The risk of dying within a month of admission to hospital is substantially lower in teaching hospitals than non-teaching hospitals. Besides being useful to anyone deciding where to seek treatment, the finding could provide clues on how we could be saving lives across the board.
Doctors have to learn many skills on the job, but most of us would probably prefer they didn't learn them while we are their patients. It's natural to prefer being treated by someone with experience than to place your health, or even your life, in the hands of a raw recruit who is still learning the ropes. Insurance companies often see things the same way.
Professor Ashish Jha of the Harvard Global Health Institute processed data from 21.5 million hospital admissions of elderly patients at 4.483 institutions across the United States – no one's idea of an insufficiently large sample. "We find that across a very wide range of medical and surgical conditions, patients receiving care at teaching hospitals have superior outcomes,” Jha said in a statement.
In the Journal of the American Medical Association Jha reports the risk of death within 30 days of admission was 9.6 percent at non-teaching hospitals, but 8.1 percent at major teaching hospitals and 9.2 percent at minor (99 beds or less) teaching hospitals. Of course, someone is much more likely to die quickly if admitted for a heart attack than a broken arm, but Jha controlled for differences in illnesses, as well as the age of the patients. The advantage for major teaching hospitals remained, and actually increased very slightly when the period of study was increased to 90 days.
Increased data availability has expanded the size of studies like this, but teaching hospitals came out better in similar research conducted decades ago. Yet the previous findings have not translated into increased respect. Some US insurers will not even cover treatment at teaching hospitals, regarding them as a waste of money, at least for the individual patient, if not for society as a whole.
Funding for the research came from the Association of American Medical Colleges, but the authors declared: “The funders had no role in the design or execution of the project and no input into the writing of the manuscript.”
The reasons for teaching hospitals' better performance is unclear, but if identified might provide clues as to what all hospitals should do better. The authors speculate on the possibility that teaching hospitals are earlier adopters of new technologies, or develop greater familiarity with rare conditions. Alternatively, doctors in training may be more attentive and less likely to miss something important.