Simply begin

Simply begin. Most innovation fails not in its execution but at its inception. Health systems are built and tended by bright people with innovative ideas, yet many of their ideas are never put into action. If perfect is the enemy of good, over-planning is the enemy of starting. When facing a monumental project, the planning is almost always easier than doing the work, and more planning can feel safer than starting a project and facing the risk of failure. Yet time spent creating plans and budgets that foresee every contingency is time that is not spent learning lessons by putting ideas into play. General Colin Powell famously said, “Once you have information in the 40-70 percent range, go with your gut.” I wanted to create an environment where our team felt loyalty to the project, not the plan.

This learn as we go approach has far-reaching benefits. I find that it offers a subtle change in mindset. Instead of focusing on a plan, our eyes are open to solutions. By simply beginning, we are freed of the need to defend a plan, and we are better positioned to receive input from our users who are normally the people on the front line of medicine. All too often, IT is viewed as the reason why nurses cannot log in or pharmacy labels do not print, or, worse, as remote people who push new systems upon an already pressed staff. When we simply begin, we are automatically part of the same team, and anyone in healthcare knows teamwork is an essential part of delivering the best patient care.

As our IT team prepared to implement a refrigerator temperature monitoring system, I thought of Powell’s sentiment and my own experiences. Years ago, the responsibilities of a hospital IT team were generally limited to supporting mainframe and desktop computing. As technology has developed, even hospital beds are connected to the network. Biomedical Engineering has aligned with IT, and we are now responsible not only for computers, printers, and televisions, but IV pumps, MRIs, and refrigerators that store vaccines, breast milk, and medicines.

To ensure that these critical medical supplies are kept between 1 and 3 degrees Celsius, nurses on hospital floors used to manually check the temperatures twice a day and log the results on a piece of paper attached to each refrigerator. Although such monitoring seems a simple task, nurses are busy, and hospital safety inspections routinely identify fluctuating refrigerator temperatures as a major problem. With the goals of freeing nurses of this responsibility and achieving reliable daily monitoring, our team installed temperature probes in each refrigerator and networked them back to a central monitoring station. No one we knew had done this before, so our team struggled with the technology, clinical workflow, and the process for addressing temperatures that fell out of the parameters for safe storage. I met with the team twice in January of 2016 as they bounced back and forth between hospitals, the data center, clinical meetings, and planning sessions. They wanted to get everything perfect for an April 1 go-live across six hospitals, but I was skeptical that we knew enough to count on a single, flawless launch. “Just turn it on now in January for one nursing unit and see what you get,” I told them. They were reluctant because they had not yet resolved problems with network communications and nurses’ notification. They were also concerned that refrigerators would alarm at the central station every time the door opened for a nurse to retrieve a vaccine. “Let’s just start,” I insisted, “and see what we get.”

So, we simply began and turned on remote monitoring for four refrigerators. One day later, we were glad we did. We learned very quickly that when a nurse checks the temperature at the same time every day, we got the temperature we expected:

 When we used IT to remotely monitor and sample the temperature every five minutes, however, we found that the temperature fluctuates wildly in between readings:

In the midst of trying to perfect the system, we did not consider the variable at the heart of this innovation – the temperature of the refrigerators themselves. Our eyes were opened. As it turns out, the refrigerators we had in place were not meeting our temperature specifications even when the refrigerator doors remained closed. Our computer-based monitoring revealed that each time the compressor started or stopped, the temperature dropped precipitously or spiked above the threshold. 

When we began the project, our focus was reducing the time nurses spent on equipment monitoring so they could spend more time with patients. By starting our work rapidly, we gained key insights early and recognized that we needed to solve a different problem entirely. Of the 1,350 refrigerators humming throughout our six hospital campuses, 780 had this issue and needed to be replaced with upgraded refrigerators at a cost $3,000 each. A little more than $2 million later, we solved the temperature fluctuation problem and resumed our focus on centralized monitoring.

By April 2016, we set up a desk in one of our former data centers and staffed it 24/7 with an analyst who monitored refrigerator temperatures across our enterprise. The formerly local and manual temperature recording process was now automated, and any time a temperature moved out of the specified range, an alert prompted the analyst to contact the appropriate nursing unit to identify the problem and follow up with an engineer as required. 

While this was clearly a step forward, we realized that centralizing alarms and data feeds presented additional opportunities. Now it was time to get creative. We took advantage of the space and the 24/7 staffing inspired by 1,350 refrigerators to create a second team dedicated to remotely monitoring the vital signs of patients in the emergency department of one of our largest academic medical centers. 

