Care intensity greater at teaching hospitals, but sicker patients are the reason
This article was first published on Thursday, May 2, 2019 in MedPage Today.
By Diana Swift, Contributing Writer
Differences in metrics of cost-effective hospital care for inflammatory bowel disease (IBD) are driven more by disease severity than hospital teaching status, according to a University of Pennsylvania study published online in Inflammatory Bowel Diseases.
The findings suggest that insurers should not exclude teaching hospitals from their network of coverage in favor of community hospitals on grounds of cost-effectiveness.
After a multivariable analysis for confounders of disease complexity, the only metric for which teaching hospitals were at any disadvantage was 30-day readmission rates for ulcerative colitis (UC), which was 1.9 percentage points higher in major teaching hospitals versus non-teaching hospitals (95% CI 0.33-3.61) after multivariable adjustment.
"This finding demonstrates that despite the time and resources needed to train new physicians, conduct pioneering research, and trial newer and potentially costly therapies, outcomes at academic centers do not suffer when compared with their community counterparts," wrote Gary R. Lichtenstein, MD, of the Perelman Center for Advanced Medicine in Philadelphia, and colleagues.
In unadjusted analysis, however, mean length of stay, mean direct costs, and 30-day readmission rates were numerically greater among teaching institutions for both UC and Crohn's disease (CD). The unadjusted mortality rate was also greater in major teaching hospitals for UC but not CD.
Incidence and prevalence of IBD are on the rise in the U.S., currently affecting more than 3 million adults.
Improving the cost effectiveness of care is a major consideration. Although teaching hospitals have overall higher costs and more often utilize medical trainees to provide care than their community counterparts, little is known about the way teaching status actually impacts hospitalization outcomes, the authors noted, stating that theirs is the first study to do so.
Although hospitalization is thought to account for more than 30% of the as much as $31.6 billion in annual costs of IBD, previous studies have not examined the impact of teaching status and the use of trainees for providing care, the researchers noted.
Medically complex patients are often referred to teaching hospitals with high-volume academic physicians, and these centers have overall higher costs and receive funding from the federal government for training and research.
"They are considered less likely to provide cost-effective care, so some insurers may narrow their treatment networks to exclude teaching hospitals in order to control costs and premiums, as has been done in cancer treatment," co-author Rahul S. Dalal, MD, of the Hospital of the University of Pennsylvania, told MedPage Today. "But we found that the higher costs were attributable to the severity of the patient populations at referral centers."
Using the Vizient clinical database, the investigators identified patients hospitalized for CD or UC from October 1, 2014 to March 31, 2018. With hospitals divided into major and minor teaching hospitals and non-teaching hospitals according to the 2016 annual survey of the American Hospital Association, they evaluated the association between teaching status and mean length of stay, mean direct cost of patient care, 30-day readmission rate, and in-hospital mortality rate. Adjustments were made for a range of patient characteristics and covariates of disease complexity that might affect resource utilization. Other variables included mean number of comorbidities per hospitalized patient, proportion of cases transferred from outside hospitals, and the proportions requiring intensive care, total colectomy or bowel resection, lower endoscopy, parenteral nutrition, or blood transfusion.
For the final analysis, the database yielded 29,863 discharges from 291 hospitals for UC and 62,698 discharges from 314 hospitals for CD. Hospitals not included by the American Hospital Association's annual survey or those with fewer than five IBD discharges or missing data were not factored into the aggregate data.
Major teaching centers had a significantly higher case mix index, reflecting resource use intensity, and their IBD patients were significantly more likely to have the following characteristics:
- Younger age (18 to 30)
- Black or Hispanic ethnicity
- Primarily Medicaid coverage
- Transfer originally from an outside hospital
- Complications such as colectomy or bowel resection, lower endoscopy
- Extreme severity of illness on admission
- Parenteral nutrition
"I think this is an important study that provides more granular adjustment for disease severity," commented Ashwin N. Ananthakrishnan, MD, MPH, of Massachusetts General Hospital in Boston, who was not involved in this research. He added that when many studies in the literature compare big categories, such as teaching versus non-teaching hospitals or small versus large hospitals, the sites with more referral cases or more complex cases usually appear to have worse outcomes.
"But this study confirms, as suspected, that this is a false association once disease severity is factored in," he said. "So as we think about paying for IBD care in the future – perhaps through bundled payments for diseases – this study highlights the importance of adjusting for severity of disease in such decision making."
The authors cited the need for studies to better characterize the factors that drive increased resource utilization in order to provide high-value inpatient IBD care at all medical centers. Access to patient-level data might shed light on these factors.
Among study limitations, the authors noted its reliance on aggregate data at the hospital level with no patient-level data. In addition, since Vizient has not been validated for the study of IBD, the analysis relied on ICD coding for diagnoses and procedures and therefore may have omitted hospitalizations coded primarily as abdominal pain or diarrhea that did not include IBD as a primary or secondary discharge diagnosis. Conversely, some hospitalizations may have been included because of miscoding of the discharge diagnosis that ultimately may have proved to be a condition other than IBD.
Furthermore, the study was unable to isolate the impact of resident teaching on outcomes for hospitalized IBD patients. Finally, it was unable to measure inpatient biologic therapy, which likely would have been utilized to a greater extent in teaching hospitals and would have increased the direct costs and hospital stays at these centers.
Lichtenstein reported ties to Salix, Shire, Axcan, Ferring, Abbott, UCB, Bristol-Meyers Squibb, Elan, and Prometheus. Co-author Lewis disclosed ties with Pfizer, Gilead, UCB, Janssen, Ortho biotech, Celgene, AbbVie, Samsung Bioepis, Bridge Biotherapeutics, Merck, Takeda, and Nestle Health Sciences. Ananthakrishnan reported no conflicts of interest.
“We found that the higher costs were attributable to the severity of the patient populations at referral centers.”
Rahul S. Dalal, MD, of the Hospital of the University of Pennsylvania
Photo credit: chrupka / Shutterstock
Although hospitalization accounts for more than 30% of the $31.6 billion in annual costs of IBD, previous studies have not examined the impact of teaching status and the use of trainees for providing care.
Medically complex patients are often referred to teaching hospitals with high-volume academic physicians, and these centers have overall higher costs.
Major teaching centers had a significantly higher case mix index, reflecting resource use intensity.