Secondary Logo

Journal Logo

Using Evidence to Change Practice: From Knowing the Right Thing to Doing the Right Thing

Sklar, David, P., MD

doi: 10.1097/ACM.0000000000002203
From the Editor

Editor’s Note: The opinions expressed in this editorial do not necessarily reflect the opinions of the AAMC or its members.

A few months ago I was working with Will (not his real name) during his second week of internship. He was excited, a bit anxious, and trying to find the right balance between asking questions and making decisions. Our first patient was a 29-year-old woman with a headache and vomiting who had a wet cloth on her forehead. Her room was dark because the light hurt her eyes. As we entered the room the woman sat up and vomited onto Will’s new white coat. Will ran out of the room to clean up and find a basin to catch any more vomit as I used a towel to help clean the patient. As I was wiping her up, I introduced myself and asked if this problem had happened to her before.

“I’m Mary Johnson. I’m so sorry. It’s my migraines,” she said. “Usually I can control them if I go home and lie down in a dark room and take some ibuprofen. But I tried to tough it out. I didn’t want to miss my staff meeting. But it got worse. So here I am. I’m sorry for the mess.”

“Oh, don’t worry about that. It happens all the time around here,” I said. “It’s part of being a doctor.” By now Will had returned with the basin and had cleaned himself up.

“I ordered a head CT scan and some morphine,” he said.

“Okay. Why do you think Ms. Johnson needs a scan?”

“To rule out head bleed,” he said reflexively.

“And the morphine?”

“For the pain,” he answered.

“Okay, well, that’s good. You’re thinking of the worst-case scenario. But before we get the CT scan and give her the morphine, let’s talk with Ms. Johnson a bit more. Why don’t you ask her a few questions? I think she’s more comfortable now.” Ms. Johnson was laying back on the stretcher and nodded her head slightly. As Will elicited her history of migraine headaches and determined that this current headache was similar to her previous migraines, he paused, and I asked if he thought we still needed the CT scan.

“I think maybe we can hold on it,” he said. “Maybe if we treat her migraine.…” We then discussed the treatment for severe migraine in the emergency department. “When I checked online, UpToDate suggested administering sumatriptan subcutaneously or an antiemetic like metoclopramide or prochlorperazine,” he said. “But at my medical school we usually gave narcotics.”

“I can’t take any triptans,” said Ms. Johnson. “I get a bad reaction.” We decided to give her intravenous prochlorperazine, and 20 minutes later she was resting comfortably and ready to go home. Later, as Will and I discussed the case, he explained that he had never seen a severe migraine headache patient in medical school, but he had heard that narcotics were the only drugs that worked for most of them. His attendings in medical school had used narcotics, and he wondered why they had not tried antiemetics like we did. I explained to him that this was a question I had been pondering for almost 30 years.

Twenty-nine years ago, a few colleagues and I published a double-blind controlled study1 of a new treatment for headache in the emergency department. At the time, many textbooks advocated the use of parenteral narcotics for patients who arrived in the emergency department with severe headache symptoms from migraine. This was because it was thought that there were no good alternatives once a migraine headache became full-blown with severe pain, nausea, and vomiting. Our study demonstrated that prochlorperazine, a medication that had been used for nausea and vomiting, was very effective compared with a placebo in reducing or completely curing the pain of migraine episodes that brought patients to the emergency department. The impetus for the study had been to identify an effective alternative to parenteral narcotics for migraine treatment. It had been our impression that narcotic administration in the emergency department for headache was not very effective and might be encouraging narcotic addiction for migraine patients.

We expected that with the publication of our study there would be rapid adoption of this alternative treatment for migraine in the emergency department. However, a study published less than three years ago2 found that narcotics have continued to be used for migraine management, making up over 50% of treatments, with hydromorphone being the parenteral narcotic most commonly used (25%). More recently, Friedman et al3 demonstrated that prochlorperazine was substantially more effective than hydromorphone for the treatment of acute migraine in the emergency department.

The publication of these study findings, combined with our study findings so many years ago, led me to wonder whether the treatment of migraine patients would finally change and, if not, what evidence would convince physicians to change practice. I did not understand why our original study had not been convincing enough, since there had been no scientific evidence for the effectiveness of narcotics for migraine. Medical education has included evidence-based medicine as part of the curriculum for some time; why weren’t its tools being used to help physicians choose evidence-based treatment rather than use a medication with significant risk of harm and lack of effectiveness?

The term evidence-based medicine (EBM), introduced by Guyatt4 in 1991, grew out of questioning the advice of “medical experts.” Using the EBM process, a physician does not rely on received wisdom but, instead, searches for published literature, analyzes and classifies the literature, scientifically applies relevant findings from the literature to the patient’s problem, and then makes a treatment decision. Later in the evolution of EBM the incorporation of patients’ values and priorities was added. Haynes5 describes EBM as “a set of tools and resources for finding and applying current best evidence from research for the care of individual patients.” He further notes that research evidence, clinical circumstances, and patients’ wishes are to be combined using clinical judgment and expertise. The current EBM approach typically involves identifying a problem or gap in knowledge that can be developed into a question, collecting research pertinent to the question, assessing the strength and validity of the research, and applying the research results to the clinical problem to make a treatment decision, taking into account the priorities of the patient and an evaluation of the treatment decision with the patient. The education of residents on shared decision making, described by Rusiecki et al6 in this issue, can encourage the incorporation of patient’s priorities in developing an evidence-based decision.

