How well is Helicobacter pylori treated in usual practice ?

1Division of Gastroenterology, University of Toronto; 2Sunnybrook Health Sciences Centre; 3Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario Correspondence: Dr Jill Tinmouth, Division of Gastroenterology, University of Toronto, Sunnybrook Health Sciences Centre, HG-40, 2075 Bayview Avenue, Toronto, Ontario M4N 3M5. Telephone 416-480-5910, fax 416-480-4845, e-mail jill.tinmouth@sunnybrook.ca Received for publication April 10, 2011. Accepted May 6, 2011 Up to 70% of the developing world’s population, and up to 30% to 40% of the developed world’s population are infected with Helicobacter pylori, a Gram-negative microaerophilic bacterium (1). H pylori is a WHO class I carcinogen, and is strongly associated with the development of gastric adenocarcinoma and MALT lymphoma. It is also linked to other nonmalignant gastrointestinal diseases including nonulcer dyspepsia and peptic ulcer disease (PUD) (2-4). Eradication of H pylori decreases the recurrence of PUD and may lead to tumour regression in low-grade and localized MALT lymphoma (5,6). On completion of treatment, the American College of Gastroenterology, based largely on expert consensus, recommends confirmation of H pylori eradication in patients with known PUD and/or MALT lymphoma, those with successfully treated early gastric adenocarcinoma and those with persistent nonulcer dyspeptic symptoms (7). Appropriate methods to confirm eradication include urea breath testing or repeat endoscopy with biopsy. Repeat endoscopy to confirm healing of gastric ulcers – regardless of whether they are H pyloriassociated – may be indicated due to the risk of concomitant underlying malignancy (8,9). Endoscopic surveillance may improve survival and may be cost effective in high-risk populations in which the prevalence of gastric malignancy is high (10). Antibiotic plus proton pump inhibitor treatment of H pylori has been shown to be both efficacious and cost effective in randomized controlled trials (5). Large systematic reviews have reported significant original article

However, consistent and recent data regarding the effectiveness of H pylori eradication -defined as treatment of H pylori in 'usual' practice -are lacking.In usual clinical practice, involvement of multiple physicians in the care of individual patients, suboptimal communication between providers and inadequate patient follow-up may compromise adherence to H pylori treatment recommendations.Reports of treatment prescription rates have varied from 10% to greater than 90% (12)(13)(14)(15).As noted above, once treated with a recognized regimen, confirmation of eradication may be indicated in certain disease states; however, in usual practice, rates of confirmation in the various disease states listed above are not known.
Due to the potential serious health sequelae of untreated H pylori infection, and because large-scale trials have shown efficacy, it is important to determine whether patients found to be infected with H pylori are routinely receiving treatment in usual practice.In the current study, our objective was to determine the rates of receipt of treatment for H pylori among appropriate patients in usual practice at Sunnybrook Health Sciences Centre (SHSC) in Toronto, Ontario.Physician and patient factors associated with receipt of treatment and clinical course following eradication were also explored.

MetHods
The present analysis was a single-centre cohort study of patients with histological evidence of H pylori infection diagnosed between January 1, 2007, and December 31, 2007, at SHSC.An English language search of all pathology reports issued during the study period using the words "Helicobacter" and "stomach biopsy" was used to identify eligible patients.Additionally, all reports coded as "stomach biopsy" -a code inclusive of all anatomical sites in the stomach -in the laboratory information system during the study period were retrieved to ensure that cases were not missed due to variation in diagnostic terminology.Pathology reports were then reviewed and only patients with histological evidence of H pylori infection were retained in the cohort.
Hospital charts and endoscopist office charts for all patients in the cohort were reviewed.Data were collected from the clinical notes and endoscopy report, and the pathology report in the charts using a standardized data collection form.From the chart review, patients who did not appear to have received treatment and, where indicated, to have had confirmation of eradication and to have received repeat esophagogastroduodenoscopy were identified.The physicians who performed endoscopy on these patients were then asked to review their records to confirm that the patient had not been treated and, where indicated, had not received confirmation of eradication or repeat esophagogastroduodenoscopy.If confirmed, the physicians were asked to provide reasons for these apparent lapses in appropriate care.
From the charts, demographic data (age, sex and inpatient/outpatient status), clinical data (medical comorbidities and admission medications [including previous nonsteroidal anti-inflammatory drug/ acetylsalicylic acid use]), endoscopic data (endoscopy indication and endoscopic findings) and pathological diagnoses other than H pylori infection were also abstracted.
Ethics approval was obtained from the Research and Ethics Board of SHSC.

study definitions
Receipt of eradication therapy for H pylori was defined as clear notation in the hospital or office chart that a standard eradication regimen had been prescribed, or that another physician was informed of the diagnosis and asked to provide treatment.An assumption by the endoscopist that another physician would treat the patient, without written documentation of a direct request, was not considered to be sufficient.Confirmation of eradication was defined as written documentation of any accepted method (upper endoscopy with biopsy or urea breath testing, but not serological testing) to verify cure of H pylori infection.For the purposes of the current study, confirmation of eradication was required only in patients with diagnoses of PUD or MALT lymphoma, while repeat endoscopy was indicated only in cases of gastric ulcer or MALT lymphoma.Complete treatment was defined as a combination of receipt of eradication therapy and confirmation of eradication, if indicated.

