January 18th, 2013

A Possible Treatment for “Methods Myopia”

As a social scientist, I did not start my career in the orbit of cardiologists, nor was it an orbit I had even contemplated entering. But for about eight years now, I have been working with leading heart doctors in this extraordinarily fast-paced, innovative, competitive specialty that often leads the healthcare field.

Another opportunity for cardiology to lead has presented itself — in the typically homeostatic arena of scientific methods. For the past decade, interest has been slowly building in the use of mixed methods in clinical and health-services research, first with a reluctant embrace and a fair amount of skepticism but more recently with growing awareness of its value.

What is mixed-methods research? How is it relevant to the everyday practice of medicine and to leadership in healthcare? For detailed answers to those questions, see an article in Circulation: Cardiovascular Quality and Outcomes that I coauthored with my colleagues. For the basics, here’s my quick overview:

Mixed-methods studies combine statistics and conversations (quantitative and qualitative methods) to elucidate not only how many” and “how much” of something is happening or should happen, but also why and how things transpire in the complex world of healthcare. For example, we know it’s a problem that 14% of patients stop taking their clopidogrel after getting a heart stent. But there are two distinct dimensions to the dilemma:

  • Quantitative: How big a problem is cessation of clopidogrel for the patient?
  • Qualitative (and just as important): Why did the patient stop taking clopidogrel?

Both types of information are needed to deliver good care to patients and to design effective interventions that improve systems.

Take another example — a cardiologist, in a leadership role at a hospital, who’s responsible for reducing readmissions for patients with heart failure. To allocate resources, she can use evidence about strategies statistically associated with readmissions, such as close coordination with post-hospital care providers (quantitative information). Yet, to be effective, she also needs to understand the nuanced features of close coordination (qualitative information).

The momentum toward mixed-methods modeling is fueled in part by the recognition that large-scale statistical computations and sophisticated multilevel modeling, though useful, cannot alone address contemporary research challenges. Those challenges present themselves when, for example, we seek to understand highly complex systems of care; the intersection among healthcare financing, delivery, and quality; nuanced interactions between social and medical dimensions of health; how to foster authentically patient-centered care; and the critical role of context in all kinds of interventions. Such issues demand new forms of measurement and ways of understanding, including mixed methods.

Methods myopia can limit our understanding to mere P values and squelch the opportunity for discovery that motivates researchers of all stripes. Mixed methods offer great potential to move cardiovascular research and practice forward, if the field is ready to lead.

What are your thoughts about mixed-methods research and what role leading cardiologists should play in advancing it?

4 Responses to “A Possible Treatment for “Methods Myopia””

  1. Karen Politis, MD says:

    A classic triumph of qualitative research is the way pulse oximetry became established as essential monitoring during anesthesia. It all began with anonymous questionnaires filled out by Australian anesthesiologists, about patient safety and what monitoring technique could enhance it. Today, who would dare practice anesthesiology, or any clinical discipline of medicine, without a pulse oximeter? And yet, as far as I know, there never has been, and now never will be, a study that proves the benefit of the oximeter in reducing mortality directly related to anesthesia. The oximeter has become a standard of practice due to qualitative research i.e. gut feelings of astute clinicians.

    • Thanks for the example, Karen! And it is exciting to see the growing interest in using rigorous qualitative methods to capture those gut feelings of astute clinicians. As a research community we need to keep investing in building the capacity of clinical and health services researchers to do this work well, and of journal reviewers and editors to publish scientifically sound papers using these methods. Only in this way can we maximize the potentially powerful contributions of instinct to improving care and outcomes for patients.

  2. Judith Andersen, AB, MD says:

    Applause for this venture: capturing “gut feelings” from “astute clinicians” is not simple or straightforward, but is very important. The good judgment of an experienced clinician – who may not be able to verbalize it in a manner that convinces colleagues in other disciplines- is very important and I am not sure that we are there yet.I don’t think that rigorous “qualitative methods” are accepted by many,, — and I do not think that all important “gut feelings” will find validation in publications in peer-reviewed journals. Partly because “rigorous qualitative methods” are to some investigators a contradiction in terms. Quantitative, not qualitative measurements are the Holy Grail–but we all know, in our hearts, that a good clinician’s judgment in an individual’s fraught situation outweighs all the freight downloaded by quantitative studies.
    How do we capture intuition? How do we characterize those clinicians who come up with “magical diagnoses and solutions” that work, despite our best reasoning to the contrary? I have no idea, but I’m a believer. Some of my best “saves” have been contrary to all published wisdom, — and I wish I knew what the learned lessons were. I have to say that there are those constantly inspired clinicians who should be listened to however crazy they may seem — and the rest of us, who have occasional, inspired recommendations that should be acceded to. How to distinguish? Track records: disaster, dont’proceed; success, pay attention and facilitate treatment. Assess outcome– and contribute it to some national database that we have yet to create. I want badly to learn from these gurus, but think I may be one of them and don’t have a clue as to how intuition works — but know that it compels me to do something to treat the patient.

  3. I agree with Leslie that mixed methods research is important and that we do need to find ways to better capture the observations and intuitions of experienced clinicians. The goal should be to first poll cardiologists as to their intuitions in a specific area of clinical work, then take those intuitions and put them to empirical test, turning the qualitative into quantitative research. In this day of evidence-based medicine, there is truly a lot of evidence out there, evidence that often provides a basis for choosing among competing intuitive treatments. Ignoring or being unaware of empirical findings is, at best, dangerous. Intuitions should be applied only when there is no quantitative evidence on which to base clinical decisions. An important double-check, in a complex case, may be to consult with a colleague to see if he/she agrees with the responsible clinician. Bottom line, I think, is that intuition is important primarily in cases where there is a lack of available scientific data.