A Discussion on the Development of a Clinical Decision Support Tool for the Management of Uncontrolled Cancer Symptoms

J Clin Pathways. 2023;9(2):9-11

Founded in 2016, the Clinical Pathways Forum is a community of pathways professionals—now total­ing over 15 institutions from across the US—who are utilizing clinical pathways in their practices and institutions to improve cancer care. Forum leader Mishellene McKinney, MHA, RN, OCN, organizes quarterly conference calls with Forum members to facilitate discussion of shared experiences and lessons learned regarding pathway use as clinical path­ways become more prevalent and evolve to meet the needs of value-based health care systems and reim­bursement models.

The mission of the Clinical Pathways Forum is to facili­tate a knowledge exchange for overcoming the chal­lenges of developing, implementing, and measuring clinical pathways to demonstrate the value of standardizing clinical care. The main activity of the Forum is to schedule quarter­ly conference calls to share experiences and lessons learned using clinical pathways.

The Forum publishes highlights from each of the confer­ence calls that occur throughout the year in the Journal of Clinical Pathways (JCP). This first installment for 2023 sum­marizes the speakers’ presentation and discussion from the January 2023 call. Health care professionals from across the continuum of care are encouraged to join in these collabora­tive discussions. Forum contact information can be found in the online version of this article.

We had the opportunity to hear from Mary E Cooley, PhD, RN, FAAN, nurse-scientist at Dana-Farber Cancer Institute in Boston, Massachusetts, and David F Lobach, MD, PhD, MS, FACMI, Vice President of Health Informatics Research for Elimu Informatics Inc in El Cerrito, California.

Drs Cooley and Lobach collaborated previously through an R01 grant. Now, Dr Lobach through Elimu Informat­ics and Dr Cooley with the Phyllis F. Cantor Center for Research in Nursing and Patient Care Services of the Dana-Farber Cancer Institute have a subsequent Small Busi­ness Innovative Research (SBIR) contract from the National Cancer Institute (NCI) to develop a clinical decision support tool that will provide evidence-based guidance to clinicians at the point of care for the management of uncontrolled symptoms in patients with cancer.

Why Cancer Symptom Management?

Dr Cooley opened the discussion by reviewing the drivers behind the project. “Most adults with cancer have distressing symptoms that are often undertreated and result in increased morbidity, mortality, and cost,” she stated. “Given the nega­tive effect of uncontrolled symptoms, improving symptom management is an important aspect of quality cancer care.” Although clinical guidelines are available and can enhance symptom management, they are often not used in practice.¹ Dr Cooley said that innovative approaches are needed to integrate evidence-based symptom management into rou­tine oncology care, and clinical decision support tools using guideline-based symptom management pathways may help in the dissemination and adherence to guidelines.

Foundational Work

Previous studies have shown that the provision of evidence-based recommendations delivered on a paper report for management of individual symptoms, such as pain and de­pression, were promising in improving care.^2,3 Drs Cooley and Lobach initially examined the use of computerized algorithm-based decision support for the assessment and management for 5 of the most common symptoms in adults undergoing treatment for lung cancer: fatigue, pain, dyspnea, depression, and anxiety.

In the first phase of their work, they evaluated differences in clinical management of symptoms and health-related qual­ity of life between a Symptom Assessment and Management Intervention (SAMI) group and an attention control group.⁴ Physicians were randomized to the intervention or the atten­tion control group in two outpatient thoracic oncology cancer centers, one of which was a comprehensive cancer center and the other was in a large safety net hospital in New England.

In the SAMI group, patients completed the computer­ized symptom assessment and clinicians received a printed symptom management report with specific recommenda­tions to enhance care for the patients. Patients also received a self-care management booklet. In the attention control group, patients completed the computerized symptom as­sessment and received the symptom management booklet at the end of the intervention.

The SAMI intervention included the collection of contextual information from the patients regarding use of supportive care medications, whether the patients were adherent to the prescribed medications, and whether the patients had hazardous alcohol use, since this issue would have implications for prescribing medications. A green, yel­low, and red traffic-light system was used to identify symp­tom severity. This approach allowed for patient-specific recommendations at the time of the visit. For example, if the patient had a moderate pain level with a somatic component, the system provided information about titrating or initiating opioids, prescribing adjuvant medications (eg, nonsteroidal, antiinflammatory agents) as appropriate, and adding a pro­phylactic bowel regimen to prevent constipation.

