Fast-Track proposals will be accepted. Number of anticipated awards: 1-2 Budget (total costs, per award):
Phase I: up to $225,000 for up to 9 months
Phase II: up to $1,500,000 for up to 2 years
PROPOSALS THAT EXCEED THE BUDGET OR PROJECT DURATION LISTED ABOVE MAY NOT BE FUNDED.
Uncontrolled symptoms during and following cancer treatment have been associated with emotional distress; diminished functional status and health-related quality of life; treatment delays, discontinuation, and non-adherence; and unplanned hospitalizations and emergency room visits. The evaluation and management of symptoms in cancer care, including multiple co-occurring symptoms (e.g., pain, depression, and insomnia), is complex.
A plethora of evidence-based clinical practice guidelines (CPG) for managing cancer-related symptoms have been developed by national organizations that include the Oncology Nursing Society (ONS), American Society of Clinical Oncology (ASCO), Multinational Association for Supportive Care in Cancer (MASCC), National Comprehensive Cancer Network (NCCN), European Society for Medical Oncology (ESMO), American Cancer Society (ACS), and the National Cancer Institute (NCI). However, implementation of these guidelines in practice has to-date been limited and haphazard, and the available guidelines are not offered to clinicians in a readily actionable format. Sifting through the options contained in multiple guidelines and determining the best approach for a specific patient takes more time than clinicians typically have available. Electronic decision-support would help to bridge this guideline implementation gap, and would allow for rapid dissemination into practice of both new guidelines and guideline updates.
Clinical Decision Support (CDS) is a health information technology designed to directly aid in clinical decision-making. CDS matches the characteristics of individual patients to a computerized knowledge base, and generates patient-specific assessments and recommendations. CDS provides clinicians and other stakeholders with pertinent knowledge and person-specific information, intelligently filtered, and delivered at appropriate times in clinical workflow to enhance health and healthcare delivery (Osheroff et al. JAMIA 2007; 14 (2), 141-145). The overall goal of CDS-Sx is to support health professionals in delivering personalized, evidence-informed, guideline-based clinical decision-making to improve the evaluation and management of cancer-related symptoms.
NCI Blue Ribbon Panel (BRP) Implementation Science Working Group Report urged an immediate strategic investment to provide actionable decision support that accelerates the implementation of evidence-based cancer symptom management guidelines. CDS for symptom management addresses recommendations made by the National Academy of Medicine (NAM), National Quality Forum, and the Coalition to Transform Advanced Care for improvements in symptom management and palliative care across the cancer continuum.
This topic requests proposals to create a system of computable algorithms to improve oncology clinicians’ evaluation and management of common cancer-related symptoms, leveraging nationally endorsed, evidence-based CPGs. The algorithms would be delivered within a CDS that also includes a small set of well-curated resources for patient self-management support, ICD-10 coding, and other features to streamline clinician workflow and support coordinated interdisciplinary symptom management such as templated progress notes and referral pathways. The algorithms will be constructed such that the sequence of evaluation and management activities is triggered by patient-reported outcomes (PRO) data derived from a variety of contemporary PRO measurement systems (e.g., PRO-CTCAE, PROMIS, and NCCN symptom indices) and would offer decision support relevant across the cancer continuum from treatment to survivorship and end-of-life. At commercial scale, the CDS would allow for information to be exported into an electronic health record. This topic is in line with the Cancer Moonshot Blue Ribbon Panel’s Recommendation to support Symptom Management Research.
Despite the plethora of evidence-based cancer CPGs for cancer symptom management, implementation is inconsistent in practice. There are few commercially available decision-support systems that provide guideline-based recommendations in interpretable and actionable ways to healthcare providers at the point-of-care. NCI investment in the development of CDS for symptom management has been sparse.
The overall objective is to develop an electronic, rule-based, clinical decision-support system for symptoms (CDS-Sx) that leverages national CPGs to improve the evaluation and management of symptoms during and following cancer treatment.
This objective will be accomplished by:
Creating and validating software-ready computable algorithms for evaluation and management of eight common
cancer-related symptoms that have associated national CPGs for cancer symptom management.
