Extending the Impact of Innovation, Everywhere

From texting new moms about blood pressure to sending geriatric patients home from the hospital with packaged meals, innovation in health care involves more than just a change in setting, a change in technology, or a change in care model. It’s a system for testing and learning which changes will succeed.

By Christina Hernandez Sherwood

Photos by Peggy Peterson

Penn Plastic Surgery had a capacity problem. It’s the country’s busiest center for “free flap” surgery, in which a patient’s tissue, usually from the abdomen, is used to reconstruct the breast after mastectomy. Demand was so high that fewer than half of new patients could be scheduled for appointments within two weeks. That bothered Michael Tecce, DO, a plastic surgery resident who worked as an auditor for five years for international accounting firms analyzing and making process improvement recommendations, before he began his medical training. At the same time, Tecce noticed how frequently postoperative free flap patients returned to the clinic: up to three times to have drains removed from their incision sites and twice more for monthly checkups at which the vast majority were cleared. “Is there a better way for the patient?” Tecce wondered. “Is there a better way for us?”

As Tecce mulled over these questions last winter, he ran into Kathleen Lee, MD, a clinical innovation manager at the Penn Medicine Center for Health Care Innovation, which shares a floor in the Perelman Center with Plastic Surgery. They chatted over slices of leftover pizza. Before long, the innovation center and Plastic Surgery partnered to study whether, and how, they could ease the burden on free-flap patients—and the strain on Plastic Surgery—by transitioning much of their post-operative care to occur at home, augmented by text-message communication and a care package tailored to their recovery needs. When they pitched a plan to test these ideas, Joseph Serletti, MD, chief of Plastic Surgery, deemed the project “brilliant,” even though it meant losing some clinical control. Serletti signed on as the project’s sponsor. “It’s doing something better for the patient,” he said. “That’s what we should be focused on.”

It was that, and more: From the initial question—can we do this better?—to the resulting pilot project and next steps that are still unfolding, the plan was emblematic of how the Center for Health Care Innovation has accelerated change across Penn Medicine. For nearly a decade, initiatives launched through the innovation center have yielded multiple benefits in patient outcomes, patient experience, and the cost of care. In many cases, crucially, these projects involve shifting the center of care toward less acute settings, from inpatient to outpatient, and from outpatient to home.

One team found they could keep postpartum women with hypertension safer by using a Penn-developed software platform to check in on their health. Another team discharged geriatric patients earlier—providing more services to keep them safe in their homes to reduce the likelihood they’d be readmitted as inpatients. And the Plastic Surgery team found that removing drains at home could save patients upwards of 20 hours of return visits to the clinic during recovery.

Innovation in health care involves more than just making a change in setting, technology, or care model, though. A lot has to go right in order for health care innovation initiatives to succeed. Not only does the project itself need to work, but it needs to be acceptable to patients, to pass muster with clinicians, and be financially viable, according to David Asch, MD, MBA’89, executive director of the innovation center. “The innovation process aims to test all of the things that must be true for success to happen.”

When a Text Message Can Save a Life

In July, Karimah Ferguson, 38, was in the midst of a move from Abington, Pa., to the West Oak Lane Philadelphia home where she planned to raise her newborn daughter, Lilliana. Ferguson had had an uneventful first pregnancy until her 32-week checkup when, before the routine exam was over, she was sent to the hospital with high blood pressure. Diagnosed with preeclampsia, Ferguson delivered Lilliana a week later via emergency C-section at the Hospital of the University of Pennsylvania (HUP).

Even after that, Ferguson was at risk for continued severe high blood pressure, one of the leading causes of maternal mortality. A few years ago, a new mom like Ferguson would have been asked to come back to HUP for a blood pressure check after discharge. Yet Penn clinicians were finding it was nearly impossible to get at-risk new mothers to show up. Despite offering follow-up phone calls, flexible scheduling and a walk-in clinic, only about 20 to 30 percent of patients would return, said Adi Hirshberg, MD, an assistant professor of clinical Obstetrics and Gynecology. “Asking women to come back to the clinic was just not working,” she said. At the same time, even patients who had their follow up were getting the benefit of only a single blood pressure reading. The timing of their visit didn’t necessarily match up with their stroke risk, which is highest for about 10 days after delivery. Clinicians just didn’t have the numbers they needed to keep these new mothers safe.

