LAUSR

We thank Kobayashi et al for their informative comments. Our published article reports on a deliberately crafted study designed to assess changes in work outcomes (employment status, hours worked, work ability, and work limitations) as patients received curativeintent chemotherapy for diverse cancers. This study is unique in following these outcomes in a US population in a prospective, longitudinal manner both during the course of curative treatment and afterward. Another unique aspect of this study is that we started with patients who reported working at diagnosis and who expressed an intent to work during treatment or return to work after treatment. The published article represents our initial goal to describe the work outcomes observed, with ongoing analyses underway to explain the patterns observed. Kobayashi et al stated, “The authors collected information before the initiation of cancer treatment (baseline information).” We would like to clarify that our baseline survey was conducted before cycle 2 of chemotherapy (not necessarily before surgery or cycle 1). As discussed in our limitations, this means that nearly onethird of the patients were not working at the time of the baseline survey because they had already undergone surgery or received cycle 1 of chemotherapy. However, they all reported working before their cancer diagnosis. Kobayashi et al correctly observed that the study population was heterogenous from the standpoint of education, job type, and income. It was homogeneous with respect to the commonality of receiving curative chemotherapy and a stated intent to return to work. They also correctly noted that the sample size was smaller. Thus, we elected to start by describing the outcomes observed and will explore the impact of additional variables (eg, the duration of therapy and the cancer type) in future analyses. Kobayashi et al noted that stage may be an important variable, especially when one is considering a stage IV or metastatic population that is being treated palliatively. In curatively treated patients rendered free of evidence of the disease, it is the acute and persistent consequences of therapy that likely drive outcomes. Therefore, we set up our study to emphasize the commonality of receiving chemotherapy and intent of treatment rather than cancer stage. We also note that stage should be interpreted with caution: for instance, the American Joint Committee on Cancer staging system changes from version 7 to 8 had a profound impact on breast cancer. Additionally, stage may not reflect what treatment will be rendered, especially in a cancer population in which treatment is driven by receptor status (eg, breast). Kobayashi et al commented that the length of the time from the baseline to the end of treatment varied, and we will account for this and other variables (eg, the cancer type and the symptom burden) in future analyses. Finally, Kobayashi et al noted that the reasons for dropout from the baseline to the end of treatment might affect the outcomes observed. We noted in the Results section that of the 111 participants with a baseline survey, 4 developed cancer progression or died. Unfortunately, we do not have data about the remaining respondents who were lost to followup or their reasons for no longer participating. In summary, we are delighted to receive these thoughtful comments and hope that our explanations address the concerns noted by Kobayashi et al. We believe that our article and the study that it outlines are a positive step in better understanding drivers of survivor work outcomes, as called for by the National Cancer Institute.2 This study was previously presented at the Cancer Survivorship Symposium in 2017.