All GP surgeries in the catchment area of a teaching hospital referral centre (Southampton University Hospital Trust) were included. All direct access referrals for gastroscopy were screened to exclude those with sinister symptoms i.e. dysphagia, vomiting, anaemia, rapid weight loss or those with history of gastric surgery. Patients were consented at the point of recruitment. Trained medical endoscopists performed the gastroscopy procedure. Patients found to have peptic ulcer, tumour, severe oesophagitis (grade C and D), Barrett's oesophagus and anatomical abnormality were excluded. Patients included were those with mild gastro-oesophageal reflux disease (GORD – non-erosive or grade A and B oesophagitis, hiatus hernia), non-ulcer dyspepsia (NUD) (mild and moderate gastritis or duodenitis) and those with normal findings.
Baseline details of socio-demographic factors, education, self-reported height and weight, smoking, alcohol (current versus non-drinker) and ulcer healing drugs (UHD) used in the past 6 months were collected by interview of all patients presenting for elective gastroscopy at Southampton University Hospitals Trust for the period between May 2002 to May 2004. All patients also completed two validated questionnaires relating to the past 6 months: the Glasgow Dyspepsia severity scores (Gladys) and the Health Status Short Form 12 (SF-12) [6, 7]. After gastroscopy, endoscopists maintained their routine practice in giving verbal and written advice to patients and documented treatment recommendation to GPs in a formal report. Patients eligible for entry after endoscopy were randomised into intervention (GNP) and control (GP) groups, with a password protected, computer generated random number table. The endoscopists telephoned a separate office to obtain the follow-up status. The 'GNP' group was given one out-patient appointment. The 'GP' cohort was discharged and advised to see their GP.
In the nurse-led clinic, a full medical history was taken. The clinical management was structured, based on national and local guidelines, with reference to each patient's predominant symptoms. Patients were given counselling and lifestyle advice, supplemented with relevant locally devised leaflets i.e. reflux, non-ulcer dyspepsia, weight control, and an individualised treatment plan agreed with them. Further investigation such as the urea breath test, motility studies and barium meal were initiated if required, as per routine clinical practice. To ensure practice consistency and reproducibility, 'history taking' and 'lifestyle advice' proformas were devised and used.
A researcher blinded to the patients' study status and diagnosis, interviewed all participants by telephone, at a pre-arranged time suitable to the patient, six months after randomisation. Data collected were Gladys dyspepsia score, SF-12 score, self-reported UHD used and weight and a patient satisfaction questionnaire.
Drug use and cost
The use of UHDs for the six months before (baseline) and for 6 months after trial entry were summed and averaged according to class: Proton pump inhibitors (PPI) and Histamine2 receptor antagonists (H2RA) and strength (half and full-dose PPI). Half-dose PPI and H2RAs were grouped, as they were equivalent in costs. Drug use by month, based on quantity of tablets, was therefore grouped: 'full-dose PPI' vs 'half-dose PPI and H2RA' vs 'no treatment'. Twice daily full-dose PPI was counted as 2 months and alternate days use counted as half a month. On demand therapy was recorded according to number of tablet/s taken per week and multiplied by 4 to give a monthly quantity. UHD prices were taken for each specific UHD from Monthly Index of Medical Speciality (MIMS 2006) and Drug Tariff (2006).
A sample size calculation indicated that a minimum of 186 patients, 93 in each group, would be needed to detect a 1 point of improvement on the Gladys dyspepsia scale with 80% power at the 5% significance level. This calculation was based on a standard deviation of 2.85 that came from a pilot study of 30 cases presenting for gastroscopy in which the Gladys scores were obtained before the procedure.
The two groups were compared by the change from baseline to month 6 in the key outcome variables – Gladys score, SF12 and overall UHDs cost, adjusted for baseline values by including the baseline levels of the outcome in the ANOVA as a covariate; p < 0.05 was taken as being significant. Intention to treat analysis was undertaken by assuming the 15 patients with no 6 month follow-up data did not change from baseline. The component scales of SF12 and Gladys were analysed with p values calculated using the Mann-Whitney or Chi-square test where appropriate; a p value of < 0.01 was taken because of multiple testing.
Local ethics committee approval was obtained (reference no. 050-02) and the study registered with the Southampton University Hospitals NHS Trust Research Department.