Why Aren’t Doctors Following Guidelines?

The Hospitalist. 2016 October;2016(10):

October 31, 2016 | The Hospitalist

New Tools of the Trade for Crafting Clinical Guidelines

The well-known GRADE system and similar tools such as Levels of Evidence and Grades of Recommendation have helped guideline writers for years, particularly in evaluating bodies of medical literature and the strength of the studies’ conclusions. Cincinnati Children’s Hospital Medical Center uses a similar strength-of-evidence pyramid to gauge the relative reliability of data: physician expertise and practice at the base, a retrospective or cohort study at a higher level, and a systematic review composed of numerous randomized controlled trials at the pinnacle.

Not every clinician has been taught how to appraise articles, however. Accordingly, Cincinnati Children’s James M. Anderson Center for Health Systems Excellence has developed another system called LEGEND (Let Evidence Guide Every New Decision) to help guideline developers know what to look for when reading a study. The system’s analysis boils down to three main questions: Is it valid? What are the results? And are they applicable to my population?

“If you want to know whether the study that you’re reading is something that should prompt you to change practice, you want to know if the study is a good one,” says Wendy Gerhardt, MSN, the hospital center’s director of evidence-based decision making.

In fact, the hospital has developed tools to assist in nearly every step of the guideline-crafting process. The tools help clinicians learn how to read studies, develop an evidence-based guideline, understand whether a guideline is solid, know where separate recommendations agree and differ, and implement new guidelines into regular practice.

One tool called REACH (Rapid Evidence Adoption to improve Child Health) uses quality improvement consultants and multidisciplinary groups to “translate evidence into point-of-care decision making by clinicians, families and patients,” according to its website. The process takes about 120 days and can result in decision aids such as prepopulated electronic order sets that default to evidence-based suggestions for, say, bronchiolitis inhalation therapies.

“It’s really helpful when you’re working in an academic center and the residents are the ones writing the orders,” says Gerhardt. “So it defaults to the right thing, and they have to actually think about not doing it that way.”

Often, it’s not enough merely to give doctors the link to a new guideline.

“If you can pull up an order set that already has the evidence embedded in it, that’s a little more compelling,” she says. “You kind of have to put the evidence at their point of care instead of in a document. And that’s what, in my mind, makes it real.”

At Cincinnati Children’s, she and her colleagues also have taught doctors how to use PubMed to seek out systematic reviews if they have a question. They have rolling computers, too: Medical librarians sometimes go on rounds with clinicians to help with on-the-spot literature searches.

“It’s however you can make it easier for them to use,” Gerhardt says. “By and large, most people just want to practice, so you have to put that evidence in their way.”

—Bryn Nelson, PhD