Results

A total of 104 H pylori positive patients were recruited (59 men; mean age 32 years, range 18–45 years). Seventy patients were smokers and seven reported an alcohol intake in excess of 21 units per week for men and 14 units per week for women. Seventy three patients had received antisecretory therapy (H₂ receptor antagonists and/or proton pump inhibitors) in the previous six months before attendance, 71 of these in the last four weeks. Six patients had previously received eradication therapy forH pylori more than four weeks before recruitment.

Fifty two patients were randomised to empirical eradication and 52 to OGD. There were no significant differences in baseline characteristics between the two groups. Tables 1 and 2, respectively show a comparison of baseline personality traits and dyspepsia scores; there were no significant differences between the two groups. Quality of life at baseline showed a significant difference in one of the eight parameters measured between the two groups: physical function was significantly better in the empirical eradication group than in the OGD group (85.42 (2.64) versus 75.10 (3.71), p<0.05; table3).

Three patients did not attend subsequent review (two in the empirical eradication group and one in the OGD group). In the OGD group 24/51 (47%) had PUD, 4/51 (8%) had PUD/GORD, 2/51 (4%) had GORD, and 21/51 (41%) had NUD. Of the 52 randomised to empirical eradication therapy, 14 (27%) subsequently proceeded to OGD because of no improvement in their dyspepsia score. Of these, eight patients were breath test positive prior to OGD; endoscopy showed a duodenal ulcer in two, and six were normal. Six patients were breath test negative prior to OGD and endoscopy showed duodenal scarring in two, duodenitis and oesophagitis in one, and three were normal. Culture and histology agreed with breath test results in all of these patients.

Eighty patients received first line eradication therapy (52 in the empirical eradication group and 28 in the OGD group). Thirty four patients noted no adverse effects (20, empirical eradication; 14, OGD). Four discontinued treatment due to adverse effects (three, empirical eradication; one, OGD). Eradication rates with this regime at 12 months were 22/28 (79%) in the OGD group and 39/50 (78%) in the empirical eradication group (two patients did not attend after receiving empirical eradication therapy). Seventeen patients received second line eradication therapy (six in the OGD group and 11 in the empirical eradication group); one patient did not attend for review following second line therapy. This regime was not successful in any of this group.

Table 2 shows the change in dyspepsia scores over the various time points of the study. Dyspepsia scores at the 12 month review showed significant improvements in both groups when compared with baseline (p<0.0001). Dyspepsia scores were significantly better in the empirical eradication group (3.37 (0.54)) compared with the OGD group (5.08 (0.62); p<0.05). Table 4 shows the change in the individual parameters of the Glasgow dyspepsia severity score measured at 12 months compared with baseline.

Quality of life at the 12 month review showed significant improvements in both groups when compared with baseline (table 3). In the empirical eradication group five of the eight parameters were improved (physical function, p=0.004; bodily pain, p=0.0001; vitality, p=0.002; social function, p=0.003; mental health, p=0.02). In the OGD group four of the eight parameters were improved (physical function, p=0.009; bodily pain, p=0.001; general health, p=0.03; social function, p=0.02). Quality of life score was significantly improved in the empirical eradication group compared with the OGD group at the 12 month review for the parameter of physical role functioning (p<0.05).

At 12 months, 21/49 (43%) of the empirical eradication group and 15/50 (30%) of the OGD group were asymptomatic. A total of 37/49 (76%) of the empirical eradication group and 32/51 (63%) of the OGD group were not taking H₂ antagonist or proton pump inhibitor therapy.

Discussion

Empirical eradication therapy for young H pylori positive dyspeptic patients is an attractive alternative strategy to conventional therapy after endoscopic diagnosis.4 Effective eradication of infection would appropriately treat patients with PUD and possibly benefit some patients with NUD without the need for further investigation. It has been suggested that this may be a cost effective method of managing the young dyspeptic patient by avoiding the expense of OGD. The young dyspeptic patient has been targeted for this strategy because of the low risk of carcinoma in this age group and hence potential suitability for non-invasive testing.1 ,11 No study to date has addressed this issue in a prospective randomised manner.5This study is the first to report on this management strategy as a direct comparison of conventional therapy.

