Patient‐reported outcome measures in multiple myeloma: Real‐time reporting to improve care (<scp>My‐PROMPT</scp>) ‐ a pilot randomized controlled trial

Abstract

Multiple myeloma (MM) is associated with a high burden of disease, compromising patients' health-related quality of life (HRQOL). With improvements in overall survival, HRQOL has become an increasingly important outcome. Patient-reported outcome measures (PROMs) are tools that collect patient-reported outcomes (PROs) - an individual's evaluation of their own health and wellbeing - and incorporate HRQOL. Routine monitoring of symptoms and other PROs to inform cancer care has been shown to improve patient outcomes including HRQOL1, 2 in solid tumors. However, few data exist for hematological malignancies,3, 4 and even fewer for MM.4 To design a randomized controlled trial (RCT) to assess the impact of real-time feedback of PROs, feasibility and acceptability to patients and clinicians of such an intervention needed evaluation. To achieve this, we designed the My-PROMPT pilot RCT to determine feasibility and appropriateness of study design, including acceptability to patients and clinicians; and to understand any change in HRQOL over time, in order to inform sample size calculation for a future trial. This parallel, non-blinded RCT (ACTRN12618001878268, Figure S1) was conducted in hematology clinics in four Australian hospitals. Table S1 provides detailed study methods. Patients were ≥ 18 years with newly diagnosed MM (NDMM) and participants in the Australian and New Zealand Myeloma and Related Diseases Registry (MRDR5), which provided patient data and infrastructure for recruitment. Intervention patients completed the Myeloma Patient Outcome Scale (MyPOS6) before 4 clinic visits: baseline, 1, 6, and 10 months (Table S2). Clinicians received summaries of MyPOS results of concern before visits. Controls completed the MyPOS at baseline and 10 months with no feedback to clinicians. Primary outcome was feasibility and acceptability of the intervention for patients and clinicians (Table 1a). Further outcomes are in Table 1b and c, and change in patients' MyPOS HRQOL score from T1 to 4 compared between groups in Figure S2. Interviews were conducted with six clinicians and seven patients. The Alfred Hospital, Melbourne, Australia provided ethics approval as the lead site. 11. What impact did it have on your workload? N = 35 My-PROMPT recruited 32 patients from May 8, 2017 to October 31, 2018, 16 to each arm (Figure S1). Median age in the intervention vs control group was 66 years (59-76) vs 69 years (62-71); baseline characteristics were well matched other than a greater proportion of males in the control arm (81% vs 25%, P = .001, Table S3). Proportion of completed MyPOS per time point was >90% (T1: 100% [32/32], T2: 94% [15/16], T3: 94% [15/16], T4: 91% [29/32]), most in hardcopy (85%, 11/13), and without help (92%, 12/13). Most patients agreed MyPOS instructions were easy to follow and understand (92%, 11/12). Median satisfaction score for patients was 5 (scale 1-5, 5 = Very satisfied, n = 13), and for 15 clinicians over all T1 to T3 visits (n = 39) was 85 (scale 1-100, 100 = Very satisfied, Table 1a). In patients' evaluations (completed in intervention arm only), patients agreed that completing the MyPOS helped convey their concerns to doctors (85% [11/13]), that their doctor had discussed MyPOS results (62% [8/13]), and that their doctor referred to the results in the consultation (46% [6/13]). In clinicians' evaluations, 80% (31/39) reported using results to discuss patient concerns, and > 40% (17/39) took action after seeing results (Table 1c). Those who did not take action and added free-text comments (n = 2), explained that no action was required or the issue was not medically related. Over 50% (21/39) of clinicians thought the intervention was ready for routine use and most (83%) thought it had reduced, or had only minimal impact on visit duration. Most clinicians thought no training (68%), or only a brief familiarization (32%) was required. With MyPOS scores the median total MyPOS score at T1 was 18 (IQR 14, 39; range 6-51) in the intervention arm and 25 (IQR 16, 29.5, range 6-42) for controls (higher score = worse HRQOL), with no significant difference in change in total MyPOS score from T1 to 4 between groups (P = .20, Figure S2A). The same held for change in MyPOS Symptoms and function (P = .66), and Emotional response subscale scores (P = .46, Figures S2B and C). Males had greater median reduction in MyPOS score from T1 to 4 than females in the same cohort (P = .03, Figure S2D). Tablets for MyPOS completion were provided to all sites, but were only used at one. Reasons reported for not using tablets were: patient preference; internet access difficulties; leaving paper copies was easier; vigilance required with a valuable device; and technical issues launching the MyPOS. Two patients who used the tablets reported they were relatively easy to use. Sites reported the following logistical challenges in delivering the intervention: ensuring MyPOS completion before visits, entering hardcopy results manually, providing summaries to clinicians before visits, and the short timeframe available to complete all these tasks. Themes and excerpts from patient interviews are summarized in Table S4. Interviewed patients had a median age of 61 years, and 57% were female. Patients reported that use of the MyPOS in the intervention helped them to prepare for their clinic visit, to raise issues, and to focus on their main issues at the visit. Several patients commented that through the intervention they now realized the range of topics