The name “Slicer Dicer” might at first conjure up images of a food-prep kitchen appliance hawked on late-night infomercials to home chefs.
But a diverse group of staff and providers at UCHealth are using a tool of the same name not to chop vegetables but rather to carve answers to health care questions from the millions of pieces of data stored in the Epic electronic health record (EHR).
Every day, providers and staff use the EHR to enter data about patient clinic visits, hospital stays, clinical trials, lab tests, and the like. The information goes into an ever-deepening data reservoir. Retrieving information about a single patient is a relatively simple matter. But what if the patient suffers from, say, asthma and takes a certain medication to treat it? A clinician might want to know how effectively the medication has worked – not only on that patient but all the other asthma patients who received it. How many exacerbations did they suffer? Were there side effects? Did they regularly refill their prescriptions? Were they hospitalized less frequently than other asthma patients?
Slicer Dicer can help to answer these kinds of questions. It enables the clinician to move from entering data to the EHR to pulling it out by crafting queries that target specific groups of patients. The tool sifts electronically through the sea of data in the UCHealth EHR. The clinical catch could drive improvements in patient care, boost efficiency, tighten medication management, increase clinical trial recruitments, and more.
A list of top UCHealth Slicer Dicer users – by number of queries run – gives a glimpse of its widening appeal. They include professional research assistants, clinical recruitment specialists, project managers, data analysts, physicians, and nurses in fields spanning neurology, pulmonary and vascular medicine, pediatrics and information technology.
That’s as it should be, said Michelle Jonjak, RN, a nurse in the Multispecialty/Renal Clinic at UCHealth University of Colorado Hospital and one of Slicer Dicer’s frequent users. Jonjak, who has experience as a database administrator, said it’s important to put “refined tools” like Slicer Dicer in the hands of nurses and other providers.
That’s because the information it produces can provide concrete help for people they care for. In 2015, Jonjak used Slicer Dicer to identify chronic kidney disease patients (CKD) who had not received their hepatitis B vaccinations, putting them at risk for liver disease. The hepatitis B virus can live on surfaces for up to seven days, Jonjak said, increasing the chances of exposure to CKD patients who may need dialysis and have blood going in and out of their bodies.
Data from the Slicer Dicer two years ago showed that only a tiny number of the Renal Clinic’s CKD patients had received the hepatitis B vaccination.
“That impacts lives,” Jonjak said. With the data, the clinic developed a protocol to ensure patients received the immunization. To date, nearly 150 additional CKD patients are now protected from hepatitis B, and the number continues to grow.
After this success, Jonjak continued to use Slicer Dicer to help both patients and providers. For example, she used it to find patients on certain high-risk medications, such as immunosuppressants. With that, she made lists to help nurses ensure these patients get necessary lab tests.
Jonjak is now working with Meg Ragland, MD, MS, a fellow in the Pulmonary Sciences and Critical Care Medicine Division at University of Colorado School of Medicine, to increase the number of patients with chronic lung disease, such as COPD, who receive pneumonia and pertussis (whooping cough) vaccinations. Jonjak used Slicer Dicer to “identify gaps” in the numbers of patients vaccinated.
The search revealed a problem. In fiscal year 2017, which ended June 30, only about 40 percent of the patients had received their initial pneumococcal 13 vaccinations. The percentage was similar for the pneumococcal 23 boosters recommended for patients 65 and older. As for the pertussis vaccine, the percentage was just 28 percent.
The information from Slicer Dicer offered a “doable” opportunity for improvement. With flu season approaching, the Pulmonary Clinic set a goal of vaccinating at least 50 percent of the targeted patients in the next year. Doing so could help to decrease hospitalizations, mortality and the cost of care.
Ragland made that case October 11 at the CU Department of Medicine’s second annual “Shark Tank” competition at UCH, where UCHealth providers present ideas for quality-improvement projects to a panel of clinicians playing the role of the skeptical entrepreneurs on the hit television series of the same name. The panel tests the strength of the ideas with questions and then huddles to evaluate the presentations. The winner receives assistance from the Department of Medicine in developing the project and implementing it in the clinical setting.
