CDS Tools and Patient Satisfaction
By Juliann Schaeffer
For The Record
Vol. 25 No. 6 P. 6
Clinical decision-support (CDS) technology, such as vaccination reminders, medication error alerts, and chronic disease management tools, is largely regarded in the medical community as a valuable asset that can assist physicians with medical treatment recommendations and optimize patient care.
Yet a new study from the University of Missouri, “Why Do Patients Derogate Physicians Who Use a Computer-Based Diagnostic Support System?” led by Victoria Shaffer, PhD, an assistant professor of health sciences and psychological sciences, and published in Medical Decision Making, found that doctors’ use of such computerized diagnostic tools led to decreased patient satisfaction and, more significantly, could eventually have ramifications for how often patients heed their doctors’ advice.
“If a patient is less satisfied with the physician’s performance, then the authority of the physician to encourage or compel compliance is diminished,” explains Hal Arkes, PhD, emeritus professor of psychology at Ohio State University and a coauthor of the study.
For the study, researchers set up three separate experiments for which participants were given vignettes of a physician-patient interaction with a diagnosis that was either unaided (for the control group) or assisted by a “computer-based diagnostic aid.” The first experiment also included a scenario in which a doctor sought a colleague’s advice before making a final diagnosis.
After reading their respective scenario, study participants evaluated the physician’s diagnostic ability, professionalism, overall satisfaction, thoroughness, and wait time at the physician’s office via a numbered scale. What did they find? “The physician who was described as using a computer-based diagnostic aid in a brief vignette was given significantly lower ratings of diagnostic ability, professionalism, and overall satisfaction than the physician who made an unaided diagnosis,” the authors wrote. Conversely, patients viewed physicians who made unaided diagnoses or those aided by a colleague as more favorable on most accounts.
But why might patients in effect punish physicians who are only utilizing such tools to provide better, more complete care? Is it the physician’s “need” to seek information elsewhere that makes patients uncomfortable? Is it a distrust of technology itself? Is this distrust felt only by a certain subset of patients and if so, what characteristics define this type of patient? The study gives no concrete answer for the whys behind patient attitudes, but the authors note that any (or all) of the aforementioned could be in play.
A Matter of CDS Semantics
Do patients really have such little faith in CDS technology? Not so fast, says Dan Riskin, MD, MBA, CEO and cofounder of Health Fidelity, who says this study should be taken in context. “Dr Shaffer offers a meaningful contribution to the literature with [this study],” says Riskin, who nevertheless takes issue with the wording used in the vignettes, which he says could cloud the results.
“The study appendix reveals the clinical stories used the word ‘aid’ to refer to clinical decision support,” he explains. “This term creates significant bias. I am a doctor and I can assure you that I would be uncomfortable seeing another physician that required an aid to treat a patient. I envision another surgeon pulling open an anatomy book during an abdominal operation to try to find the liver. Concern over doctors that require aid to do routine work is justified.”
According to Riskin, CDS technology should not be referred to as an aid but as technology that provides “support to consistently follow standards of care and improve outcomes. Who would object to that?
“If there is a ‘timeout’ before an operation to ensure that consent is in place and antibiotics have been given, that’s just good medicine and has been shown to significantly reduce complication rates,” Riskin continues. “Can this be considered an aid? Certainly. But it is better considered a quality-focused approach to medicine. And anyone who has read the literature would strongly prefer to be seen in a system that requires such approaches.”
Jonathan Teich, MD, PhD, chief medical informatics officer for the health sciences division of Elsevier and a physician at Brigham and Women’s Hospital in Boston, agrees: “One really important thing to say up front [is that this] study was not about doctors who use any kind of clinical decision support; it was specifically about doctors using computer-assisted diagnostic aids. This wasn’t about a doctor who checks for drug interactions or who uses order sets or who has a computer provide reminders about important health maintenance tests; those often are associated with good outcomes and good satisfaction. The study was about a doctor who relied on a computer to decide what the likely diagnosis was and what test to do next. And in the experiment, the doctors made a big point of saying that they were going to ask the computer what to do.”
According to Teich, when a similar experiment was conducted using real patients, physicians were rated lower on diagnostic ability but not on other factors such as overall satisfaction or professionalism. “So if the question is ‘Would I guess that my doctor’s personal diagnostic skill is lower if she makes a big point of asking the computer what to do about the diagnosis?’ then sure, I’m not surprised about that finding,” he says.
Patient Education Key
The study authors, Teich, and Riskin agree about one thing regarding CDS: There’s a definite need for patients to be educated on the topic—both on what these tools are as well as what they do.
First on the agenda, according to Teich, is clarifying how CDS differs from diagnostic aids. “We know that CDS systems are intended to support the doctor, to be a part of the decision-making process, to do what the computer does best: remind, alert, present information, never forget,” he says. “[Whereas] clinical diagnostic-support programs are more complex and can sometimes be perceived as replacing, not supporting, the doctor’s judgment.”
“In the electronic record, clinical decision support and even diagnostic support similarly strive to create consistent guideline-based care,” Riskin says. “Perhaps patients need to be educated that this strongly benefits their outcome.”
Or is it possible that changing the vernacular could thereby change patients’ perceptions of the tools physicians use to support their decision making? Riskin believes it’s a possibility. “Perhaps we as an industry need to avoid referring to this as aid. If we refer to it as ‘clinical decision support to create consistency and improve outcomes’ perhaps patients will not only feel comfortable but will actually come to expect it,” he says.
Arkes agrees: “Perhaps the CDS might be described as a tool, much like an X-ray or MRI. Patients don’t object to the use of diagnostic tools. [Or physicians] might describe the CDS as a second opinion, which I suspect most patients would not object to.”
Teich says part of the problem may come down to the fact that many patients still want to believe their doctor is all knowing. “I think even in this day and age, some patients still want to see their doctors as superintelligent beings who have—as they said in the study—‘an aura of omniscience.’ There are subtle and smooth ways to incorporate CDS into the workflow; the scripts used here, where the doctors made a point of saying they needed to ask the computer to make the call, would certainly lead you to the opposite impression,” he says, noting that proper implementation can correct that. “The implementation and use of CDS is every bit as important as its design.”
Whatever wording physicians use to describe the CDS technology they’re utilizing, Teich says the most important point is that they use it—it being whatever technology allows them to provide the end point of the best quality care.
“Hopefully, the real question for many of us is ‘Did I come out of the encounter with the best care and the best plan going forward?’ We all know that doctors aren’t omniscient and that they make mistakes,” he says. “For me, I’d rather know that by whatever means necessary, I got the best advice and the best result.”
— Juliann Schaeffer is a freelance writer and editor based in Allentown, Pennsylvania.