A limitation of this study is the assumption that patients randomised to empirical eradication therapy had similar pathology to those randomised to OGD. In order to minimise any differences between the two groups the randomisation procedure was stratified to take into account such variables as sex, smoking, and alcohol intake. These variables were taken into account as PUD has been more commonly reported in men, smokers, and conditions associated with an excess alcohol intake—that is, cirrhosis and pancreatitis.12-14 These factors may also contribute to dyspeptic symptoms.15 The groups were well matched for these variables. The baseline dyspepsia scores, quality of life (except for the parameter of physical function, p<0.05), and personality traits were similar in the two groups.

In the selection of patients for entry into this study, patients with reflux symptoms were excluded in an attempt to exclude patients with GORD, as H pylori is not known to play an aetiological role in GORD.16 While six (12%) in the OGD group had GORD, only two of these had GORD in isolation, the other four having a combination of PUD and GORD. This highlights some of the difficulties in making the clinical diagnosis of GORD, as localised epigastric pain without classic retrosternal radiation may be the presenting symptom and some patients may have combined pathology.

In the empirical eradication group, 14/52 (27%) patients subsequently proceeded to OGD. Of these, 6/13 were breath test negative confirming effective H pylori eradication, and only one of these H pylori negative patients had evidence of ongoing mucosal disease—that is, duodenitis and oesophagitis. Follow up of patients by symptom assessment is an important aspect of any non-invasive policy and ensured that while the repeat breath test was negative in this patient, the patient proceeded to OGD because dyspepsia failed to improve.

Our first line eradication regime produced similar eradication rates in both groups (78%) which is less than that found by other centres with this regime.17 ,18 Recent guidelines have suggested that acceptable eradication regimes should produce eradication rates of greater than 80% on an intention to treat basis.4Disappointingly second line therapy did not eradicate any patients. While treatment in other centres with a similar regime has produced eradication rates of 79% as first line therapy,19 our results may have been influenced by use of tinidazole in first line therapy and potential development of nitroimidazole resistance and suboptimal dosage of metronidazole. One of the difficulties with use of a nitroimidazole and clarithromycin in first line therapy is the choice of second line therapy if this fails. Quadruple therapy with a proton pump inhibitor and bismuth based triple therapy as advised by the Maastricht consensus report may be an alternative.4

We can only postulate as to why the empirical eradication group had improved dyspepsia and quality of life scores compared with the OGD group. The empirical eradication group would have comprised patients with PUD and NUD; H pylori positive patients with NUD may benefit from eradication if follow up is prolonged.20 ,21 O’Morain and Gilvarry showed symptom benefit in patients with NUD at one year follow up that had not been seen at four weeks in patients in whom H pylori had been eradicated.20 More recently the UK MRC trial, a double blind, placebo controlled study, has shown significant resolution of dyspeptic symptoms at one year follow up in patients with NUD who received eradication therapy.21 The reassurance of a normal OGD and potential benefits on symptoms and quality of life may be overestimated. Lucock et al have shown that medical reassurance following a normal OGD reduced worry about health and illness belief22; however, this was only in the very short term in patients with high anxiety levels. Our two groups showed no significant differences in personality profiles although their anxiety levels were not assessed to the same extent as in this study. Knowledge of persistingH pylori positive status in the normal OGD patients may also have affected outcome, as these patients may have been less reassured by a normal OGD which may have subsequently affected their perception of dyspepsia and quality of life.

Over the 12 month period 73% of OGDs were avoided in the empirical eradication group by implementation of this strategy. While the majority of patients in both groups continued to complain of some dyspepsia at 12 months, 76% of the empirical eradication group compared with 63% of the OGD group were not taking any antisecretory therapy. The failure of complete resolution of symptoms is not surprising as several studies have shown that while eradication therapy has been successful, dyspeptic symptoms may persist.23 ,24This study is the first report of a randomised, prospective study examining empirical eradication therapy versus treatment based on endoscopic diagnosis in the young dyspeptic patient and supports the use of empirical eradication therapy. Our results show that in the short term, OGDs can be avoided. Patients whose H pylori are empirically eradicated have improved dyspepsia scores and quality of life when compared with patients managed by endoscopy.

The true impact of this strategy in dyspepsia management requires a longer term follow up period of assessment. Our results apply to patients seen at a secondary referral centre and should not be extrapolated to primary care. A similar strategy would need to be assessed in this setting before implementation was considered.