Helicobacter Pylori and Peptic Ulcer

A systematic review of effectiveness and an overview of the economic benefits of implementing what is known to be effective

R A Moore MA DPhil


Pain Research
The Churchill
Headington
Oxford

December 1994


Executive Summary
Helicobacter pylori - background
Introduction
H pylori
Evidence implicating H pylori in gastritis and peptic ulcer
H pylori and ulcers
How does H pylori cause gastritis and ulcers?
Methods of diagnosing H pylori infection
Invasive methods
Culture
Histology
CLO test
Non-invasive testing
Breath testing
Antibody measurement
Quantitating antibody tests
Antibody testing after H pylori eradication
Prevalence & incidence of peptic ulcer
Prevalence
UK prevalence
Endoscopy studies
Dyspepsia in seropositive and seronegative populations
Incidence
peptic ulcer bleeding
Helicobacter pylori eradication and peptic ulcer
Antibiotics and H pylori eradication
H pylori eradication and relapse rates
Effect of treatment on rebleeding rates
Effect of eradication therapy on ulcer healing

H pylori and dyspepsia
Systematic review
Conclusions

H pylori and gastric cancer
Background
Gastric cancer and H pylori
Epidemiological considerations

Clinical guidelines on H pylori
NIH Consensus Conference
Endoscopy and H pylori testing
Economic consequences of H pylori testing and eradication
Background data
Epidemiology
Prescribing information
Treatment costs
Endoscopy costs
Calculation of economic impact
Patients on acid suppressing maintenance therapy
Newly diagnosed patients with peptic ulcer disease
Dyspepsia
Economic effects on endoscopy

H pylori - demand and place for testing
NPT for H pylori - impact on a Health Authority
H pylori - research agenda

Bibliography
Acknowledgements


Executive Summary


This report "Helicobacter pylori and peptic ulcer" is a systematic review of the effectiveness of curing infection of H pylori and an overview of the possible economic benefits of implementing what is known to be effective.

This executive summary highlights the main conclusions of the report.
  1. The background evidence of the association between H pylori and peptic ulcer is reviewed, together with methods of diagnosing H pylori infection. Serological testing is considered to be simple, cheap and effective. At least one authority considers serological tests for H pylori to be the "gold standard" against which all others should be measured.
  2. Peptic ulcers are found in 25% of dyspeptic patients whose blood tests positive for H pylori, compared with only 3% of similar patients who test negative. Combining data from three separate studies shows that rates of gastric and duodenal erosions, and gastric cancer, are also higher in patients who test positive for H pylori.
    This indicates that testing blood for H pylori can be a useful way of determining which patients would benefit from conventional conservative therapy (acid-suppressing medicines) and those who would benefit from curing H pylori infection.
  3. The data from randomised controlled trials of H pylori eradication therapy (triple therapy plus acid suppressing medicine or amoxycillin plus omeprazole) has been subjected to a meta-analysis for the dichotomous outcomes of H pylori eradication, ulcer healing at 6-10 weeks and ulcers remaining cured at one year. Using single-arm comparisons with acid-suppressing medicine and numbers-needed-to-treat methods, the results were:-
    H pylori eradication was highly efficient using eradication therapy. Ten patients were cured of their infection for every eleven or twelve treated.
    Ulcer healing at 6-10 weeks was significantly better with eradication therapy than with acid suppressing drugs alone. Compared with acid suppressing treatment, curing H pylori infection resulted in one extra ulcer cured for every five patients treated.
    Significantly more ulcers were cured at one year with eradication therapy compared with short term use of acid suppressing medicine. Ten extra ulcers were healed at one year by curing H pylori infection for every 18 patients treated.
  4. There is no conclusive data concerning benefits of H pylori eradication in non-ulcer dyspepsia. There is no evidence that H pylori eradication is beneficial in preventing development of gastric cancer.
  5. About 0.8% of the population of the UK is presently taking acid suppressing medicines for an average of more than seven years, and about 0.5% of the population is taking these medicines with diagnosed peptic ulcers; they would benefit now from H pylori eradication. This would mean that most would not have to continue taking long-term acid suppression medicine. It would save some £90 - £106 million for the NHS, perhaps as much as £2 million per million of population. Such a course of action has been recommended by the American Medical Association.
  6. New patients would benefit from eradication therapy. Diagnostic protocols involving testing by serology and endoscopy need to be developed. Eradicating H pylori rather than using long term acid suppression therapy is likely to save additional sums and prevent the inexorable rise in acid suppressing medicine costs - perhaps to the extent of £40 million per year for the NHS (almost £1 million per million of population). Such a course of action has been recommended by the American Medical Association.
  7. Use of H pylori testing as a triage before endoscopy has been proven to reduce the numbers of endoscopies performed by as much as 25%, and save additional sums of about £10 million per year.
  8. Near patient testing for H pylori has been suggested as a way for Health Authorities to implement improved care for patients with peptic ulcers. The ways in which such a policy might impact on a Health Authority of 650,000 population has been modelled using two options to produce a range of effect.
    Using NPT to identify quickly all those patients already diagnosed with peptic ulcer and who are H pylori positive, curing the H pylori infection and stopping long-term acid suppressing medicine would save the Health Authority up to £800,000 a year.
    Using NPT to identify H pylori infections in patients newly presenting with dyspepsia, and using H pylori eradication therapy in those who tested positive could reduce endoscopies for ulcer diagnosis from 1,000 a year to 400 a year, and reduce the overall number of endoscopies performed by 25%. This would save another £250,000 a year, as well as reducing significantly waiting times for endoscopy.

Helicobacter pylori - background

Introduction


Our knowledge about the cause of peptic ulcer disease, of the aetiology of gastric cancer, and perhaps of some forms of dypepsia is in a state of revolution. The discovery of the infective organism Helicobacter pylori and its involvement in these diseases has begun to change our views on how to approach diagnosis and treatment. As yet the story is still unfolding. Even so, enough is now known to consider getting some of the fruits of a decade of research into practice - to cure peptic ulcer disease with antimicrobial therapy rather than put patients on expensive long term acid-suppressing therapies.

It is not intended here to give a definitive review of all the information that exists on Helicobacter pylori and the associations with peptic ulcer disease. Rather, this section is designed to provide sufficient background information to familiarise the reader with the subject. There are a number of excellent reviews which describe various aspects of peptic ulcer disease, dyspepsia and H pylori (Blaser, 1990; Soll, 1990; Petereson, 1991; Graham, 1993; Pope, 1994).

As usual, an excellent overview comes from the non-science literature. The Economist of March 5 1994 produced a remarkably concise review of the H pylori story with some punchy comment. It refers to an economic analysis by Amnon Sonnenberg of the Medical College of Wisconsin, demonstrating that a 15-year treatment of ulcers with current drugs would cost $11,500 per patient, to which could be added at least $17,500 if surgery was needed. Eradicating H pylori would cost less than $1,000. Says Sonnenberg "From an economic perspective, antibiotic therapy is the treatment of choice"; says the Economist "It also makes people healthier".

H pylori


The presence of spiral-shaped micro-organisms in stomach mucosa was described almost 100 years ago (quoted by Madan et al, 1988). Their presence was not really taken seriously until the late 1970s, when John Warren, a pathologist in Perth, Western Australia, noted the appearance of spiral bacteria overlaying gastric mucosa, and chiefly over inflamed tissue. Warren and Barry Marshall cultured these organisms in 1982 from 11 patients with gastritis; the story of the early part of the discovery of H pylori is related by Marshall and is worth reading because of that (Marshall, 1989).

Originally called Campylobacter pyloridis, the name was changed to Campylobacter pylori and then later to Helicobacter pylori as specific morphologic, structural and genetic features indicated that it should be placed in a new genus. The organism is a motile, gram-negative, curved rod which is oxidase, catalase and urease positive.

Marshall and Warren were able to demonstrate a strong association between the presence of H pylori and the finding of inflammation on gastric biopsy (Marshall & Warren, 1984). People who did not have gastritis did not have the organism, a finding confirmed in a number of studies (see below). Marshall elegantly fulfilled Koch's postulates for the role of H. pylori in antral gastritis with self administration of H. pylori, and also showed that it could be cured by use of antibiotics and bismuth salts.

Evidence implicating H pylori in gastritis and peptic ulcer


There are a number of pieces of evidence:-
  • Voluntary ingestion of H pylori resulted in chronic gastritis
  • Experimental animal challenge simulates human infection and gastritis.
  • Antimicrobial therapy that clears infection clears gastritis.
  • H pylori only overlies gastric epithelium.
  • H pylori is associated only with certain types of gastroduodenal inflammations (notably type B gastritis almost universally present in duodenal ulcers), not all types.
  • There is a systemic immune response to H pylori infection.
  • H pylori antibodies diminish with effective antimicrobial therapy.
  • Clearance of gastric inflammation with bismuth salts (bismuth has antimicrobial properties).
  • Association of H pylori with epidemic gastritis and hypochlorhydria.

Another compelling piece of evidence comes from the epidemiology of gastritis and H pylori infection. The pattern of H pylori acquisition with age is identical to that of gastritis. Serological tests for H pylori infection (circulating IgG and IgA antibodies measured by immunological methods) show that H pylori infection is low in children, but rises dramatically in the fifth and subsequent decades, and that more than half of the population over 50 years is infected. This seems to be due to a continuous risk of infection (Veldhuyzen van Zanten et al, 1994). The prevalence of chronic gastritis with age is given alongside the serology results below:-



H. pylori and ulcers

A further early finding was that infection with H pylori was almost 100% in those patients with duodenal ulcers, and approached that figure in patients with gastric ulcers. Collected data from studies published up to 1990 are shown below (from Blaser, 1990).



People who have developed peptic ulcer are much more likely to have had H pylori infection up to 10 years previously (Nomura et al, 1994), and studies of twins in Sweden have shown that there is a considerable genetic component in susceptibility to infection (Malaty et al, 1994).

How does H. pylori cause gastritis and ulcers?

H pylori are motile, even in the highly viscous mucus layer in which they live. This may allow the organisms to evade both gastric motility and peristalsis, and to some extent gastric acidity also. The organism seems specifically to overlay gastric-type epithelial cells, whether in the stomach or metaplastic in the duodenum; they will not overlay absorptive-type duodenal cells, even when these are metaplastic in the stomach. Although it is motile, it also may adhere to the gastric mucosa through specific adhesion mechanisms.

The secretion of large amounts of urease results in any urea in the environment being converted into ammonia - with the result that the intense acidity of the stomach may be ameliorated in the microenvironment surrounding the bacterium. The Michaelis constant of H pylori urease is 0.4 mM, making it one of the lowest known for this enzyme, and allowing significant conversion of urea to ammonia at very low urea concentrations and, therefore, to work efficiently in the stomach.

About 50% of H pylori strains produce cytotoxins (Crabtree et al, 1991b), of which some have been specifically linked to active gastritis and peptic ulceration. The strains isolated from patients with the most severe disease tend to be more likely to secrete these cytotoxins than strains isolated from asymptomatic patients. These cytotoxins can cause local inflammation, though other secretions by the organism, such as proteases and phospholipases, can attack and damage mucosal cell membranes. Weakening the gastric-mucosal barrier permits back-diffusion of hydrogen ions resulting in further tissue injury, as well as causing local immune responses to the organism.

There is also evidence that H pylori infection is responsible for reducing the levels of ascorbic acid in the gastric juice; levels in infected patients were only 25% of those in non-infected subjects (Banerjee et al, 1994). Moreover, eradication of H pylori resulted in a large increase in gastric juice ascorbate. The reversible lowering of gastric juice ascorbate may predispose to gastric cancer and peptic ulceration (Banerjee et al, 1994).

It is entirely likely that assays could be made available which detected antibody response to the aggressive strains of H pylori or their associated damaging cytotoxins. Such a development would allow for a much more simple therapeutic approach - if a test for such an aggressive strain proved positive in symptomatic or asymptomatic individuals, eradication would be a sensible first course of action.

Methods of diagnosing H pylori infection



There are several methods which can be used to diagnose whether a patient is infected with H pylori. They differ in being invasive or non-invasive, simple or difficult, cheap or expensive. Though far from an exhaustive coverage, the main points of each method are given below.

Invasive methods

Culture


H pylori can be cultured only when a specimen containing the pathogen has been obtained, and in this case that means obtaining a biopsy specimen at endoscopy. Methods of culture vary, but in general they involve homogenising the biopsy specimen and culturing the homogenate on a variety of specialised agar plates at elevated temperatures for at least seven days.

Culture is generally regarded as the 'gold standard' for detecting a bacterium. For H pylori, however, the success of the technique depends on local technique and access to facilities, and can be regarded as being no more than 60 - 90% sensitive, though being 100% specific; the cost of each test is high (DeCross & Puera, 1992; Perez-Perez et al, 1988).

Histology


Histological examination of tissue biopsy samples (usually four, taken from different parts of the stomach lining) permits detection of the bacterium together with evaluation of tissue damage. Most infection can be detected with haematoxylin & eosin (H&E) stain of gastric tissue, but special stains like Giemsa can be used if H&E results are not conclusive.