Enter the Innovation Accelerator Program, the innovation center’s most intense involvement in transformation initiatives. In 2014, Hirshberg, along with Sindhu K. Srinivas, MD, MSCE’08, director of obstetrical services at HUP and now vice chair for quality and safety in Obstetrics and Gynecology, earned a spot in the program. Their idea: Could they close the gap by sending patients home with a blood pressure cuff and using text-message interactions to make risk monitoring both more consistent and convenient? They called it Heart Safe Motherhood.

Through a series of small, quick pilots—with Hirshberg manually texting patients—the team tested what frequency, timing and wording would work best for the texts. “If we tried to focus on the [technology] platform, we would have looked at all these vendors, checked with legal, gone through all these processes,” said Shivan Mehta, MD’06, MBA’06, MSHP’12, associate chief innovation officer for the center, “and we wouldn’t have been able to test what we really cared about.” Once they proved what worked, the team scaled up with Way to Health, Penn Medicine’s automated software platform for engaging with patients. Way to Health made it easy for patients, who continued to receive a regular text message at the other end—no need to download a dedicated app.

Innovation isn’t always about adding more technology or choosing the high tech solution, Asch noted. “People think it’s innovative if you use Apple watches,” he said. “To me, it’s innovative if it succeeds.” In fact, the Heart Safe Motherhood team chose to give patients traditional blood pressure cuffs for home monitoring instead of high-tech cuffs that connect wirelessly with Way to Health. The wireless cuffs were clunky and unnecessary, Asch said, and additional complexity tends to result in less participation. “I’d like the lowest tech,” he said. “I’d like the easiest solution.”

Fast forward to this summer. Once Ferguson was discharged with a blood pressure cuff, she received twice-a-day text messages for about 10 days asking her to check her pressure and reply with the reading. Though she was anxious about what would happen if her reading was high, Ferguson said she appreciated not having to go to the hospital for a check. “I was over seeing doctors after being in the hospital for two weeks and poked and prodded every day,” she said. “I was glad to be in my own home.”

The Way to Health platform automated review of Ferguson’s blood pressure numbers. When her numbers were normal, she received a reply letting her know everything was okay. If any had been abnormally high, her care team would be alerted and she would receive instructions about what to do. (Clinicians can also review a patient’s full history of blood pressure readings at any time.)

Ferguson wasn’t the only patient who found it easier to monitor her blood pressure remotely. Since Heart Safe Motherhood went live in 2017 at HUP, 85 to 90 percent of the more than 2,500 participants reported their blood pressure at least twice. Readmission of postpartum hypertension patients to HUP has decreased by 80 percent, from about five to only one out of every 100 patients. It turned out that even though providers don’t get to see their patients in person, Srinivas said, they’re “actually more effectively obtaining the data that’s needed to act in a timely fashion.”

Plus, in stark contrast to the way medical technology can sometimes widen inequality, Heart Safe Motherhood “actually narrowed disparities because they used technology that was very patient centered,” Mehta said. Black women are three times more likely to die of preeclampsia than white women, but Heart Safe Motherhood has shown evidence of tightening the gap in follow-up care. While only 33 percent of black patients returned to the office for a blood pressure check after discharge, a whopping 93 percent texted at least one blood pressure reading, Hirshberg, Srinivas and Penn biostatistician Mary D. Sammel, ScD, reported recently in the American Journal of Obstetrics and Gynecology.

Extending the program’s impact is the next step. Heart Safe Motherhood has been the standard of care at HUP since 2017 and at Pennsylvania Hospital since 2018, and the program is gearing up to expand locally, and perhaps across the country. Efforts are underway to determine whether the model can be translated to other medical issues, such as postpartum depression, blood sugar testing, and breastfeeding support.

“We’re not looking for projects that just show early evidence,” Mehta said. “Innovation is the identification, evaluation and, ultimately, implementation and scale of those projects.”

Such innovations in health care are hardly a simple matter of changing technology. The innovation center is continuously learning alongside clinicians that when it comes to revolutionizing health care, one size doesn’t fit all: clinical context matters. The right approach for young mothers isn’t necessarily the right approach for keeping older adults out of the hospital.

When Recasting Roles Can Aid Recovery

The hospital can be a dangerous place for older adults. Some geriatric patients become delirious or lose functional status. Others develop a secondary complication. All told, about one-fifth of patients 65 and older are readmitted to the hospital within a month of discharge. “What happens over the course of a five- or seven-day hospital stay could alter the course of that older adult’s life,” said Rebecca Trotta, PhD, RN, director of Nursing Research and Science at HUP. “It’s a highly vulnerable time.”

