Acceptability and accuracy of patient-reported outcome measures (PROMs) for surveillance of breathlessness in routine lung cancer care: A mixed-method study

Background Breathlessness in lung cancer negatively impacts on quality of life but often goes undetected and undertreated in clinical practice. There is a need for routine surveillance for early identification and proactive management of breathlessness using patient reported outcome measures (PROMs) in clinical care but it is unclear what PROMs should be used or are accurate for use in routine care. Methods We used mixed-methods (quantitative surveys and qualitative interviews) to examine the predictors of breathlessness in 339 lung cancer participants and acceptability of PROMs. In addition to multivariate analysis to examine predictors of dyspnea, participants completed an acceptability survey and themes were derived for the qualitative data (n = 26) to explore patient experience of PROMs. We also tested the accuracy of PROMs using a Receiver Operating Characteristic and Area Under the Curve analysis. Results A total of 339 patients completed the breathlessness PROMs and acceptability survey and 26 patients participated in an in-depth interview to investigate their experiences of breathlessness and its PROMs. Prevalence of breathlessness was 51.9 % (n = 176) and 70.5 % of patients preferred the Medical Research Council (MRC) scale followed by the Breathlessness Intensity (BI) scale (63.7 %) among the five measures for breathlessness- Modified Borg Scale (MBS), Cancer Dyspnea Scale (CDS), MRC, BI, and Breathlessness Distress (BD). The finding showed wide variation in the MRC grades across the BI rating even among patients with the same BI score. AUC scores for the Borg scale was 0.71 (using MRC cut-off score of < 2), for CDS, 0.72, for BD, 0.70, and for BI 0.79. For an MRC score of 2, the Borg score cut-off was 0.8 (optimal sensitivity, 50 %; specificity, 93.3 %); the cut-off score of CDS, BD, BI score was 1.4 (optimal sensitivity, 67.1 %; specificity, 70 %), 1.5 (optimal sensitivity, 57.5 %; specificity, 73.3 %), and 1.5 (optimal sensitivity, 72.6 %; specificity, 83.3 %) respectively. AUC by ROC analysis for breathlessness and modest concordance among five PROMs showed important gaps between the individuals’ experience and PROMs data. Three main themes from qualitative data included 1) Making sense of symptom reporting, 2) Valuing the reported data, 3) Managing the symptom of breathlessness. Conclusion This study examined measurement of breathlessness using PROMs for routine clinical care and showed that severity measures alone do not accurately detect this symptomnor the experiential dimensions of breathlessness that are critical to guide appropriate intervention.