The sensitivity of Wright-Giemsa and Brown-Hopps stains has been shown to be 100%, compared with H&E (Madan et al, 1988), though histological detection of the organism has generally been considered to have lower sensitivity at 80 - 95% with 100% specificity. The need for a number of biopsy specimens to be examined by experienced pathologists renders histology expensive, and it requires an invasive procedure (Table 3).

CLO test


The CLO test is for Campylobacter-like organisms. At the time of endoscopy and biopsy, a pinch of tissue is removed and placed in a test solution containing urea, a pH colour reagent and a bacteriostatic agent. The presence of urease from H pylori results in hydrolysis of neutral urea to alkaline ammonia, together with a pH change and a change in colour (usually from yellow to red).

In infected tissues the change occurs within about one hour, so that the results of the CLO test are often available while the patient is still at the endoscopy unit, meaning that therapy decisions can be made immediately. Though the test itself is inexpensive, it still requires an invasive procedure to obtain the sample. It has 90 - 95% sensitivity and high specificity, and the overall cost is moderate (Table 3).

Non-invasive testing

Breath testing


Urea breath testing is non-invasive, and a specific indicator of H pylori infection (Atherton & Spiller, 1994). The procedure requires a patient to attend a centre fasting, to eat a standard meal and then ingest urea labelled with C-14 or C-13. The presence of large amounts of H pylori urease results in the production of labelled carbon dioxide, which is absorbed into the blood and excreted in the breath. Samples of breath are collected before and some time after the ingestion of labelled urea and the amount of labelled carbon dioxide measured by mass spectrometry or radioactive counting.

While quite specific, few detection systems are readily available, and they do have a significant capital cost. Breath samples can be sent by post to measuring centres, though this involves a necessary delay before results are available. Breath testing is both sensitive (at least 95%) and specific (at least 98%) with a moderate cost.

Breath testing has the important advantage of being both non-invasive, and being able to confirm H pylori eradication about one month or so after treatment has ended, should that be necessary.

Antibody measurement


Patients infected with H pylori have immunoglobulin antibodies to the organism. Tests for the detection of antibodies to H pylori circulating in blood, or found in saliva, have excellent sensitivity and specificity of above 95% and are cheap and simple compared with invasive techniques (DeCross & Puera, 1992). They can give very quick results even within minutes of the first consultation, and are the only tests which are not likely to give false negative results in patients who have taken antibiotics, bismuth compounds or omeprazole in the recent past (NIH Consensus Conference, 1994).

Because there are different strains of H pylori, antigen for antibody manufacture is generally prepared by using preparations from several different strains. Antibody assays in blood have measured IgG and IgA antibodies which have been shown to be specific for H pylori and not other gram-negative organisms (Perez-Perez et al, 1988). Where both IgG and IgA assays have been compared with other testing methods, like culture and/or histology, IgG assays tend to have slightly higher sensitivity and specificity, and so anti-IgG methods tend to be favoured (Perez-Perez et al, 1988; Glassman et al, 1990; Crabtree et al, 1991a).

The commercially available assays are of two sorts: either microtitre-plate assays for use in a laboratory, or near-patient testing devices. Both types of assay usually have a cut-off value set with control sera so that they differentiate patients with H pylori infection from those who do not, rather than quantify the concentration of circulating anti-H pylori immunoglobulin. Laboratory based microtitre assays (Crabtree et al, 1991a) and near-patient testing devices (Pronovost et al, 1994) perform equally well compared with standard techniques.

Tests for antibodies to H pylori in saliva are equally effective as those for antibody in serum (Patel et al, 1994). In saliva IgG immunoglobulins must be measured, as measuring IgA antibodies does not distinguish positive from negative cases.

Antibody tests have high sensitivity and specificity when compared with other methods (Table 3). They also have the advantage of being non-invasive, and near-patient versions are available which can produce results from a small finger prick of blood within a few minutes. The costs of antibody tests (bedside or laboratory) are much lower than any other method.

At least one commentator has suggested that for H pylori infection, serology testing may be the "gold standard" (Blaser, 1990). Essentially, every person whose gastric biopsy can be shown to harbour H pylori has evidence of a systemic humoral immune response. Those in whom the organism cannot be detected by other methods have a low false positive rate by serology. However, gastritis can be a patchy phenomenon and histologic biopsy and culture assess only a small area of the stomach, while serology in essence assays the entire stomach. "False-positive" serology may well, in fact, reflect falsely negative biopsy results.


Quantitation of antibody tests


Knowing how much antibody was present in the blood has not generally been thought to be helpful. Very high concentrations have been found in infected patients who also have gastric cancer (Forman et al, 1991), and more recently Nomura and colleagues (1994) found that there was a higher degree of association between H pylori infection with higher levels of circulating
antibody. The extent of the elevation in this study was much lower than that seen by Forman and colleagues for gastric cancer.

Almost all studies have concentrated on demonstrating the presence of antibodies to H pylori as a means of diagnosis of infection. Systematic attempts to correlate levels of antibody to particular disease states have not been conducted. It is interesting to speculate that tests which discriminate actual circulating concentrations of antibodies to H pylori might increase significantly the diagnostic yield for both peptic ulcer disease and gastric cancer, thus making screening a viable option for these diseases.

Antibody testing after H pylori eradication


Titres of IgG and IgA in serum fall relatively slowly after H pylori has been successfully eradicated. Thus six weeks after eradication titres fall by about 20-30%, but by six months, about 97% of patients have titres reduced by 50% or more from pre-treatment levels (Kosunen et al, 1992). Because of this slow reduction in antibody titres after successful eradication, serology testing is not generally useful in confirming the success of eradication within a few weeks of treatment; urea breath tests and histology or culture after repeat endoscopy can make that confirmation, if needed. However, since H pylori reinfection or regrowth can take place place within the first six months after treatment, waiting at least six months before retesting may be a better option.

However, there is one report, albeit of a preliminary nature, which indicates that in saliva, antibody titres to particular H pylori membrane antigen fall much more rapidly (Clancy et al, 1993). In a small number of patients it was shown that 83% had falls in salivary antibody titres of more than 50% within one month of successful eradication therapy. As an indicator of successful treatment, this is similar to standard methods for monitoring elimination.

Prevalence & incidence of peptic ulcer

Prevalence


There are a number of problems in assessing the prevalence of peptic ulcer. In particular, many ulcers are asymptomatic and the accurate diagnosis of duodenal ulcers by radiographic methods is difficult. Mortality from peptic ulcers is low. Despite these difficulties, a number of methods have been used to measure the prevalence of peptic ulcer in the community; such methods have included post mortem studies, the use of the incidence of perforation as an index, the number of prescriptions for histamine receptor antagonists and more recently endoscopic surveys.

UK prevalence

In the UK, a number of studies (reviewed by Shearman, 1989) have indicated that the prevalence of peptic ulcers in the population is of the order of 6-13% for men and 2-5% for women between the ages of 15 and 64 years.

Endoscopy studies

These data generally date from surveys conducted in the 1950s and '60s, but more recent studies from elsewhere tend to confirm these estimates. An Italian study (Vaira et al, 1994) which looked at 1010 residents of Bologna who donated blood over the period 1990-91 showed that 42% had antibodies to H. pylori. Of the 422 seropositive donors, the first 180 were sent a postal questionnaire, and 128 consented to endoscopy, of whom 121 were confirmed to be still seropositive. The findings at endoscopy are given in Table 4.



The implications of these findings are an overall prevalence of peptic ulcer of 8.5% in the adult population of Bologna (blood donors were representative of the population as a whole). The prevalence of peptic ulcer was 21% in seropositive patients.

Dyspepsia and seropositive and seronegative populations

Vaira et al (1994) also looked at the prevalence of peptic ulcer in a population of 219 patients presenting with dyspepsia. The H pylori positive patients (55% of the total) had a distribution of endoscopy findings and prevalence of peptic ulceration almost identical to that of H pylori positive blood donors shown above. In the H pylori negative dyspeptic patients (45%), 51% had normal findings on endoscopy and only three (3%) had duodenal ulcers. Thus the prevalence of peptic ulcers in patients seropositive for H pylori was seven times greater than in those who were seronegative.

These results are similar to those found in a study from Leeds in 1991 (Table 5; Sobala et al, 1991). In a prospective study of 293 consecutive patients referred to a dyspepsia clinic that offered general practitioners open access to diagnostic endoscopy, the overall prevalence of peptic ulcer discovered by endoscopy was 14%; the prevalence in seropositive patients was 23% and in seronegative patients 2.5%. Moreover, the small number of malignancies found in this study were all seropositive. The distribution of diagnoses found at endoscopy in the seropositive and seronegative patients is shown below in Table 5.



Very similar results can be calculated from a study (Patel et al, 1994) where salivary antibody tests were used to detect the presence of H. pylori infection before endoscopy in 119 consecutive patients, with prevalence of peptic ulcer of 25% in H. pylori positive patients, and only 3.4% in H pylori negative patients.

These three studies (Vaira et al, 1994; Sobala et al, 1991; Patel et al, 1994) are generally similar in design, in that they all have looked at consecutive patients with dyspepsia presenting for endoscopy, and report their results in terms of endoscopic finding with regard to antibody status for H pylori. Combining the studies provides a much larger sample of over 600 patients, and the results are shown in Table 6.



The overall prevalence of peptic ulcer in asymptomatic subjects and patients with dyspepsia who were H pylori positive was 24%, and in those who were H pylori negative the prevalence was 3.6%.

Incidence


The annual incidence of duodenal and peptic ulcers has been reviewed by Shearman (1989). On average over several studies the mean annual incidence per 1000 inhabitants aged 15 years and over for duodenal ulcer is 2.2 for men and 0.6 for women; for gastric ulcer, the figures are 0.5 and 0.3 respectively. The overall average incidence is 1.8 per 1000 inhabitants per year. The preponderance of duodenal over gastric ulcer is the norm for European countries and the USA, but in Japan gastric ulcer is five to ten times more frequent than duodenal ulcer.

A more recent study of 21,440 persons in the municipality of Tromsø in Norway (Johnsen et al, 1992) gives detailed sex and age-related incidences over the age range 20-49 years, corrected for death and migration. The overall incidence rates are similar, if slightly higher than those above, and detailed results are shown in Figure 1.


Figure 1: Age and sex-related incidence of peptic ulcer







Data are taken from Johnsen et al (1992), describing the age and sex specific incidences of duodenal and gastric ulcers censored for death and migration in a population of 21,440 persons aged between 20 and 54 years in the municipality of Tromsø. Insufficient data were available for women aged over 49 years.

Peptic ulcer bleeding

A recent investigation in Germany has examined the incidence of peptic ulcer bleeding in the population in and around Düsseldorf (Ohmann et al, 1992); given the very significant changes in treatment of ulcers in recent years, with the use of histamine receptor antagonists and proton pump inhibitors, the pattern of peptic ulcer bleeding might have been expected to undergo significant change. The evidence is that no such change has been seen.

The data in the Ohmann report covers the period from March 1989 to February 1990. The crude peptic ulcer bleeding rates found are given below. Important sociodemographic variables for increased incidence of ulcer bleeding were age (with large increases in incidence above age 49 years), widowed status, elementary school education only and being self-employed. Age was the most important variable, though, and risk factors like widowed status are much less remarkable when age-adjusted. The predominant effects of age are shown in Table 7.



All the ulcers in the Ohmann study were proven endoscopically, and full clinical histories were available for this thorough investigation.

Helicobacter pylori eradication and peptic ulcer


There is a growing body of evidence concerning the way in which antibiotics can be used to eradicate H pylori in the stomach and duodenum, and the way in which that eradication affects the course of peptic ulcer disease. The early literature in the 1980s contained few randomised controlled trials (RCTs), and this section will concentrate mainly on the RCTs published in the 1990s, where the evidence is much stronger.

A systematic overview of trials looking at H pylori eradication between 1983 and 1992 has recently been published (Veldhuyzen van Zanten & Sherman, 1994). This well conducted overview used solid methodology and identified eight trials involving duodenal ulcer, of which five were considered high quality. For the purposes of this review, trials from 1992 were sought by use of a MEDLINE search, together with an approach to a number of pharmaceutical companies known or thought to be conducting clinical research in this area with a request for published studies which involved peptic ulcer and H pylori eradication. This strategy located further trials reported as major papers in peer-reviewed journals; all RCTs identified are detailed in Table 8. Most studies treated duodenal and gastric ulcers together, though some did not; no differentiation was made for this review.

There are two main issues to be addressed: whether antibiotic regimens are successful in eradicating H pylori infections, and whether eradication has any effects on the subsequent course of the disease. These are treated separately.

Antibiotics and H pylori eradication

A number of studies have now demonstrated the efficacy of antibiotic therapy in H pylori eradication compared with conventional therapies. Tables 8 - 10, and Figure 2 summarise the results from published randomised controlled trials.

