© 2007 American Thoracic Society
One More Prediction Rule: Has Anything New Been Added?From the Authors:The main point of the rule presented in our recent article (1) is to help identify at admission patients with severe community-acquired pneumonia (SCAP), or at risk for developing it. This is the first prediction rule that achieves this goal, which makes it different and worthy. Regarding Dr. Bhatt's letter, we will try to clarify some points. Patients with diabetes and chronic renal failure were included in the study. Our rule was not created to identify patients who must be admitted to the ICU directly. It was developed to identify patients who are at risk for an adverse outcome and therefore who need additional monitoring and more aggressive treatment after the first evaluation in the emergency department. We believe that our SCAP rule is likely to be useful for less experienced physicians to alert them to a high-risk group of patients for whom consultation with more experienced clinicians is required. We measured the average time elapsed between hospital admission and initial diagnosis and treatment with vasopressors (8 h), the onset of mechanical ventilation (29 h), and hospital discharge (3.92 d). Therefore, our rule can identify at admission patients developing SCAP within and beyond the first 24 hours. Our rule was compared with three other commonly used rules, two of them not specifically designed to evaluate SCAP. We evaluated several cutoff points for PSI, CURB65, and our rule, but only those with the best sensitivity and specificity were included in the article. Our rule did better than the other scores in the internal validation cohort, obtaining statistically significant differences. Regarding the external cohort, the results obtained by the four rules were similar, mainly because in this cohort some data were missing and did not include outpatients. As stated in the article, the best use of our rule is as a stratification tool to be used as an adjunct to clinical judgment. Nevertheless, a prospective validation is required to assess the generalization of this finding. In relation to the request for information on the results of the R2, we must indicate that the R2 is obtained in linear regression models. In our case, we used logistic regression models exclusively, where the AUC parameter provides similar information or, alternatively, the –2logL would be the appropriate parameter. We would like to remark that approaches to severity assessment for CAP are evolving. Every effort and improvement in this route must be welcomed.
Hospital de Galdakao, Galdakao, Bizkaia, Spain FOOTNOTES Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
