The application of patient reported outcomes (PROs) to build personalized patient education aids

Abstract

Background After diagnosis, cancer patients receive a large volume of information in a short period of time when they are usually feeling anxious and stressed. In particular, prostate cancer patients can be overwhelmed by multiple treatment options and the expectation of making a life-altering personal choice during shared decision-making. Consent forms attempt to summarize potential side effects but are rarely personalized to each patient. Clinicians can underestimate symptoms/side effects experienced by patients, compared to patient-reported outcomes (PROs). PROs are the best estimate of how our patients will feel during/after treatments. Methods We used real world data including PROs and clinical/radiation data from the Enterprise Data Warehouse and our departmental radiation database to explore patient characteristics associated with PRO changes over time. We created several different visualizations showing how PROs changed over time, and tried to pick visualizations that were most accessible to patients with different backgrounds, languages, data/health literacy. In the future, we hope to conduct patient interviews to improve visualizations, and to automate real-time generation of personalized patient education aids for use at consultations. Results Our cohort included ∼3,900 patients who completed >35,000 PROMIS-10, >5,000 EPIC-26, and >11,000 weekly radiation treatment visit questionnaires about urinary, bowel, erectile, and overall quality of life. Among significant predictors of toxicity were initial symptom severity. Discussion Patient reported outcomes (PROs) improve outcomes and have become standard in clinical trials and are gradually being incorporated into more cancer clinics. The next phase of PRO research is continuing to expand applications into general clinical practices and routine daily workflows using real world data (collected from each clinic because clinical practices/outcomes can vary). It is time to use our PROs work back to directly benefit patients. We are working on real-time generation of personalized patient education resources in the routine clinical workflows. Resources include data visualizations that aim to be accessible to patients with diverse backgrounds, languages, data/health literacy. Prostate cancer patients could especially benefit from personalized resources for standard-of-care shared decision-making after initial diagnosis. This work in progress describes how to combine PROs and clinical data to create personalized PRO visualizations for patient education, and potential mechanisms for real-time generation of these resources. After diagnosis, cancer patients receive a large volume of information in a short period of time when they are usually feeling anxious and stressed. In particular, prostate cancer patients can be overwhelmed by multiple treatment options and the expectation of making a life-altering personal choice during shared decision-making. Consent forms attempt to summarize potential side effects but are rarely personalized to each patient. Clinicians can underestimate symptoms/side effects experienced by patients, compared to patient-reported outcomes (PROs). PROs are the best estimate of how our patients will feel during/after treatments. We used real world data including PROs and clinical/radiation data from the Enterprise Data Warehouse and our departmental radiation database to explore patient characteristics associated with PRO changes over time. We created several different visualizations showing how PROs changed over time, and tried to pick visualizations that were most accessible to patients with different backgrounds, languages, data/health literacy. In the future, we hope to conduct patient interviews to improve visualizations, and to automate real-time generation of personalized patient education aids for use at consultations. Our cohort included ∼3,900 patients who completed >35,000 PROMIS-10, >5,000 EPIC-26, and >11,000 weekly radiation treatment visit questionnaires about urinary, bowel, erectile, and overall quality of life. Among significant predictors of toxicity were initial symptom severity. Patient reported outcomes (PROs) improve outcomes and have become standard in clinical trials and are gradually being incorporated into more cancer clinics. The next phase of PRO research is continuing to expand applications into general clinical practices and routine daily workflows using real world data (collected from each clinic because clinical practices/outcomes can vary). It is time to use our PROs work back to directly benefit patients. We are working on real-time generation of personalized patient education resources in the routine clinical workflows. Resources include data visualizations that aim to be accessible to patients with diverse backgrounds, languages, data/health literacy. Prostate cancer patients could especially benefit from personalized resources for standard-of-care shared decision-making after initial diagnosis. This work in progress describes how to combine PROs and clinical data to create personalized PRO visualizations for patient education, and potential mechanisms for real-time generation of these resources.