Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
Clinical significance of test seems dubious
- Ronald Sultana (16 July 2001)
-
muddy waters
- Alexander Williams (19 July 2001)
-
Statistical discrepancies
- Adam Jacobs (26 July 2001)
-
H pylori test and treat - problems with selection by dyspeptic symptoms.
- Kenneth E L McColl (9 August 2001)
|
|
|||
|
Ronald Sultana, Emergency Physician Department of Emergency Meidicine, The Royal Melbourne Hospital, Parkville, Australia
Send response to journal:
|
Dear Editor, I read with interest the article by Weijnen et al on testing for helicobacter pylori in dyspeptic patients. The authors stated that the test provided additional diagnostic information to those patients deemed to have a high risk of peptic ulcer disease. Closer scrutiny of Table 4 shows that the test actually performed similarly in the low and high risk groups. Positive likelihood ratios calculated for the two groups are 1.8 (95% CI 1.1-3.1) and 1.8(1.2-2.7) respectively. The negative likelihood ratios were 0.7(0.4-1.1) and 0.6(0.3 -1) respectively. Assuming a prevalence of 16% for peptic ulcer disease this equates to a post-test probability of 18.6%-34% if one utilises the confidence limits of the positive likelihood ratio in the high risk group. Note that if the test were negative in this group the post-test probability would be 5.4%- 16%.Based on this the test adds very little further information to that of history-taking. The authors do not state whether the changes in predictive values(16% to 26% and 16% to 10%)were statistically significant however their clinical significance,in terms of diagnosing peptic ulcer disease,seems dubious even for the high risk group. |
|||
|
|
|||
|
Alexander Williams, General Practitioner St Thomas Health Centre
Send response to journal:
|
Dear editor, The suggestion by Weijnen(1)to test and treat patients athigh risk of peptic ulceration seems to muddy the waters in the debate about testing for H- Pylori infection in primary care In the 38 patients they identified as having a peptic ulcer only 22(58%) had a positive non invasive H-Pylori test, although their rate of detection overall was increaced in all patients from 31% to 41% by using invasive tests of culture or histology suggesting serological testing is not as sensitive. It has previosly been reported(2) that the sensitivity of H-Pylori elisa kits gave an average sensitivity of 85%(low and high extremes 495 and 99% respectively). Why are their rates of detection so low even though 33 of the 38 patients had a duodenal ulcer, when it has been shown that patients with a DU have almost 100% infection with H-Pylori(3). This apparent discrepancy in their results will make it difficult to generalise their results into primary care. As a practice we still follow the recommendations in Guidelines(4),a free publication to all general practitioners three times a year,that summarises current evidence.This makes a suggestion from the Primary Care Society for Gastroenterology that routine testing of patients with uninvestigated dyspepsia is not recommended at first presentation.At subsequent presentations testing and referral for endoscopy is appropriate. 1.Weijnen CF,Numans ME,J de Weit N,SmoutJPM,Moons GM,Verheij TJM,Hoes AW. Testing for Helicobacter pylori in dyspeptic patients suspected of peptic ulcer disease in primary care:cross sectional study.BMJ 2001;323:71 -5 2.Agreus L,Talley N.Challenges in managing dyspepsia in general practice.BMJ1977;315:1284-8 3.Blaser M.Helicobacter pylori and the pathogenisis of gastroduodenal ulceration.J of Infect Dis 1990;162:623-33 4Guidelines-summarising clinical guidelines for primary care. Medendium Group Publishing 1999 |
|||
|
|
|||
|
Adam Jacobs, Director Dianthus Medical Limited
Send response to journal:
|
I am confused by Weijnen et al's description of their statistical methods and how they fit with the data they present. The methods section states that all variables found to be univariate predictors of peptic ulcer with p < 0.25 were entered in the multivariate regression model. However, the results section says that age was included in the model, although Table 2 shows that it was not predictive (p = 0.67). Table 2 also shows that 'hiatal hernia' and 'pain after meal' both had p = 0.24, and so should have been included in the multivariate model, but neither of them was. Are these discrepancies due to a typing mistake, or is there another explanation? |
|||
|
|
|||
|
Kenneth E L McColl, Professor of Gastroenterology University Dept. of Medicine & Therapeutics, Western Infirmary, Glasgow, G11 6NT.
Send response to journal:
|
Dear Sir, In their recent paper, Weijnen et al recommend that the H pylori test and treat strategy should be restricted to dyspeptic patients with a history indicating high risk of underlying ulcer disease. This would include patients with a past history of peptic ulcer or those who were both smokers and experienced pain on an empty stomach. Their recommendation is based on their finding that the prevalence of underlying ulcer disease in such patients with a positive H pylori test is 26% compared to only 7% in their other dyspeptic patients with a positive H pylori test. We agree that the proposed strategy is attractive in reducing the number of patients treated with antibiotics per ulcer cured. However, due to the insensitivity of clinical history in predicting ulcers it will deprive a substantial proportion of dyspeptic patients of a simple long- term cure of their underlying ulcer disease. Indeed, the data in the Weijnen paper show that their strategy would leave 36% of their H pylori positive ulcer patients uncured of their chronic disease, at risk of subsequent ulcer complications and requiring long-term acid inhibitory medication. We also disagree with their assertation that the likelihood of underlying ulcer disease is the only factor in favour of treating H pylori infection in dyspeptic patients. There are several benefits of treating the infection in dyspeptic patients without ulcer disease. These include: (i) curing symptoms in 9% such subjects(2), (ii) removing their recognised increased risk of subsequent ulcer disease(3), (iii) removing a recognised risk factor for gastric cancer and lymphoma(4) and (iv) removing the risk of developing atrophic gastritis with subsequent proton pump inhibitor therapy(5). For the above reasons, it seems inappropriate to restrict the H pylori test and treat strategy to patients whose history indicates a higher risk of ulcer disease. KENNETH E.L. McCOLL LILIAN S. MURRAY REFERENCES 1. Weijnen C F, Numans M E, de Wit N J, Smouth A J P M, Moons K G M, Verheij T J M, Hoes A W. Testing for Helicobacter pylori in dyspeptic patients suspected of peptic ulcer disease in primary care: cross sectional study. BMJ, 2001; 323: 71-75. 2. Moayyedi P, Soo S, Deeks J, Forman D, Mason J, Innes M, Delaney B, on behalf of the Dyspepsia Review Group. Systematic review and economic evaluation of Helicobacter pylori eradication treatment for non-ulcer dyspepsia. BMJ, 2000; 321: 659-664. 3. McColl K E L. Should we eradication Helicobacter pylori in non-ulcer dyspepsia? Gut, 2001; 2001; 48: 759-761. 4. Blum A L, Talley N J, O'Morain C, Veldhuyzen van Zanten S, Labenz J, Stolte M, Louw J A, Stubberod A, Theodors A, Sundin M, Bolling-Sternevald E, Ed D N, Junghard O. Lack of effect of treating Helicobacter pylori infection in patients with non-ulcer dyspepsia. N. Engl. J. Med., 1998; 339: 1875-1881. 5. Kuipers E J, Lundell L, Klinkenberg-Knol E C, Havu N, Festen H P M, Liedman B, Lamers C B H W, Jansen J B M J, Dalenback, Snel P, Nelis G F, Meuwissen S G M. Atrophic gastritis and Helicobacter pylori infection in patients with reflux esophagitis treated with omeprazole or fundoplication. New Eng. J. Med., 1996; 33: 1018-1022. |
|||