The critical illness volume-outcome relationship: Updates and implications

The volume-outcome relationship in critical care is one of the more replicated global findings in health services research: patients admitted to higher volume hospitals and ICUs that treat greater numbers of critically ill patients tend to demonstrate higher survival rates. Much like the well-established volume-outcome surgical relationship, in which higher surgical volumes regularly translate to superior operative outcomes, the theory is that high(er) patient volume allows for accumulated institutional and clinician experience, standardized approaches, and iterative quality improvement.

An important conceptual and analytic question to this volume-outcome relationship in critical care is how volume itself is defined and measured. Many (but not all) prior studies have measured volume using aggregate annual ICU admissions, a single, undifferentiated count that does not consider ICU case-mix differences. There is, however, substantial variation between hospitals in terms of patient acuity and ICU utilization—the threshold at which patients of a given acuity are admitted to the ICU—making a straight measure and comparison of ICU volume lack important nuance. Further, there is among-hospital variation in how sensitive ICU admission decisions are to capacity strain, with some hospitals more substantially than others increasing the acuity threshold for ICU admission (i.e., admitting fewer patients at a given level of acuity) in response to increased hospital capacity strain, which could further accentuate the above missing nuance.

In added complication, many (but not all) foundational studies establishing this association in critical care relied predominantly on data from the 1990s and 2000s—a period that partially predates the widespread global adoption of evidence-based approaches and standardized expert consensus guidelines, such as lung-protective ventilation and sepsis protocols, and the rapid dissemination and availability of data and treatment protocols, which now define ICU practice across institutions of diverse size and resource level. Whether the standardization of critical care has diminished, preserved, or amplified the traditional volume-outcome relationship, and in what way, remains not fully understood, with the answer having potential implications for how today’s health systems and regions strategically organize and concentrate or distribute the delivery of critical care services.

Robin M Wood, George L Anesi