they could raise with clinicians, apart from treatment. Patients indicated that those with more severe disease or those less forthcoming could potentially benefit more from the intervention; and some expressed a desire or expectation that MyPOS data would be used to track progress across consultations. Themes and excerpts from clinician interviews are summarized in Table S5. One or two clinicians per site were interviewed; median age: 41 years, median years of clinical experience: 19, and 50% were female. According to one clinician, their focus is very treatment-centered (C5, Table S5), another found that modifying their behaviour to patients' needs before visits helped optimize communication (C1). Clinicians indicated the intervention enhanced and streamlined patient/clinician communication by summarizing patients' self-reported status, which helped to focus on priorities (C1-3). While the intervention resulted in some action being taken (C6), not all concerns raised required additional resource use (C2). Clinicians reported the intervention had minimal impact on, or reduced workload by identifying patients' main issues to focus on upfront, however it had increased workload for study staff (C1). Clinicians coincided with patients, that less forthcoming patients could potentially benefit more from the intervention (C4). Our findings from patient and clinician My-PROMPT evaluations and interviews suggested the intervention enhanced patient/clinician communication, in line with comparable interventions in solid tumors.2 A RCT providing summaries from an interactive ePROM system to clinicians treating patients with blood cancers, showed symptom outcomes were improved in intervention patients.3 Our pilot results are not powered to detect such a difference; however, patient and clinician evaluations and interviews revealed the potential for the intervention to improve patients' symptoms by prompting discussion, and by focusing visits on patients' priorities. Furthermore, by making clinicians more aware of the extent of patients' problems through the intervention (Table S4 and S5), the possibility for symptom improvement is enhanced. Some ePROM systems allow remote online completion of PROMs on demand,1, 4 and generate reports for patients and clinicians to view and monitor progress. Some such systems have shown promise3 and conferred clinical benefit1 for patients. Incorporation of such a system with funding in a future definitive trial would increase patients' involvement in their own care, could improve outcomes, and warrants consideration. Our study sample was small but adequate for its purpose - to assess feasibility of this intervention for use in a larger RCT. The small sample size contributed to imbalance in baseline characteristics, mainly sex (Intervention: 25% males v Controls: 81%, P = .001, Table S3). Males had greater median reduction in MyPOS score (lower score = better QOL) from T1 to 4 than females: −10 (−14, −2.5) v −2 (−6, +1), P = .03, suggesting that over-representation of males in the control arm may have biased results for change in overall MyPOS score between groups (Figure S2A and D). Furthermore, the small sample size meant the comparison of change in MyPOS HRQOL score between groups was underpowered to detect a difference. This is the first study looking at feedback of PROs to clinicians in a MM-only sample.3, 4 Focusing on a homogenous group may help demonstrate the impact of the intervention and its relevance in MM. In addition, with median age representative of the disease (Table S3), and patients sourced from hospitals similar to those at which many patients in a larger trial would be recruited, the generalizability of this feasibility trial to a future definitive trial in MM is enhanced. My-PROMPT sites found some logistical aspects challenging, particularly ensuring patients completed the MyPOS, and providing a summary to clinicians before patient visits in an often-limited timeframe. To reduce workload for site staff in a future trial, embedding a more sophisticated ePROM system in our registry would help ensure feasibility and sustainability of the intervention by streamlining remote online MyPOS completion and generating easily-accessed summaries for clinicians. The intervention could be scalable to current and future MRDR sites, unlike hospital-based ePRO systems that are built into existing hospital electronic medical record systems site by site, limiting scalability and increasing cost. Development and uptake of systems to feed back ePROM results to clinicians treating cancer patients has been slow in Australia despite promising supportive evidence of benefits.1, 2 In addition, blood cancers have not been included in an important local trial in this area,7 so the data presented here shows the importance of My-PROMPT and a future definitive trial in myeloma. This pilot RCT confirms the feasibility and acceptability of real-time reporting of PROMs for patients with MM to their treating clinicians, and supports the rationale for proceeding, with some modifications, to a larger, definitive trial in this area. We thank participating patients, clinicians and sites for supporting this study. The Australian and New Zealand Myeloma and Related Diseases Registry (MRDR), which provided data and facilitated recruitment to this trial, has received funding from: Amgen, Celgene, Gilead, GSK, Janssen, Novartis, Sanofi, and Takeda. This work was supported by a competitive Gilead Australia Fellowship Research Grant and a grant from Takeda Pharmaceuticals Australia Pvt Ltd. Appendix S1 Supplementary Information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts