“If we can put the right information, at the right time, into physicians’ hands, we can change two things. [We can] lower costs and improve the quality of care by enabling physicians to make the
right decisions with the right information.” —Dan Michelson, CMO, Allscripts-Misys Health-Care Solutions Inc.
The Problem
Many health care providers and economist are chasing after the dream described by Michelson (above) and Pfizer decided to investigate ways to make this happen by investing in efforts to help enhance the quality of health care services by bringing cutting-edge research into the clinic. They awarded funding to a team of researchers to learn more about how physicians can use the latest evidence-based research in the clinic in real time.
The Approach/Methods
Implementing a mixed methods approach, the researchers conducted a series of focus groups and intensive interviews with a range of practicing physicians to better understand their perspective on the value o evidence-based research results and what potential benefits and obstacles exist in creating such a system. EthnoNotes (Dedoose’s predecessor) was the research team’s choice for managing and analyzing their data. The fact that Weisner got the grant had nothing to do with it. Project resources included transcripts of focus groups which were then coded. Resource descriptors included physician age, gender, specialty, # of years in practice, type of practice setting and size, comfort with computers, online medical resources, and PDAs, and resources consulted in and out of the clinic.
Preliminary Findings
The research team used the code-by-code matrix—a feature that filters the qualitative data codes to search for pairs of codes used in an excerpt—and the code-by-descriptor charting
engine—which runs codes by each descriptor variable (e.g., by gender, specialty, …). A number of important findings were quickly identified.
• Clinicians in particular areas of specialty see a greater need for evidence farming
“Since most of the random controlled trials are done with adult, in pediatrics you have to make a judgment call…so that’s bad”
• Desire for up-to-date information on patient compliance with prescribed medication
“Compliance is a big issue with hypertension…we’ve actually ordered the drug [but] we don’t capture whether the person has actually picked it up and certainly we don’t know if they’re taking it”
• Obstacles to system adoption include issues of the confidentiality of the data, patient-doctor relation management—discomfort in searching for information in front of a patient or leaving exam room to do
search, trust in system accuracy, and time-consuming interface.
“Now, we get the abstract data out of systems into a data warehouse that have the same model…and by that we mean the diagnosis of discharge. That’s the only one we store. And if everybody is doing
that, regionally, then we can collect our diagnosis regionally and be reasonably certain that we’re all talking about the same thing. Now, if [X] gave us the admitting ones and UC Davis, you know,
sent in discharge ones, and then Kaiser did something different, then it wouldn’t make any sense”
“I mean if it’s more than a couple of clicks or keystrokes, it’s not going to get used...and sensitivity to the workflow is going to be a real key…in the right place at the right time for the right patient”
• Great value in quickly being able to diagnose and prescribe medication for ‘tricky’ ailments
“Particularly, yeah, for people with chronic pain I think that’s very useful simply because your medication management becomes so tricky with issues with dependence and whatnot. Also insomnia would be
another one where it’s very confusing as to what to do, what the first line should be, and how to really select amongst the medications available out there, you know, what the rationale is. There just
doesn’t seem to be much guiding evidence at all. And then it would be most I guess in cases where I feel like my clinical experience is less than any data whatsoever”
Phase 2—Video Scenarios:
Findings from Phase 1 were used to shape the content and production of brief videos depicting scenarios of an electronic evidence-farming system being used within a clinical setting.
A number of physicians and medical students viewed these videos and then participated in intensive interviews and/or focus groups.
Phase 2 Findings:
Overall, the participants envisioned a great value if having detailed and scientific information available in the clinic. Yet they also astutely pointed out a number of issues that
demanded consideration in the development and implementation of a viable system. Among the most illuminating issues raised:
• “Buy-in”—many physicians, particularly those with more extensive clinical experience, may be reluctant to use evidence based findings in the clinic.
“I mean as much as we want to go evidence based on everything, you gotta be a little bit careful. Sometimes our instinct and what we go by is very simple and it might be the best option”
• Potential Limits on usefulness of Randomized Control Trial (RCTs) Findings
“You don’t know how controlled it is, you don’t know how many people are actually making it in or not…[the clinic is] like a study. That’s why these studies are done, because it’s hard to get the patients. But now you’re just making up your own little study and extrapolating data and it just doesn’t seem… very safe”
• Tremendous value in reducing costs and patient discomfort if possible to share and explore diagnostic data in real-time—particularly for uncommon or poorly understood conditions and medication effectiveness “Or like the one with the leg pains, the weird viral thing? Oh that was wild. It was a viral thing. My son, literally, could not walk for 2 days because his legs hurt so much,
and we had like 3 or 4 [cases], and it was like, just, you know, some poor kid went through a whole workup with a lumbar puncture and everything, and by the time it got to my side,
it was like, ‘Oh, it’ll be over in a day and that’s it.’… If we put it in [to an evidence farming system] and then other people started seeing that…[it’s] something that’s not just
isolated to one person…It makes the physician feel a little more comfortable… and then you see it resolve, and what works”
“I’ll say, in 2 days if you’re not better, then you can go ahead and get it filled… and they’ll say, ‘Okay, that’s fair’ and then they’ll try it. And most of the time people get
better with time, whether the antibiotics whether they’re taking it or not, but it’s nice to have a backup. To say, ‘Here, look, only 10 people out of 50 or 80 used it’ I think that could be helpful”
“I think even something like adhesive medicines like say Aderol, say, and compare complications you’ve had and complications which you can’t understand, is there a theme and why and
so forth? Something like that, something we could say is why we use this medicine a lot more. Why is this used a lot more, why is the price more, why is it used for osteoporosis
or whatever it is, then you can have some idea of what education for the community is needed”
“For me I’d really look for something to really track my outcomes as well, whether it’s having people come in and have a reexamination in a year or at three years or even just sending
a questionnaire that they can fill out on-line and let you know how it’s going. It would be easy to track that data”
• The characteristics of patient populations must be considered
“I think with our patient population it probably would work because our patient population is very computer savvy and generally well-educated, and isn’t overwhelmed by charts and graphs, and doesn’t
feel intimidated by it, and I think perhaps with some patient populations that might get in the way, like it would be too information when they’re not feeling well. It’s easier with Pediatrics
because you have the parent whose not, well hopefully not ill, as opposed to, I guess with you guys, because the patient who’s coming in is also the person you’re explaining to, and if they’re
feeling miserable they may or may not hear all you have to say”
• And so many more benefits:
o Teaching—“ I think it would be very useful in a teaching environment for a medical student. I could definitely foresee the attending kind of serving what’s going on in the clinic,
the residency and everybody and then the medical student for in addition to everything that they’ve done there. The medical student pulling up data to further convince the patient if
the patient argues with them. I mean I see that very much in a teaching setting”
o Added confidence—“a lot of time there’s obvious surgical cases, obvious non-surgical cases, but there’s a fair amount that fall in the grey area where you have to really weigh the
risks and benefits of the procedure and I think it would come into play quite often in those cases. Either in convincing yourself or convincing your patient…yes or no”
o Direction—“90% have osteoporosis, then I’d say, ‘Maybe we should all start screening everybody, or maybe we should cut down the age group, or, you know, or are they all Japanese?
Or are they all, are they white? Are they…?’ You know, it just, it would be very helpful, on screening people, I mean, to figure out what you should be looking at”
o Access to information—“Once you step back from that training environment, you just don’t have that much access as much anymore because, you know, you’re in private practice or even
the county, like being in a satellite clinic… So something like this you could actually plug into the data and know what’s going on right away”
EthnoNotes helped bring this study’s findings to light—quickly, systematically, and with results that were clear and to-point.