The Sapphire Fully Integrated Cancer Symptoms Management Tool

The conversation then transitioned to Dr Lobach, who dis­cussed the move from the SAMI paper-based system to a more sophisticated, algorithm-driven system that is fully integrated into the electronic health record (EHR), called Sapphire Cancer Symptom Management. The goal of the Sapphire project is to generate actionable, detailed, patient-specific guidance based on a combination of patient-reported information and EHR data processed by decision-support algorithms, which have been developed through expert panels that Dr Cooley coordinated with clinical experts at Dana-Farber Cancer Institute. Dr Lobach em­phasized, “The tool makes recommendations, but a clinician always has the final say in what should be done.”

Sapphire Cancer Symptom Management, which is fund­ed by a Phase II SBIR contract from the NCI, focuses on 9 symptoms—fatigue, pain, constipation, depression, anxiety, insomnia, nausea/vomiting, diarrhea, and skin reactions— with direct data collection from patients to assess symptom severity and obtain contextually relevant information that impacts symptom management. Workflows facilitate data review by clinicians and provide explicit, patient-tailored, evidence-based guidance for symptom management. In ad­dition to providing recommendations for symptom man­agement, the Sapphire application is designed to support encounter documentation and guidance for billing, and to provide issue-relevant educational materials for patients. The product uses SMART on FHIR technology, which allows the application to seamlessly integrate into the elec­tronic record workflow.

Workflow

Patients with an upcoming visit that meet specific criteria are identified to receive an SMS text message invitation to complete an electronic patient-reported outcome (ePRO) questionnaire prior to the visit. The team uses the Patient- Reported Outcome Common Terminology Criteria for Adverse Events (PRO-CTCAE) as the initial assessment tool. The questionnaire leads the patient through ePRO screening questions, and branching logic identifies the need for collect­ing additional data from patients whose responses show mod­erate or severe symptom intensity. For example, if a patient has a positive screen for severe constipation, the application will collect additional contextual information from the patient about their bowel movement pattern and the remedies they have tried to treat constipation to better inform and custom­ize the guidance recommendations to the patient’s situation.

During the visit, the clinician opens the symptom-management app in the EHR and reviews symptom sever­ity and trends, selects the symptoms he/she wants to ad­dress during the encounter, reviews and validates the data for those symptoms, and submits those data to the Sapphire Clinical Decision Support Engine to generate guidance. This complex decision engine generates patient-specific, evidence-based symptom management recommendations. The clinician reviews and selects the preferred management recommendations, chooses education materials to provide for the patient, and generates a management plan for the visit note, which includes suggested billing codes.

The conversation then opened up to the Forum participants for questions.

Have you gotten any feedback from clinicians as to whether this tool has been valuable, and what have you used to measure value to the clinician? People are particularly concerned about the growing number of messages in the inbox.

Dr Lobach: We don’t go through the inbox. This system is designed to be part of the clinical encounter. Because we use SMART on FHIR technology, the clinician will see cancer symptom management as a tab in the EHR. We also want to clarify that this product is still in development; we have not yet deployed this version.

Dr Cooley: We’ve done a lot of usability testing. We en­gaged clinical experts to help us update the algorithms by developing use cases that show them what the recommenda­tions would be like. We had half a dozen clinicians for each of the symptoms walk through the algorithms step by step to get feedback. Across the board, universally, people were excited about it. We engaged with both academic medical center providers and community-based providers. We think that an important target for a tool like this is a communi­ty-based center. These centers may not have as many rich resources available. One of the big concerns that came up is just making sure that it’s integrated into the workflow in a meaningful and easy way. So, the workflow is going to differ depending on the clinic or disease center.

Is the PRO-CTCAE required?

Dr Cooley: This product does not require the use of a spe­cific ePRO, because many organizations may have ePROs implemented already and we did not want to be limiting.

Can you give a sense as to how you envision this tool being funded by institutions? Would this be part of Epic’s application programming interfaces to be purchased as part of the App Orchard?

Dr Lobach: We are planning to have the application in the App Orchard for Epic and for the equivalent in Cerner, at least for our starting rollout. There would be a licensing fee. We feel, based on the literature we’ve seen, that there is a fa­vorable return on investment by doing symptom management that’s supported with decision support. The literature shows that effective symptom management decreases utilization like emergency department use and hospitalizations and improves quality of life and even longevity. So, we think we can make a case that there is a favorable return on investment.

The Sapphire Cancer Symptom Management application is 100% funded using Cancer Moonshot funding through a $1.5 million contract from the National Cancer Institute, Contract #75N91020C00019.

The next Clinical Pathways Forum call will be held on May 23, 2023, at 12:00 pm EDT/9:00 am PDT.