Recently, my colleague John Lynn conducted a video interview with healthcare consultant and “recovering CIO” Drex DeFord (@drexdeford) on patient engagement and care coordination. During the interview, DeFord made a very interesting observation: “When you finally have a process leaned out to the point where [tech] can make fewer mistakes than a human, that’s the time to make big technology investments.”
This makes a lot of sense. If a process is refined enough, even a robot may be able to maintain it, but if it remains fuzzy or arbitrary that’s far less likely. And by extension, we shouldn’t automate processes until they’re clearly defined and efficient.
Honestly, as I see it this is just common sense. If the way things are done doesn’t work well, who wants to embed them in their IT infrastructure? Doing so is arguably worse than keeping a manual process in place. It may be simpler — though not easy — to change how people work than to rewrite complicated enterprise software then shift human routines.
Meanwhile, if you do rush ahead without refining your processes, you could be building dangerously flawed care into the system. Patients could suffer needless harm or even die. In fact, I can envision a situation in which a provider gets sued because their technology rollout perpetuated existing care management problems.
Unfortunately, CIOs have powerful incentives to roll ahead with their technology implementation plans whether they’ve optimized care processes or not.
Sometimes, they’re trying to satisfy CEOs pushing to get systems in gear no matter what. They can’t afford to alienate someone who could refuse to greenlight their plans for future investments, so they cross their fingers and plunge ahead. Other times, they might not be aware of serious care delivery problems and see no reason to let their implementation deadlines slip. Or perhaps they believe that they will be able to fix workflow problems during after the rollout. But if they thought they could act first and deal with workflow later, they may get a nasty surprise later.
Of course, the ultimate solution is for providers to invest in more flexible enterprise systems which support process improvements (including across mobile devices). To date, however, few big health IT platforms have strayed much from decades-old computing models that make change expensive and time-consuming. Such systems may be durable, but updating them to meet user needs is no picnic.
Eventually, you’ll be able to adjust health IT workflows without dispatching an army of developers. In the meantime, though, providers should anything they can to perfect processes, especially those related to care delivery, before they’re fixed in place by technology rollouts. Doing so may be a bit disruptive, but it’s the kind of disruption that helps rather than hurts.