47 The overdiagnosis of chest sepsis in children: a quality improvement project

Abstract

Objectives The aim of this project was to test our clinical impression that a significant number of children and young people are diagnosed with and treated for ‘chest sepsis’ unnecessarily. We also wanted to investigate the hypothesis that children received unecessary investigations, including Chest X-rays (CXRs) and measurement of acute phase reactants such as C-Reactive Protein (CRP), which are not recommended by British Thoracic Society (BTS) guidelines and would have been treated inappropriately with intravenous (IV) antibiotics. Our main objectives were thus: To design a Quality Improvement Project to investigate whether we overdiagnose chest sepsis, over investigate children presenting with possible signs and symptoms of chest sepsis and over use IV antibiotics To evaluate adherence to BTS and NICE sepsis guidelines To suggest measures to implement to prevent over investigation and unnecessary use of antibiotics in children with chest sepsis. Method An audit was undertaken within the Paediatric Department in an NHS District General Hospital setting in the United Kingdom. Data was collected retrospectively from May 2017 to January 2018. The inclusion criteria were any child >10 days old admitted to the ward and treated with antibiotics for suspected lower respiratory tract infection (LRTI). Patients with underlying chronic lung disorders (except asthma/wheeze), cardiac conditions, immunodeficiency or those who are immunosuppressed were excluded. Patients were categorised by age ( 12) and then stratified according to sepsis risk, as per NICE and BTS guidelines. The data were then analysed assessing adherence to these guidelines during clinical assessment and management following presentation with signs and symptoms of a LRTI. Results • Data from 156 patients were collected: 132 aged under 5, 21 aged 5–11 and 3 aged 12–17. • 10% were low, 21% moderate and 69% high risk for sepsis on presentation. • 36 patients were noted to have low oxygen saturations on admission, of which 64% had co-existing wheeze. • 66% of patients, of which 73% were high risk and 27% were low or moderate risk for sepsis were treated with IV antibiotics, most commonly Ceftriaxone. • 33% were treated with oral antibiotics, most commonly Co-amoxiclav. Only 9.8% of this group were treated with oral Amoxicillin as per BTS guidelines. • The majority (88%) had a CXR. Half were reported as normal, of which 71% were treated with IV antibiotics. 78% of those treated with oral antibiotics had a CXR. • 78% of patients had blood tests, which all included a CRP, which is not recommended to be routinely measured. Conclusions First phase audit results support the hypothesis that children presenting with signs and symptoms of a possible LRTI are overdiagnosed and over investigated. 61% of children were under the age of two and treated with antibiotics even though BTS guidelines suggest this age group is much less likely to have a bacterial infection. Over half of those with low oxygen saturations had co-existing wheeze, suggesting over-treatment of patients with wheeze for LRTI. 88% of our patients had CXRs, which is contrary to BTS guidelines, that suggest routine CXRs are unnecessary. 71% of children reported to have a normal CXR received IV antibiotics. We plan to implement a change of practice within the Paediatric department to prevent unnecessary investigations, notably CXRs and blood tests and to reduce unnecessary antibiotic use that comes with long term risks for the wider population. Following this we plan to re-audit.