| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
Electronic Letters to:
|
|
Electronic letters published:
![[Read eLetter]](/icons/misc/sectionDOWN.gif){kind=icon}
Untitled
|
|
|||
|
James M O'Brien, MD, MS The Ohio State University Medical Center, Naeem A. Ali and Scott K. Aberegg
Send letter to journal:
obrien-2{at}medctr.osu.edu James M O'Brien, et al.
|
In their recent article, Ray et al.(1) examine the association between body mass index (BMI) and outcomes from critical illness. The authors contend that their report is “of maximal usefulness to critical care clinicians” as “actual rates of outcomes are easier to interpret than odds ratios based on multivariate techniques.” In observational studies, risk-adjustment is required to determine the independent association between the explanatory and outcome variables. The authors provide no information about the calibration of their risk-adjusting model using APACHE II. Considering the advances in critical care since its publication twenty years ago(2), it is not surprising that the APACHE- standardized mortality rates predicted a greater number of deaths than actually occurred. Methodologic rigor for observational studies includes describing the strategy of model estimation, the form (i.e. continuous versus categorical) of the included covariates, and the calibration of the model(3-5). We also wonder about differences in provided care among the BMI groups. While the authors present data on complications, this is based on a per-patient basis, rather than a per-procedure basis. If obese patients received disparate care, this may obscure a true difference in rates adjusted for this variation. Lack of attention to processes of provided care also allows for misclassification of excess body weight. We found that adjustment for fluid balance changed the BMI categorization of approximately one in every seven subjects in a clinical trial(6). Patients with BMIs in the normal range might be classified as obese because of aggressive volume resuscitation occurring prior to weight measurement for the study. These pseudo-obese subjects might have a lower risk of mortality than other subjects with normal BMIs who do not receive adequate resuscitation and, therefore, remain in the normal BMI category and die because of inadequate care. Without a standard assessment of weight and exploration of the effect of fluid balance, this analysis remains limited. Finally, the authors criticize the study by Goulenok et al.(7) because it did not use the WHO classifications of BMI. Since the relationship between BMI and outcome remains obscure, Goulenok’s analysis uses a reasonable approach and should not be dismissed unless directly tested. This study simply argues against an association between WHO BMI categories and outcome. Alternate classification for critically ill patients can only be tested by greater statistical exploration with adequate adjustment for confounding, including variations in provided care. 1 Ray DE, Matchett SC, Baker K, et al. The effect of body mass index on patient outcomes in a medical ICU. Chest 2005; 127:2125-2131 2 Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity of disease classification system. Crit Care Med 1985; 13:818-829 3 Altman DG, Goodman SN, Schroter S. How statistical expertise is used in medical research. Jama 2002; 287:2817-2820 4 Goodman SN, Altman DG, George SL. Statistical reviewing policies of medical journals: caveat lector? J Gen Intern Med 1998; 13:753-756 5 Moss M, Wellman DA, Cotsonis GA. An appraisal of multivariable logistic models in the pulmonary and critical care literature. Chest 2003; 123:923-928 6 O'Brien JM, Jr., Welsh CH, Fish RH, et al. Excess body weight is not independently associated with outcome in mechanically ventilated patients with acute lung injury. Ann Intern Med 2004; 140:338-345 7 Goulenok C, Monchi M, Chiche JD, et al. Influence of overweight on ICU mortality: a prospective study. Chest 2004; 125:1441-1445 |
|||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
eLetters: