NEW YORK (GenomeWeb) – Even with electronic alerts and hard stops built into the process of test ordering, some clinicians will forge ahead with inappropriate diagnostic testing and antibiotic use. In the case of Clostridium difficile testing, researchers at Johns Hopkins University are now bringing in cultural anthropologists and human factors engineers to figure out why some doctors persist in ordering tests for patients that don't fit criteria and guidelines, and they hope to ultimately develop improved computerized support tools.
C. diff bacteria have certain unique features that require more judicious testing than average infectious diseases. They are encapsulated in a spore that is released in the copious diarrhea produced by a patient experiencing an active infection. The spores are particularly hard to kill on surfaces, contributing to C. diff being a common and reportable hospital-acquired infection and requiring the hospital to undertake special contact isolation proceedures for infected patients.
But C. diff can also be part of a healthy person's intestinal microbiome. It is only when the bug grows out of control that it produces toxins, potentially causing severe intestinal damage.
Clare Rock, an infectious disease expert and epidemiologist, is part of a JHU team that recently won approximately $600,000 in funding from the Centers for Disease Control and Prevention's National Center for Emerging and Zoonotic Infectious Disease to research the problem of over-testing for C. diff.
The project – called the Clostridium difficile over testing improvement initiative, or CDOT – includes the participation of nine other medical centers and also involves cultural anthropologists and human factors engineers associated with JHU's Armstrong Institute for Patient Safety and Quality. The group is also funded by the CDC's Epicenters Program.
Rock speculated in an interview with GenomeWeb that the etiology of inappropriate ordering is complex.
It could be, in part, that clinicians think "more is better," or that they don't understand the potential risks to a patient of disrupting the intestinal microbiome with unnecessary antimicrobials, Rock said. It may also be that more senior clinicians recommend testing during rounds, and more junior clinicians do not feel empowered to overrule those recommendations in the light of new information discovered when they are entering the test orders in later.
Unfortunately, diarrhea is a common problem in hospitals, and testing for the presence of bacterial DNA in a person who is a carrier of C. diff bacteria will lead a positive result, even if C. diff is not actually the cause of the symptoms. On the other hand, testing for the presence of the toxin or for metabolic enzymes produced by the bacteria is less sensitive, so molecular testing is now more frequently done, to the point that the increased sensitivity is leading to higher reported rates of infection, as previously reported.
In general, the choice of using molecular or non-molecular C. diff testing should be based in part on how strict the lab or hospital is in deciding whom to test, according to guidelines issued this year by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America.
In order to routinely use molecular testing, the guidelines specify that the hospitals must have strict criteria, only testing "unformed stool" from patients who have had more than three episodes of diarrhea in 24 hours, and excluding any formed stool samples from testing as well as samples from patients who have been given laxatives within the past 48 hours.
If hospitals are not screening out samples in this strict manner, then the guidelines suggest they are better off using an algorithm involving screening out negative cases with non-molecular tests for toxins or C. diff metabolites, followed by additional tests to confirm positive results.
Rock said the JHU project is an extension of research that found, among other things, that many C. diff tests were being performed on patients receiving laxatives, she said.
And, incredibly, even face-to-face feedback and education from a team of infectious disease specialists was not able to completely ameliorate the problem of inappropriate testing, as described in a recent publication of the pilot study.
Rock said the JHU team initially integrated a computerized clinical decision support tool into the electronic medical record that would alert a clinician if, for example, a patient had received laxatives or was otherwise not a good candidate for testing. Then, the tool would provide links to educational material related to best practice for C. diff testing.
If, after reading up on the subject, the provider still thought that there was some reason to send the C. diff test, he or she would have to call the microbiology lab to get a secret passcode to enter into the electronic medical record to allow the test ordering process to proceed.
In the published study, a team of experts also reviewed hospital records to find cases of inappropriate testing, then went to the offending clinicians with case-specific teaching points and recommendations.
The research found a high proportion of positive cases were asymptomatic carriers who had subsequently received C. diff therapy. After discussing the inappropriateness of testing and treating these carriers, prescribers were only willing to discontinue the inappropriate therapy in about half of the cases.
