“Wearable physical activity monitors can inform clinical practice and provide a measure of patient functional status that is free of patient and provider bias.”— M. Shaalan Beg, MD, MS
Tweet this quote
COMMERCIALLY AVAILABLE wearable physical activity monitors have been making their way into clinical research in recent years; however, most studies on these devices have been related to non-cancer conditions including obesity, depression, and physical activity. Their application in the field of oncology has not been established, but it should be for a number of reasons.
Wearable physical activity monitors can inform clinical practice and provide a measure of patient functional status that is free of patient and provider bias. Wearable devices potentially offer investigators involved in interventional cancer trials access to a degree of detail about the lives of trial participants not possible with current quality-of-life tools. As a result, this would allow an improved ability to assess real-world tolerability of investigational agents. The future application of “quantified self-data” in the cancer field relies on the feasibility of these devices in the cancer patient population and the clinical relevance of collected data.
Research presented at the 2017 ASCO Annual Meeting has begun to address these fundamental questions, so wearable technology can be incorporated into routine oncology clinical practice or research. Advanced age, frailty, lack of technologic know-how, along with treatment and cancer symptoms, can affect the ability of the average cancer patient to keep up with two crucial steps upon which wearable devices rely: syncing and charging.
Feasibility of Wearable Devices
STUDIES PRESENTED at the 2017 ASCO Annual Meeting sought to determine the feasibility of commercially available devices and correlate their readouts with clinician-assessed performance status. In the study by our group,1 81% of patients used the wearable monitor for 50% or more of the observation period.
Clinician-determined functional status assessments are fraught with inaccuracies and prone to multiple biases. The most commonly used performance status measures—Eastern Cooperative Group (ECOG) and Karnofsky Performance status (KPS)—classify patient activity into broad categories based on the level of functioning, ability of patients to care for themselves, and physical activity. These methods are broadly accepted measures of how disease and treatments impact a patient’s daily activity and are incorporated as a standardized measure in clinical trials.
Gresham et al demonstrated that monitor-derived step counts associated with clinician-assessed performance status.2 Average daily steps correlated with clinician-assessed performance status (r = 0.73). Similarly, our group’s study demonstrated that the average number of steps differed significantly between patients with an ECOG status of 0, 1, and 2.1 This is an essential first step before wearable derived data are used as a surrogate for legacy performance status instruments such as ECOG and KPS.
Patient-Reported Outcome Tools
THE ASSOCIATION of wearable data with patient-reported quality-of-life outcome tools needs to be established. There are multiple patient-reported outcome assessment tools available for use in cancer patients. These tools are inherently prone to overestimation and underestimation from recall bias. Survey tools require a face-to-face visit and provide a snapshot around the time of the encounter and also demand significant research staff (and patient) time; however, they are well validated for cancer indications and therefore routinely incorporated in clinical trials.
Several teams have studied patient-reported outcome tools and demonstrated interesting trends. Shinde et al used the National Institutes of Health Patient-Reported Outcomes Measurement Information System (PROMIS) programs.3 They found that (1) average daily steps and (2) the number of floors climbed correlated with PROMIS Physical Function, Pain, Fatigue, Sleep, and Emotional tools. Similarly, our group found a minimum, but not average, number of steps correlated with the Brief Fatigue Index, Functional Assessment of Cancer Therapy–General, and the Quick Inventory of Depressive Symptomatology.1 These relationships between wearable-derived and clinician-assessed measures are important to establish, as wearable data and electronic medical records become more integrated. Clinicians and investigators will have access to unprecedented data for clinical decision-making at their disposal.
Nuances of Wearable Data
FUNG ET AL DETERMINED the feasibility of applying a platform that integrates data from wearable activity monitors with electronic medical records to allow providers to track activity in testicular cancer survivors.4 An ongoing clinical trial by Melisko et al is evaluating how to use wearable devices in breast cancer patients undergoing chemotherapy to measure physical activity and sleep throughout chemotherapy, describe patterns of physical activity, and explore associations of activity and sleep with quality of life.5
Given the daily, longitudinal, minute-by-minute data captured by these devices, analysis of data to look at trends and variability related to wearable-derived measurements is possible. This process can allow deep learning algorithms to begin to look at subtle effects of treatment, which can predict long-term toxicity, or to identify departures from baseline, which may be the initial signs of clinical decompensation. These studies will form the foundation, as we start to understand the nuances of wearable data. ■
DISCLOSURE: Dr. Beg reported no conflicts of interest.
REFERENCES
1. Beg MS, et al: Feasibility of wearable physical activity monitors in cancer patients (PAMCaP). 2017 ASCO Annual Meeting. Abstract 6577. Presented June 5, 2017.
2. Gresham GK, et al: Assessing performance status and clinical outcomes with wearable activity monitors. 2017 ASCO Annual Meeting. Abstract 6571. Presented June 5, 2017.
3. Shinde AM, et al: Correlating wearable activity monitor data with PROMIS detected distress and physical functioning in advanced cancer patients. 2017 ASCO Annual Meeting. Abstract e21689.
4. Fung C, et al: Feasibility of utilizing a novel mhealth platform to deliver an evidence-based exercise intervention among testicular cancer survivors. 2017 ASCO Annual Meeting. Abstract e21608.
5. Melisko ME, et al: Objective assessment of physical activity during chemotherapy for breast cancer. 2017 ASCO Annual Meeting. Abstract TPS6626. Presented June 5, 2017.
