A comprehensive new study has found that major teaching hospitals in the United States outperformed non-teaching hospitals in the most important of all health care outcomes: reducing mortality rates.
Using a traditional measure of surgical quality, the study analyzed mortality rates for 21 million Medicare patients who were hospitalized with one of the 15 most common medical diagnoses or who underwent one of the six most common surgical procedures.
It found that the 30-day mortality rate for such patients was 1.5 percent lower in absolute terms (8.1 percent in teaching hospitals vs. 9.6 percent in non-teaching hospitals), or nearly 15 percent lower percentage-wise.
There would be 58,000 fewer deaths per year among those patients if non-teaching hospitals achieved the same mortality rate as teaching hospitals, according to Dr. Laura Burke, lead author of the study, published in the Journal of the American Medical Association in late May.
This finding was not surprising for those of us who practice at major teaching hospitals and who are especially concerned about the decision of some insurance plans to offer “narrow networks” that exclude teaching hospitals, which can be more expensive.
The study cuts through one of the central clouds of confusion plaguing health care: how to measure health care outcomes. This is difficult because it is often subjective. As dean of a major medical school, I know very well the challenges of delivering high-quality care while also preparing the next generation of doctors.
Well-being hard to measure
We live in a world of metrics, particularly in business. Peter Drucker, an icon in the development of modern business principles, is credited with coining the phrase, “If you can’t measure it, you can’t manage it.” Or improve it.
This has led to the proliferation of measurements of quality and cost. It has helped create an entire industry – from Consumer Reports magazine to TripAdvisor to Yelp – that is very useful for those buying a washing machine or planning a vacation.
This approach is not as helpful in health care, because so much important information cannot be measured. It’s hard to quantify the things that make a patient have a better experience. Did the new medicine make you feel better? Are you able to resume the lifestyle you seek after surgery?
For this reason, although modern health care is awash in metrics, many of these outcomes, while measurable, are not particularly relevant. In so many ways, the metrics used to assess quality in health care today are reminiscent of a much older proclamation, “It’s hard to see the forest for the trees.”
Using the gold standard
The brilliance of the recent study is that the metric measured was mortality: How many patients from each group were still alive at 30 days after their hospital discharge? Mortality is a gold standard for objectivity – “the body count,” as it is often referred to in clinical trials. Just as important, the study accounted for the fact that teaching hospitals often treat a sicker mix of patients.
Given that academic medical centers often treat patients who are sicker, some people may have been surprised by the results. Non-teaching hospitals, after all, have clinical care as their single mission. Major teaching hospitals, on the other hand, also perform research to develop new therapies and to educate trainees in many medical fields. Therefore, the clinical care mission of academic hospitals is diluted by these other missions, their thinking goes.
In addition, we know that care at academic medical centers can be more fragmented. This is the result of a larger treatment team that includes trainees. The larger the team, the harder communication can sometimes be. And, the role of the attending physician can sometimes be weakened as a result. These are issues that academic centers have long worked on.
That mindset misses a key point. Yes, physicians and leaders in major teaching hospitals have a different approach to health care. Their focus is not solely on the care they can deliver today but also on how they can make it better tomorrow. How can outcomes be improved? A crude but effective formula for measuring health care outcomes is: Value = Quality/Cost.
Complex cases, complex care
Major teaching hospitals, with faculty who focus on this formula, are not only thinking about how to provide care but how to improve health care value every day. For instance, in medicine, many guidelines exist for the treatment of different diseases. These guidelines are based on high-quality clinical studies. New and expensive technologies and drugs are likely to be used differently in major teaching hospitals given their mission to provide the best care today and to develop even better care in the future.
This blend of practicality with philosophy – of research, care and training – is a distinct difference between these two different types of hospitals. That approach has helped medical schools and teaching hospitals pioneer some of the greatest advancements, including the first polio vaccine, the first successful pancreas transplant, the first human genome treatment for cystic fibrosis and the first successful surgery on a fetus in utero.
That success is an important reason why about half of the NIH’s extramural grants support research conducted at medical schools and teaching hospitals.
Here at the University of Michigan, physicians and researchers have collaborated to develop new and better treatments to reduce mortality rates for a range of deadly conditions including skin cancer and congenital heart failure. Cutting-edge use of 3-D printers, for example, helped a pediatric surgeon and biomedical engineer pioneer a new technique to treat infants whose collapsed windpipes were nearly sure to be fatal. A splint, made from a bio-absorbable polymer, allows such infants to breathe on their own and eventually develop a normal trachea.
Where will you get best care?
For consumers, does this mean they should seek out care at teaching hospitals? Of course, as the leader of a premier academic medical center, I know there are some advantages. The combination of researchers and practitioners, cutting-edge facilities and a commitment to both care and knowledge help produce first-rate care. But the real takeaway from this study is that it validates for consumers that the standard of care will not be lower at a teaching hospital than a non-teaching hospital.
Even as those of us at teaching hospitals trumpet the positive results and lower mortality rates pinpointed by this new study, we should acknowledge that all measurements serve some purpose. Metrics do not predict the future, but to the degree that past performance informs the future, they do have value.
However, they do not tell the whole story. We do – and will continue to – live in Peter Drucker’s world of statistics and metrics. But that should not blind us to the larger fact that so many medical outcomes have a very real subjective component, too.