Greenhalgh et al7 raise concerns about the current state of EBM, including the involvement of those with vested interests, such as drug companies, deciding which studies will be funded and thus what evidence will be produced and available for review. They also mention other concerns such as the challenge of sorting through a sometimes-overwhelming volume of evidence, the potential for advocating approaches with statistical but marginal clinical benefit, the potential inflexibility of guidelines to account for patient preferences, and the limited information about how to apply an evidence-based guideline in circumstances of comorbidity where a treatment of one problem may worsen another. Their suggestions included individualizing evidence-based guidelines to the specific circumstances and preferences of the patient, and using expert judgment guided by evidence and supported by relationships with patients and communities. They further propose that training in EBM should go beyond simple vignettes to include application of the EBM principles in actual clinical settings.

Maggio et al8,9 have studied the teaching of EBM through a review of the literature and through interviews with faculty involved in teaching it. In their review of the literature on teaching EBM,8 they concluded that such teaching should occur in clinical settings that provide relevant learning opportunities, and should take place longitudinally throughout undergraduate training. In their interviews of teachers,9 they noted challenges that include suboptimal role models, students’ lack of willingness to admit uncertainty, lack of clinical context, and students’ difficulties in mastering EBM skills, and suggested that to overcome these challenges, EBM skills should be integrated with other courses and content. They further urged faculty development in EBM so that faculty would be able to apply the tools to actual patients, carry out exercises with students that include an entire EBM task, and longitudinally integrate EBM into the curriculum.

I believe that such an approach might benefit a population that has not been adequately considered for it: the growing number of patients with chronic diseases, discussed in this issue by Shi and Nambudiri.10 They describe a chronic care clerkship that offers the opportunity for students to make longitudinal decisions shared with patients, which I believe could also be an opportunity to learn and apply EBM principles to chronic care management. Patients with acute recurring pain such as Ms. Johnson with her migraines are examples of persons with a chronic disease where EBM could be utilized as part of a chronic care clerkship to improve the quality of care.

While EBM provides the tools to analyze the research related to a problem, it does not explain how, why, or when physicians use EBM to make decisions on individual patients or change their practice patterns. Unfortunately, there is a substantial time lag from the appearance of a discovery or innovation to its adoption. Morris et al11 studied the time from appearance of a new discovery in the medical literature to its general adoption and found that the time varied depending on the clinical problem. While 17 years has been identified as a general time frame of such a lag, Morris et al explain how that frame will often vary depending on the type of discovery. In addition, not all providers adopt an innovation at the same time. Finally, Leggott et al12 demonstrated the timeline of adoption of an innovation using the example of peripheral nerve blocks to replace general anesthesia for orthopedic surgery. There was an initial slow adoption followed by a rapid acceleration and finally a general acceptance. They noted that communication, leadership, and trust between anesthesiologists and orthopedic surgeons facilitated the decision-making process. They suggested that not only the evidence of efficacy but also social and environmental factors contributed to the adoption of an innovation.

Trust and the engagement of the community are also important for the success of community-based initiatives, such as programs promoting better diets to prevent hypertension, diabetes, and chronic kidney disease, as described by Wesson and Kitzman13 in an Invited Commentary in this issue. EBM can be applied to population health problems and innovations using the same tools as those used for individual patients. But instead of shared decision making with individual patients, a relationship of trust between health providers and the community can facilitate the implementation of the evidence to provide community health solutions. However, Rowland and Kumagai14 in this issue caution that choosing a person to represent a community of patients and be its voice is fraught with hazards, and the purpose of the representation must be carefully considered to obtain the best result. Clearly, this same care must be taken in issues representing the wider community. For example, if the community is asked to provide input into a research protocol, it is critical that its representative be someone who can speak to the real concerns of the community, which could include issues of exploitation of the community during previous research.

What does all of this have to do with the problem I posed earlier about the failure to substantially change migraine headache treatment despite clear evidence that change was needed? Unfortunately, evidence is sometimes simply ignored or resisted. Many physicians persist in their habits of treatment just as scientists persist in their defense of theories, even when evidence opposes the treatments and theories.15 In this issue, Johnson et al16 advocate the adoption of high-value care at academic medical centers that would utilize EBM guidelines but include quality improvement principles to implement EBM and reduce unnecessary care. “Education begins with sharing evidence, typically including systematic reviews, consensus recommendations, and society guidelines.” They conclude that

although academic medical centers comprise only a small percentage of all medical centers in the United States, they must integrate high-value-care education into medical school curricula and residency training, and they can lead widespread improvements by publishing implementation guides with outcomes data to facilitate and inspire high-value care.