outcomes and statistical analyses
The primary outcome measure was the proportion of appropriate patients in the entire cohort who received eradication therapy.Patients were considered 'inappropriate' for treatment and excluded from the cohort if they died before treatment could be prescribed or if their life expectancy was so short that the treating physician believed that treatment was not warranted.Secondary outcome measures were the proportions of appropriate patients who received confirmation of eradication, repeat endoscopy and complete treatment, if indicated.
Depending on the nature of the data, proportions or median values with interquartile ranges were used to describe the baseline characteristics and outcomes of the entire cohort.Similarly, outcomes were assessed in the following subgroups: inpatients versus outpatients, and according to type of follow-up (follow-up appointment with any type of physician versus none; follow-up appointment with gastroenterologist versus any other specialty).ORs with 95% CIs were used to compare subgroups in a univariate fashion, and P<0.05 was considered to be statistically significant.
In the subgroup analyses, outpatients were more likely than inpatients to receive eradication therapy (OR 10.3 [95% CI 2.6 to 40.4]) and complete therapy (OR 13.2 [95% CI 3.8 to 45.7]).Patients with any follow-up appointment were more likely to receive eradication therapy (OR 12.0 [95% CI 3.0 to 47.5]) and complete therapy (OR 25.7 [95% CI 6.5 to 99.1]) than those without documented follow-up.Patients who had follow-up with a gastroenterologist, as opposed to another physician, were more likely to receive eradication therapy and complete therapy (OR 10.8 [95% CI 1.3 to 88.5]); however, these results were not statistically significant (Table 3).

disCussion
In the present study of usual clinical practice, approximately 90% of appropriate patients received treatment.Inpatients were less likely to receive eradication therapy and complete treatment.Similarly, we found that patients with formal follow-up were more likely to receive eradication therapy and complete therapy.There was a nonsignificant trend toward receiving eradication therapy and complete therapy in patients who specifically had follow-up with a gastroenterologist.Our study, unlike others in this area, also examined care following H pylori treatment and found that a significant proportion of patients did not receive confirmation of eradication.
Other studies that addressed a similar question have reported varying rates of eradication therapy.A recent study investigating H pylori treatment rates in largely outpatient Asian immigrants (16) reported suboptimal treatment, with rates of 57%.Similarly, an earlier study in Medicare patients in the United States (14) found treatment rates of 53% to 60%.In contrast, one other study from London, Ontario, found higher treatment rates comparable with ours (17).
Certain subgroups may be at risk of suboptimal treatment.We found, as have others (17), that inpatients were at risk of failing to receive treatment.One explanation may be that inpatients, as a whole, tend be more ill than outpatients and have multiple comorbidities.Therefore, other possibly more pressing health issues may take precedence, and treatment of 'lesser' conditions, such as H pylori infection, may not be a priority.Second, at our institution, as in many other hospitals, inpatients often undergo endoscopy while admitted under nongastrointestinal physicians for other reasons.At the SHSC, pathology reports are often not available before discharge of a short-stay patient and are routinely sent to the endoscopist only (and not the admitting or family physician).Among the inpatients in our study, only nine of 24 (38%) had reports available before discharge.Conversely, if H pylori infection is determined by serological testing, that result is sent as a hard copy report to the admitting physician and not to the endoscopist.If a formal follow-up appointment with the endoscopist is not arranged, as may be the case for some endoscopists at our institution, treatment of H pylori may be overlooked by the admitting or family physician.
A similar study also confirmed that having a 'regular doctor' was associated with higher rates of H pylori treatment completion (16).With a formal follow-up appointment, physicians have the opportunity to ensure that patients have completed their eradication therapy as prescribed and to arrange confirmation of eradication, if indicated.Gastroenterologists, as opposed to other following physicians, might be expected to focus more on gastrointestinal diseases and to treat H pylori infection if present.In our study, the trend toward higher treatment rates in patients who followed up with their gastroenterologists was nonsignificant -probably due to the small sample size.
Although our overall rates of H pylori treatment are reasonable, there is still room for improvement, particularly among certain subgroups such as inpatients.Based on our findings, several quality of care interventions could be considered such as routine follow-up appointments and modifying the way that results are reported or distributed.In terms of the latter, pathology departments may wish to consider mandatory copying of reports to referring and primary care physicians  to improve communication and to reduce the likelihood that the result will be overlooked.Increased patient involvement, such as access to one's own test results on a secure web-based system, might also improve these processes of care.Finally, educational programs for general physicians and gastroenterologists may be useful in addressing the deficiencies we have identified in post-treatment care, including confirmation of eradication and complete therapy, when indicated.Formal evaluation of these quality of care interventions would be useful in identifying practices that can actually be implemented to increase rates of complete therapy.
Because the present analysis was a single-centre study, the generalizability of our results may be in question.However, we believe that the challenges in the treatment of H pylori are common, and that our findings may be applicable to other centres and to other areas of patient care, particularly where procedures and pathology reports are involved.For example, there may be similar deficiencies in the diagnosis and care of patients with other conditions that involve communication between multiple physicians across different locations such as follow-up surveillance colonoscopies after the diagnosis of tubular adenoma or ongoing management of cervical dysplasia.

ConCLusion
Our one-year analysis at SHSC showed that H pylori treatment rates were good, particularly in the outpatient population and in patients with regular follow-up.However, it is clear that there are opportunities for improvement in the administration of treatment and follow-up of patients with H pylori infection.The lessons learned from H pylori may be applicable to other areas of patient care, especially where conditions are easily identifiable, treatable or preventable.Additional controlled studies are required to identify quality interventions that are effective in ensuring best practices in patients infected with H pylori.

Table 3 OR of primary and secondary outcomes according to patient subgroup
*Follow-up with family physician, general internist or referring physician.EGD Esophagogastroduodenoscopy; N/A Not applicable