My View: A SWOT analysis of the government’s health reforms

VIEW - Johan Van Schalkwyk, Auckland DHB Anaesthesia

The key issue when assessing the government’s plan to reform the health system is whether we and those driving the change have the required breadth of vision. Let’s do a quick SWOT analysis.

Strengths
This change is likely what we need. The DHB system is overtly dysfunctional and past its sell-by date, as shown by the heterogeneity of practice (and level of innovation) in adjacent DHBs. If we can make a truly national healthcare system (think of the establishment of the NHS in the UK all those years ago) that delivers and is loved and respected, we have the potential to deliver quality care to everyone, particularly those who are currently disadvantaged. Maori Health is tasked to keep us honest, a good thing. To their credit, the architects of this bold new plan have articulated the need for good epidemiology; and we already have a recent history of government listening to epidemiologists—with good results.

Weaknesses
It’s very easy to mouth the words “patient centric” and “clinically focussed” and “preventive medicine” but actually quite difficult to articulate design that will bring these into being. It’s also easy to overreach, and try to get healthcare to do things it’ll never do, because the main driver for the development of non-communicable diseases will forever be out of clinicians’ hands—the flow of commodities like alcohol and salt and nicotine and high-energy-density foods into communities. Health is widely acknowledged to be a ‘wicked’ problem, although a lot of this may be related to bad design, poor communication, failure to disseminate successful innovation, and bureaucratic inertia.

Opportunities
The key opportunity here is to simplify. For those marinated in relational terminology, we can build simple systems around a single source of truth, in third normal form. We have the potential to prevent duplication of work—something that I see clinically every day. We however need to do several things.

We need to re-jig the underlying structures so that they are data-centric and not document-centric, as they largely are at present. From a clinical perspective, we need problem lists that work—and that are reusable, so that when someone comes back to the clinician, we don’t go through the tedium of documenting things yet again. We can instead review the problems and ask how things have changed. But more than this, we need UIs that make it trivial to do things like make causal associations, and drill down to the evidence supporting and refuting the problem, and work together on problems. We need simple, reliable, robust prescribing, a key source of error and harm. We need seamless integration between supporting services in and out of hospital.

Another opportunity is to disseminate things that work. I know of two hospitals side-by-side where one is at EMRAM 6 and the other, EMRAM 3. I know that pretty much everyone at the former is happy to share how they’ve got there—including sharing code, even sharing licences—but the other is stubbornly going its own way. With luck, this sort of silliness will cease.

A further opportunity is to institute continuous quality improvement—to listen to Deming. We can build a system where clinicians understand how to use simple measures like control charts to re-engineer systems that work.

• We have more time to think.
• Cut and paste has disappeared, as has “copying others’ work”. There is one source of truth.
• We have more time to talk to patients.
• We spend less time in front of the computer.
• Things just work—within and even more importantly, across regions. Errors become less frequent
• Everyone is happier in their work.
• Crazy ideas like targets, league tables, annual performance reviews and traffic light systems are a thing of the past. (A predictor of success will be how many of those reading this post understand why all of these are bad!)

Threats
National are already whining, and might therefore try to pull back before any useful change is made, but they are still in disarray, so it’s very possible they won’t be able to wreck things. Perhaps they’ll even come to the party?

A bigger threat is likely the phrase “data driven”: it will be a disaster if some bright bugger decides that “we need to let the data speak”, rather than taking a scientific approach where we use the data to test, refute and refine our models.

Another threat is failure to appreciate the need for evolutionary change, growing the small things that work in a way that supports a bigger long-term vision.

As I see it, a further, pernicious threat is that those who build the new and revised systems we require will make them too complex. There are several ways this can fall over:

• We’ve already seen the massive, £14 billion pounds wasted stuffup that was NPfIT; Obamacare blew $29bn on systems that overburden clinicians with extra administrative duties to the extent that they have to employ new people to follow them around and capture data; cut and paste is rampant in many such systems. These failures all represent excessive complexity combined with a failure to understand the basics of how good medicine is done.
• An insidious danger is “everything as an app”. This represents a failure to impose a coherent structure on the data—something you can’t fix with FHIR (I’m happy to explain why).
• Some current trends are inimical to the creation of good systems that persist. These include inadequate documentation (a widespread issue); unthinking use of XML and some OO structures that are in direct conflict with 3NF; a failure to establish simple, consistent data dictionaries; and failures of code transparency. Those who have built good databases and written good code all understand how easy it is to tack on another module or table, and how difficult it is to normalise properly and refactor—or better still, get the initial design right so all of this is unnecessary!
• If clinicians are disempowered and put upon, they will break things, one way or another. They need to be empowered—but this doesn’t mean listening to the clinician with the loudest voice.
• If all participants don’t understand, some of them will break it.

Another threat is that the DHBs will still maintain their silos, while paying lip service to change. Stifling bureaucracy may hang in there, and prevent dissemination of good ideas using all sorts of stalling tactics. Clinicians and managers need to gain a new found respect for one another, and work together with IT experts to build more functional systems that just work.

As I’ve suggested already, the greatest threat is a paucity of vision.