• The terrible horrible no good very bad challenge: Changing patient and provider behaviour

    AlexanderI thank Judith Viorst and her classic bedtime story for the words to express my feelings about the report of an RCT from Anne Kennedy and her colleagues, published in the BMJ.

    Kennedy et al. conducted a large effectiveness trial of an intervention designed to help chronically-ill patients manage their illnesses. Self-management is the simplest and most familiar medical practice: consider your daily habits of tooth care. In this study the patients had diabetes, chronic obstructive pulmonary disease, and irritable bowel syndrome, but the idea of promoting consistent daily self-care disciplines is the same. Self-management can allow patients to achieve a high quality of life while avoiding the side effects associated with medications or surgery. Self-management of chronic illness is one of the great hopes for improving patient outcomes while reducing costs.

    The researchers were hoping to pull off a three-bank shot: The intervention was to train primary care practices in techniques for training patients in self-care behaviors that would improve the patients’ health. The well-designed experiment included 5,599 patients in 43 practices, so it had plenty of statistical power to find a moderate effect. This was an ‘effectiveness trial,’ which meant that its goal was to determine how well the intervention worked under everyday clinical conditions. The interventions had already been successfully tested in ‘efficacy’ trials. An efficacy trial is a clinical proof of concept study: Can you make the intervention work under well-controlled conditions using doctors who are committed to applying the treatment rigorously and, often, with patients who have been selected to make them ideal candidates for success? If the intervention works in an efficacy trial, the next question is: can you make it work in a typical practice using whatever patients they happen to be treating?

    So what Kennedy and her colleagues found was that in the real world… training the primary care providers didn’t help patients at all.

    With one exception, patients attending intervention practices and those attending control practices did not differ significantly on any primary or secondary outcome. The exception was shared decision making at the six month follow-up (P=0.05), with the difference favouring the control group… The lack of effect applied equally to the intermediate outcomes of shared decision making, self efficacy, enablement, and self care activity—which might reasonably be expected to be most directly affected by increased support for self management—as it did to health related outcomes.

    So much for the great hope of empowering patients, improving outcomes, and reducing costs. As Teddy KGB said in Rounders:

    It hurts doesn’t it? Your hopes dashed, your dreams down the toilet. And your fate is sitting right besides you.

    Why didn’t it work? It seems that despite the training they received, the primary care providers changed their behaviour only to a limited degree.

    Questionnaire data (a low response rate 48%) obtained from clinicians showed varying levels of implementation in routine practice… Across the 12 months of the study, similar percentages of intervention and control patients reported on each type of support, including receiving a guidebook (25% v 24%) and encouragement to use community programmes (19% v 20%) and patient support groups (11% v 12%).

    So what do these results mean? I do not think they provide a reason to give up on patient self-management. What these results mean is that training alone rarely changes provider behaviour.

    This has nothing to do with the character or personality of primary care docs, who are no better or worse than anyone else. The problem is that primary care providers work in highly constrained environments. Their routines of care have been set by years of training and decades of tradition, and are adapted to cope with a fire hose of patients and tasks. It takes more than brief training to change what docs do.

    What is required, I think, is reorganization of the health care system to integrate primary care practices more tightly into medical systems. In that new system, promotion of patient self-management becomes the primary responsibility of a nurse or other physician extender. Patient self-management will not become a priority unless it becomes a permanent priority for the medical system and that system exerts incentives — carrots and sticks — to make sure that it happens.


    • The passive patient and the expert fix-it-all doctor are fairly pervasive parts of our medical culture. It is interesting, in this context, that so many people are involved in self-help diets, exercise, etc. but don’t do much when the medical system encourages it.

      On another point, I wonder how many of our interventions, including medication interventions, would survive effectiveness trials.

    • I suspect part of the problem is that if the doctor explains that the best results come from the patient doing certain things, many patients take that to mean the doctor can’t do anything (and maybe they should find a better doctor) and the doctors become cynical when patients don’t reform.

    • “On another point, I wonder how many of our interventions, including medication interventions, would survive effectiveness trials.”


    • -Is there a consensus in the literature derived from independent, large scale studies that suggest that any cohort of health care providers in any context can inspire/cajole/motivate/etc… patients into permanently taking better care of themselves?

      Based on insights from front-line providers, my guess is that the answer is a resounding “No.” Doctors can help people who come to their office who *want* to change their own behaviors for their own reasons, but their capacity to make people who lack the desire and the will to take better care of themselves to do so is virtually nill.

      • JayB,
        This is a great comment — thanks for the challenge. I think I may try to put the answer into a follow-up post.

        • It’ll be interesting to see what you find. From what I can recall the consensus from employer wellness programs is that they simply don’t work, and employers are beginning to mix in significantly greater sticks into the incentive mix.

          My prediction is that making people pay even part of the price of their unhealthy behaviors will generate significantly more compliance than the all-carrot incentive programs that they are replacing, or enough resentment to prompt the people who don’t change their habits to quit, both of which will be net gain for the employer and the rest of the employees who no longer have to foot the bill for their co-worker’s unhealthy lifestyles.

          Speaking of sticks, if their are no patient-level sticks incorporated into the care delivery model that you are envisioning above, one potential hazard is that physicians, recognizing that they are being asked to do the impossible, will do whatever they can to avoid caring for patients who have unhealthy lifestyles, and thus be rewarded for their capacity to recruit and retain healthy patients, or at least those that are predisposed to change their habits.

          • Responding briefly:
            a) there is a lot of evidence about self-management. See http://www.cochrane.org/search/site/self%20management (thanks, @Brad_Flansbau). Not all programs work and there is not enough data yet to evaluate all of them. But some work.
            b) Also, consider the example of dental flossing. This behaviour did not exist when I was a child, but it prevalent (although far from universal) now. This was partially the work of the dentists.
            c) I have read some discouraging reports on employee wellness and also have read reports of increasing use of sticks. I have argued in favor of sticks, but I have to say that I am not sure that is the right view.

    • In re the last few sentences, integrating the physician into the medical system will create other, potentially even more significant problems like productivity issues and provider concentration (i.e., prices will rise and incentives for quality decline). Understand the urge from Bill Gardner to assume that scaling up will solve the problem of alignment but definitely think that this post understates the potentially adverse impacts associated witht that.

      • V,
        I agree that scaling up by itself may have significant risks. Your comment makes me realize that I am usually imaging that the scaled up system is an Accountable Care Organization, or something similar, that is continuously monitoring quality.

    • Thanks for posting, Bill.

      “What these results mean is that training alone rarely changes provider behaviour.”

      Substitute ‘human’ for ‘provider’ and you have another reason why the trial didn’t work. Behavior change is difficult and multifactorial. Not impossible. Not unscalable. But not simple.

    • I’m puzzled by most of these comments. The research seems to show that patients are capable (little difference between efficacy and intervention participants), but that clinician behavior was unchanged by training. And yet most of the comments reflect on the unwillingness of patients to change unhealthy lifestyles.

      Primary care docs have a lot on their plates and, as Bill points out, work within constrained systems. There may be ways to address this, though financing and delivery reform and use of other types of providers, as Bill also notes. But clearly, part of the problem lies with physician training and culture. Until we recognize and accept that, it certainly won’t change.