A Q&A with the Population Health Lab
The Population Health Lab, led by CHIBE leadership team member Dr. Shivan Mehta, evaluates and implements evidence-based interventions to improve health outcomes at scale.
CHIBE spoke with Dr. Mehta to learn more about the lab’s work.
What does the Population Health Lab seek to do, and what are some specific health areas that your team targets?
Our goal is to improve health on a population level by working with health systems, meaning we focus on outcomes for large groups of patients.
To do that, we design care delivery interventions informed by behavioral science principles, partnering closely with clinicians and operational leaders in the health system to ensure that changes can be feasibly implemented and sustainable in the long-term.
We evaluate these interventions with pragmatic clinical trials and scale those that lead to better clinical outcomes or increase health-promoting behaviors like completing screenings or vaccinations.
Our portfolio spans efforts to prevent or better manage:
- several types of cancer (colorectal, breast, and liver)
- hepatitis C (HCV)
- influenza
- cardiovascular disease
Can you tell us about a project you’re particularly proud of?
We’re particularly proud of our breast cancer screening outreach work in partnership with Penn Medicine. After the initial surge of the pandemic, the health system was looking for ways to invite women back for mammograms. During breast cancer awareness month in 2021, we partnered with primary care to send outreach to over 24,000 eligible women in the Philadelphia and Princeton regions, while also conducting 2 concurrent pragmatic trials to evaluate if behavioral interventions could increase response rate.
In these trials, we found that bulk ordering and text messaging significantly increased patient response to breast cancer screening outreach by 2.7% and 2.1%, respectively.
There was a small (1%) but non-significant difference in completion when the message was signed by the PCP, compared to the primary care practice.
The opt-out framing embedded within the trial (via bulk ordering) sets mammogram completion as the default behavior and reduces the friction of taking the desired action by removing steps in the process (having an office visit or requesting an order) toward completion.
Text messaging adds an additional nudge beyond the basic outreach and bulk ordering, also allowing for patients to get additional support or information about screening and is highly scalable to large populations.
These trials led to an additional 5,714 women completing breast cancer screening, 76 new cancer diagnoses, and to a new population level outreach program within Penn Medicine primary care aimed at mammogram completion among women 40+ identified as overdue for mammogram.
This program was implemented in 2022 by primary care with similar response rates, and the breast cancer screening rate has increased from 70% to 76% as a result of this and other efforts.
Could you give us an example of a behavioral science tool you use to achieve health goals?
One technique we frequently use is opt-out framing, which sets the clinically recommended choice as the default instead of requiring the effort to opt in.
For example, we found that mailing patients an at-home test kit tripled screening rates for colorectal cancer among eligible patients compared to offering it if they opted in.
On the clinician side, we’ve also implemented default hepatitis C screening orders in hospital admission order sets, which nearly doubled screening rates among at-risk hospitalized patients.
What else do you want people to know about the Population Health Lab?
Our group operates at the intersection of research and operations, and we have conducted over 30 pragmatic trials of health care delivery interventions.
We leverage the latest insights from behavioral science and new technology to evaluate which interventions work (and do not work) through rigorous methods such as randomized controlled trials and interrupted time series analysis.
By pragmatically embedding these interventions in real world clinical settings, they are more relevant to practice, and they are better positioned to scale when effective.
Finally, we are committed to working with our clinical partners after the trial is completed to implement successful interventions across all relevant settings across the health system.