Computable algorithms will be iteratively developed by panels comprising clinical experts and experts in rule-based CDS system design, leveraging nationally endorsed symptom management guidelines, and aligned to an existing data standard to ensure interoperability with downstream systems.
Algorithms will be tailored to different levels of symptom severity and interference, as well as to other clinical and demographic factors such as concurrent symptoms (e.g., depression in the setting of pain and fatigue), disease site, treatment type, age, concurrent medications, comorbid conditions, and allow tailoring to patient goals and preferences.
Options for site-specific customization of the algorithms, particularly with respect to resources available at the practice site (e.g., referrals to specialized consult teams) will be included.
Included with the algorithm will be resources for patient self-management support; ICD-10 codes to facilitate billing for symptom management services; and features to streamline clinician workflow and support interdisciplinary symptom management, such as templated progress notes and referral pathways.
Branching logic within the algorithms should allow for tailoring to different places on the cancer continuum from treatment to survivorship and end-of-life.
Designing a clinical-decision support software system to deliver the algorithms to clinicians at point-of-care
The system will allow for the entry and encoding of the CPGs in a user-friendly manner.
The system will present the clinical decision workflow to the clinician using straightforward medical language in an intuitive web-based graphical user interface (GUI) on a tablet, desktop, or laptop computer.
The system may be developed using existing software applications that are configured and/or integrated to achieve the desired functionality, or it may be developed as a custom application. However, if the offeror proposes development of a custom application, offeror must provide compelling justification for doing so versus configuration of off-the-shelf solution(s).
At commercial scale, the CDS-Sx should have the capacity to interoperate with EHR systems commonly used in oncology settings (e.g. data extractable in HL7CDA or similar format) and be compliant with applicable FDA regulatory guidance for CDS software.
Conducting iterative cycles of usability testing, CDS-Sx refinement, and user acceptance testing, with a multidisciplinary panel of clinicians, cancer patients and survivors. Clinicians should reflect the breadth of settings where cancer care is delivered, including specialty care, community-based, home-based, and primary care settings.
Activities not supported by this topic:
Applications that do not leverage national practice guidelines, do not incorporate iterative development of the computable algorithms by expert panels, and approaches that do not address the complexities of cancer symptom management (e.g. co-occurring symptoms) will be not be considered for funding.
Activities and Deliverables
Expected CDS-Sx system functionalities include:
Presentation of the CPG content in an intuitive user interface that fits into the clinician workflow.
Graphical user interface with branching logic that allows for the clinician to quickly select responses and arrive at
clear and specific clinical guidance for patient evaluation and management.
Ability to make both minor and major revisions to the CDS-Sx system, such as adding new symptom management guidelines, or updating content when guidelines are revised, through the user interface and without changes to software code.
Ability for clinician to print patient self-management materials or send them via email or text message.
Phase I Activities and Deliverables:
Establish a project team with expertise in the areas of clinical decision-support, cancer symptom management, cancer care delivery, knowledge translation and implementation science, human factors engineering, and software design.
Develop a replicable consensus-based methodology to synthesize and transform evidence-based guideline recommendations from their narrative prose formulation into algorithms for symptom assessment and treatment, converting the content into computable language, decision-points, and logic flows. Algorithms should reflect health IT standards, including Health Level 7 (HL7) and the Clinical Quality Framework (CQF) Initiative (http://cqframework.info).
Develop algorithms for evidence-based evaluation and management of two symptoms (specifically, constipation and fatigue) that reflect the anticipated spectrum of algorithm complexity with respect to the number of decision nodes and separate pathways, based on prior research Algorithms for evaluation and management should be evidence-based, and should leverage symptom management guidelines and patient self-management support materials that are offered and updated regularly by organizations such as ONS, ASCO, NCCN, AHRQ, ACS, ESMO, and NCI-PDQ. Algorithms should reflect recommendations for specific pharmacological and behavioral interventions, including recommendations to initiate medications or explicit adjustments for medication doses, laboratory tests, supportive care referrals, and behavioral self-care suggestions. The curation of guideline materials into the algorithms should include an approach to annotate the material to convey the source(s) of the evidence, and to address any intellectual property issues. Algorithms must clearly address the complexities of cancer symptom management (e.g., algorithms must integrate with one another to address multiple co-occurring symptoms, and must consider the multiple factors that may aggravate and/or alleviate symptoms).
