HOSPITALS have been criticized, including by me, as wildly inefficient. Yet two new pieces of evidence suggest that the hospital market may be more efficient than conventional wisdom suggests.
Hospitals practice medicine in drastically different ways, and the higher-spending ones don't seem to generate any better results than those that spend less. Because of third-party insurance and other distortions, though, the market doesn't punish the inefficient hospitals. That has been the traditional critique.
A new study suggests, however, that hospitals actually do gain patients when they provide better value.
Economists Amitabh Chandra of Harvard University, Amy Finkelstein and Adam Sacarny of the Massachusetts Institute of Technology, and Chad Syverson of the University of Chicago examined how 5,000 hospitals treated some 3.5 million patients who suffered heart attacks.
About a third of the patients died within a year of their heart attack, and the researchers used the variation in survival rates across hospitals to approximate the quality of care provided. The average cost of treatment was $16,000, but that also varied from place to place.
The researchers found that hospitals with higher survival rates net of the cost of treatment were rewarded with more patients.
More specifically, if in a given year hospital A had 10 percent higher productivity than hospital B, hospital A tended to have 25 percent higher market share that year and to experience 4 percent more growth over the subsequent five years.
The team also found that the variation in survival rates was more important in driving market share than the variation in cost was.
In other words, patients seem to seek out hospitals with better survival rates but not ones with lower costs.
As the researchers note, "It could be that competitive market forces re-allocate market share to higher productivity hospitals, or it could be that higher productivity hospitals happen to have other features - such as beautiful lobbies or good managers - which separately increase demand.