Use of histamine receptor antagonists (cimetidine, ranitidine) or the proton pump inhibitor omeprazole as a single therapy resulted in high levels of ulcer healing of about 78% at 10 weeks after therapy commenced, but eradication of H pylori was low, at 5% or less (Table 9; Figure 2).

Triple therapy studies (in combination with ranitidine or cimetidine) covering up to 768 patients demonstrated eradication rates of about 96%, with ulcer healing in 93% of patients at 10 weeks (Table 9; Figure 2). A recent study (Hosking et al, 1994) examined the need for an acid suppressing medicine (omeprazole) in combination with triple therapy, and found comparable eradication rates and ulcer healing; they concluded that addition of omeprazole was not needed, and that one week of treatment was sufficient. Most other studies continued acid suppressing medicine for some weeks longer than eradication therapy.

Use of amoxycillin with omeprazole, the other major treatment regimen, produced eradication rates of 70-86%, with ulcer healing of 89-95%. The numbers of patients in these studies is smaller, and complicated by an incompletely reported study with high numbers but apparently with less good results; the data in Table 9 reflect the results both with and without this trial (Unge & Ekström, 1993). Omeprazole was usually used for some weeks longer than amoxycillin.

Combining the single arms of randomised controlled trials identified (Tables 9 and 10) and comparing active eradication therapies with results of ulcer healing therapies using cimetidine, ranitidine or omeprazole, it can be seen that the eradication therapies are highly effective. The improved eradication rates are highly significant, and the numbers needed to treat (NNT; Sackett et al, 1991) are 1.2 - 1.5. Thus 1.3 patients needs to be treated with eradication therapy to get a positive H pylori eradication; this is a very efficient treatment - 77 patients out of 100 would benefit.

The use of adequate antibiotic therapy results in eradication of H pylori together with cure of the ulcer.

H pylori eradication and relapse rates


Patients with ulcers who are treated with histamine antagonists or proton pump inhibitors to cure the ulcer, and who then stop taking their medicines, have a significant risk of having the ulcer recur within one year. Table 9 and Figure 2 show that over several studies, the relapse rate for ranitidine and cimetidine is over 80% and for omeprazole it is over 50%. For this reason, many patients with ulcers are prescribed anti-ulcer drugs for long periods to prevent relapse.

With triple therapy, cure of the ulcer and eradication of H pylori leads to a one year relapse rate of under 10%. With omeprazole and amoxycillin, the one year relapse rate is 8-22%. This is achieved without maintenance therapy with acid suppressing drugs.

Combining the single arms from the identified randomised controlled trials (Table 10) and comparing active eradication therapies with results of ulcer healing therapies using cimetidine, ranitidine or omeprazole, it can be seen that the eradication therapies are highly effective in promoting low relapse rates at one year. The relapse rates are significantly lower with eradication therapy and NNTs are of the order of two patients treated to prevent one relapsing within one year. This is highly efficient in comparison with an active treatment - 50 patients out of 100 treated would benefit.

Two studies have reported relapse rates at periods longer than a year. Schütze et al (1993) reporting two-year results of the study of Hentschel et al (1993), indicated over two years there were 102 recurrences of ulcers among 53 patients in whom H pylori persisted, as compared with 1 recurrence among 46 patients in whom H pylori had been eradicated. Another study (Forbes et al, 1994) reported results seven years after H pylori eradication in patients treated in 1985/6; 32/35 in whom the bacterium had initially been eradicated remained free of the infection for a median of 7.1 years. Only 1 of 38 patients who were H pylori negative was found to have an ulcer.

The eradication of H pylori results in patients who have low relapse rates, and who do not need to take maintenance histamine antagonists or proton pump inhibitors.

Effect of treatment on rebleeding rates

Patients whose ulcers have bled once have an increased risk of further complications, including rebleeding. The natural history of peptic ulcers in patients not undergoing surgery was established several decades ago; recurrent bleeding was found in 19-46% of those with one prior bleed and 24-62% of those with two prior bleeding episodes. Prospective data indicated that ulcer complications (with bleeding being foremost) occurred at the rate of 5% a year and recurrent bleeding in up to 1% per month.

The effect of H pylori eradication has been studied in patients with ulcers who had presented with a major upper gastrointestinal bleed requiring a hospital stay, were positive for H pylori and who had no other cause for the bleed (Graham et al, 1993). They were randomised to ranitidine alone or ranitidine plus triple therapy, and were followed up with endoscopy every 3 months for up to 2 years; ranitidine was withdrawn when the ulcer healing was documented endoscopically.

Rebleeding was seen in 4 of 31 patients: all were in the ranitidine alone group. Overall the frequency of rebleeding was 29% over a median of 7 months in the ranitidine group, and 0% over a median of 9 months in the ranitidine plus triple therapy group.

Eradication of H pylori in patients who have had a bleeding episode with an established ulcer appears to result in elimination of further episodes of bleeding.

Effect of eradication therapy on ulcer healing

The meta-analysis described in Table 10 showed one unexpected result: significantly better short-term ulcer healing occurred with eradication therapy compared with conventional treatment. For every eight patients treated with eradication therapy, one more had an ulcer healed compared with conventional therapy. This result has not been described previously, and cannot be obtained from single studies with smaller patient numbers.

Eradication therapy confers an immediate benefit on ulcer healing as well as having long-term benefits on ulcer relapse.

The most effective treatment


There is little to choose between the two main types of eradication therapy in terms of their effectiveness in eradicating H pylori, in promoting short-term ulcer healing or in preventing long-term relapse.

Triple therapy involves taking four different tablets - bismuth, metronidazole, amoxycillin or tertracycline and ranitidine, and perhaps taking as many as 14 tablets each day. This makes it complicated for patients to comply with the treatment regimen. In addition, there is an incidence of adverse effects, including gastrointestinal disturbance and bad taste which some authors have thought reduces the compliance with, and hence effectiveness of the therapy.

Amoxycillin and omeprazole is much simpler, with combinations of tablets being taken twice a day, and with perhaps somewhat fewer adverse effects, though this is not easy to quantify from the information quoted in the studies.

A number of different therapeutic regimens are currently being tested, and perhaps the most effective and most acceptable has yet to emerge. Individual prescribing choice and local cultural conditions are perhaps most important in making a decision concerning which therapy to use.











H pylori and dyspepsia


There is at present no satisfactory information on which to base any conclusions about whether H pylori has some causative role in dyspepsia, or whether eradication of H pylori results in clinically significant improvements.

Systematic review


A critique of the published trials in H pylori-positive functional dyspepsia has elegantly dissected all the RCTs published in full between 1984 and 1993 (Talley, 1994). A total of 16 reports was identified by the search strategy, half of which reported that anti-H pylori therapy was effective and the other half could find no statistically significant benefit.

However, each study had at least one major design flaw, and it can be persuasively argued that not one adequately addressed the problem in a way which would allow sensible conclusions to be drawn. Some of the most important specific points where the trials failed are as follows:-
  • Only one of the studies included a medication which is known unequivocally to result in effective H pylori eradication (triple therapy; Pratchett et al, 1991, though here allocation of patients to treatment was not random), and while most papers report H pylori suppression, few addressed H pylori eradication in dyspepsia. The majority of the studies used bismuth alone or a single antibiotic alone. None had a histamine antagonist or proton pump inhibitor added.
  • Five of the sixteen studies were not placebo controlled. There is a high rate of resolution of symptoms of patients with dyspepsia without treatment; studies which lack a placebo arm therefore become uninterpretable.
  • Most only looked at H pylori positive patients. Including H pylori negative patients allows a determination of whether active treatment alters symptoms independently of H pylori status.
  • No objective outcome measures are available for functional dyspepsia (unlike ulcer healing or recurrence). The various approaches used in the published trials were not tested for reliability before the study commenced for their sensitivity, stability or validity. Veldhuyzen van Zanten has produced a systematic overview of outcome measures in H pylori gastritis and dyspepsia trials (Veldhuyzen van Zanten, 1993); a useful discussion of symptom measurement in dyspepsia is found in Talley, 1994.

Conclusion


Taking all these criticisms together, it can be seen that making any ex-cathedra statements about the role of H pylori eradication in dyspepsia is extremely dangerous in the present state of knowledge. What is required is a large randomised controlled trial of sufficient power to detect a clinically important symptom improvement in dyspeptic patients, which examines patients who are H pylori positive and those who are negative, which has a placebo control arm together with at least one active treatment which is known to be effective in eradicating H pylori, and which uses validated outcome measures.

What is certain is that patients presenting with dyspepsia who are H pylori positive have a much higher incidence of peptic ulcer and gastric cancer than do those who are H pylori negative. This may have important implications for making management decisions when patients first present with symptoms.

H pylori and gastric cancer

Background


Gastric cancer is the fourth most common cause of death from cancer in the UK (after lung, colorectal and breast cancer) and accounts for about 10,000 deaths each year in the UK. Five year survival after resection is not particularly good, being perhaps 5% for all patients and 20% for those having a potentially curable resection.

Sue-Ling et al (1993) showed that gastric cancer can be diagnosed at an earlier stage through more widespread use of endoscopy, and that treatment of early stage cancers results in excellent five-year survival rates (above 70% for stages I and II, compared with 30% for stage III cancers). A prospective study investigating dyspeptic patients over 40 years at their first visit to a general practitioner with early endoscopy covering 10 general practices in Birmingham (Hallisey et al, 1990) demonstrated a 16% prevalence of peptic ulcer and a 2% prevalence of gastric cancer; H pylori status was not assessed. A high proportion of the gastric cancers found were operable and a significant proportion discovered sufficiently early to be curable.

Gastric cancer and H pylori

An association between H pylori infection and an increased risk of gastric cancer was suggested first by histopathological examination of gastric biopsy specimens showing that H pylori infection was more common in patients with gastric cancer than in those with no pathological lesions. A case-control study (Forman et al, 1991) which looked at serum antibodies from a relatively small number of index cancers showed not only was there a higher rate of seropositivity in cancers than controls (69% vs. 47%), but that antibody concentrations in the cancer patients were considerably higher (90 µg/mL vs. 3.6 µg/mL). The Forman study estimated an odds ratio for risk of gastric cancer in those infected with H pylori of about 2.8 (95% confidence intervals 1.04 to 7.97).

The most important evidence to date is the EUROGAST study (1993), an international study of associations between H pylori and gastric cancer. A total of 17 populations in 13 countries were examined, again using serum antibody tests all performed in a single centre as the index of infection with H pylori. Prevalence rates of H pylori seropositivity were related to local gastric cancer incidence, and there was a statistically significant relation between the prevalence of seropositivity and cumulative incidence rates for both gastric cancer incidence and mortality. The conclusion was that a 100% infection rate with H pylori conferred about a six-fold risk of gastric cancer compared with a similar population with no infection.

A small meta-analysis (Forman et al, 1994) of three studies of the association between H pylori and gastric cancer (Forman et al, 1991; Parsonnet et al, 1991; Nomura et al, 1991) examined time trends, and concluded that an approximate 9-fold relative risk may be a more accurate estimate of the association between H pylori infection and gastric cancer.

It should be noted that not all types of gastric cancer may show this high level of association. Wotherspoon et al (1991) concluded that gastric mucosa-associated lymphoid tissue was a direct result of H pylori infection and perhaps provided the background to the development of lymphoma in some cases. By contrast, Tatsuta et al (1993) were unable to confirm an association between H pylori infection and undifferentiated-type early gastric cancer, but did for differentiated-type early gastric cancers.

Epidemiological considerations

Though there is increasing evidence that H pylori may play some part in the development of gastric cancers, it should be stressed that about 50% of the population aged 45 years and over are infected with H pylori, with perhaps fewer than 1% ever going on to develop gastric cancer. That said, there are some obvious correlations between the two conditions, such as both having elevated rates in low socioeconomic groups and in some ethnic groups. There are also disparities between the epidemiological findings: gastric cancer is more common in men, but H pylori infection rates are similar in each sex. Some populations have a high rate of H pylori infection but low rates of gastric cancer.

Non-Hodgkin's lymphoma of the stomach accounts for about 3% of gastric malignancies, with mucosa-associated lymphoid tissue lymphomas as a subset. There is preliminary epidemiological evidence to suggest a stronger relationship of H pylori with these malignancies.

Clinical Guidelines on H pylori



There are two main sets of 'guidelines' which have been proposed and which refer to H pylori and peptic ulcer disease.

NIH Consensus Conference


In the United States, the National Institutes of Health brought together specialists in gastroenterology, surgery, infectious diseases, epidemiology and pathology, as well as the public in a Consensus Development Conference; the results of their deliberations were published in JAMA in July 1994.

Six questions were asked of the conference:-
1. What is the causal relationship of H pylori to upper gastrointestinal disease?

2: How does one diagnose and eradicate H pylori infection?

3: Does eradication of H pylori infection benefit the patient with peptic ulcer disease?

4: What is the relationship between H pylori infection and gastric malignancy?

5: Which H pylori infected patients should be treated?