In her presentation, Ragland emphasized the value of increasing vaccination rates to patients with chronic lung disease. She noted that hospitalizations related to COPD alone account for more than $6 billion in medical costs annually in the United States. She presented the Pulmonary Clinic’s vaccination data, calling it ripe for improvements that could protect patients while also reducing costs.
The Pulmonary Clinic did not win the Shark Tank competition, but Ragland said the panel provided valuable feedback. The effort to increase vaccination rates will move forward, she added.
“We need projects that will make an impact on patients that we care about,” Ragland said.
The next step is to evaluate the clinic’s workflows and staffing, with an eye toward finding barriers to vaccination and minimizing or eliminating them.
“The data is motivating,” Ragland said. “If we find we can do better at vaccinating patients than we were, then we will keep going. If the things we are doing aren’t working, then it’s better to know that and ask ourselves why.”
The more Slicer Dicer encourages health care providers in all fields to answer these kinds of real-life questions, the more valuable it will become, said Jeff Sippel, MD, MPH, an associate professor of Pulmonary Sciences and Critical Care Medicine at CU and an Epic physician champion for UCHealth. Sippel devotes a portion of his time to encouraging colleagues like Ragland to take advantage of Slicer Dicer.
“It’s a fast and efficient tool that people can use to test ideas they think of,” Sippel said. “It can help people understand what reality looks like in the patient data.”
Sippel, for example, is interested in developing treatments and therapies for high-risk pulmonary patients, such as those with COPD and emphysema. Slicer Dicer helps him to identify patients with those diagnoses quickly, and then add refining criteria, such as oxygen and high-risk medication use. The idea is that by isolating these patients, Sippel and others can evaluate their care and possibly find new, evidence-based ways to protect them while making the best use of limited resources.
That’s the general idea behind population health, which centers on finding ways to standardize care to groups of patients, especially those with chronic conditions, rather than delivering fragmented, episodic services to individuals – an approach that has been much criticized as wasteful and inefficient, yet difficult to change.
Slicer Dicer is an important part of the movement to population health management, Sippel said, but he cautioned that its value is only as great as the ingenuity and creativity of the people using it. Simply finding diabetes patients who need foot or eye exams, for example, is only the first step toward improving care. The next question lies in how providers and communities make that happen.
“We might think that population health follows a well-worn script,” Sippel said. “There is nothing close to that. It’s actually an emerging area that needs a lot of people to think about it. Slicer Dicer can help it to evolve more rapidly. It’s very efficient for people who want to just sit down and start learning more about their patients.”
Work in progress
The number of users is relatively small, at least for now, Sippel conceded. That’s attributable to several factors, including unfamiliarity with the tool; the habitual lack of free time available to clinicians; and an initial sense of intimidation in looking at the seemingly bottomless depths of a half-million patient records at the beginning of a search for a relative handful.
Slicer Dicer also has its limitations. For example, yes-or-no queries – have patients received their vaccinations or had their A1C blood sugar levels checked, for example – return clear-cut data. But Jonjak noted the tool falters at times on more nuanced queries. She cited creatinine blood levels – a key measure of kidney function – as an example. Slicer Dicer can reveal patients with normal or abnormal levels, but it can’t as yet separate patients by multiple ranges of values.
In addition, the data Slicer Dicer retrieves is only as good as the data that goes into the EHR. That highlights the importance of good documentation, Sippel said, notably entering the correct ICD-10 code for disease diagnoses.
Ultimately, the value of Slicer Dicer will be determined by the role it plays in improving patient care, Jonjak said. For providers, that means using the tool to identify issues that can be reasonably addressed with available resources.
“If less than half our patients are getting the pneumonia vaccinations, for example, it’s worth our time to do something about it,” Jonjak said. “It’s a place we can devote our energy.”