In a paper published online in JAMA Internal Medicine on May 27, 2013, Tak and colleagues presented the results of a study describing the association between hospital costs and length of stay, and hospitalized patients’ answers to a question about whether, “I prefer to leave decisions about my medical care up to my doctor.” The roughly one-third of participants in the study who “somewhat” or “definitely” disagreed with this statement had significantly higher mean length of stay (by about a quarter of a day) and total hospital costs (by about $900).
While the paper itself presented the results cautiously, pointing out the study’s limitations, a press release was entitled, “When doctors and patients share in decisions, hospital costs go up.” The message from the press release was picked up by many news outlets.
We think this broader conclusion is inappropriate given the methods and findings of the study for the following reasons:
First, two questions about decision-making preferences were asked, but only one was analyzed. The other question asked whether, “I prefer my doctor gives me choices and asks my opinion.” We think that’s a better question to get at patients’ preferences to share in medical decision-making, and interestingly, only about 4% of these patients somewhat or definitely disagreed with that statement, quite a difference from the response to the question above. Poor survey questions about patients’ desire for involvement in medical decision making have been a major source of confusion in the past (see “Foundation Perspectives: Patients Want to be Involved“).
Second, we have no basis for thinking there was actually any shared decision-making during these hospitalizations for the patients or even whether it was appropriate. Shared decision making involves clinicians and patients working together to make decisions when there are more than one reasonable management strategy, such as whether to treat prostate cancer with surgery, radiation, or observation; or whether to have a hip replacement or pursue medical treatment for arthritis of the hip. The patients in this study were hospitalized for acute medical conditions such as acute kidney failure, pneumonia, and bleeding from the intestine. Many of these patients needed and likely got one particular appropriate treatment…bacterial pneumonia requires antibiotics, for example. Shared decision making wouldn’t play a big role for these problems.
Finally, as noted in the paper, the study was conducted at one urban hospital. To generalize the findings to all 35 million hospitalizations in the United States, as was done in the press release, is inappropriate.
Shared decision making between clinicians and patients using decision aids has been shown to increase the quality of medical decisions in appropriate situations in a synthesis of 86 randomized trials, the strongest study method (see “Decision Aids for People Facing Health Treatment or Screening Decisions”). In another randomized trial of a shared decision making approach for involving patients in preference sensitive decisions, total costs were reduced by about 5% and hospitalizations by 12% (see “Enhanced Support For Shared Decision Making Reduced Costs Of Care For Patients With Preference-Sensitive Conditions”). Another recent study of the effect of giving patients decision aids about knee and hip replacement surgery for arthritis found costs reduced by about 20%, due to reduced patient desire for surgery (see “Introducing Decision Aids At Group Health Was Linked To Sharply Lower Hip And Knee Surgery Rates And Costs”). Total costs for a population are driven much more by hospitalization rates than by length of stay or in-hospital costs. Involving informed patients in their own preference sensitive health care decisions is not only the right thing to do; it also has the potential to identify the right rates of medical interventions. The right rate may be lower in situations of baseline overutilization, or higher in situations of baseline underutilization. No one should take away from this study that informing and involving patients in their health care decisions is inadvisable because it will inevitably increase costs.