6: What are the most important questions which must be addressed by future research in H pylori infections?

The answers to these questions are summarised in the JAMA paper, and found elsewhere in this report. The results of the deliberations of the conference led to three main conclusions:-
Ulcer patients with H pylori infection require treatment with antimicrobial agents in addition to antisecretory drugs whether on first presentation with the illness or on recurrence.

The value of treatment of patients with nonulcerative dyspepsia and H pylori infection remains to be determined.

The interesting relationship between H pylori infection and gastric cancers requires further exploration


The latter two conclusions quite rightly focus on areas where there is insufficient information on which to base judgements, and also rightly seek to prevent widespread introduction of H pylori eradication in areas where there is no evidence for its effectiveness or safety.
The first conclusion has been commonly interpreted as a "test and treat" dictum. The main practical consideration then comes in the definition of a diagnosis of peptic ulcer, and for that endoscopy is seen as the gold standard in the USA, though serology is thought by some to be best (Blaser, 1990).

Endoscopy and H pylori testing


Screening patients with dyspepsia for H pylori infection as a triage to reduce endoscopy workloads was put forward by Sobala et al (1991). The rules they formulated were simple:-

1: Endoscope all patients aged 45 years or over.
2: Endoscope patients under 45 years who are H pylori positive.

This strategy was based on a perceived need to screen for malignancy in patients in older age groups, where the disease is much more prevalent, and to provide a sensitive method for detection of peptic ulcer. The results indicated that the strategy had a sensitivity of 97.5% (CI 94.5-99.1%) for peptic ulcer, and missed no malignancies, while reducing endoscopy workload by 23% compared with present policies.

A similar strategy, but using salivary H pylori antibody testing, has estimated that savings for endoscopy workload would be about 20% compared with present policies (Patel et al, 1994).


Economic consequences of H pylori testing and eradication


Eradication of H pylori is presently justified in two circumstances: the newly diagnosed patient with peptic ulcer disease, and patients with previously diagnosed peptic ulcer disease who are receiving maintenance therapy with acid-suppressing medicines. There is, at present, insufficient evidence to justify the widespread use of eradication in patients with dyspepsia or other symptomatic or asymptomatic conditions, with the possible exception of certain rare gastric lymphomas.

Even so, there may be considerable economic and health gains to the community in testing for H pylori infection and eradicating H pylori where the evidence indicates that to be appropriate.

Background data

Epidemiology


Key information needed to begin this economic analysis are epidemiological: how many patients present each year with peptic ulcer (and with dyspepsia, with which it linked), and how many patients previously diagnosed as having peptic ulcer disease and being treated for it now are in the community, and how are they being treated? Information on the former comes from the prevalence and incidence studies reported earlier.

The most important information on the latter comes from a recent study of long term acid suppressing treatment in general practice in the UK (Ryder et al, 1994). Seven general practices in the Harrow area which refer to Northwick Park Hospital, with 60,000 patients on their lists, were recruited to identify patients receiving long term treatment through repeat prescribing data, followed by a survey of patients' notes for indications and investigations. 'Long term' was taken to be continuous therapy for six months or more.

A total of 492 patients was identified as taking long term acid suppressing treatment - 0.82% of the population, with 95% confidence intervals of 0.80-0.84%. The age range of these patients was 4 to 93 years, with a mean of 64 years; there was a slight preponderance of men over women. The largest part of this group was aged over 60 years (65%), with 25% over 75 years. 75% of the patients had been taking acid suppressing treatment for more than 5 years.

There were a number of diagnoses associated with this group. The largest group comprised duodenal (183; 37%) and gastric (25; 5%) ulcers, together being 42% of the total. A further 93 patients (19%) had no diagnosis made, but had abdominal pain or gastritis on endoscopy. Oesophagitis or oesophageal stricture accounted for 16% and hiatus hernia for 8%. The remainder were receiving acid suppressing treatment as prophylaxis for NSAID or steroid administration, or for some other medical condition.

The drugs prescribed were, for the most part, ranitidine (80%), cimetidine (12%) and omeprazole (6%). For ranitidine, most patients were taking 150 mg once or twice daily.

The implication of this report is that about 61% of patients receiving long-term acid suppressing medicine would be eligible for H pylori eradication - some 0.5% of the population.
There is one report (Cottrill, 1994) of the effects of a policy of H pylori eradication in a general practice. This is from in a single practice (9,200 patients, five practitioners); 206 patients were identified from the repeat prescribing register as having taken cimetidine or ranitidine either continuously (58%) or intermittently in the previous year, with an average duration of nearly six years.

Patients eligible for eradication were those with peptic ulcer proven by radiology or endoscopy, or who had endoscopically proven duodenitis or gastritis, and who had given informed consent. There were 79 such patients out of 161 assessed - 0.86% of the practice population, somewhat higher than the proportion found in Harrow. Patients were randomised to triple therapy or omeprazole/amoxycillin, with both giving eradication rates approaching 70% in the 64 who entered the study.

At the end of one year, 10 patients were still taking histamine receptor antagonists for a variety of reasons, including concurrent NSAID therapy. The majority of patients (64%) who had received H pylori eradication did not take acid suppressing drugs other than occasional antacid use.

This study demonstrated that H pylori eradication therapy could be used successfully for a significant proportion of users of long term acid suppressing treatment - 0.55% of the population.

Prescribing information


Since its introduction some years ago, prescribing of the proton pump inhibitor omeprazole has become very common, though surprisingly the use of cimetidine and ranitidine has hardly changed.

The number of prescriptions for cimetidine, ranitidine and omeprazole in the Oxford Region (four counties of Oxfordshire, Buckinghamshire, Berkshire and Northamptonshire) for the years 1992 and 1993 is shown in Figure 3; omeprazole use has increased three-fold over these two years.



The effects on costs (Figure 4) is similar, with omeprazole increasing three-fold and approaching spending on ranitidine, whilst spending on cimetidine has tended to drop, mainly because of generic prescribing. There is no indication of a plateau being reached for omeprazole prescribing.



On a national basis, more prescriptions were given for antacids than for acid suppressing drugs in 1991 (Figure 5), but in subsequent years, mainly because of the growth of omeprazole prescribing, this has changed.



A pattern seen nationally is similar to the Oxford Region. Spending on histamine anatagonists has remained static at about £180 million over several years (Figure 6), while spending on omeprazole and other proton pump inhibitors has rocketed from £20 million to £94 million in two years.


Treatment costs


The comparative costs of different treatment regimens in peptic ulcer disease have been investigated in detail (MeReC Bulletin, 1994). The various costs are presented in summary form in the table below: in general eradication therapy is less expensive than treatment with acid suppressing drugs, and much less expensive than maintenance therapy.

Endoscopy costs

The cost of an endoscopy in the NHS varies, and the 'true' cost is probably not known accurately. Costs as low as £120 have been reported (Patel et al, 1994), whilst others have suggested costs as high as £400. The number of endoscopies performed in the UK is about 280,000 per year, so the total cost of endoscopy to the NHS is probably in the region of £40-70 million a year.

Calculation of economic impact

Patients on acid suppressing maintenence therapy

The key assumption here is that eradicating H pylori in patients who are long term users of acid suppressing medicines will save money.

There are several issues which influence these calculations:-

1: The proportion of the population in which this is appropriate. Two studies above seem to agree that this is appropriate and effective in about 0.5% of the population - some 5,000 patients per million population.

2: The medicines now being taken and their cost. There is considerable scope for this affecting the calculations, as the table on costs of treatment above shows. The cost of a year's treatment could be as low as £50 and as high as £474 - a tenfold range. The mix of treatments is likely to vary with geography and time, but three things are known from the Ryder study (published in 1994, but conducted at some unreported time previously) :- · Only 12% of patients took cimetidine, the cheapest option at £50 per year. · Ranitidine was taken by 80%, and by using costs of the range of doses taken, the average cost per patient for the 252 patients taking ranitidine was £258 a year. · Omeprazole was taken by 6% at a cost of £474 a year each. Taking an average weighted for percentage use, the cost of acid suppressing treatment is of the order of £246 per patient per year. This is very close to the average saving per patient quoted in one report (Gwynn Owen, 1993) of £250 per patient per year in a practice in Devon.

3: The medicines being taken in the future, and their cost. It is clear from the figures from England and the Oxford Region that omeprazole prescribing is rising very fast. Data on prescribing above is at least one year, if not more, out of date, and newly diagnosed patients are likely to be given omeprazole therapy. It seems likely that the average cost of maintenance therapy is likely to be higher not lower.

4: The assumptions are for a full year of savings, excluding the cost of eradication therapy. Savings would be reduced by about £40 per patient in the first full year if the cost of a course of omeprazole and amoxycillin eradication was taken into account.
Given these assumptions, the following calculations can be made:

For a GP practice with 10,000 patients, 50 would be treated with a full year saving of: 50 x £246 = £12,300 per year

For a health district of 250,000 patients, 1250 would be treated with a full year saving of: 1250 x £246 = £307,500 per year · For 1,000,000 population, 5000 would be treated with a full year saving of: 5000 x £246 = £1,230,000 per year

For the UK population of 55 million, 275,000 would be treated with a full year saving of: 275,000 x £246 = £67,650,000

This identifies savings of at least 24% of the amount spent on acid suppressing drugs in 1993 in England, and represents about 0.2% of all NHS spending. There are reasons to believe that this is a conservative estimate:-
A: The data on which the calculations are based are at least one year old, and probably two. More patients would have been placed on long term acid suppressing therapy - peptic ulcer has an incidence of 1.8 per 1000 per year, almost all of whom are H pylori infected, so in two years perhaps another 0.1-0.2% of the population would have joined the patients appropriate for eradication therapy, even allowing for a proportion leaving. This growth could increase the numbers treated by perhaps 20-30%. Patients placed on long-term treatment in 1993/4 would be more likely to be prescribed omeprazole than cimetidine, and the average cost of treatment for newer patients may be substantially higher - perhaps increasing the average cost of treatment by a further 10-20% (and in part accounting for the rise in omeprazole sales). The lower estimate would increase savings to £90 million and the upper would increase savings to £106 million.

B: The calculations have been limited to conditions for which there is evidence that H pylori eradication is effective. A proportion of patients treated with acid suppressing drugs for conditions such as chronic dyspepsia may also be proved to benefit from H pylori eradication in the future. This would lead to further gains.

C: The rapid rise in prescriptions for acid suppressing drugs is in contrast to the static incidence of peptic ulcer disease. The implication is that considerable repeat prescribing is going on for patients who may not take these medicines continuously. This also may represent a substantial number of patients, the majority of whom are likely to be infected with H pylori, and for whom eradication therapy would lead to improved health and reduced spending.

The conclusion is, therefore, that substantial savings can be made and increased health benefits obtained by following an aggressive policy of identifying patients for whom there is solid evidence of effectiveness for H pylori eradication, and by piloting well designed audits and studies of H pylori eradication therapy in groups of patients where effectiveness is possible but not yet proven.

Newly diagnosed patients with peptic ulcer disease

The incidence of peptic ulcer disease in the population over 15 years old is about 1.8 per 1000 of population per year. This would suggest that the typical GP with 1,800 patients of all ages would see one or two new cases of peptic ulcer each year. The evidence is that most of these will be treated with a course of ranitidine or omeprazole, and that an unknown proportion will go on to have long-term treatment with ranitidine or omeprazole. Given that each GP has about 15 patients on long-term acid suppressing medicine, it can be estimated that about half of newly-diagnosed patients will go on to have long-term treatment.

There are two cost comparisons. The first is between a course of H pylori eradication and a course of cimetidine, ranitidine or omeprazole. As the table of comparative costs shows, there is little difference between a course of eradication therapy and a course of acid suppression therapy. Medically, the eradication therapy is superior because it leads to superior ulcer healing and reduced relapse.

That brings in the second cost comparison which looks at events following the first course. With acid suppression, about 0.5 patient per GP per year would be expected eventually to receive long-term acid suppression therapy; it is most likely to be ranitidine or omeprazole, at a cost of between £70 and £240 per year (average likely to be about £150). Because eradication therapy is highly effective in preventing relapse, almost all of this additional cost can be saved.

The amount spent by the NHS on acid suppressing drugs has been growing at 14% per year (£40 million). It is likely that a large proportion of this could be saved - and perhaps all of it - if newly diagnosed patients with peptic ulcer disease were treated with H pylori eradication therapy.

Introduction of H pylori eradication for patients with newly diagnosed peptic ulcers would not have an immediate cost advantage, but would result in a slow-down in the rate of growth in spending on acid suppressing medicines. The year-on-year saving could be £40 million in year 1.

Dyspepsia

There is no satisfactory evidence that treating patients with dyspepsia with H pylori eradication is or is not effective. Patients attending general practitioners with dyspepsia represent a significant portion of a GP's workload. A positive demonstration that H pylori eradication cures dyspeptic symptoms in a proportion of sufferers, and that the effect was as long-lived as that for peptic ulcer would have major economic consequences for intermittent prescribing of acid suppressing medicines and of GP's time.