Trotta worked to establish the role of the geriatric nurse consultant to support older inpatients at HUP in 2016. But to ensure that added support had a lasting impact after the patient’s hospital stay, Trotta wanted to do more. She was inspired by England’s “flipped discharge” model, which involves discharging a geriatric patient immediately after his acute medical problem is resolved and sending an interdisciplinary team—including a nurse, social worker, therapists and nutritionist—to his home to assess what he needs to return to baseline and prevent readmission. She brought her idea to life as part of the innovation center’s 2017 accelerator cohort.

In the accelerator’s first phase, the innovation center worked with Trotta and her team to study the problem of geriatric hospitalizations. “Any time we go into a project, even if people have solutions in mind, we still want to understand more about the problem,” Mehta said. Then the team performed a series of rapid cycle experiments to explore potential solutions in cohorts of 10 to 20 patients. What’s required to coordinate all the activities involved in an earlier discharge? How do they do so while ensuring the patient has what they need to be safe at home? “We explore different directions and test those concepts very quickly,” Mehta said. “Those learnings can help inform what our actual intervention is.” Finally, the most promising solutions were rolled out in pilots of 50 or more patients to test outcomes.

Trotta’s project, known as SOAR (supporting older adults at risk), kicks off while the patient is still in the hospital. They are followed by the geriatric nurse consultants and receive additional interventions to prevent cognitive and functional complications. On the morning of discharge, geriatric nurse consultants have a handoff phone call with Penn Medicine Home Health to discuss patient needs, instead of relying solely on the electronic health record to share information. The patient is then discharged in the morning—moved up from late afternoon—and leaves with his prescriptions, lunch for himself and his caregivers, and a case of nutritional shakes. “They often don’t have fresh food at home, given they were in the hospital for a few days,” Trotta said. A nurse visits that same day, she said, cutting the time between discharge and a home visit from two days to about three and a half hours while also ensuring the patient has the correct medications in hand days earlier than previously observed. Home health providers use a secure texting app to maintain contact with hospital providers as needed. After the transition home, SOAR continues to support patients, with the help of a geriatric-certified virtual case manager nurse, by ensuring they follow their care plan, attend follow-up medical appointments, and get the supplies they need.

About 120 patients have come through the SOAR pilot at HUP, Trotta said, and as the program has been refined, the readmission rate has decreased from 20 percent to 17 percent. This fall, the SOAR team is moving into a rigorous program evaluation of its playbook (including steps, templates for consistent messaging, and documentation) with the aim of ultimately expanding from the hospital medicine service at HUP to other departments caring for geriatric patients.

While SOAR uses some backend technology, such as text messaging between providers, it’s primarily focused on redesigning providers’ roles. Similarly, IMPaCT, another accelerator alumnus, found success developing a replicable care model that employs community health workers, people who are already known and trusted in their communities, to connect with and provide social support and care navigation for the sickest and most vulnerable patients. Likewise, Penn Medicine’s Center for Opioid Recovery and Engagement (CORE) leverages certified recovery specialists—in recovery themselves—to provide peer support for people struggling with opioid use. “There’s so much potential for improving care coordination and rethinking the roles people play,” said Roy Rosin, MBA, Penn Medicine’s chief innovation officer. “It’s not just technology. It’s care redesign.”

When Less Time in the Clinic Means More CARE

Karen Towell, 56, of Havertown, Pa., left the hospital after her free-flap breast reconstruction in May with four surgical drains tethered to her incision sites. She also brought home a care package of recovery supplies. It included a sleek bathrobe with a belt for holding surgical drains and a plush seatbelt cover (to protect her surgical site while driving or riding in a car), along with other items.

Towell was one of about 50 patients in the first three months of the pilot program sparked by Plastic Surgery resident Michael Tecce to redesign care following free-flap surgery—a model they called “Care, Reimagined,” or CARE in a stylized, shortened form. 

When Tecce first teamed up with clinical innovation manager Kathleen Lee, innovation manager Lauren Hahn, and Plastic Surgery Business Director David Okawa to study the problem, one of their most striking findings was that it took free flap patients an average of four to five hours, including travel, each time they had a drain removed at the office. The drain removal itself lasted three to five seconds.

