Study of recovery in patients hospitalised with community acquired pneumonia using Clinical Practice Research Datalink (CPRD), linked to Hospital Episode Statistics

Date of Approval: 
2018-07-23 00:00:00
Lay Summary: 
Community acquired pneumonia (CAP) is a common condition. Each year in the UK, over 100,000 adults are hospitalised because of CAP. Most (85 % - 90%) survive and are discharged after an average hospital stay of 7 days. Following discharge, patients report slow recovery lasting many weeks (~50% still have symptoms at 4 weeks). Some develop new heart problems. Many (>60%) reconsult their GPs after discharge, and a third receive more antibiotics. The frequency at which post-discharge complications and reconsultation occurs are poorly described. Who suffers most and why are not understood. We propose using data from the Clinical Practice Research Datalink linked to Hospital Episode Statistics to determine the: • frequency of reconsultation following a hospitalisation with CAP and the reasons thereof • frequency and type of antibiotic prescription at reconsultation • frequency of new heart and memory problems in the months following hospitalisation. We will investigate who is more likely to develop problems during recovery from CAP, what those problems are and why these problems develop. We will also explore the value of additional antibiotic use after hospital treatment. These findings will aid identification of strategies to improve the care of patients recovering from CAP.
Technical Summary: 
Objectives To determine the incidence and reasons for reconsultation following hospitalisation with CAP, including antibiotic usage at reconsultation. Methods Adults with a first episode of hospitalised CAP between July 2002- June 2017 as recorded in CPRD linked to HES based on ICD-10 codes (J12- J18) will be included. Data analysis Statistical analyses will be performed using Stata 15. Incidence of CAP and other diseases (cognitive decline and cardiac complications) following CAP will be estimated using the whole CPRD as the denominator population. Incidence rates per 100,000 person-years, adjusted incidence rate ratios and 95% confidence intervals will be described. The independent association between patient characteristics and rate of reconsultation (overall/ patients without co-morbidities /patients with underlying respiratory disease) will be calculated using a competing-risks regression with death and readmission as competing events; adjusted for age, gender, smoking, alcohol consumption, practice region, primary care consultation in the previous year, social deprivation, co-morbidities, vaccine status, length of hospital stay, and admission year. Causes of reconsultation will be divided into either respiratory or non-respiratory (cardiac symptoms and cognitive decline) symptoms. We will measure the number of antibiotic prescriptions at reconsultation and where possible, the type of antibiotics prescribed. Association of antibiotic prescription at reconsultation with further reconsultation episodes will also be analysed. Proportion of patients who received smoking cessation advice before and after the index pneumonia episode will be calculated. The rate of pneumonia recurrence (per 100 person-years) will be determined by smoking status. Effect of smoking on hospitalization for recurrence of pneumonia will be determined using competing-risks regression (death as a competing event), adjusted for variables determined using directed acyclic graph (DAG).
Health Outcomes to be Measured: 
Primary outcomes • Reconsultation rate following hospitalisation for CAP, stratified by time: within the first 7 days, 8- 14 days, 15- 30 days, 31- 60 days • Causes of reconsultation; respiratory versus non-respiratory (cardiac, cognitive impacts) • Antibiotic prescription rate at reconsultation Secondary outcomes • Types of antibiotics prescribed at reconsultation • Association of antibiotic prescription at reconsultation with further reconsultation episodes within 30 days • Assocation of reconsultation with underlying comorbid illnesses • Incidence of incipient cognitive decline and cardiac disease following CAP at 30 days, 90 days and 1 year.
Application Number: 

Tricia McKeever - Chief Investigator - University of Nottingham
Vadsala Baskaran - Corresponding Applicant - University of Nottingham
Christos Chalitsios - Collaborator - University of Nottingham
Wei Shen Lim - Collaborator - Nottingham University Hospitals

HES Accident and Emergency;HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Carstairs Index for 2011 Census (Excluding Northern Ireland);Practice Level Index of Multiple Deprivation