It is commonly believed by some that the cost of care in teaching hospitals, the sites where we train future physicians, is higher than that of community hospitals that do not have the additional burden of trainees. The thought is that trainees order more tests and provide care less efficiently. Of course, since these trainees are all nominally under the supervision of fully trained physicians, excess testing or other inefficiencies should accrue to the teachers, not the students. A new study from JAMA Open Network seeks to see whether the common belief about more expensive care at teaching hospitals is true.
The study made use of CMS payment data for 23 months beginning in January 2014 looking at the actual payments for the initial or index hospitalization and subsequent care for the next 30 or 90 days. These after hospital costs include readmissions, post-acute care like visiting nurses or rehabilitation services, and other outpatient claims. Physician professional claims, the expenditure for the care provided by the physicians were not included, nor were outpatient drug costs calculated. The hospitals were characterized as major teaching hospitals, those that belong to the Council of Teaching Hospitals, minor teaching hospitals – hospitals affiliated with a medical school, and nonteaching hospitals - your typical community facility. They looked at 1.2 million admissions
After adjusting for the severity of the illness being treated the researchers found that
- The 30-day standardize cost of care was $575 less at a major teaching hospital than a community hospital. The majority of that disparity has been previously shown due to the cost of post-discharge services, not the care rendered during hospitalization.
- Physician costs were greater in community hospitals by about $50
- The cost of readmissions was higher in community hospitals by about $244
- By 90 days, there was no statistically significant difference in costs between hospital types; community care cost $61 more.
So the first take away is that the cost of care varies based both on the type of hospital and the length of care considered. CMS's bundled care program uses 90 days as the covered interval, and in that setting, all hospital are comparable. In the 30-day window teaching hospitals seem to be a bit of a bargain, saving the government $575. But with any study, there are some nuances that only the insiders see; it's a bit like reading between the lines.
Some nuances of cost
Physician charges are driven by the number of consultants involved in care, and my personal experience as well as reviewing CMS payment data is that community care engages more consultants – consultants are added out of an “abundance of caution,” or perhaps as a way of lubricating the wheels of care. Teaching hospitals have less of a problem in the wheel lubricating area because they have trainees which brings us to the most glaring problem with the data, they did not include the additional payments teaching hospitals receive for training residents (one of the last forms of indentured servitude). These are termed indirect medical education payments, IME, and including them in the cost study was felt to be “controversial.” After all, they are payments for education, not care; so while they might be external costs in this study or for an individual patient, they are definite costs to the taxpayer. When IME costs were factored in, the savings by teaching hospitals quickly disappeared and became additional costs in the range of $1200 more for that 30-day care in a teaching hospital.
And there are no calculations of a patient’s out of pocket expenditures, another cost that is different from the one presented and that bears significance in both policy discussions and beneficiaries personal spending. As it turns out, the cost of care is only a little more transparent than the cost of medications; and we have to be very careful when we compare costs because there are many meanings to the term.
Source: Comparison of Costs of Care for Medicare Patients Hospitalized in Teaching and Nonteaching Hospitals JAMA Open Network DOI: 10.1001/jamanetworkopen.2019.5229