The real reasons not to rely on severity scores*

    loading  Checking for direct PDF access through Ovid


In this issue of Critical Care Medicine, Vincent et al (1) provide a thoughtful review of the risks associated with the use of scoring systems to guide treatment decisions or to determine eligibility to participate in clinical trials. The authors acknowledge the value these systems have to the critical care community but also identify examples of their misuse. They articulate 10 reasons not to use severity scoring systems for either clinical decision-making or for entry into clinical trials. Whereas the 10 explanations are interesting, there is significant overlap of the various reasons they describe. In addition, whereas the purpose of the discussion was meant to provide a general overview of all scoring systems, the discussion points relate primarily to the Acute Physiology and Chronic Health Evaluation II score, albeit one of the most commonly utilized severity scoring systems both for clinical decision-making and as part of clinical trials (2–4). Despite these limitations, the message is important for every critical care provider and clinical investigator to understand. Severity scoring systems were developed and are used for a variety of reasons, but none was designed to determine when an individual patient should receive a specific therapy or, more importantly, when therapy should not be provided. As importantly, as articulated by the authors, severity scores have significant limitations when they are used to define the patient population to be included in a clinical trial. Unfortunately, although the authors suggest that there might be better measures to define therapeutic options, they offer few real alternatives. Biomarkers, genetic profiling, and other measures might be helpful in the future, but at the present time the options are limited (5). Perhaps the absence of obvious alternatives reflects our limited ability to predict outcome for the individual critically ill patient and the critical need to broaden entry criteria for clinical trials to better-understand which patients are most likely to respond and under what circumstances.In evaluating the arguments brought forth by the authors when they recommend not using severity scores for treatment decisions, there are three critical points that should be emphasized. First, the scoring systems were developed or refined to integrate multiple data points, including both chronic and acute clinical conditions to predict outcomes of care and assess the risk of death for populations (2–4, 6), not to dictate clinical decision-making for individual patients, nor to determine which patients should be enrolled in clinical trials. Despite this reality, the facts that clinicians use severity scores to classify patients, make clinical decisions, and advise families are understandable. Critical care providers are looking for straightforward answers about the most appropriate way to treat their patients, when to introduce new therapies and for whom, and, as importantly for those patients who do not respond to interventions, when to withdraw them. The concept of looking at a computer screen for the results of a scoring system that integrates clinical and demographic data to help us make clinical decisions is comforting. It offers a theoretical framework to objectify our clinical assessment and support our clinical judgment. Fortunately (or unfortunately), clinical skills and judgment have yet to be replaced by severity scoring systems or, for that matter, clinical guidelines or check lists. Each of the scoring systems can provide some guidance to the clinician, patient, and family, but cannot and will never be the ultimate determinant of what care is provided, when care should be withheld, or when a patient is a candidate for entry into a clinical trial. In fact, the scoring system is often misleading or subject to error (7, 8).

    loading  Loading Related Articles