There is no present indication that savings could be made. Demonstrating that H pylori eradication was effective in dyspepsia would be likely to yield savings of a high order.

Economic effects on endoscopy

Two recent papers have shown that screening patients for antibodies to H pylori before endoscopy, following the guidelines outlined previously, can result in a reduction of 20-25% in the number of endoscopies performed (Sobala et al, 1991; Patel et al, 1994).

The number of endoscopies performed in the UK is probably about 300,000 a year and the cost of each is about £150-200 (though figures as low as £120 and as high as £400 are quoted occasionally - the actual cost probably depends upon exact definitions used). Overall, the likely range of NHS spend on endoscopies is of the order of £45 - £60 million.

The systematic use of H pylori testing and doing fewer endoscopies as a result would be likely to save £10 - £15 million each year, even including the cost of the testing. These savings are equivalent to £600 per practice, or £200 per GP per year.

Should guidelines change to allow endoscopy only on patients who were H pylori positive, these savings could be doubled.

Performing endoscopies only on H pylori positive patients under 45, but on all patients over 45 would result in considerable savings on endoscopy services, and would probably result in much reduced waiting times for endoscopy.

H pylori - demand and place for testing



The UK does not have generous provision for testing for H pylori, though this is a general impression obtained through discussion with various healthcare professionals and company representatives rather than through systematic enquiry. Provision for serological testing seems particularly poor.

There will always be a need for specialist advice on biopsy samples obtained at endoscopy, and CLO testing may become more popular as some kit devices are becoming available. Breath testing is unlikely to make rapid headway because of the capital expense and time constraints on patient and system. Serological testing has been suggested as the gold standard (Blaser, 1990), and the main question here is the balance between near patient and laboratory testing. The following paragraphs seek to explore the benefits of each, and likely size of provision required.

The cost of a near-patient serology test (NPT) for H pylori antibodies will be about £12, and the cost of a laboratory test about £3-6 not including the laboratory overheads and cost of transport and handling. The differential cost will £6-9, though it would be expected that laboratories would provide results in about one week compared with a time to result of a few minutes with NPT.

Testing of patients on maintenance acid suppressing medicines


The evidence is that just under 1% of the population of the UK takes acid suppressing medicine, and that about 0.5% of the UK population would benefit from stopping this long term maintenance and eradicating H pylori if present. The implication, then, is that 0.5% of the population of the UK should be tested - some 250,000 persons, even though the indication from known evidence is that a high proportion would be positive for H pylori.

For all these patients a diagnosis of peptic ulcer has been made, or a successful therapeutic trial of acid-suppressing medicine indicated peptic ulcer was present. Testing by the GP would allow the immediate institution of management change, without recourse to a second appointment while laboratory-generated results were being obtained. This would be cost effective - though the cost of the NPT is higher, the difference is actually less than one week's typical treatment with acid suppressing medicine (about £250 a year or £10 a week). The cost of a second GP appointment would, of course, make the use of laboratory testing strategy much higher.

Testing of newly presenting patients


About 10% of patients attending GP surgeries complain of dyspepsia, and on incidence data each GP (average 1800 persons) should see 1.8 new cases of peptic ulcer per 1000 persons aged over 15 years. Thus each GP should see one or two new cases of peptic ulcer per year.

The problem faced by the GP is deciding which of the many patients presenting with dyspepsia may have peptic ulcer, and what protocol to follow in making a diagnosis.

The former question may be answered to some extent by simple clinical decision making and the availability of diagnostic procedures. Thus the availability of endoscopy services may determine how GPs proceed; open access endoscopy with short waiting times opens the door to a process whereby patients in whom there is a high index of suspicion are seen early by a gastroenterologist who can then make a definite diagnosis on the results of endoscopy.


However, in many areas the waiting time for endoscopy is long (3 months is quoted in some counties), and here treatment may be initiated well before endoscopy, making it of questionable use. In these circumstances, an alternative procedure may be to test for H pylori infection to aid clinical diagnosis, with a therapeutic trial of eradication therapy. Such a protocol may be justified by the high prevalence of peptic ulcer disease (about 24%) in symptomatic or asymptomatic patients who are seropositive for H pylori antibodies.

The relative effectiveness and safety of these or other protocols is hampered by the lack of evidence of effectiveness or safety of H pylori eradication in dyspepsia.

In any event, it is likely that local considerations of availability of testing and other services, as well as individual GP preferences, will determine practice in the absence of agreed strategies within Health Districts, especially as to whether H pylori positive patients should preferentially be offered endoscopy.

With the known incidences of dyspepsia and peptic ulcer disease, it is not unlikely that each GP would need about 10 serology tests for each patient with peptic ulcer disease, making a total of up to 40 tests per GP, and about 800,000 tests each year, divided between NPT and laboratory based services.

Conclusion


The total number of tests required is about 1 million each year, with some 250,000 - 500,000 being near patient tests.

It should be borne in mind that these figures represent 'best guess' estimates based on limited information concerning decision making by individual practitioners, and in the absence of nationally or locally agreed guidelines, which could increase or decrease the scale of testing.

H pylori offers a unique opportunity to develop an evidence-based testing strategy combined with evidence-based guideline implementation. An economic case could be made for FHSAs paying GPs to perform H pylori tests, or at least to reimburse the cost of the reagents, as part of a study of implementing locally agreed guidelines.


NPT for H pylori - impact on a Health Authority



There seems to be little doubt that treatment of peptic ulcers by eradicating H pylori can dramatically alter treatment regimens and the lives of patients. The evidence is that the treatments are effective, and could be delivered at low cost by general practitioners if the patients who would benefit could easily be identified.

The following pages take some actual figures from an actual Health Authority and model the possible impact of introducing an active policy of H pylori eradication using near patient testing for H pylori by GPs. The cost impacts are on medicines prescribed, and on endoscopies performed. Much of the calculation on medicines costs has been described in a previous chapter (page 26 et seq), but are included here with calculations of how different strategies could impact on endoscopy rates.

Patients previously diagnosed with ulcers


There are few problems here. The patients have already been diagnosed, and most will be taking long-term acid suppressing medicines. Almost all will be infected with H pylori, but a minority will not be infected. Eradicating H pylori in the infected patients and stopping their long-term acid suppressing medicines will save about £250 per patient, and will have a significant and immediate economic impact on drugs spending of GP practices and of Health Authorities.

A strategy of determining from practice records which patients to test and treat should be a relatively straightforward task for most practices. Near patient testing by the GP to confirm H pylori infection using an antibody-based test is cheap and simple, and eradication could start immediately. Since about 0.5% of the population would be involved, careful preparation of information leaflets for patients which explained the reasons for the change in treatment, the short term consequences and long-term benefits would be of undoubted help in getting a policy of treatment change successfully enacted.

An outline schematic of how such patients are likely to be treated is shown in Figure 7 . Those with sinister symptoms (bleeding, weight loss) would be referred for urgent consultation. Those with proven ulcer could be tested, and if positive a course of eradication followed. Failure of eradication in patients with proven infection might lead to a further referral for endoscopy, or possibly a second eradication course.

Patients who are taking long-term acid suppressing drugs where no ulcer had been demonstrated by endoscopy or radiology, or where no diagnosis had been made, and who were not taking NSAID drugs could also be tested. A positive test might lead to a course of eradication, and a negative test might lead to a trial or retrial of simple antacids, reversion to previous therapy or referral for endoscopy.

In the Health Authority with a population of 650,000, some 3,250 people would be expected to be taking long-term acid suppressing medicines where a diagnosis of ulcer had already been made (Ryder et al, 1994) at an average cost of £250. Savings on the cost of medicines to these patients would be of the order of £810,000 a year, though may be less where eradication therapies have actively been pursued.

Patients newly presenting with dyspepsia


This is more difficult. An outline schematic of how such patients are likely to be treated now is shown in Figure 8. Patients with any sinister symptoms (bleeding, weight loss) will be referred for urgent consultations and (usually) endoscopy. Those without sinister symptoms whose clinical history shows some possible lifestyle cause such as heavy drinking or smoking, or poor diet, or prolonged NSAID use, will often first have a trial of lifestyle change, perhaps with antacids or alginates.

Failure of this is likely to lead to older patients (over 40, 45 or 50 years, depending on local policy) being referred for consultation or endoscopy, often because of the possibility of gastric cancer, where the incidence rises rapidly above age 45 years. Because access to endoscopy services can often take a significant time (up to three months), it is likely that all patients in whom antacids have failed will be given a course of acid suppressing medicine for symptom relief.

For some of these patients, this therapy will be continued indefinitely (Ryder et al, 1994), though for others where a short course fails and symptoms recur, referral for endoscopy would be a next step. The finding of an ulcer at endoscopy in the past has led to use of maintenance treatment with acid suppressing medicines. Very recently, however, the practice of gastroenterologists has begun to change, and most would now biopsy for H pylori at endoscopy and institute a course of eradication if found.

How might NPT for H pylori alter this?


First the evidence:-
  • Serology testing is an accepted method for diagnosing H pylori, and some authorities consider serology testing to be the most effective method (Blaser 1990).
  • NPT for H pylori antibodies has been shown to have high sensitivity, specificity and predictive value compared with other testing methods, and more extensive information on populations is being generated.
  • Surveys using serology tests have shown that about 25% of patients presenting with dyspepsia and who are H pylori positive have ulcers, compared with fewer than 3% of patients with dyspepsia who are H pylori negative (Table 6). Gastritis and gastric cancer are also more common in H pylori positive patients than in those who are H pylori negative. This concentration of ulcer patients in the seropositive group suggests the use of NPT as triage at initial consultation.
  • Eradication therapy is effective in peptic ulcers.
  • Eradication therapy has not been shown conclusively to be useful in non-ulcer dyspepsia. Though these studies did not use effective eradication regimens, and most had some methodological flaws, eight showed a benefit and eight did not (Talley, 1994). None showed any potential harm.
  • Eradication therapy is not known to be a sensible treatment for gastric cancer.
  • Eradication therapy does not seem to be helpful in ulceration caused by prolonged NSAID use.
  • Eradication therapy is associated with adverse effects. For the antibiotics used in eradication these are understood, and for triple therapy there may be additional effects from high doses of bismuth.
  • Deaths from gastric cancer are relatively uncommon. In the Health Authority during the period 1986-92, there were 103 deaths, and the risk of death rose from 1 in 44,000 at age 40-44 to 1 in 5,000 at age 60-64 years.

A scheme which suggests itself, therefore, might be similar to that outlined in Figure 9. In this scheme, patients with any sinister symptoms (bleeding, weight loss) would be referred for urgent consultations and (usually) endoscopy. Those without sinister symptoms would, of course, have a history taken where important lifestyle issues and NSAID consumption could be addressed.

In patients where these are not overriding factors, a test for H pylori could be performed immediately.
  • A negative test would demonstrate that the chance that the patient had an ulcer or malignant condition was very low. Symptomatic treatments and lifestyle changes should be tried, and only if they failed would a referral for consultation or endoscopy be considered.
  • A positive test would indicate that the patient had a high chance of having an ulcer, or gastritis or duodenitis; there would be a 2% chance of gastric cancer.

Some options



At this point there are several options, and two extreme options are considered here to demonstrate the likely range of economic impact.

One (Option A) is to endoscope all the H pylori positive patients quickly, and where appropriate institute eradication therapy. This has the benefit of being a conservative policy, and would have the benefit of acting as a more effective screen for early and perhaps treatable gastric cancer. Its effectiveness would be impaired where the delay before endoscopy was more than a week or two; where waiting time for endoscopy was up to three months it would be unacceptable.

A second option (Option B) is to begin immediately a course of eradication therapy in all patients who are H pylori positive. This would be of proven benefit in the 25% of patients with ulcers, and, since common eradication therapies have acid suppressing medicine as a component, would provide symptom relief for the others as well as being a therapeutic trial of eradication. Early symptom recurrence in patients who had completed an eradication course would be a sinister event, and such patients should be considered for urgent referral. This option has the demerit of treating patients with non-ulcer dyspepsia with a non-proven therapy, but it could also be considered as an N-of-1 trial. This option would not, perhaps, be appropriate where rapid open-access endoscopy was available, but would be appropriate where waiting times for endoscopy were long.

Assessing their effects

The situation now


The present situation in the Health Authority with a population of 470,000 adults aged more (Table 12) than 20 years is that there are about 7,500 presentations per year to GPs for dyspepsia. This estimate is obtained from the known prevalence of dyspepsia (about 25% of the adult population) and assuming that about 7% of those affected present in any one year.



There are 2755 endoscopies performed each year in the Health Authority by three gastroenterologists, and the diagnostic categories under which they are performed are given in Table 12. This is not entirely helpful in assessing which are associated with peptic ulcer diagnosis, and there are no data concerning the proportion which are normal. However, the two most frequent diagnostic groups are 'Oesophageal/gastrointestinal' and 'abdominal pain', which together comprise nearly 2,000 endoscopies a year, over 70% of the total. It has been assumed that about half of these, about 1,000 endoscopies a year, are performed for diagnosis of peptic ulcer, and the maximum number must be 2,000 (Table 12).

Future possibilities: Option A


This option has NPT for H pylori as triage. Taking the proportion of H pylori positive and negative shown earlier in Table 6 for the results of endoscopies performed on 631 patients presenting with dyspepsia, and whose serological status for H pylori was established, some 56% (4,200) of the 7,500 patients presenting to GPs with dyspepsia will be seronegative and 44% (3,225) will be seropositive.

This option would have all seropositive patients being endoscoped, and if it is assumed that about 5% of seronegative patients will eventually present for endoscopy because of failed therapy, then it is clear that the result would be 3,435 endoscopies a year for peptic ulcer ( Figure 10 ).

This is much more than the absolute maximum now, and instituting this option would result in doubling the number of endoscopies performed in the Health Authority. The cost (if each endoscopy cost £400) would be almost £1 million more than at present. The cost of the near patient testing would be about £100,000 per year in addition. It is unlikely that Health Authorities could bear this additional burden to endoscopy services.

However, such a strategy would ensure that almost all patients with peptic ulcer would be found - about 850 patients a year in a Health Authority with an adult population of 470,000 (Table 12). Curing the H pylori infection in the majority of these patients would be a significant factor in preventing additional increases in the medicines bill for acid suppressing drugs.

Future possibilities: Option B


This option also has NPT for H pylori as triage. Taking the proportion of H pylori positive and negative shown earlier in Table 6 for the results of endoscopies performed on 631 patients presenting with dyspepsia, and whose serological status for H pylori was established, some 56% (4,200) of the 7,500 patients presenting to GPs with dyspepsia will be seronegative and 44% (3,225) will be seropositive.

The diagnoses for those who were seropositive were given, and the proportions can be applied to the population with dyspepsia in the Health Authority. Making some assumptions about the percentage in each group who eventually present for endoscopy because of failed therapy, the following calculations can be made:-



If it is assumed that about 5% of seronegative patients will eventually present for endoscopy because of failed therapy, then the total number of endoscopies performed each year would be 410 - at least 600 fewer than at present ( Figure 10 ). This would reduce the number of endoscopies performed each year in the Health Authority for ulcer diagnosis by about 60%, and the total number of endoscopies by 25%. Savings would be £250,000 per year, and cut waiting times for endoscopy. Against this saving would need to be set the £100,000 cost of near patient testing.

In the Health Authority, given the annual incidence of peptic ulcer of about 1.8/1000 adults, some 850 patients a year would be expected to be identified as having peptic ulcers. Standard treatment has been acid suppressing drugs, which for many have become effectively a lifelong treatment.

Options A and B follow a strategy of initial identification of patients with H pylori infection which can subsequently be eradicated. This means a saving on future prescribing if these patients do not take long term acid suppressing drugs. If 50% of patients with ulcers were saved from this, the drugs budget would increase by some £110,000 per year less than would otherwise have occurred.

Summary


Option A looks expensive, whilst option B would provide the immediacy of testing and treating the patient with dyspepsia at an early stage, and delivering cuts in endoscopy numbers and waiting times. Without taking into account savings on costs of future medicines, Option A would cost an additional £290,000 a year, while Option B would save £960,000.




NOTE

It should be noted that these are outline proposals designed to be indicative of possible consequences of particular policy decisions. The proposals made are not guidelines, nor are they intended to be guidelines. Individual authorities or practices should determine their own strategies depending upon the local provision of services, but the suggestions in this chapter may be useful in pointing out the directions most likely to produce results.


H pylori - demand and place for testing



The UK does not have generous provision for testing for H pylori, though this is a general impression obtained through discussion with various healthcare professionals and company representatives rather than through systematic enquiry. Provision for serological testing seems particularly poor.

There will always be a need for specialist advice on biopsy samples obtained at endoscopy, and CLO testing may become more popular as some kit devices are becoming available. Breath testing is unlikely to make rapid headway because of the capital expense and time constraints on patient and system. Serological testing has been suggested as the gold standard (Blaser, 1990), and the main question here is the balance between near patient and laboratory testing. The following paragraphs seek to explore the benefits of each, and likely size of provision required.

The cost of a near-patient serology test (NPT) for H pylori antibodies will be about £12, and the cost of a laboratory test about £3-6 not including the laboratory overheads and cost of transport and handling. The differential cost will £6-9, though it would be expected that laboratories would provide results in about one week compared with a time to result of a few minutes with NPT.

Testing of patients on maintenance acid suppressing medicines


The evidence is that just under 1% of the population of the UK takes acid suppressing medicine, and that about 0.5% of the UK population would benefit from stopping this long term maintenance and eradicating H pylori if present. The implication, then, is that 0.5% of the population of the UK should be tested - some 250,000 persons, even though the indication from known evidence is that a high proportion would be positive for H pylori.

For all these patients a diagnosis of peptic ulcer has been made, or a successful therapeutic trial of acid-suppressing medicine indicated peptic ulcer was present. Testing by the GP would allow the immediate institution of management change, without recourse to a second appointment while laboratory-generated results were being obtained. This would be cost effective - though the cost of the NPT is higher, the difference is actually less than one week's typical treatment with acid suppressing medicine (about £250 a year or £10 a week). The cost of a second GP appointment would, of course, make the use of laboratory testing strategy much higher.

Testing of newly presenting patients


About 10% of patients attending GP surgeries complain of dyspepsia, and on incidence data each GP (average 1800 persons) should see 1.8 new cases of peptic ulcer per 1000 persons aged over 15 years. Thus each GP should see one or two new cases of peptic ulcer per year.

The problem faced by the GP is deciding which of the many patients presenting with dyspepsia may have peptic ulcer, and what protocol to follow in making a diagnosis.

The former question may be answered to some extent by simple clinical decision making and the availability of diagnostic procedures. Thus the availability of endoscopy services may determine how GPs proceed; open access endoscopy with short waiting times opens the door to a process whereby patients in whom there is a high index of suspicion are seen early by a gastroenterologist who can then make a definite diagnosis on the results of endoscopy.


However, in many areas the waiting time for endoscopy is long (3 months is quoted in some counties), and here treatment may be initiated well before endoscopy, making it of questionable use. In these circumstances, an alternative procedure may be to test for H pylori infection to aid clinical diagnosis, with a therapeutic trial of eradication therapy. Such a protocol may be justified by the high prevalence of peptic ulcer disease (about 24%) in symptomatic or asymptomatic patients who are seropositive for H pylori antibodies.

The relative effectiveness and safety of these or other protocols is hampered by the lack of evidence of effectiveness or safety of H pylori eradication in dyspepsia.

In any event, it is likely that local considerations of availability of testing and other services, as well as individual GP preferences, will determine practice in the absence of agreed strategies within Health Districts, especially as to whether H pylori positive patients should preferentially be offered endoscopy.

With the known incidences of dyspepsia and peptic ulcer disease, it is not unlikely that each GP would need about 10 serology tests for each patient with peptic ulcer disease, making a total of up to 40 tests per GP, and about 800,000 tests each year, divided between NPT and laboratory based services.

Conclusion


The total number of tests required is about 1 million each year, with some 250,000 - 500,000 being near patient tests.

It should be borne in mind that these figures represent 'best guess' estimates based on limited information concerning decision making by individual practitioners, and in the absence of nationally or locally agreed guidelines, which could increase or decrease the scale of testing.

H pylori offers a unique opportunity to develop an evidence-based testing strategy combined with evidence-based guideline implementation. An economic case could be made for FHSAs paying GPs to perform H pylori tests, or at least to reimburse the cost of the reagents, as part of a study of implementing locally agreed guidelines.


NPT for H pylori - impact on a Health Authority



There seems to be little doubt that treatment of peptic ulcers by eradicating H pylori can dramatically alter treatment regimens and the lives of patients. The evidence is that the treatments are effective, and could be delivered at low cost by general practitioners if the patients who would benefit could easily be identified.

The following pages take some actual figures from an actual Health Authority and model the possible impact of introducing an active policy of H pylori eradication using near patient testing for H pylori by GPs. The cost impacts are on medicines prescribed, and on endoscopies performed. Much of the calculation on medicines costs has been described in a previous chapter (page 26 et seq), but are included here with calculations of how different strategies could impact on endoscopy rates.

Patients previously diagnosed with ulcers


There are few problems here. The patients have already been diagnosed, and most will be taking long-term acid suppressing medicines. Almost all will be infected with H pylori, but a minority will not be infected. Eradicating H pylori in the infected patients and stopping their long-term acid suppressing medicines will save about £250 per patient, and will have a significant and immediate economic impact on drugs spending of GP practices and of Health Authorities.

A strategy of determining from practice records which patients to test and treat should be a relatively straightforward task for most practices. Near patient testing by the GP to confirm H pylori infection using an antibody-based test is cheap and simple, and eradication could start immediately. Since about 0.5% of the population would be involved, careful preparation of information leaflets for patients which explained the reasons for the change in treatment, the short term consequences and long-term benefits would be of undoubted help in getting a policy of treatment change successfully enacted.

An outline schematic of how such patients are likely to be treated is shown in Figure 7 . Those with sinister symptoms (bleeding, weight loss) would be referred for urgent consultation. Those with proven ulcer could be tested, and if positive a course of eradication followed. Failure of eradication in patients with proven infection might lead to a further referral for endoscopy, or possibly a second eradication course.

Patients who are taking long-term acid suppressing drugs where no ulcer had been demonstrated by endoscopy or radiology, or where no diagnosis had been made, and who were not taking NSAID drugs could also be tested. A positive test might lead to a course of eradication, and a negative test might lead to a trial or retrial of simple antacids, reversion to previous therapy or referral for endoscopy.

In the Health Authority with a population of 650,000, some 3,250 people would be expected to be taking long-term acid suppressing medicines where a diagnosis of ulcer had already been made (Ryder et al, 1994) at an average cost of £250. Savings on the cost of medicines to these patients would be of the order of £810,000 a year, though may be less where eradication therapies have actively been pursued.

Patients newly presenting with dyspepsia


This is more difficult. An outline schematic of how such patients are likely to be treated now is shown in Figure 8. Patients with any sinister symptoms (bleeding, weight loss) will be referred for urgent consultations and (usually) endoscopy. Those without sinister symptoms whose clinical history shows some possible lifestyle cause such as heavy drinking or smoking, or poor diet, or prolonged NSAID use, will often first have a trial of lifestyle change, perhaps with antacids or alginates.

Failure of this is likely to lead to older patients (over 40, 45 or 50 years, depending on local policy) being referred for consultation or endoscopy, often because of the possibility of gastric cancer, where the incidence rises rapidly above age 45 years. Because access to endoscopy services can often take a significant time (up to three months), it is likely that all patients in whom antacids have failed will be given a course of acid suppressing medicine for symptom relief.

For some of these patients, this therapy will be continued indefinitely (Ryder et al, 1994), though for others where a short course fails and symptoms recur, referral for endoscopy would be a next step. The finding of an ulcer at endoscopy in the past has led to use of maintenance treatment with acid suppressing medicines. Very recently, however, the practice of gastroenterologists has begun to change, and most would now biopsy for H pylori at endoscopy and institute a course of eradication if found.

How might NPT for H pylori alter this?


First the evidence:-
  • Serology testing is an accepted method for diagnosing H pylori, and some authorities consider serology testing to be the most effective method (Blaser 1990).
  • NPT for H pylori antibodies has been shown to have high sensitivity, specificity and predictive value compared with other testing methods, and more extensive information on populations is being generated.
  • Surveys using serology tests have shown that about 25% of patients presenting with dyspepsia and who are H pylori positive have ulcers, compared with fewer than 3% of patients with dyspepsia who are H pylori negative (Table 6). Gastritis and gastric cancer are also more common in H pylori positive patients than in those who are H pylori negative. This concentration of ulcer patients in the seropositive group suggests the use of NPT as triage at initial consultation.
  • Eradication therapy is effective in peptic ulcers.
  • Eradication therapy has not been shown conclusively to be useful in non-ulcer dyspepsia. Though these studies did not use effective eradication regimens, and most had some methodological flaws, eight showed a benefit and eight did not (Talley, 1994). None showed any potential harm.
  • Eradication therapy is not known to be a sensible treatment for gastric cancer.
  • Eradication therapy does not seem to be helpful in ulceration caused by prolonged NSAID use.
  • Eradication therapy is associated with adverse effects. For the antibiotics used in eradication these are understood, and for triple therapy there may be additional effects from high doses of bismuth.
  • Deaths from gastric cancer are relatively uncommon. In the Health Authority during the period 1986-92, there were 103 deaths, and the risk of death rose from 1 in 44,000 at age 40-44 to 1 in 5,000 at age 60-64 years.

A scheme which suggests itself, therefore, might be similar to that outlined in Figure 9. In this scheme, patients with any sinister symptoms (bleeding, weight loss) would be referred for urgent consultations and (usually) endoscopy. Those without sinister symptoms would, of course, have a history taken where important lifestyle issues and NSAID consumption could be addressed.

In patients where these are not overriding factors, a test for H pylori could be performed immediately.
  • A negative test would demonstrate that the chance that the patient had an ulcer or malignant condition was very low. Symptomatic treatments and lifestyle changes should be tried, and only if they failed would a referral for consultation or endoscopy be considered.
  • A positive test would indicate that the patient had a high chance of having an ulcer, or gastritis or duodenitis; there would be a 2% chance of gastric cancer.

Some options



At this point there are several options, and two extreme options are considered here to demonstrate the likely range of economic impact.

One (Option A) is to endoscope all the H pylori positive patients quickly, and where appropriate institute eradication therapy. This has the benefit of being a conservative policy, and would have the benefit of acting as a more effective screen for early and perhaps treatable gastric cancer. Its effectiveness would be impaired where the delay before endoscopy was more than a week or two; where waiting time for endoscopy was up to three months it would be unacceptable.

A second option (Option B) is to begin immediately a course of eradication therapy in all patients who are H pylori positive. This would be of proven benefit in the 25% of patients with ulcers, and, since common eradication therapies have acid suppressing medicine as a component, would provide symptom relief for the others as well as being a therapeutic trial of eradication. Early symptom recurrence in patients who had completed an eradication course would be a sinister event, and such patients should be considered for urgent referral. This option has the demerit of treating patients with non-ulcer dyspepsia with a non-proven therapy, but it could also be considered as an N-of-1 trial. This option would not, perhaps, be appropriate where rapid open-access endoscopy was available, but would be appropriate where waiting times for endoscopy were long.

Assessing their effects

The situation now


The present situation in the Health Authority with a population of 470,000 adults aged more (Table 12) than 20 years is that there are about 7,500 presentations per year to GPs for dyspepsia. This estimate is obtained from the known prevalence of dyspepsia (about 25% of the adult population) and assuming that about 7% of those affected present in any one year.



There are 2755 endoscopies performed each year in the Health Authority by three gastroenterologists, and the diagnostic categories under which they are performed are given in Table 12. This is not entirely helpful in assessing which are associated with peptic ulcer diagnosis, and there are no data concerning the proportion which are normal. However, the two most frequent diagnostic groups are 'Oesophageal/gastrointestinal' and 'abdominal pain', which together comprise nearly 2,000 endoscopies a year, over 70% of the total. It has been assumed that about half of these, about 1,000 endoscopies a year, are performed for diagnosis of peptic ulcer, and the maximum number must be 2,000 (Table 12).

Future possibilities: Option A


This option has NPT for H pylori as triage. Taking the proportion of H pylori positive and negative shown earlier in Table 6 for the results of endoscopies performed on 631 patients presenting with dyspepsia, and whose serological status for H pylori was established, some 56% (4,200) of the 7,500 patients presenting to GPs with dyspepsia will be seronegative and 44% (3,225) will be seropositive.

This option would have all seropositive patients being endoscoped, and if it is assumed that about 5% of seronegative patients will eventually present for endoscopy because of failed therapy, then it is clear that the result would be 3,435 endoscopies a year for peptic ulcer ( Figure 10 ).

This is much more than the absolute maximum now, and instituting this option would result in doubling the number of endoscopies performed in the Health Authority. The cost (if each endoscopy cost £400) would be almost £1 million more than at present. The cost of the near patient testing would be about £100,000 per year in addition. It is unlikely that Health Authorities could bear this additional burden to endoscopy services.

However, such a strategy would ensure that almost all patients with peptic ulcer would be found - about 850 patients a year in a Health Authority with an adult population of 470,000 (Table 12). Curing the H pylori infection in the majority of these patients would be a significant factor in preventing additional increases in the medicines bill for acid suppressing drugs.

Future possibilities: Option B


This option also has NPT for H pylori as triage. Taking the proportion of H pylori positive and negative shown earlier in Table 6 for the results of endoscopies performed on 631 patients presenting with dyspepsia, and whose serological status for H pylori was established, some 56% (4,200) of the 7,500 patients presenting to GPs with dyspepsia will be seronegative and 44% (3,225) will be seropositive.

The diagnoses for those who were seropositive were given, and the proportions can be applied to the population with dyspepsia in the Health Authority. Making some assumptions about the percentage in each group who eventually present for endoscopy because of failed therapy, the following calculations can be made:-



If it is assumed that about 5% of seronegative patients will eventually present for endoscopy because of failed therapy, then the total number of endoscopies performed each year would be 410 - at least 600 fewer than at present ( Figure 10 ). This would reduce the number of endoscopies performed each year in the Health Authority for ulcer diagnosis by about 60%, and the total number of endoscopies by 25%. Savings would be £250,000 per year, and cut waiting times for endoscopy. Against this saving would need to be set the £100,000 cost of near patient testing.

In the Health Authority, given the annual incidence of peptic ulcer of about 1.8/1000 adults, some 850 patients a year would be expected to be identified as having peptic ulcers. Standard treatment has been acid suppressing drugs, which for many have become effectively a lifelong treatment.

Options A and B follow a strategy of initial identification of patients with H pylori infection which can subsequently be eradicated. This means a saving on future prescribing if these patients do not take long term acid suppressing drugs. If 50% of patients with ulcers were saved from this, the drugs budget would increase by some £110,000 per year less than would otherwise have occurred.

Summary


Option A looks expensive, whilst option B would provide the immediacy of testing and treating the patient with dyspepsia at an early stage, and delivering cuts in endoscopy numbers and waiting times. Without taking into account savings on costs of future medicines, Option A would cost an additional £290,000 a year, while Option B would save £960,000.




NOTE

It should be noted that these are outline proposals designed to be indicative of possible consequences of particular policy decisions. The proposals made are not guidelines, nor are they intended to be guidelines. Individual authorities or practices should determine their own strategies depending upon the local provision of services, but the suggestions in this chapter may be useful in pointing out the directions most likely to produce results.


H pylori - farsight and research agenda



There will be a considerable number of different research issues involving H pylori in coming years. This section looks only at some of the potentially large areas of public health and cost of healthcare delivery which have been highlighted by the report and which may form part of a research agenda.

Implementing H pylori eradication strategies


Any strategy which seeks to address the question of implementing H pylori eradication therapy in the 250,000 persons taking long-term acid suppressing therapy should be evaluated as part of a randomised controlled trial. This is a unique opportunity to investigate the effectiveness of management strategies in primary care.

Here is a therapy that is known to be effective, that confers benefits on the patients and on society by means of reducing costs. How is a strategy best devised and implemented? The GRiP initiative (Getting Research into Practice) has demonstrated that a number of management strategies can be attempted, and time will tell as to their effectiveness. Each, however, was directed towards a different end-point (glue ear, D&C, stroke rehabilitation and steroids in pre-term delivery); H pylori eradication in established peptic ulcer disease offers an opportunity to evaluate a number of different management strategies in a single context.

Use of antibody quantitation in diagnosis of peptic ulcer disease and gastric cancer


Where quantitation has been attempted, the degree of elevation of antibodies to H pylori has been associated with higher prevalence of peptic ulcer disease and gastric cancer. In the absence of a specific centre of excellence in H pylori research, Health Authorities might seek a collaboration with commercial manufacturers to test the hypothesis in a blind fashion.

Thus, a bank of serum samples collected from patients with a mixture of upper gastrointestinal disorders, perhaps as part of a larger study in primary care, could be used for testing in a blind manner to assess the value of antibody quantitation.

Screening for cancer and peptic ulcer.


Earlier, the prevalence of gastric cancer and peptic ulcer was described in symptomatic populations with dyspepsia and in asymptomatic blood donors. The prevalence was also divided by whether patients were seropositive or seronegative with respect to H pylori infection.

The message was simple - irrespective of symptoms, subjects who were infected with H pylori had a high prevalence of peptic ulcer (about 24%) and a moderate incidence of gastric cancer (about 2%). These figures were much higher than those in seronegative subjects.


There are several questions which need to be answered:-
  • What is the health gain for a population of screening for H pylori infection, endoscoping those who are positive, and treating those with peptic ulcer with eradication therapy, even when asymptomatic?
  • What is the health gain for a population of screening for H pylori infection, endoscoping those who are positive, and treating those with gastric cancer even when asymptomatic?
  • What is the health gain for a population of screening for H pylori infection and treating those who are positive without endoscoping first?


The last of these is perhaps unethical to consider without considerably greater prior knowledge than presently available. The first and second should be assessable. Large studies of open-access endoscopy have been undertaken and this has been suggested as a way of screening for early, treatable gastric cancer. Usually it is limited to those patients with dyspepsia.

Performing a study with GPs and gastroenterologists which followed a simple set of rules would give the following yields for a total population of 100,000

Assume - Population 100,000 with 50,000 over the age of 40.
Test population over 40 attending surgery for whatever reason for H pylori antibodies.
Assume 45% positive and 80% of those take up offer of endoscopy.

Then - Total of 18,000 H pylori positive available for screening.
Find 4320 peptic ulcers
360 gastric cancers

Sufficient numbers would be available with a fraction of this population to produce a study of sufficient power and randomise asymptomatic peptic ulcers to eradication/no treatment and follow the natural history of the disease. A similar study could be performed for gastritis. This would be the definitive study to elucidate effects of H pylori eradication on health.

Gastric cancers should mostly be small and curable. The incidence and presentation of gastric cancer in the treated and untreated arms above would be one of the major outcomes of the study.

For gastric cancer, finding this number would substantially increase the number detected in this population. It would increase the numbers detected by about 5 - 10 fold, initially, at least.

Treatment of dyspepsia


A major uncertainty is whether eradication of H pylori is of benefit to patients with dyspepsia. As has been described, all the studies to date have an inadequate design, though of the 16 studies eight appeared to show some benefit.

A well designed study which examined H pylori positive and negative patients with dyspepsia in a randomised double-blind fashion against placebo, and which produced positive evidence of benefit in H pylori infected patients would add very considerably to the numbers of patients tested and treated.


