DOI: 10.1097/01.CCM.0000277516.04507.B7

,  

,  

PMID: 17667246

Issn Print: 0090-3493

Publication Date: 2007/08/01


Print

Higher volume isn't always better*

Mark A. Malangoni

    loading  Checking for direct PDF access through Ovid

Excerpt

The medical literature and lay press are replete with articles about the relationship between hospital volume and outcomes. This association was first recognized by Luft and colleagues (1), who identified a direct relationship between mortality and the number of operations done at a hospital, primarily in cardiac surgery. The same association was reported for other high-risk operations and medical conditions (2). The ability to analyze this relationship by examining large patient databases and concerns by the lay public about healthcare delivery has enhanced interest in this topic. There has been a proliferation of reports purporting the association of high volume with improved outcomes and reduced mortality among a potpourri of medical conditions, some of which are commonly performed at the majority of hospitals in the United States, and this has fostered the debate about whether all emergency medical care should be regionalized (3). Many believe that the volume-outcome relationship is a surrogate for quality health care, although others recognize the shortcomings of this assumption (4).
In this issue of Critical Care Medicine, Dr. Clark and colleagues (5) report on an analysis of mortality among Medicare beneficiaries using Medicare Provider Analysis and Review data from 1999. These authors examined 30-day mortality among the geriatric population sustaining injury during that 1-yr period and determined that there was no relationship between mortality and hospital volume. They corrected for the severity of the index injury, comorbid conditions, rural or urban location, and interhospital transfer. The authors also compared their results with previous reports by Rzepka et al. (6) and Marcin and Romano (7), who found opposing relationships between mortality among older patients treated at designated trauma centers. Each of these studies has its faults, as does the current report.
The authors reached two main conclusions. The abstract concludes that “existing systems of trauma care result in similar survival for older patients with serious injuries seen first at low-volume or high-volume hospitals.” This conclusion is clearly supported by their results. However, analyses of most existing large databases are encumbered by issues regarding the validity of using data collected for financial purposes to measure patient outcome and the reliability of coding for comorbidities. More important, critical information about the physiologic condition of the patient at the time of hospital arrival, which has a critical correlation with mortality, is missing. The presence of hemorrhagic shock, an admission base deficit of ≤−6, a Trauma Score <7, and a respiratory rate of <10 are associated with increased mortality in geriatric trauma patients (8, 9), yet this vital information is absent in the Medicare Provider Analysis and Review database. The inability to provide risk-adjusted mortality in geriatric patients is a major limitation of this study and its comparators.
The concluding sentence of the paper states that “existing systems of trauma care thus appear to be effective in the triage of older patients with serious injuries.” This is a speculative assumption at best, since Dr. Clark and colleagues (5) did not evaluate the mortality of patients who were taken to designated trauma centers. MacKenzie et al. (10) demonstrated that the risk of death is significantly lower when care is provided in a trauma center than in a facility other than a trauma center. Although the benefit of trauma center care was not statistically significant for patients who were ≥55 yrs of age, there was a 10% reduction in mortality at 30 days after injury for the older injured patients treated at trauma centers. The relatively small number of older patients with serious injury in their study resulted in a wide confidence interval and lack of interaction (10).
    loading  Loading Related Articles