HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): A. David Slater Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Edmonds, H. L. Articles by Shields, C. B. Search for Related Content PubMed PubMed Citation Articles by Edmonds, H. L., Jr. Articles by Shields, C. B.

J Thorac Cardiovasc Surg 1994;107:1378
© 1994 Mosby, Inc.

LETTERS TO THE EDITOR

Use of the odds ratio in outcome prediction

Department of Anesthesiology
Information Technology and Biostatistics^a

To the Editor:

In a recent JOURNAL report, ¹ we described the use of electroencephalographic (EEG) monitoring to predict neurologic outcome after myocardial revascularization The customary false-positive and false-negative rates were used to characterize the effectiveness of the EEG as a predictor of postoperative disorientation. However, this approach has a number of limitations. First, it may lead to ambiguity because these terms have no universally accepted definition. ² Second, these rates, as well as therelated terms sensitivity and specificity, are most appropriately used to describe the probability that a diagnostic test will correctly detect an existing condition. Unfortunately, this situation does not apply to the preoperative or intraoperative prediction of postoperative outcome. Third, only a point estimate of these inferential statistical quantities is available. Probability-based estimates such as a significance level or the more desirable confidence interval, ³ are not used with false-positive or false-negative rates.

These limitations may be overcome through application of a simple statistical alternative, the odds ratio. ⁴ This measure provides an unambiguous estimate of predictor efficacy. For example, with data from phase one of our study, ¹ we compared the information provided by the two approaches. In our original analysis, the intraoperative EEG abnormality had 68% false-positive and 8% false-negative rates in predicting disorientation. Because of the previously mentioned limitations of these measures, readers may reach varying conclusions regarding the prognostic value of the EEG.

In contrast, the odds ratio provides a clearer estimate of the association between the EEG change and postoperative disorientation. The ratio of 11.9 implies that one is about 12 times more likely to see postoperative disorientation in patients with a prominent EEG change than in patients without such a change. The ratio is significant (p 0.01) and has a 95% confidence interval of 2.7 to 53.7. This wide interval indicates considerable interpatient variability, the effects of a relatively small sample size, or both. Nevertheless, even readers unfamiliar with statistical analysis can appreciate the magnitude of predictor effectiveness. Because outcome prediction forms an essential ingredient in many JOURNAL articles, we encourage other authors to consider use of the odds ratio.

References

1. Edmonds HL Jr, Griffiths LK, van der Laken J, et al. Quantitative electroencephalographic monitoring during myocardial revascularization predicts postoperative disorientation and improves outcome. J THORAC CARDIOVASC SURG 1992;103:555-63.[Abstract]
   
2. Fleiss JL. Statistical methods for rates and proportions. New York: John Wiley, 1981:1-13, 61-74.
   
3. Simon R. Confidence intervals for reporting results of clinical trials. Ann Intern Med 1986;105:429-35.
   
4. Matthews DE, Farewell VT. Using and understanding medical statistics. 2nd rev. ed. Basel: Karger, 1988:212-5.
   

This article has been cited by other articles:

				D. B. McElhinney, V. M. Reddy, P. Moore, and F. L. Hanley Bidirectional Cavopulmonary Shunt in Patients With Anomalies of Systemic and Pulmonary Venous Drainage Ann. Thorac. Surg., June 1, 1997; 63(6): 1676 - 1684. [Abstract] [Full Text]

				A. Carotti, B. Marino, G. Oppido, and C. Marcelletti Biventricular repair in patients with left isomerism J. Thorac. Cardiovasc. Surg., October 1, 1995; 110(4): 1151 - 1152. [Full Text]

This Article Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): A. David Slater Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Edmonds, H. L. Articles by Shields, C. B. Search for Related Content PubMed PubMed Citation Articles by Edmonds, H. L., Jr. Articles by Shields, C. B.