Healing at Home

Supporting new mothers in the fourth trimester

Opportunity

In 2017, the average postpartum length of stay (LOS) for a routine vaginal delivery was three days at the Hospital of the University of Pennsylvania (HUP). However, many new parents want to be in the comfort of their own homes, recovering and bonding with their babies as quickly as possible after delivery. In fact, a 2018 survey at HUP found that 80 percent of low-risk patients who had vaginal deliveries were amenable to going home one day sooner.

Regardless of when discharge occurs, numerous challenges can arise in the weeks following birth, also known as the fourth trimester. The CDC estimates that approximately one-third of pregnancy-related deaths happen within this period. Clinical considerations in the fourth trimester are complex – with patients experiencing everything from typical recovery symptoms to potentially dangerous complications. In addition to their own medical conditions, many parents struggle with navigating infant feeding and care. 

In 2018, when this project began, the American College of Obstetricians and Gynecologists (ACOG) released new guidelines advising that postpartum care should become an ongoing process, rather than a single encounter, with services and support tailored to each patient's individual needs.

Intervention

Healing at Home supports the postpartum needs of parents and babies in the setting they prefer using a two-pronged approach.

First, the program gets eligible patients home sooner by implementing an expedited discharge process. Historically, health screenings were completed for new babies 36 hours after birth. Healing at Home moves that screening to occur at 24 hours, enabling patients and their babies to be ready for discharge one day after delivery rather than two.

Next, Penny, a text-based automated postpartum chatbot, bridges the gap in fourth-trimester care by providing new parents with around-the-clock access to clinical guidance. In addition to answering parents' concerns and providing just-in-time education, Penny enables increased and more efficient lactation support and postpartum depression screening and facilitates triage to high-value care if necessary.

Impact

During the initial pilot at HUP, Healing at Home safely decreased LOS by 40 percent, with zero postpartum emergency department visits or readmissions among participants. This reduction in LOS translates to 18 fewer hours in the hospital for new parents and babies and more time at home to recuperate in a comfortable and familiar setting.

Healing at Home was implemented as a clinical program at HUP in March 2020. It is currently available to patients who complete vaginal delivery with recommended discharge around 30 hours of life. 

Since launching, more than 380 patients have enrolled in the program. Penny correctly and automatically answers 80 percent of patient inquiries - unburdening busy providers from answering routine questions so that they can focus their attention on more acute issues. The tool has also proven to work as an early-warning system for postpartum conditions. For example, of parents completing the postpartum depression screening on Penny, 25 percent have scored as at risk. When scores such as these are registered, Penny notifies providers to enact early clinical intervention.

The team is currently working to establish Healing at Home as the standard of care at HUP and expand to other Penn Medicine hospitals. Penny's clinical content is expanding to include support for parents who had a cesarean section or whose babies were born preterm, with the goal to make Penny accessible to all parents delivering at HUP in 2021.

At scale, Healing at Home has the potential to improve the postpartum experience for more than 16,000 Penn Medicine patients annually, setting the stage for the long-term health and well-being of new parents and their babies.

COVID-19

In March 2020, patients started requesting to leave the hospital earlier to reduce possible exposure to COVID-19. In response, the Healing at Home program expanded enrollment to include patients who had higher risk vaginal delivery. This ensured that patients remained connected to care while recovering at home.

Phase 2: It does work
Collaborators

Kirstin Leitner, MD
Lori Christ, MD
Joana Parga-Belinkie, MD
Laura Scalise, MSN, RN
Jessica Gaulton, MD, MPH
Penn Medicine Information Services

Innovation leads

Emily Seltzer, MPH
Davis Hermann, MiD
Christina Mancheno, MPH
Ryan Schumacher
Lauren Hahn, MBA
Roy Rosin, MBA
Raina Merchant, MD, MSHP, FAHA

Funding

Innovation Accelerator Program
Johnson & Johnson Maternal Health QuickFire Challenge
Penn Medicine Women’s Health Leadership Council

External partners

Memora Health

Innovation Methods

A day in the life

One of the best ways to learn more about a problem area is to experience it yourself. Immerse yourself in the physical environment of your user.

Do the things they are required to do to gain a firsthand experience of the challenges they face. Completing a day in the life exercise will enable you to uncover actionable insights and build empathy for the people you're hoping to help.

A day in the life

We spent several hours on postpartum floors talking to patients to understand their pain points.

A common theme was that most individuals we spoke to, 80 percent in fact, wanted to leave the hospital sooner than they were scheduled to.

We also spent several hours observing new mothers in their homes to better understand what type of support was necessary for the first few weeks after delivery.

Fake back end
It is essential to validate feasibility and understand user needs before investing in the design and development of a product or service.
 
A fake back end is a temporary, usually unsustainable, structure that presents as a real service to users but is not fully developed on the back end.
 
Fake back ends can help you answer the questions, "What happens if people use this?" and "Does this move the needle?"
 
As opposed to fake front ends, fake back ends can produce a real outcome for target users on a small scale. For example, suppose you pretend to be the automated back end of a two-way texting service during a pilot. In that case, the user will receive answers from the service, just ones generated by you instead of automation.
Fake back end
We piloted a fake back end to evaluate whether postpartum parents would utilize text-based support.
 
To test this in a low-cost way, we manually interacted with the patients, mimicking what interactions with a chatbot might look like.
 
Over two months, 90 patients sent more than 2,000 messages to the text message support team. The pilot revealed the most common questions and concerns of patients - providing us with a framework for building an automated chatbot to support families in the fourth trimester.

Videos

Pitch Day 2019

Penn Medicine and Memora Health: Penny Demonstration