Surviving Transfer for Sepsis: Size Matters—Or Does It?
Knowledge of geographic and economic disparities led to discussions of regionalization of healthcare at least a half-century ago (2). Regionalization can be achieved in different ways, but at its heart the term means the establishment of an administrative and governance system for organizing and delivering healthcare to a defined population, usually geographically defined. The concept encompasses organized transfer of patients from smaller or more remote facilities to larger facilities with more capacity, expertise, and equipment. Under fee for service payment models, regionalization has been difficult to achieve, but consolidation of hospitals and physicians into healthcare systems in anticipation of value-based reimbursement has resulted in de facto regionalization for some patients in some regions of the country. Because of the need for specialized expertise and equipment, discussions of regionalization of critical care services appeared in Critical Care Medicine at least 2 decades ago (3). In recent years, the discussions have often invoked the “success” of the regional trauma networks, supported by data demonstrating reductions in regional and even statewide mortality in mostly rural states after institution of a regionalized trauma system (4, 5). Success is a relative term, however, as rural residents are still more than twice as likely to die if they are involved in a motor vehicle crash than are urban residents (6).
Severe sepsis, as perhaps the most common critical illness, has the potential to benefit from regionalization and could, in fact, serve as a prototype illness for it because of its ubiquity. However, if regionalization for patients with severe sepsis were to be accomplished, it would be important to determine which hospitals should function as regional centers of expertise, much as trauma centers are designated with levels of expertise. In this issue of Critical Care Medicine, Ofoma et al (7) address this question directly by performing an analysis of outcomes of severe sepsis patients who were transferred from one facility to another for their care.
The authors used the Nationwide Inpatient Sample (NIS) from 2002 through 2011 to evaluate outcomes of patients with severe sepsis who were received in transfer and treated at referral hospitals. The NIS is a stratified sample of administrative data from approximately 20% of discharges from American hospitals, capturing hospitals of all sizes and all patient volumes. The authors sought to determine whether the total annual volume of septic patients treated at the receiving hospital is predictive of the outcome of transferred septic patients as has been shown for overall sepsis outcomes. They divided the referral hospitals into quintiles, according to the annual volume of septic patients treated, with the hypothesis that in-hospital mortality would be lower in the high volume hospitals. Secondary outcomes included length of stay (LOS) and hospital charges. Age, sex, number of organ dysfunctions, All Patient Refined-Diagnosis Related Group severity of illness, and other covariates were controlled for in the multivariable logistic regression models for outcomes.