Berwick17 has identified five factors related to the adoption of an innovation: the innovation’s perceived benefit, its compatibility with values and beliefs, the complexity of the innovation, the opportunity to try out the innovation, and the chance to observe others using it. I believe that these factors contributed to the long lag in adoption of prochlorperazine, but there may also have been effects from the lack of financial incentives for the pharmaceutical company that produced the drug to study and publicize this new indication, since the patent for prochlorperazine (1957) had long expired by the time our research was published. Although communication through the published literature and through discussion with peers can influence physicians’ behaviors, the advertisements from drug companies continue to affect physicians’ decisions about adoption of new treatments.

While EBM provides a valuable tool for analysis of complex health questions and helps to educate students about how to use scientific methods to review the literature and arrive at the best options for patients, its widespread adoption into the flow of clinical care and medical education will require a more comprehensive commitment to high-value care such as that suggested by Johnson et al.16 Rather than being satisfied with the results of an evidence-based analysis, we should endeavor to understand the barriers in our health system that limit the adoption of evidence-based treatment and work to overcome them.

EBM holds great promise for our future health care system as large patient care databases become available that will augment the published medical literature. Advances in information technology will undoubtedly give us new tools to help us combine the information from these two sources to provide patients more useful, in-depth EBM answers to clinical questions specific to individual patients during a clinic visit. But these tools will be useless if we do not adopt them and integrate them into shared decision making. We must encourage the use of these tools through our classroom teaching of EBM with our students and through our implementation of EBM in clinical practice. Our patients should not have to wait 30 years for the next generation of advances in care, as occurred with acute migraine treatment. By combining the tools of EBM, encouraging the diffusion of innovation, and using principles of quality improvement, we can teach our students to provide the scientifically rigorous, financially responsible, and ethically appropriate care our society needs and our patients expect.

Back to Top | Article Outline


1. Jones J, Sklar D, Dougherty J, White W. Randomized double-blind trial of intravenous prochlorperazine for the treatment of acute headache. JAMA. 1989;261:1174–1176.
2. Friedman BW, West J, Vinson DR, Minen MT, Restivo A, Gallagher EJ. Current management of migraine in US emergency departments: An analysis of the National Hospital Ambulatory Medical Care Survey. Cephalalgia. 2015;35:301–309.
3. Friedman BW, Irizarry E, Solorzano C, Latev A, Rosa K. Randomized study of IV prochlorperazine plus diphenhydramine vs IV hydromorphone for migraine. Neurology. 2017;89:1–8.
4. Guyatt GH. Evidence-based medicine. ACP J Club. 1991;114:A16.
5. Haynes RB. What kind of evidence is it that evidence-based medicine advocates want health care providers and consumers to pay attention to? BMC Health Serv Res. 2002. Accessed February 3, 2018.
6. Rusiecki J, Schell J, Rothenberger S, Merriam S, McNeil M, Spagnoletti C. An innovative shared decision-making curriculum for internal medicine residents: Findings from the University of Pittsburgh Medical Center. Acad Med. 2018;93:937–942.
7. Greenhalgh T, Howick J, Maskrey N; Evidence Based Medicine Renaissance Group. Evidence based medicine: A movement in crisis? BMJ. 2014;348:g3725.
8. Maggio LA, Tannery NH, Chen HC, ten Cate O, O’Brien B. Evidence-based medicine training in undergraduate medical education: A review and critique of the literature published 2006–2011. Acad Med. 2013;88:1022–1028.
9. Maggio LA, ten Cate O, Chen HC, Irby DM, O’Brien BC. Challenges to learning evidence-based medicine and educational approaches to meet these challenges: A qualitative study of selected EBM curricula in U.S. and Canadian medical schools. Acad Med. 2016;91:101–106.
10. Shi CR, Nambudiri VE. Time for acute focus on chronic care in undergraduate medical education. Acad Med. 2018;93:835–838.
11. Morris ZS, Wooding S, Grant J. The answer is 17 years, what is the question: Understanding time lags in translational research. J R Soc Med. 2011;104:510–520.
12. Leggott KT, Martin M, Sklar D, et al. Transformation of anesthesia for ambulatory orthopedic surgery: A mixed-methods study of a diffusion of innovation in healthcare. Healthc (Amst). 2016;4:181–187.
13. Wesson DE, Kitzman HE. How academic health systems can achieve population health in vulnerable populations through value-based care: The critical importance of establishing trusted agency. Acad Med. 2018;93:839–842.
14. Rowland P, Kumagai AK. Dilemmas of representation: Patient engagement in health professions education. Acad Med. 2018;93:869–873.
15. Tatsioni A, Bonitsis NG, Ioannidis JP. Persistence of contradicted claims in the literature. JAMA. 2007;298:2517–2526.
16. Johnson PT, Alvin MD, Ziegelstien RC. Transitioning to a high-value health care model: Academic accountability. Acad Med. 2018;93:850–855.
17. Berwick DM. Disseminating innovations in health care. JAMA. 2003;289:1969–1975.
© 2018 by the Association of American Medical Colleges