| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ann Thorac Surg 1995;60:404
© 1995 The Society of Thoracic Surgeons
St. Louis University Medical Center, 3635 Vista Ave, PO Box 15250, St. Louis, MO 63110
This report has all the inherent disadvantages of any study in which the data are acquired retrospectively from hospital records, including incomplete data. By various adjustments Hartz and associates have attempted to compensate for these weaknesses. This technique of analyzing retrospectively acquired data is common at all levels of outcomes analysis today. The search for additional significant risk factors that may be obtained retrospectively from hospital records has continued as this report demonstrates. Sometimes these factors may be relevant, but at other times determination of clinical relevance may be difficult. However, if one is left with this form of data collection, this type of analysis may be all that will be universally available. It is important to understand, however, that extensive extrapolation from relatively soft data may be inappropriate as addressed recently by review of data analysis from the State of New York [1].
In this review of four clinical data sets Hartz and associates have found elevated blood urea nitrogen level to correlate with increased coronary artery bypass operative mortality. In their comments they address possible reasons for this correlation. They postulate that blood urea nitrogen level elevation is the result of decreased renal blood flow due to reduced cardiac output. There are several causes of prerenal azotemia, of which this is one. Reduced intravascular volume as the result of intense diuresis is another. This, of course, indirectly may be the result of reduced cardiac output producing the congestive failure that prompted the diuresis.
Although it is important to continue to search for additional risk factors for outcomes analysis, it is essential that we not rely on data points that may be weak or clinically irrelevant. A risk factor extracted from a record may be especially seductive if it is numeric and has an accepted normal range. The mere fact that a risk factor is numeric may be regarded by those unsophisticated in clinical medicine as verifying its scientific significance.
Furthermore, it is important for us to continue to analyze data that are acquired prospectively by trained personnel as demonstrated initially by the Coronary Artery Surgery Study and as being carried on today by the Continuous Improvement in Cardiac Surgery Study of the Veterans Administration.
Hartz and associates are to be commended for their effort to further our knowledge of risk stratification and outcomes analysis. However, we as physicians must be certain that our patients are not denied evaluation and treatment by those relying solely on a numeric value. A numeric data point, just as the other aspects of a patient's clinical evaluation, should be interpreted in the context of the entire patient.
Reference
Related Article
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
