Computers that Show Recognition of Patients' Symptoms

The goal of automated symptom management has often been narrowly defined as collection and reporting of data. Yet, an automated system can do more. These s ystems can mimic h uman reassurance f or patients th at their symptoms have been recognized. Patient engagement has long been recognized as central in symptom management. An electronic system was b uilt to co llect the Treatment Related Symptom Checklist (TRSC) and to reassure patients. In this paper we di scuss the design of the s ystem and procedures t aken to show recognition of pa tients’ s ymptoms. The lit erature o n cl inician-patient in teractions was reviewed selectively. Methods used by clinicians to promote patient engagement during medical hi story were identified. Similar methods w ere i ncorporated into the symptom management s ystem in order to facilitate understanding by t he pa tient. These i ncluded ( a) co nversational data co llection as opposed t o s urvey s tyle or s tandardized questionnaires, ( b) s hort response phrases indicating understanding of the reported symptom, (c) use of open ended questions to reduce asking long lists of symptoms, (d) leading questions that ask fo r confirmation of expected symptoms, (e) review of symptoms at designated stages, and (d) alerting patients when computer has informed the clinician about patient-reported symptoms. We are in the process of pilot testing the system among oncology pa tients in one hospital. Examination of existing informatics knowledge e ases i ntegration o f well es tablished p aper-based tools into automated systems.