Hospital investors want to know who owns implementation before a startup signs
Northwestern Medicine looks for an operating sponsor. Cedars-Sinai can invest before a pilot, but wants founders to know who controls the budget and how quickly a hospital can see a return.
John Bass had a startup through hospital contracting when it learned the information technology team had a yearlong queue for implementation. By then the company had done the work of winning the account. It still could not put its product to use, collect evidence from clinicians or show its next investor that the sale was going anywhere. Bass, founder and CEO of Hashed Health, recounted the experience while moderating a November 17 discussion with four health system investors in Las Vegas.
The investors disagreed, in useful ways, about what a company needs before a health system will write it a check. Northwestern Medicine wants an operator who will take responsibility for the change. Cedars-Sinai Health Ventures can invest before a pilot exists. Tampa General Hospital’s Rachel Feinman urged founders to consider whether the fastest route to credible evidence might run through a smaller hospital while talks with a large academic center continue. Dr. Matthias Kleinz of UPMC traced a surprising parallel between getting hospital software live and getting a complex therapy to a patient.
The person who has to change the work
“We can’t move forward with anything without operator engagement in our world,” said Kali Arduini Ihde, Northwestern Medicine’s director of ventures and innovation. Her next question was “Do we have somebody who wants to sponsor it?” The distinction is significant. A clinician can testify that a problem exists. A sponsor has to persuade colleagues to change how they work and take responsibility for what happens when an old process and its replacement briefly operate side by side.
Arduini Ihde described changes that can reach beyond a screen or a shift in staffing. Teams may need to revisit who owns a decision or which department handles an exception. That helps explain why an investor at Northwestern can admire a company and still judge that its timing is wrong for the organization. Her team has developed relationships with other institutions, in the Midwest and beyond, that may be ready to try something sooner. Northwestern’s innovation program describes how its clinical and administrative leaders help identify priorities.
Maureen Burgess, a partner at Cedars-Sinai Health Ventures, can proceed on a different basis. “We don’t have a requirement that there is a commercial agreement or a pilot in place,” she said. A stakeholder’s interest can support investing in a business whose role at the health system may be several years away. The founder still needs a way to earn revenue in the interim. Burgess raised smaller clinics, other healthcare organizations and, in some circumstances, sales to consumers as possibilities.
That flexibility changes once a company seeks an actual deployment. Burgess wants clarity on the budget owner, the incentives, the resources and who will work with legal. Bass’s account of an unexpected year in the IT queue illustrates why those questions belong early in the sales conversation. Someone with authority to sign a contract may have little say over when the engineers needed to fulfill it become available.
What would count as a return in the first year?
Burgess put a question to founders working against tight provider budgets. In her words, “how do you articulate and prove a 12 month ROI?” A business may eventually create value for clinicians, patients, payers and drugmakers. The hospital contemplating a purchase still needs to know what benefit it can expect to capture within its own budget cycle. Savings elsewhere in the system can be substantial without helping the department that has to pay the invoice.
Feinman, Tampa General’s senior vice president for innovation, ventures and digital solutions, stressed how much the answer depends on the people who use the product. A proof of concept may start with a small group of enthusiastic superusers. An enterprise rollout includes clinicians who did not volunteer and staff who have other priorities that week. Extrapolating the pilot’s performance to every department ignores the time and effort that those teams need. For Feinman, “it’s all about alignment at the outset.”
That is also why she saw value in a community hospital or private practice as an early customer. A recognizable academic center can open doors, but “getting those proof points is important,” Feinman said. A smaller institution that can begin promptly may give a founder usable evidence while a larger system works through governance. She cited a startup newly willing to put fees at risk because its own customer data had become good enough to tell it what it could reliably deliver.
Kleinz, executive vice president and head of strategic partnerships at UPMC Enterprises, urged founders to “think about the last mile.” A complex therapy requires someone to identify an eligible patient, reach that person at the right time and arrange for treatment. The digital analogue is the tool that solves the right clinical problem but sits in an implementation queue. In either case, an investor waiting for proof of use can become impatient while the organization works out the last few steps.
When a hospital builds something it cannot sell
The same scrutiny applied to software created inside hospitals. Kleinz acknowledged that UPMC had sometimes built products that fit its own way of delivering care better than they fit another system’s needs. “It’s another thing to build a successful business,” he said of the work beyond creating a useful tool. The developers’ access to a large and demanding first customer can conceal the fact that the second customer handles the same task quite differently.
Arduini Ihde described researchers and staff bringing forward more inventions as development tools become accessible. Faster development raises the burden of proving that a product is different enough to commercialize. A hospital may get plenty of value by keeping an internal tool for its own staff. If another institution can readily build its own version, forming a company adds an entirely different set of costs and customers to serve.
Bass also asked what happens when a proposed product needs patient data from several institutions. Kleinz said some questions require more diverse data than even UPMC can supply. He described giving partners access through a federated model that seeks to keep data on the organization’s premises. Arduini Ihde described Northwestern investing in engineering expertise for de-identification and using explicit agreements to define what partners can do. A founder budgeting for that partnership must account for the people who prepare the data and safeguard its use.
By the close of the Health System Risk Takers panel, the investors had offered founders a way to read a promising meeting. Find the person who will change the workflow, the person who controls the money and the team with time to implement. Then ask those people what result they would recognize a year later. In Bass’s experience, one conversation about the IT calendar could have revealed a year’s delay while the startup still had time to plan around it.