Beiwe Case Study Q&A: Panda Nikhil MD MPH of Ariadne Labs

Can you give a brief description of your medical practice and research?

I am currently a general surgery resident and cardiothoracic surgery fellow at Massachusetts General Hospital/Harvard Medical School. My training path is focused on general thoracic surgery, a field that provides surgical care to patients with both benign and malignant diseases of the lungs, esophagus, airways, and chest wall. Like any area of healthcare, the effect of these diseases on patient physical and psychosocial health, as well as health-related quality of life is substantial. Unfortunately, there are few scalable tools that allow patients and their surgical teams to prioritize the patient perspective throughout phases of surgical care. My area of research attempts to improve the ways health systems measure the outcomes that matter most to patients as they enter surgical care cycles, with an interest in harnessing technology to do so.

What research question were you looking to answer through the use of Beiwe?

We sought to evaluate the feasibility of applying smartphone-based digital phenotyping within a cohort of patients undergoing surgery for cancer. We harnessed both the passive data collection and survey capabilities of Beiwe to collect data on pre-and postoperative health-related quality of life, decision quality, and recovery expectations. The aim of the study was to understand how smartphone-based digital phenotyping and other similar mobile health technology could (1) enhance recovery monitoring and (2) improve shared decision-making.

How many patients were enrolled and for how long?

100 cancer patients who were enrolled for 1 week preoperatively and 6 months postoperatively. The study lasted for 2 years.

What was the main result of the study?

The study generated rich data for each patient enrolled. At the conclusion of the enrollment period, we performed several analyses based on the aims of the study. Below are three examples:

  1. Enhanced recovery monitoring: using passively-collected accelerometer data, we compared the physical activity levels of individuals experiencing a postoperative event (e.g., complication after surgery) compared with uncomplicated recovery patterns. We found that there were substantially lower activity levels among patients experiencing events; in some cases, this included before the event was diagnosed clinically.
  2. Association between health-related quality of life and GPS-derived summary measures: we compared passively-collected GPS-derived home time and distance traveled (e.g., surrogates for aspects of recovery) with self-reported health-related quality of life scores among patients undergoing different operations for breast cancer. We found that while there were similar health-related quality of life scores regardless of the type of operation, there were substantial differences in the GPS-derived measures.
  3. Preoperative decision conflict: we harnessed the Beiwe platform to deliver preoperative surveys to measure patients’ decision conflict scores after they had decided to undergo surgery, but prior to the date of surgery. We then compared the recovery patterns – as defined by passively-collected sensor data – among conflicted patients versus those with no clinically significant conflict. We found that among patients with preoperative conflicts, they experienced lower activity levels postoperatively compared with patients who expressed no conflict, even after considering the clinical difference in the recovery.
What made Beiwe stand out over other platforms that you researched? What feature of Beiwe was most appealing?

Features of Beiwe that stood out compared with other platforms:

  1. The data that is generated remains in unprocessed “raw” form. This eliminates the undisclosed processing of data to generate a summary measure (e.g., step count) often applied in other platforms. The latter approach limits generalizability and pooling of data across patients or studies.
  2. Passive data collection means that there is a large data yield with minimal burden to patients and surgical care teams.
  3. The most important feature was the security of data collection and storage.
What plans do you have for Beiwe in the future?

Our team is currently evaluating how to scale tools like Beiwe in surgical care.

Share:

Recent Posts

Beiwe Service Center Overview

Beiwe Service Center (BSC) studies are run on the Onnela Lab production instance of Beiwe. This includes the latest version of the software and the

Welcome to Metadata Blog

Welcome to the Metadata blog, a blog about digital phenotyping and statistical network science by the Onnela Lab. We will use this blog to share