period of 24 months was divided into two 12-month periods to allow year-on-year comparison – April 2015 to March 2016 (Year 1) and April 2016 to March 2017 (Year 2). The mean access block (SD) for the total study period was 109% (17.7) ranging from 69% to 139%, with a Mean Access Block of 106% for Year 1 and 113% for Year 2. Patients of the Internal Medicine Department made up the largest proportion of boarders in each month of the study period. The comparison of the month-to-month variation of MAB over the study period to the variation of the EC monthly head count showed a weak correlative trend between the two variables with regards to large month-on-month changes, but with no absolute correlation for discrete monthly comparisons (r(22) = .14, p = .53). Across the study period there was a direct correlation between BOR and PD (r(22) = .90, p< 0.001), and neither showed a significant correlation with the Mean Access Block. The Mean Access Block was 110% prior to the appointment of a dedicated Bed Manager in August 2016 and increased to 115% afterwards. The Bed Occupancy Rate also increased from 89% prior to 92% after the appointment. Conclusions: The severity of access block was demonstrated using a basic system of recording Mean Access Block for a 24-month period and demonstrated that, on average, all available space was occupied by boarding patients. Whilst the hospital's patient ‘flow' system should address increased access block, the systems employed mostly failed. Although it had no direct positive effect on the Mean Access Block, a dedicated bed manager appeared to make a positive change in how the inpatient system compensated for access block