Response to: “Enhancing Geriatric Assessment in Cancer Rehabilitation: Suggestions for Future Research”

We thank the authors of the letter (Su and Zeng) [1] for their questions and thoughtful insights. We agree that this study is one of the first to describe referral patterns to rehabilitation in community oncology settings in the United States (U.S.) using the geriatric assessment (GA) [2]. We welcome the opportunity to respond to the suggestions.

The first suggestion was to evaluate individual comorbidities and their relationship to referrals to rehabilitation services. In the original trial, “Communicating about Aging and Cancer Health” (COACH) study (clinicaltrials.gov identifier: NCT02107443), comorbidity as an aging-related domain was impaired for patients that had 3+ comorbidities or one that significantly influenced quality of life [3]. We previously found that impairment in the comorbidity domain was associated with impairments in functional status and physical performance [4]; impairments in the functional and physical performance guided rehabilitation recommendations in the COACH intervention [3]. For this analysis, given the large number of variables that influence functioning and recommendations for rehabilitation services and due to the rather modest sample size, an analytical decision was made to include a composite variable for comorbidities. We agree that future research should strive to understand the individual and cumulative impact of comorbid conditions on rehabilitation referral.

The second recommendation was to classify patients with lymphoma as a single subcategory. Of note, there was no difference by cancer subgroup in outcomes of the original COACH trial [3]. For this secondary analysis, the number of patients with lymphoma was small (n = 20) and thus did not allow for a separate analysis. We are not aware of literature to support that patients with solid versus hematologic malignancies differ in terms of uptake of rehabilitation services in older adults with advanced cancer. This is a future direction to be considered in population-based datasets.

The third suggestion was to involve geriatricians in the care of older adults with advanced cancer, as oncologists may have limited knowledge of GA. The original trial provided GA training to all participating oncologists, and primary results showed improved aging-related communication in the intervention arm, demonstrating that the oncologists effectively utilized the GA results in clinical practice [3]. In the U.S., the availability of geriatricians is quite low; one study reported that in 210 oncology practice groups, geriatricians were available in only 5% of the oncology clinics [5]. Thus, educating oncologists on the use of GA and how to appropriately apply the intervention recommendations will be critical to support the growing pool of older adults who are diagnosed with cancer.

Fourth, we believe that this study is a necessary first step to understanding how GA and related referrals could be supported in community cancer centers, which treat the majority of patients in the U.S. In this analysis, we tested whether differences by practice site influenced results and found that including a random intercept for practice site did not explain a significant portion of the variability in the outcome [2]. This suggests that estimates of associations between patient factors and referral were largely consistent across practice sites included in the COACH study. We agree that additional work should evaluate the influence of GA and guided referrals in health care systems in other national contexts.

Finally, we agree that future work could consider rationales for the uptake of referral. The integration of GA into clinical care would support communication/discussions as well as uncover and improve awareness of disability among patients and clinical teams. In this study, willingness to undergo rehabilitation services should not affect the primary aim (i.e., oncologist-initiated discussions about and/or referral to rehabilitation services). Studies have shown that factors associated with rehabilitation utilization are multi-level and complex. In the U.S., older adults with cancer may be unaware of the potential benefits of rehabilitation and concerned with costs [6]. Further, across national contexts, social determinants of health (e.g., education, access to transportation, and socioeconomic status) can be associated with physical therapy utilization [7].

In sum, we appreciate the opportunity to respond and agree that this study's results inform many important future research directions examining the referral patterns and benefits of rehabilitation services for older patients with cancer.

Concept and design: all authors. Acquisition of subjects and/or data: all authors. Analysis and interpretation: all authors. Manuscript writing and approval: all authors.

The findings and conclusions in this report are those of the authors and do not represent the official position of the National Institutes of Health. The article was prepared as part of some of the authors' (Streck) official duties as an employee of the US Federal Government.

The authors declare no conflicts of interest.