Between-practice variation in chronic obstructive pulmonary disease diagnosis guideline compliance: an observational study.

Abstract

Early chronic obstructive pulmonary disease (COPD) diagnosis is vital, but little is known about compliance with relevant diagnostic guidelines or variation in primary care. To quantify between-practice variations in guideline compliance and over time. An observational study in English primary care. The Clinical Practice Research Datalink was used to assess the use of four pre-diagnostic investigations (spirometry, chest X-ray, full blood count [FBC], and body mass index [BMI]) by GP practices for patients with COPD recorded first in primary care, in three time periods: 2006-2007 (cohort 1), 2016-2017 (cohort 2), and March-August 2020 (cohort 3). Multilevel logistic regression models quantified the non-random variation between GP practices in spirometry around diagnosis. Funnel plots counted the proportion of outliers. Cohort totals were 31 676 (cohort 1), 37 393 (cohort 2), and 3368 (cohort 3). Overall, the mean age was 68.3 years (standard deviation 12.0), with 46.1% female. The use of pre-diagnosis spirometry improved a little in cohort 2 (74.2%) on cohort 1 (62.8%) but fell back for the COVID-19-era group (61.1%). In contrast, chest X-ray, FBC, and BMI all improved after cohort 1 and were maintained for the COVID-19 cohort; almost all patients received one of these investigations. The proportion receiving all four investigations before diagnosis jumped from 26.6% in cohort 1 to 46.7% in cohort 2 and was maintained in cohort 3 (43.0%). Modelling and funnel plots showed considerable non-random variation in spirometry use by practice, although with some improvement since cohort 1. The recording of spirometry and chest X-rays warrants further and consistent improvement in the context of COPD care.