Similarly, recent research at a health system in Washington found that 70 percent of patients with positive C. diff tests had been tested inappropriately because had been taking laxatives or did not have diarrhea, and 95 percent of these patients were subsequently treated with antibiotics. In this case as well, a computerized support tool and education program did not fully eliminate inappropriate testing.
The JHU study hypothesized a few reasons for the reluctance to change behavior, including an expectation that a test will detect subclinical disease, a poor understanding of the limitations of the test, so-called "defensive medicine," and an overestimation of signs and symptoms and subsequent classification of patients as having severe disease.
Given all the prior work, the JHU team's goal now is to send the anthropologists to do real-time one-on-one interviews of clinicians who have chosen to override the best practices, in the hopes of understanding the motivations.Â
"We're going to have a team of human factors engineers and cultural anthropologists, or qualitative researchers, and they are going to get to the front-line providers," Rock explained. This front line includes physicians, nurse practitioners, physician assistants, and the residents and attending physicians as well. The goal is reach them in the moment as they are ordering these tests to try to get insights into their thought processes.
The team will also follow up with semi-structured interviews to understand if the inappropriate ordering is due to a perception of risk, a fear of missing infection, or perhaps not understanding the potential disruption of the colon microbiome associated with treating an asymptomatic C. diffcarrier, Rock said. With this data in hand, Rock hopes they may be able to design better ways to target overprescribers.
Finally, the project involves usability testing, with the human factors engineering group adjusting the current computerized clinical decision support to enhance compliance, but also integrating it into the clinical workflow so it blends and is not intrusive, Rock said.
Although some doctors may perhaps think more testing is better, even advocates promoting public awareness of C. diff know that is not the case.
"There are many people who have C. diff bacteria in their gut who are not sick, and that is one of the challenges of dealing with this disease — the presence of the bacteria itself does not necessarily mean you're ill," said Christian Lillis, co-founder of the Peggy Lillis Foundation, an advocacy organization he started with his brother after their mom died of C. diff infection.
Lillis agreed that perhaps particularly with an older, hospitalized patient population, clinicians don't want to miss anything. Also, "It has only been a couple of years since hospitals and professional associations have begun to push out stronger guidelines," Lillis added.
Using the recommended two-step algorithm may have an increased cost associated, but "another C. diff test to then not give someone a multi-thousand-dollar course of antibiotics that will make them sicker is worth the two steps," Lillis said.
Perhaps one explanation for the resilience of the inappropriate test ordering behavior is that doctors, like all of us, tend to do things out of habit, Lillis said. "If you have a strategy to get people to stop ordering C. diff tests and you haven't accounted for the all of the cultural incentives and beliefs that make them order tests, it's never going to work — culture eats strategy for breakfast," he said.
Lillis himself has served on an expert panel to help develop antibiotic stewardship tools for healthcare workers, and he said that in the age of smartphones, "things are dinging at us all day, every day," and perhaps it is also automatic to tune out alerts.
More gatekeeping, including strict rejection protocols – typically involving not testing so-called formed stool, for example – could reduce over treatment, but "the downside of that is you could miss people who actually have C. diff infections," Lillis said.
He said people who don't fit the typical C. diff profile — who may be younger or who have not been recently hospitalized — struggle to get diagnosed when physicians assume their diarrhea is due to food poisoning and don't order a C. diff test. In fact, only about 30 percent of Americans have ever heard of C. diff, Lillis said, so patients aren't asking for testing. His group wants to increase awareness through outreach to help change that.
"All of this is trying to thread a needle, where you're not underdiagnosing and you're not over-treating," he said.
Ultimately, appropriate diagnostic use is better for the hospital, since C. diff rates are reportable and contact precautions are expensive, but it is also best for the patient, Rock said. "If they are inappropriately diagnosed, it is exceptionally difficult for physicians to not treat a positive test," she said, so inappropriately tested patients invariably end up getting inappropriate treatment.
The study is expected to last two years, and the final tool kit will be able to be used by other hospitals to help them direct their physicians and providers to more appropriate C. diff testing, Rock said.