| Sign In to gain access to subscriptions and/or personal tools. |
Questionable Performance of a Questionnaire
1 Boston University School of Dental Medicine, Department of General Dentistry. (Dr. Janket is supported by a grant from The American Heart Association, # 0635351N.) To the Editor: According to a recent study, self-reported periodontitis was justified as a proxy for the clinical assessment of periodontitis (Pitiphat et al., 2008). The authors reported sensitivity and specificity of 18.4% and 98.7%, respectively, when validated against dental radiographs. Further, they declared the questionnaire to be valid because the sum of predictive values was larger than 120, citing a study published in the JDR (Blicher et al., 2005). We disagree with this affirmation for several reasons. Sensitivity and specificity are two different properties of a diagnostic test that should be assessed independently but simultaneously. Sensitivity/specificity and predictive values (positive/negative) being complementary, one property can vary, causing opposite changes in the other. Therefore, summing these properties would be inappropriate. Moreover, since predictive values measure usefulness of the test, not accuracy, and since accuracy supersedes usefulness, predictive values should be considered only after accuracy has been established (Zweig and Campbell, 1993). We present evidence that, in the presence of extreme levels of sensitivity/specificity, the use of predictive values (positive/negative) is irrelevant.
Let us consider a hypothetical cohort of 1000 persons (summarized in the Table
Although the questionnaire failed to identify over 80% of periodontitis cases, the ‘predictive value positive’ is 79% (37/47) and ‘negative’ is 83% (790/953). The performance of a questionnaire that failed to identify over 80% of the disease it represents is unacceptable, and using predictive values to validate its accuracy is misleading.
REFERENCES
Journal of Dental Research, Vol. 87, No. 10,
e1 (2008)
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