References


  • Atherton JC, Spiller RC. The urea breath test for Helicobacter pylori. Gut 1994;35:723-5.
  • Banerjee S, Hawksby C, Miller S, Dahill S, Beattie AD, McColl KEL. Effect of Helicobacter pylori and its eradication on gastric juice ascorbic acid. Gut 1994;35:317-22.
  • Bayerdorffer E, Mannes GA, Sommer A et al. High dose omeprazole treatment combined with amoxycillin eradicates Helicobacter pylori. European Journal of Gastroenterology and Hepatology 1992;4:697-702.
  • Blaser M. Helicobacter pylori and the pathogenesis of gastroduodenal inflammation. Journal of Infectious Diseases 1990;161:626-33.
  • Borody TJ, Brandl S, Andrews P, Ferch N, Jankiewicz E, Hyland L. Use of high efficacy, lower dose triple therapy to reduce side effects of eradicating Helicobacter pylori. American Journal of Gastroenterology 1994;89:33-8.
  • Clancy RL, Cripps AW, Taylor DC, McShane L, Webster VJ. The clinical value of a saliva diagnostic assay for antibody to H. pylori. In "Helicobacter pylori - basic mechanisms to clinical cure" (Hunt RH & Tytgat GNJ, eds), KLUWER, Proceedings of the International Symposium 1993:342-50.
  • Cottrill MRB. The prevalence of Helicobacter pylori infection in patients receiving long-term H2-receptor antagonists in general practice: clinical and financial consequences of eradication using omeprazole plus amoxycillin or "triple therapy". British Journal of Medical Economics 1994;7:35-41.
  • Coughlan JG, Humphries H, Dooley C et al. Campylobacter pylori and recurrence of duodenal ulcers - a 12 month follow-up study. Lancet 1987;ii:1109-11.
  • Crabtree JE, Shallcross TM, Heatley RV, Wyatt JI. Evaluation of a commercial ELISA for serodiagnosis of Helicobacter pylori infection. Journal of Clinical Pathology 1991a;44:326-8.
  • Crabtree JE, Taylor JD, Wyatt JI, Heatley RV, Shallcross TM, Tompkins DS, Rathbone BJ. Mucosal IgA recognition of Helicobacter pylori 120kDa protein, peptic ulceration and gastric pathology. Lancet 1991b;338:332-5.
  • DeCross AJ, Puera DA. Role of H pylori in peptic ulcer disease. Contemporary Gastroenterology 1992;May:18-28.
  • Eurogast Study Group. An international association between Helicobacter pylori infection and gastric cancer. Lancet 1993;341:1359-62.
  • Forbes GM, Glaser ME, Cullen DJE, Warren JR, Christiansen KJ, Marshall BJ, Collins BJ. Duodenal ulcer treated with Helicobacter pylori eradication: seven-year follow-up. Lancet 1994;343:258-60.
  • Forman D, Newell DG, Fullerton F, Yarnell JWG, Stacey AR, Wald N, Sitas F. Association between infection with Helicobacter pylori and risk of gastric cancer: evidence from a prospective investigation. British Medical Journal 1991;302:1302-5.
  • Foreman D, Webb P, Parsonnet J. H pylori and gastric cancer. Lancet 1994;343:243-4. letter.
  • Glassman MS, Dallal S, Berzein SH, Bostwick HE, Newman LJ, Perez-Perez GI, Blaser MJ. Helicobacter pylori-related gastroduodenal disease in children. Diagnostic ultility of a enzyme-linked immunosorbent assay. Digestive Diseases and Sciences 1990;35:993-7.
  • Glynn Owen O. More research needed before belly-achers will be heard. Healthcare Management 1993; November:38-41.
  • Graham DY, Lew GM, Evans DG, Evans DJ, Klein PD. Effect of triple therapy (antibiotics plus bismuth) on duodenal ulcer healing: a randomised controlled trial. Annals of Internal Medicine 1991;115:266-69.
  • Graham DY, Lew GM, Klein PD et al. Effect of treatment of Helicobacter pylori infection on the long-term recurrence of gastric or duodenal ulcer. A randomised, controlled study. Annals of Internal Medicine 1992;116:705-8.
  • Graham DY. Treatment of peptic ulcers caused by Helicobacter pylori. New England Journal of Medicine 1993;328:349-50.
  • Graham DY, Hepps KS, Ramirez FC, Lew GM, Saeed ZA. Treatment of Helicobacter pylori reduces the rate of rebleeding in peptic ulcer disease. Scandinavian Journal of Gastroenterology 1993;28:939-42.
  • Graham DY, Lew GM, Klein PD, Evans DG, Evans DJ, Saeed ZA, Malaty HM. Effect of treatment of Helicobacter pylori infection on the long-term recurrence of gastric or duodenal ulcer. A randomised controlled study. Annals of Internal Medicine 1992;116:705-8
  • Hallissey MT, Allum WH, Jewkes AJ, Ellis DJ, Fielding JWL. Early detection of gastric cancer. British Medical Journal 1990;301:513-5.
  • Hentschel E, Brandstätter G, Dragosics B, Hirschl AM, Nemec H, Schütze K, Taufer M, Wurzer K. Effect of ranitidine and amoxycillin plus metronidazole on the eradication of Helicobacter pylori and the recurrence of duodenal ulcer. New England Journal of Medicine 1993;328:308-12.
  • Hosking SW, Ling TX, Yung MY et al. Randomised controlled trial of short term treatment to eradicate Helicobacter pylori in patients with duodenal ulcer. British Medical Journal 1992;305:502-4.
  • Hosking SW, Ling TKW, Chung SC, Yung MY, Cheng AFB, Sung JJY, Li AKC. Duodenal ulcer healing by eradication of Helicobacter pylori without anti-acid treatment: randomised controlled trial. Lancet 1994;343:508-10.
  • Johnsen R, Straume B, Førde OL, Burhol PG. Changing incidence of peptic ulcer - facts or artefacts? A cohort study from Tromsø. Journal of Epidemiology and Community Health 1992;46:433-6
  • Kosunen TU, Seppälä K, Sarna S, Sipponen P. Diagnostic value of decreasing IgG, IgA, and IgM antibody titres after eradication of Helicobacter pylori. Lancet 1992;339:893-5.
  • Labenz J, Börsch G. Evidence for the essential role of helicobacter pylori in gastric ulcer disease. Gut 1994;35:19-22.
  • Labenz J, Gyenes E, Rühl GH, Börsch G. Amoxycillin plus omeprazole versus triple therapy for eradication of Helicobacter pylori in duodenal ulcer disease: a prospective, randomised, and controlled study. Gut 1993;34:1167-70.
  • Labenz J, Gyenes E, Rühl GH, Börsch G. Efficacy of omeprazole and amoxycillin to eradicate HP. American Journal of Gastroenterology 1992;87:1271.
  • Labenz J, Gyenes E, Rühl GH, Börsch G. Omeprazole plus amoxycillin: efficacy of various treatment regimens to eradicate Helicobacter pylori. American Journal of Gastroenterology 1993;88:491-5.
  • Labenz J, Rühl GH, Bertrams J, Börsch G. Clinical course of duodenal ulcer disease one year after omeprazole plus amoxycillin or triple therapy plus ranitidine for cure of Helicobacter pylori infection. European Journal of Gastroenterology & Hepatology 1994a;6:293-7.
  • Labenz J, Rühl GH, Bertrams J, Börsch G. Medium- or high-dose omeprazole plus amoxycillin eradicates Helicobacter pylori in gastric ulcer disease. American Journal of Gastroenterology 1994b;89:726-30.
  • Madan E, Kemp J, Westblom U, Subik M, Sexton S, Cook J. Evaluation of staining methods for identifying Campylobacter pylori. American Journal of Clinical Pathology 1988;90:450-3.
  • Malaty HM, Engstrand L, Pedersen NL, Graham DY. Helicobacter pylori infection: genetic and environmental influences. Annals of Internal Medicine 1994;120:982-86.
  • Marshall BJ, Warren JR. Unidentified curved bacilli on gastric epithelium in active chronic gastritis. Lancet 1984;i:1311-5.
  • Marshall BJ, Goodwin CS, Warren JR et al. Prospective double-blind trial of duodenal ulcer relapse after eradication of Camylobacter pylori. Lancet 1988;ii:1437-42.
  • Marshall BJ. History of the discovery of C pylori. In "Campylobacter pylori in gastritis and peptic ulcer disease" (Blaser MJ, ed) New York, Igaku-Shoin, 1989:7-24.
  • MeReC Briefing. Dyspepsia. 1994;8.
  • NIH Concensus Conference. Helicobacter pylori in peptic ulcer disease. Journal of the American Medical Association 1994;272:65-69.
  • Nomura A, Stemmermann GN, Chyou P-H, Kato I, Perez-Perez GI, Blaser MJ. Helicobacter infection and gastric carcinoma among Japanese Americans in Hawaii. New England Journal of Medicine 1991;325:1132-6.
  • Nomura A, Stemmermann GN, Chyou P-H, Kato I, Perez-Perez GI, Blaser MJ. Helicobacter pylori infection and the risk for duodenal and gastric ulceration. Annals of Internal Medicine 1994;120:977-81.
  • Ohmann C, Thon K, Hengels K-J, Imhof M. Incidence and pattern of peptic ulcer bleeding in a defined geographical area. Scandinavian Journal of Gastroenterology 1992;27:571-81.
  • Parsonnet J, Freidman GD, Vandersteen DP, Chang Y, Vogelman JH, Orentreich N, Sibley RK. Helicobacter pylori infection and the risk of gastric carcinoma. New England Journal of Medicine 1991;325:1127-31.
  • Patchett S, Beattie S, Leen E, Keane C, O'Morain C. Eradicating Helicobacter pylori and symptoms of non-ulcer dyspepsia. British Medical Journal 1991;303:1238-40.
  • Patel P, Mendall MA, Khulusi S, Molineaux N, Levy J, Maxwell JD, Northfield TC. Salivary antibodies to Helicobacter pylori: screening dyspeptic patients before endoscopy. Lancet 1994;344:511-2.
  • Perez-Perez GI, Dworkin BM, chodos JE, Blaser MJ. Campylobacter antibodies in humans. Annals of Internal Medicine 1988;109:11-17.
  • Peterson WL. Helicobacter pylori and peptic ulcer disease. New England Journal of Medicine 1991;324:1043-8.
  • Pope CE. Acid-reflux disorders. New England Journal of Medicine 1994;331:656-60.
  • Pronovost AD, Rose SL, Pawlak JW, Robin H, Schneider R. Evaluation of a new immunodiagnostic assay for Helicobacter pylori antibody detection: correlation with histopathological and microbiological results. Journal of Clinical Microbiology 1994;32:46-50.
  • Rauws EJ, Tytgat GNJ. Cure of duodenal ulcer disease associated with eradication of Helicobacter pylori. Lancet 1990;335:1233-35.
  • Ryder SD, O'Reilly S, Miller RJ, Ross J, Jacyna MR, Levi AJ. Long term acid suppressing drugs in general practice. British Medical Journal 1994;308:827-30.
  • Sackett DL, Haynes RB, Guyatt GH, Tugwell P. Clinical Epidemiology: a basic science for clinical medicine. 2nd edition. Boston: Little, Brown; 1991.
  • Schütze K, Hentschel E, Brandstätter G. More on the eradication of Helicobacter pylori and the recurrence of duodenal ulcer. New England Journal of Medicine 1993;X:1356.
  • Shearman NDC. Diseases of the gastrointestinal tract and liver. Churchill Livingstone, Edinburgh, 1989:215-6.
  • Sobala GM, Crabtree JE, Pentith JA, Rathbone BJ, Shallcross TM, Wyatt JI, Dixon MF, Heatley RV, Axon ATR. Screening dyspepsia by serology to Helicobacter pylori. Lancet 1991;338:94-6.
  • Soll AH. Pathogenesis of peptic ulcer and implications for therapy. New England Journal of Medicine 1990;322:909-16.
  • Sue-Ling HM, Johnston D, Martin MF, Lansdown MRJ, McMahon MJ, Axon ATR. Gastric cancer: a curable disease in Britain. British Medical Journal 1993;307:591-6.
  • Sung JJY, Chung SCS, Ling TKW, Yung MYY, Cheng AFB, Hosking SW, Li AKC. One-year follow-up of duodenal ulcers after 1-wk triple therapy for Helicobacter pylori. American Journal of Gastroenterology 1994;89:199-202.
  • Talley NJ. A critique of therapeutic trials in Helicobacter pylori positive functional dyspepsia. Gastroenterology 1994;106:1174-83.
  • Tatsuta M, Iishi H, Okuda S, Taniguchi H, Yokota Y. The association of Helicobacter pylori with differentiated-type early gastric cancer. Cancer 1993;72:1841-5.
  • Unge P, Ekström P. Effects of combination therapy with omeprazole and an antibiotic on Helicobacter pylori and duodenal ulcer disease. Scandinavian Journal of Gastroenterology 1993;28 Suppl 196:17-18.
  • Vaira D, Miglioli M, Mulè P, Holton J, Menegatti M, Vergura M, Biasco G, Conte R, Logan RPH, Barbara L. Prevalence of peptic ulcer in Helicobacter pylori positive blood donors. Gut 1994;35:309-12.
  • Veldhuyzen van Zanten SJO. A systematic overview (meta-analysis) of outcome measures in Helicobacter pylori gastritis trials and functional dyspepsia. Scandinavian Journal of Gastroenterology 1993;28 Suppl 199:40-43.
  • Veldhuyzen van Zanten SJO, Pollak PT, Best LM, Bezanson GS, Marrie T. Increasing prevalence of Helicobacter pylori infection with age: continuous risk of infection in adults rather than cohort effect. Journal of Infectious Diseases 1994;169:434-7.
  • Veldhuyzen van Zanten SJO, Sherman PM. Indications for treatment of Helicobacter pylori infection: a systematic overview. Canadian Medical Association Journal 1994;150:189-98. systematic review.
  • Wotherspoon AC, Ortiz-Hidalgo C, Falzon MR, Isaacson PG. Helicobacter pylori-associated gastritis and primary B-cell lymphoma. Lancet 1991;338:1175-6.




Acknowledgements

A report prepared with the financial assistance of Cortecs Diagnostics and the
Anglia & Oxford Regional Health Authority R&D Directorate


These pages were written following an article on Helicobacter pylori published in the second edition of Bandolier in March 1994. The response to the article from healthcare professionals and policy makers was intense, with a call for more information. The American Medical Association guidelines published in July indicated that no detailed economic analysis had been attempted, and it seemed worthwhile to try to draw together several different aspects of the subject - effectiveness of treatment, economic impact, and testing. Cortecs Limited made the work possible.


A number of people have helped to create this book. They include Henry McQuay and Muir Gray, who were enormously encouraging, Phil Whiffen from the Churchill pharmacy who provided much background information, André Tomlin at the library of the Oxford Regional Health Authority who traced many of the papers, and Dawn Carroll and Alex Jadad helped with many of the ideas. I am particularly indebted to Shirley Holton and Sharon Hart and their colleagues in Buckinghamshire who gave freely of their time to discuss the difficulties faced in implementing healthcare policy in a Health Authority.


Many other people contributed by reading manuscripts for errors of omission, commission and simple mistakes; they include Rachel Cross, Paul Phull and Maura Moore, as well as many of those mentioned above. Thank you all.

Oxford January 1995


The Old Blue School, Lower Square, Isleworth, Middlesex TW7 6RL.

Copyright Cortecs Limited and Health Technology Evaluation Association.

First published 1995

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical , photocopying, recording, or otherwise, without prior permission of Cortecs Limited and the Health Technology Evaluation association.

This book is sold subject to the conditions that it shall not, by way of trade or otherwise, be lent, re-sold, hired out or otherwise, circulated without the publishers' prior consent in any form of binding or cover other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchase.


ISBN 1 899137 25 4