Influencing referral practice using feedback of adherence to NICE guidelines: a quality improvement report for dyspepsia

Problem: Rising demand and increasing waiting times for upper gastrointestinal endoscopy (gastroscopy). Design: Quality improvement study with pre- and post-intervention data collection. Setting: Three endoscopy units in two hospital trusts (Singleton, Morriston and Baglan Hospitals endoscopy units), UK. Key measures for improvement: Number of gastroscopy requests from general practitioners (GPs) and hospital doctors; their adherence to dyspepsia referral guidelines and the referral-to-procedure interval for upper gastroscopy. Data collected for six months before and for five months after the intervention. Strategy for change: Referrals were assessed against the National Institute for Health and Clinical Excellence (NICE) guidelines for the management of dyspepsia by two part-time GPs and feedback sent to clinicians where requests did not adhere to the referrals criteria. Effects of change: Adherence to guideline criteria increased significantly among GPs after the intervention (from 55% to 75%). There was no similar effect for hospital doctors, although their adherence rate (70%) was at a higher level than that of GPs before the intervention. The number of gastroscopy referrals for dyspepsia declined after the intervention, particularly from hospital doctors where a drop of 31% was observed, from 26.6 to 18.4 referrals per week. With the inclusion of seasonal effects, an estimated drop of 3.2 referrals per week from general practice was not significant (p = 0.065) while an estimated drop of 10.0 referrals per week for hospital doctors was very significant (p<0.001). Lessons learnt: Referral assessment can be successfully introduced and shows promise as a way of improving the quality of referrals and reducing demand. Hospital clinicians are more resistant than GPs to referral assessment but nevertheless responded to the feedback by reducing their endoscopy gastroscopy requests. Most such referrals are generated in hospitals rather than in primary care: this finding has important implications for demand management.