Home Icon
Search Icon

Development and Evaluation of Patient-Reported Outcome Score Visualization to Improve Their Utilization (PROVIZ)

Project Final Report (PDF, 1.06 MB) Disclaimer

Optimizing PRO data visualization with clinicians’ and patients’ input will improve clinicians’ ability to effectively synthesize and communicate complex data to provide patient-centered clinical management.

Project Details -
Completed

Assessing the patient’s experience of illness and wellness is key to informing their goals of care and clinical management. Patient-reported outcome (PRO) measures may lead to improvements in clinical management, health outcomes, and patient engagement. However, even when PROs are integrated into electronic health records (EHRs), clinicians do not always know how to interpret and use the data to inform patients and subsequent clinical management. To address this challenge, this research iteratively designed, tested, and evaluated multiple prototypes of PRO visualizations for the National Institute of Health’s PRO measurement information system (NIH PROMIS) scores.

The specific aims of the research were the following:

  • Evaluate patient preferences, understanding, usability, and acceptability of modes of PRO score presentation in the patient portal.
  • Evaluate clinician preferences, understanding, usability, and acceptability of modes of PRO score presentation in the EHR.

The team used a user-centered, mixed methods approach, conducted in three phases: 1) qualitative interviews with clinicians and patients, 2) development of the prototypes to meet participants’ preferences, and 3) formal evaluation and subsequent iteration of the visualizations with stakeholders. Participants included patients with hip or knee pain, or hip or knee osteoarthritis, and who were using a patient portal. Clinician participants were practicing orthopedic surgeons, physical and occupational therapists, or rheumatologists.

A variety of visualizations were shown to participants. “Smiley-timelines” visualizations have five circles ranging from a frown to a full smile placed along a timeline, along with stoplight coding going from red to green. These PRO score visualizations resulted in significantly quicker response times than all other line charts for finding a negative extreme and explaining how to use the graph. Slower response times occurred when participants were asked to identify the steepest rate of change, to estimate numerical score values, and to identify data points beyond a given threshold. For traditional line charts, quicker response times and lower error rates occurred when the y-axis used words versus just position to show beneficial change. It was found that directional and semantic mismatches confuse patients who read PROMIS scores that traditionally are shown on line charts. Encoding PRO scores via iconography instead of direction is a viable method of addressing this problem but does not outperform traditional line charts that doubly encode scores with color. The research showed that these novel visualizations for NIH PROMIS scores overcome some issues with traditional line charts, but not all. The team plans further research in these areas.

Optimization of Patient-Reported Outcome Data Visualization to Improve Shared Decision Making

Optimization of Patient-Reported Outcome Data Visualization to Improve Shared Decision Making

Optimizing PRO data visualization with clinicians’ and patients’ input will improve clinicians’ ability to effectively synthesize and communicate complex data to provide patient-centered clinical management.

Principal Investigator: Gold, Heather, Bertini, Enrico
Organization: New York University School of Medicine
Research Profile: Development and Evaluation of Patient-Reported Outcome Score Visualization to Improve Their Utilization (PROVIZ)
Funded Amount: $293,981

Patient-reported outcome measures are not always used to inform discussion with patients

The use of PRO measures to assess patients’ experiences of illness and wellness has been increasing, and may lead to improvements in clinical management, health outcomes, and patient engagement. However, even with PROs integrated into EHRs, clinicians may not always know how to interpret and use the data to inform discussions with patients for shared decision making and clinical management. Equally as important is the patient’s ability to understand the data and engage in decisions about their care.

How PRO data are interpreted is impacted by how they are visually presented

Visually presenting PRO data via EHRs can enhance clinicians’ ability to understand and utilize the data. There are several factors that can impact visual interpretation of data. One factor is that some data metrics represent improvements in patient symptoms when their value increases (e.g., physical function), whereas other metrics indicate improvements by decreasing values (e.g., pain). Another factor is that PRO visualizations often do not provide useful contextual information to help the viewer interpret the scores (e.g., information about changes over time or comparison to other patients).

Co-principal investigators Drs. Heather Gold and Enrico Bertini, with a team of investigators at New York University School of Medicine, are developing and testing PRO visual presentations in the EHR to improve PRO interpretability, usability, and data completion with the ultimate goal of improving clinical care and shared decision making.

The purpose of this work is to figure out how we make these scores interpretable and usable so that clinicians are actually applying them as part of their clinical care. And in turn, these data visualizations would help clinicians have conversations with patients about the scores and improve shared decision making about their care.”
– Dr. Gold

Optimizing data visualizations based on clinician and patient input

The team is conducting research based on engineering and human-computer interaction principles to inform optimal data visualization and presentation of orthopedic PROs, including interference of pain in daily life, pain intensity, and limitations of physical function. Focus groups and interviews will be conducted with clinicians and their patients with hip and knee pain or osteoarthritis to understand current perceptions of PROs and preferences for data presentation. This input will inform the development of several prototypes of data visualizations that will then be evaluated by their usefulness, acceptability, and understanding by clinicians and patients.

Ultimately, the investigators hope that this study will improve patient outcomes by accelerating the ability of clinicians to effectively synthesize and communicate complex data to inform patient-centered clinical management for patients, families, and healthcare professionals. While the investigators chose PROs in musculoskeletal condition patients, the study can inform the optimal design of data visualizations for any type of health data.