Abstract
INTRODUCTION: In hospital settings, patient isolation is used to limit transmission of certain pathogens (e.g. M. tuberculosis [TB], antibiotic-resistant bacteria and viruses causing respiratory and enteric infection). Data is lacking on utilization of paediatric isolation facilities in low-resource, TB-endemic settings. METHODS: Prospective weekday observation of 18 paediatric isolation rooms at Tygerberg Children's Hospital, Cape Town, South Africa, was conducted between 1 May 2014 and 31 October 2014 documenting: occupancy rate; indication for isolation; duration of isolation; application of transmission-based precautions and infection prevention (IPC) behaviour of personnel. Potential under-utilization of isolation rooms was determined by cross-referencing isolation room occupancy with laboratory isolates of antibiotic-resistant bacteria, M. tuberculosis and selected viral pathogens. RESULTS: Six percent (335/5906) of hospitalized children were isolated: 78 % (260/335) for IPC purposes. Most IPC-isolated patients had community-acquired infections (213/260; 82 %), including tuberculosis (130/260; 50 %) and suspected viral infections (75/260; 29 %). Children (median age 17 months [IQR 6-50]) spent 4 days (IQR 2-8) in isolation. Isolation occupancy was 66 % (2172/3294 occupied bed days), but varied significantly by month. Laboratory data identified an additional 135 patients warranting isolation with 2054 extra bed-days required. Forty patients with 171 patient days of inappropriate isolation were identified. During 1223 weekday visits to IPC-isolated patient rooms: alcohol-based handrub was available (89 %); transmission-based precautions were appropriately implemented (71 %); and personal protective equipment was provided (74 %). Of 358 observed interactions between paediatric staff and isolated patients, hand hygiene compliance was 65 % and adherence to transmission-based precautions was 58 %. CONCLUSION: Patients isolated for TB (under airborne precautions) accounted for more than half of all isolation episodes. Missed opportunities for patient isolation were common but could be reduced by implementation of syndromic isolation. Demand for isolation facilities was seasonal, with projected demand exceeding available isolation beds over winter months.