In the CARE pilot, Towell received text messages about her drain output and a visit from a home health nurse starting the day after she left the hospital. The team alerted home health when Towell’s first drain was ready for removal. (As it turned out, home health nurses visiting free flap patients were perfectly capable of removing drains—they just hadn’t been trained to do so.) The nurse removed Towell’s drain as she relaxed on her living room recliner. “For me to go down to Penn… it would have been my whole day,” Towell said. “Being at home is just so much more comfortable.”

Towell was initially less comfortable responding to the text messages she received about her recovery, however. In a post-surgery haze when the concept was first introduced in the hospital, Towell said she wasn’t sure what to make of the texts she received each morning after she went home. “I was like, ‘Wow, this is so nice, but I don’t know who you are,’” she said. “At first it kind of threw me as being impersonal.”

The innovation center is experimenting with more casual tones for communications sent via automated systems, Asch said, but patient experiences with these remote platforms vary—most patients in the CARE pilot were enthusiastic about the text messages from day one.

Once the text-messaging service was re-explained at an in-person postoperative visit, Towell grasped that even though the check-in questions she received were automated, it was her care team interfacing through Way to Health—not a robot—responding when she asked more personal questions about her own recovery. The very next day, she sent a text asking if it was safe to start using deodorant again. Towell’s two-month postoperative visit was also handled remotely. Instead of going to the clinic, she answered a series of survey questions and sent photos of her incision sites via a two-day text message check-in beginning on a Monday. On Wednesday, Towell received a text completing her check-in and confirming that she was cleared to return to work.

Transitioning care to the home setting for patients like Towell offered benefits to other patients, too. The CARE model eliminates the need for four out of five follow-up office visits, which frees up a significant number of appointments for new patients, who can now get in faster. Automating the most rote aspects of the text check-ins further frees up clinicians’ time to spend on more personal aspects of patient care. The CARE team has fully automated monitoring of drain outputs and scheduling home health care visits to pull drains at the appropriate times, and is working on automating responses to non-urgent frequently asked questions.

Gaining these kinds of benefits through innovation projects requires constantly refining processes while embedded in a natural clinical context, Rosin said, even though it risks some hiccups along the way. “So many times in clinical medicine, people never get past the white board,” he said. But innovation is about blowing right past the white board. “There’s no way to learn something without doing something. You need to try it.”

It’s also a delicate balance to see patients less frequently in a clinical setting, while maintaining care excellence and a patient-provider relationship. Other Penn innovation projects that are finding success in this area include Advanced Heart Care at Home, which provides IV diuretics to heart failure patients at home, and BreatheBetterTogether, which provides individualized care for patients with chronic obstructive pulmonary disease who are at high risk for hospital readmission.

Even though CARE patients aren’t going to the clinic as often, just like new mothers using Heart Safe Motherhood, they’re interacting with Penn Medicine more frequently. In the old model, a patient received a visit from a home health care provider and a call from the surgery team a few days after discharge. Then, it was crickets until the first drain removal. Now, CARE patients rarely go a day without receiving a text message or seeing a home health nurse. “We’ve pushed all of our contact with the patient to be more personal,” Tecce said, “and earlier in their postoperative care. By connecting with patients through texting and post-operative support garments, we show patients in a tangible way that, even though they are making few in-person office visits, we are still supporting them in their recovery.” The selection of items in the care package was informed by the experience of previous Penn patients who found these items provided comfort in the slow healing process.

Not all innovation projects are designed to shift care away from acute clinical settings. But those that do are built on the premise that the total amount of time patients spend with a health care provider is only a small slice of their lives. Even for the sickest patients with chronic illness, it’s only a few hours each year. It’s what happens during a person’s other 5,000 waking hours—whether they take their medication, if they eat healthy and exercise, whether they smoke, and whether the demands of a new baby make it difficult to get to a clinic—that matters most to their health.

But it is not merely that time spent in clinical settings is limited. It’s also costly and inconvenient for patients who are more comfortable than ever handling the most personal aspects of their lives remotely. “Would you rather write a check made out to cash, go to a bank during banking hours, engage with a teller and get $50? Or would you rather go to an ATM? Or would you rather skip it entirely and Venmo?” Asch said. “Other businesses have successfully made those transitions from what would be called a more acute site of care. It’s time for health care to get past that.”