Role of Patient-Reported Outcome Measures in the Inpatient Setting

Abstract

To the Editor With great interest, I read the randomized clinical trial by Nipp et al. 1 This trial represents an important step in examining the role of patient-reported outcome measures (PROMs) in the inpatient setting. However, I believe that the results need to be taken within the larger context of PROM use in oncology for the following reasons. First, it is important to note that patients in the usual care arm completed the PROMs in contrast to the trial by Basch et al, 2 where patients in the usual care arm had a standard assessment. Evidence supporting PROMs suggests that better patient-clinician communication leads to an enhanced symptom identification compared with the standard clinical assessment, which might change the management leading to improved outcomes. 3 Administering the PROM in the control group, despite not being used by the clinicians for assessment, might have primed patients to better express their symptoms, including severity and patterns, 3 thus inadvertently narrowing the potential gap of symptom identification between the intervention and control.Second, enhanced communication in the context of moderate to severe symptoms might not result in a change of outcome unless symptoms were addressed adequately by clinicians with the potential involvement of palliative care. Similar to this study, Newcomb et al 4 reported a high moderate to severe symptom burden in inpatients with unplanned admissions. However, a variation in palliative care consultations was reported, which suggests that symptom severity alone does not drive management, and clinicians consider other factors. Accordingly, it is plausible that the lack of a known approach in this study 1 for addressing symptoms between the intervention and usual care arms might have resulted in the lack of difference in outcomes.Third, not all symptoms during unplanned admission (eg, fever related to infection) 5 can be captured through PROMs, and such symptoms can influence the duration of admission, thus potentially affecting the denominator and the primary outcome of this trial. 1 For example, the primary outcome for a patient would be 40% if symptoms improved for the first 2 days of enrollment, then were unchanged for 3 more days before discharge if the reason for hospital stay was completing workup, as opposed to 67% for the same patient if they were discharged on the third day owing to earlier completion of workup.