G594 Tackling high mortality in a kenyan district hospital

Abstract

Context Completed in a Kenyan hospital paediatric Emergency Department. Clinical officers, Nurses and administration team involved. Problem Stabilisation of paediatric admissions High mortality levels within first 24 h of admission. Often no identifiable cause of death from note keeping. Lack of documentation of stabilisation procedures. Clinically unstable children on ward Assessment of problem and analysis of its causes Liasion with the clinical staff identified lack of full stabilisation of patients prior to ward transfer. Assessing patient files revealed poor note keeping of stabilisation/ lack of early stabilisation interventions. Intervention Five point checklist designed to standardise early care for all paediatric admissions. Filled out by nurses. The 5 points covered were *Airway, Breathing and Circulation stable *Check glucose if not alert *Take haemoglobin and cross match samples if pale *Medications written up regulalry on drug chart *First dose of medications given The checklist informed staff that each of the points must be completed before the child was to be transferred to the paediatric ward and what to do if these were not performed. Patient demographics, staff details included. Study design Retrospective analysis of mortality cases over a 3 month period. End point examined was change in mortality rates within first 24 h of admission in intervention and non intervention populations. Strategy for change 'Safety checklist' designed and trialled by nursing team in the Emergency Department. This was refined and expanded to cover 5 key clinical points. Several meetings with nursing staff were undertaken. Nurses designated to complete the checklist for new admissions. Local consultant paediatrician agreed to trial of checklist in notes. From September 2014 administration staff were able to put the checklist in every child's file prior to Emergency Department clerking. Results of the study have been discussed with the teams involved and both nurses and clinical officers in the emergency department have agreed to continue working with the checklist. Further data from the hospital has been requested to monitor for ongoing change. Measurement of improvement All available mortality files over a three month period July-September 2014 were reviewed. Groups with a checklist and without were compared at different time intervals from admission. All availabel notes of paediatric admissions to the hospital during the same period were reviewed to identify the uptake of the checklist prior to it being put in to all medical files. A 24% reduction in mortality within the first 24 h was observed in the checklist group compared with the non-checklist group. Effects of changes Initial results reveal that a lower percentage of deaths were in the first 24 h of admission in the checklist versus non checklist group. Further data is required to allow month on month comparisons now that every child is receiving a checklist in their notes to see if there is a direct effect on total mortality figures. Objectively fewer numbers of unstable patients seen on the ward. There were initial difficulties with getting people to fill in the checklist and insert it in to the patient's file. This was much improved by getting the checklist put in the patient notes by the clinical records team before the child was admitted. Nurses team nominated for a Kenyan award for their using checklist to improve clinical governance measures. Lessons learnt To liaise early with nursing team regarding additions to improvement project. To get one group of staff to take responsibility for checklist completion as avoids confusion of roles. To get records department to include checklist in admission clerking sheets to allow ease of use and continuity of provision of checklist. Message for others Simple checklist intervention can have a substantial influence on child mortality within the first 24 h of admission and lead to better level of standardisation of care for acutely unwell children. (Figure Presented).