It seems counterintuitive that even when a patient’s vital signs are being monitored in the emergency department, there are times when no one is directly looking at the screens or listening for the alarms above the noise of the environment. More importantly, even when the nurse or physician is looking at the monitor, it is often challenging to identify a patient whose condition deteriorates slowly and unremarkably over time. Sometimes the recognition of a patient’s frail condition comes too late. By having a clinician sit at an offsite location and monitor 50+ patients at once without the distraction of being in the emergency department, we added a layer of safety. Instead of adding to the demands of an already busy environment, we now had an experienced clinician whose only responsibility was monitoring the data of each patient. When the remote clinician identified a significant change in the condition of a patient, the patient’s assigned nurse was immediately alerted to the situation. Within the first few weeks of remote monitoring, we had more than six of what we called “great catches” in which the patient’s deterioration was identified early and in time.

Instead of building an entirely new remote patient monitoring center, we had taken the former data center, added minimal equipment and staffing, and created what we called our Clinical Operations Center or CLOC. Other innovations followed – we found that in our health system, new ideas sometimes would not be put into practice because of a lack of space or resources to make them happen. So, we began to expand our capabilities at the CLOC.

We next moved operators from four of our hospitals into the CLOC and sought ways to maximize their productivity. Formerly, they had simply answered phones and transferred calls. When a patient called for a prescription refill, our operators queried the patient for data, completed a form, then faxed it to the appropriate physician’s office.  We saw that these talented people were underutilized and could help us improve and streamline the process. We taught them to use our electronic medical record, trained them to coordinate telemedicine video calls, then promoted all of them from the role of Call Center Agents to Account Representatives with an average 13 percent increase in pay. In their new roles, they became active participants in the clinical care process, increased first-call resolution by solving issues themselves, and even reduced the time patients waited for prescription refills from an average of eight hours to only five minutes. Patient care improved, and our staff advanced their training. Our innovation in this case was simply to take the people and resources we had and make them more efficient by focusing on their training and their workflow. IT teamed up with patient experience, and we were all better for it.

We took this simple approach to innovation to new heights in the summer of 2016 when our Chair of Emergency Medicine said he would like to try telemedicine visits within the busy emergency department of one of our hospitals. While urgent care visits by video are nothing new, he thought we could make physical visits to the ED more efficient by offering patients with less serious conditions the option of a virtual visit with a physician. Like the CLOC, we decided to simply begin. 

Within two weeks, we installed video equipment in an existing room in our ED, added a camera to an existing physician office 200 yards away and started seeing patients in what we call NYP OnDemand Emergency Department (ED) Telehealth Express Care. Patients are greeted at the front door and triaged by a nurse. Those whose injuries or illness are mild are presented the option of a traditional visit (about 2.5 hours) or a video visit (about 30 minutes). The patients who opt for a virtual visit enter a private room and immediately begin a one-on-one consultation with a physician by video. At the end of the visit, the remote physician sends discharge instructions to a printer in the patient room and can send prescriptions electronically to the pharmacy of the patient’s choice, when necessary.

We simply began by leveraging people, space, and technology we already had on hand. By starting rapidly, we learned what worked and what needed improvement within two weeks of our initial idea. Patients of all ages loved the service – when a 21- and an 89- year-old patient were seen back-to-back, it was hard to tell whose comments (“This was the coolest,” and “I am going to tell all of my friends.”) were whose. We started with 10 patients, rapidly tweaked the process, saw 50 more patients, evaluated the process, then opened it up broadly. We have now seen more than 10,000 patients through the service and reduced ED wait times and revisits. 

Net Promoter Score (NPS) is a way to gauge the satisfaction and engagement of customers. Ikea has an NPS of -9, Lego an NPS of 6, and Apple an NPS of 72. Our ED Telehealth Express Care Service is well-liked by our patients and has an NPS in the mid-90s.

This service is one of 10 different telemedicine modalities we now offer patients. In each case, the offering was not the result of years or even months of planning, but a rapid response to an identified need. When we realized that nursing homes would send patients to our ED in the middle of the night “to be safe,” we created a TeleNursing Home service that allows nursing home staff to connect with our ED physicians 24/7 to determine if a patient really needs to be transferred. When we realized that patients in our EDs sometimes waited up to 24 hours to be seen by a psychiatrist or transferred to a facility with psychiatry coverage, we used existing equipment and our own fellows to create a TelePsych service which shortened that wait to within two hours.

Innovation is not consistent with multi-year plans. Rapid implementation and real-world testing quickly demonstrate what innovation has value. While clinical interventions that have direct impact on patients can and should undergo rigorous clinical trials and testing, operational process improvements can be put into place rapidly and refined using real-world feedback. The mantra to fail fast can only be followed once work has started and has a process in place that can succeed or fail. By simply beginning, we create an environment which is free from the stigma of mistakes. As hospitals strive to identify risk and prevent mistakes, it is important for IT to follow physician leadership and do the same. With loyalty to project before plan, a willingness to embrace rapid change can significantly improve not just operational efficiency but patient and employee experiences as well.

This article appears as a chapter in the forthcoming book Voices of Innovation: Fulfilling the Promise of Information Technology in Healthcare from CRC Press and edited by @Marxtango