Identify the approach and specific standards (e.g., CDISC, SDTM, and ICD) for standardization and encoding of data points in the algorithms to provide for computability and interoperability with downstream systems. At commercial scale, the CDS-Sx would allow for information to be exported into the electronic health record in a standard format to strengthen documentation, support metrics of care quality and value, and document re-evaluation of symptoms and intensification of management as warranted.
Formalize CDS-Sx design considerations, including proposing novel features for CDS that enhance usability in busy practice settings.
Curate patient self-management support material for inclusion.
Demonstrate the algorithm development and curation methodology using two common symptoms (constipation and fatigue).
Pilot test the CDS-Sx algorithms with a multidisciplinary panel of clinicians (including physicians, nurse practitioners, physician assistants, registered nurses, social workers, psychologists, and physical therapists). While the pilot test group need not be large, it should reflect the breadth of settings where cancer care is delivered including specialty care settings, comprehensive cancer centers, community-based cancer settings, and primary care. Refine the prototype algorithms in an iterative manner based on results of pilot testing.
For the CDS-SX:
Specify the user requirements for the system.
Provide an analysis of available open source, off-the-shelf (OTS) software systems, including options developed in the academic setting such as SEBASTIAN (Lobach et al 2016; JMIR Med Inform; 4 (4), e36) in terms of their suitability for use, and customization and integration requirements.
Recommend a design approach (i.e., custom development, integration of OTS software, or a combination).
Phase II Activities and Deliverables:
Using the methodology developed in Phase I, create the assessment and management algorithms for six (6) additional symptoms (specifically pain, insomnia, nausea/vomiting, diarrhea, dermatologic toxicities, and psychological distress [anxiety and depression]).
Specify detailed functional and non-functional requirements for the system.
Specify technical requirements (add more detail to specification created in Phase I).
Maintain a traceability matrix that details the relationship between user, functional, and technical requirements.
Using the wireframe developed in Phase I build a production system CDS-Sx using an agile methodology.
Develop test scripts for functional and user testing.
Prior to usability testing, validate the accuracy of the CDS-Sx recommendations against the algorithms; identify and correct any logical inconsistencies or non-agreement in the symptom evaluation and management recommendations generated using the CDS-Sx employing test cases that include simulated patient data targeting boundary conditions for
each decision node and multiple branches in the algorithm’s decision logic. Documentation of successful tests must be
Conduct preliminary usability testing with a small, diverse group of early adopter clinicians to test CDS-Sx in their clinic settings and to provide feedback to be used to iteratively improve the design, UI, and workflow.
Conduct usability testing with a multidisciplinary panel of clinicians (i.e., including physicians, nurse practitioners, physician assistants, registered nurses, social workers, psychologists, and physical therapists), cancer patients, and survivors. The clinician testing group should reflect the breadth of settings where cancer care is delivered including specialty care settings, comprehensive cancer centers, community-based cancer settings, and primary care settings.
Conduct final user acceptance testing with a diverse group of end-user clinicians. The sample for end-user testing should also be diverse with respect to practice site and patient population served (e.g., diverse tumor sites, radiation, medical and surgical oncology, and active treatment and survivorship), and should include clinicians in comprehensive cancer centers, community-based settings, home-based care, and primary care settings.
Create standardized data extract (e.g., HL7 C-CDA or CDA/Progress Note) that can be imported/integrated into
existing EHR solutions (e.g., Epic).
Present Phase II findings and demonstrate the technology to an NCI evaluation panel via webinar.
In the first year of the Phase II contract, provide the program and contract officers with letters of commercial interest.
In the second year of the Phase II contract, provide the program and contract officials with a letter(s) of commercial commitment.
Create a dissemination/publication plan that outlines potential presentations at national meetings and publications
resulting from this scientific development work.