It's often said that death rates and costs are lower and outcomes are better at facilities that perform high volumes of a procedure. But a report this week challenges that assumption and says more research is needed.
Clinicians are using a flawed statistical model to correlate volume and outcome, according to a study in the journal Applied Health Economics and Health Policy.
Hospital-specific factors that could affect outcomes are ignored in many models, and not every type of surgery is riskier at a low-volume facility, said Vivian Ho, a health economist at Rice University's Baker Institute for Public Policy.
She and researchers from Baylor College of Medicine in Houston warned that pushing surgical patients to high-volume centers could lead to unintended consequences, such as less marketplace competition.
The study used hospital discharge data for more than 164,000 patients who underwent six types of major cancer surgery in Florida, New Jersey and New York between 2000 and 2011.
They looked at the relationship between volume and mortality following these surgeries using three statistical models: logistic regression, random-effects and fixed-effects.
The first one, the most simple and frequently used in clinical journals, does not account for potentially unobserved hospital-specific factors that could be affecting outcomes, as do the other two models.
According to the analysis, no matter what model was used, there was no relationship between doing more esophagus cancer surgeries and lower mortality. However, on two of the three models studied, there was a significant correlation between how many colon and lung cancer surgeries were performed and better outcomes.
For the fixed-effects model, the most complicated statistical model that takes into account more unobserved factors, there “was no volume effect anywhere,” Ho noted.
The medical literature may overemphasize the role that hospital volume plays in patient outcomes, and could mislead physicians looking for evidence on where to send patients, the researchers said.
Factors such as organizational structure, culture or infection policies can impact surgical outcomes.
This week's analysis comes as healthcare policy, quality and safety leaders point to studies that show high-volume hospitals tend to have overall better outcomes and lower costs.
For example, a 2015 analysis found that more than 60% of patients undergoing cancer surgery in California went to a low-volume facility, even when a high-volume facility was within 50 miles. That report said low-volume facilities had longer lengths of stay, higher postoperative complications and a greater number of patients readmitted for follow-up care.
Some are calling for hospitals to face tougher scrutiny before they launch new surgical specialty programs, and others would like to regionalize where patients are sent for certain high-risk surgeries.
A few notable health systems are taking what they call “volume pledges,” where they plan to cut back on surgeries at low-volume locations.
The Joint Commission, the Oakbrook Terrace, Ill.-based hospital accreditation organization, declined to comment, calling the study “a complex methodological paper that is probably best left to statisticians to discuss.”
However, the group is critical of volume pledges and its president, Mark Chassin, recently told Kaiser Health News that volume should never be used as a measure of quality because the measure is imperfect.
Statisticians from U.S. News & World Report, a group that provides healthcare quality ratings and recently began looking into outcomes of common high-volume surgeries and procedures, said this week's study raises several interesting points, including why volume is surfacing as an important topic.
“So volume does appear to be relevant in most instances,” said Geoff Dougherty, the senior health services researcher who conducts statistical modeling for U.S. News. “Shifting patients away from low-volume hospitals could improve outcomes, even if volume does not cause quality.”
