Feasibility of Collecting Patient‐Reported Outcomes for Inpatient Rehabilitation Quality Reporting

Abstract

Collection of PROs from rehabilitation patients is feasible. Results inform policy makers regarding feasibility of PRO data in evaluating rehabilitation quality.

Feasibility of PROMIS using computerized adaptive testing during inpatient rehabilitation

“…In one study, only 7% of the eligible patients completed the CAT-administered PRO as they were using the internet or telephone after being discharged from inpatient rehabilitation [ 32 ]. In addition, completion rates for other feasibility studies during inpatient (51%) and outpatient (41%) were almost the same as ours (39% and 67% in the initial and modified phase respectively) [ 21 , 76 ]. Although our rate appears to be on the low end of estimates compared to one study [ 77 ], the response rate reflects our ability to provide sufficient staffing to collect PRO data.…”

“…Several other factors have contributed to the rapid increase in the adoption and implementation of PROs in clinical settings, including the application of item response theory (IRT) measurement to the development of PROs [ 18 – 20 ], and progress within the technological infrastructure, allowing for wider use of information technology for PRO administration, scoring, display of results, and interpretation via tablet computers [ 21 – 23 ], smartphone apps [ 24 – 26 ], and electronic medical records [ 27 , 28 ]. In addition, there is an increased demand by payers, accreditors, professional organizations, and clinicians to measure and improve PRO-measured outcomes at the patient, clinic, and healthcare system levels [ 29 ].…”

“…The assessment of hospitalized patients may pose additional challenges and require additional resources, as these patients are likely to require assistance in completing measures [ 36 ]. Depending on the clinical population being assessed, there may be cognitive, communication, and physical challenges ranging from intravenous lines in arms to functional limitations that limit PRO data collection [ 21 , 37 ]. Patients in acute care settings typically have short stays, requiring consideration of the timing and frequency of PRO administration [ 1 , 38 ].…”

How to implement routine electronic patient‐reported outcome monitoring in oncology rehabilitation

“…Over the whole course of cancer treatment, in which rehabilitation and follow‐up play an increasingly important role, 12 PRO can be used as a standardised screening to promote efficient use of time in medical consultations, improve the communication between patients and their therapists, ensure treatment continuity and facilitate patient‐centred interventions as well as participatory decision‐making 13 . Hence, similar to their benefits for routine active anti‐cancer care, PROs can also be used in rehabilitation to screen patients for special care needs, tailor interventions to screening results, evaluate the treatment progress and collect data for performance measurement, which can as well feed into Big Data analyses 14,15 . It is reported that especially vulnerable cancer survivors (older age, lower education, depressive symptoms) are at a higher risk of dropping out longitudinal follow‐up PRO assessments post‐active treatment 16 .…”

“…13 Hence, similar to their benefits for routine active anti-cancer care, PROs can also be used in rehabilitation to screen patients for special care needs, tailor interventions to screening results, evaluate the treatment progress and collect data for performance measurement, which can as well feed into Big Data analyses. 14,15 It is reported that especially vulnerable cancer survivors (older age, lower education, depressive symptoms) are at a higher risk of dropping out longitudinal follow-up PRO assessments post-active treatment. 16 This might be associated with major problems in at least two respects: First, these patients might have less access to the care they actually need (especially if PRO data are used to support follow-up care) and second, this creates a considerable positive bias in data possibly used for Big Data analyses.…”