Reducing Knowledge Management Failures

It was the best presentation on knowledge management I have seen in over a decade.

He began, as all good presentations do, with a story that set the context, engaged the audience, and created a sense of urgency. It was the story of Vanessa Young, who died seven years ago at the age of fifteen as a result of a reaction to a stomach drug called Prepulsid she was prescribed. The heart damage this drug can cause to patients with eating disorders was known to some, but alas, not to her doctor. Vanessa died of knowledge management failure.

Doctors are a conservative and sometimes ornery group. They balance what they’ve learned in medical school, personal experience, colleagues’ experiences and judgements, their own instincts, and whatever they can glean from current reading and research they can fit into their schedule. They do their best, though some do much better than others. Dr. Davis’ goal is to help them do better.

Traditional KM lore has it that you buy and deploy appropriate knowledge content, processes, and technologies to bring about ‘culture change’ and hence make people more effective in their work. Davis takes a different approach: He starts by trying to understand why doctors aren’t already figuring out how to do their best with what’s available. They are, after all, smart, motivated people.

So he starts by looking for objective measures of the quality care ‘gap’: the measurable difference between what is reasonably achievable in a complex health system and what is actually being achieved. This gap is analyzed into components:

• underuse of knowledge and tools (e.g. poor diagnosis and treatment of depression, alcoholism, pneumonia, and diabetes)
• overuse (e.g. of antibiotics, tranquilizer prescriptions to seniors)
• frequent misuse or error

Then, the possible sources and causes of the gap are identified:

• problems originating with the clinician (e.g. age, training, disincentives, poor self-directed learning skills, inability of some clinicians to self-assess their knowledge well); there is a model called the Pathman model that analyzes these problems into four sequential components: awareness, agreement, adoption and adherence, and identifies reasons for failures in each component
• problems with the continuing medical education system (e.g. ‘predispositional’ just-in-case training that tells you ‘what you should do if’ is ineffective, training that isn’t patient-mediated, doesn’t have known thought leaders behind it, or isn’t reinforced at point-of-care rarely gets deployed)
• problems in the health care system (we all know about them)
• problems with the evidence/knowledge (e.g. quality, useful format, credibility, consistency, complexity of understanding and applying, cost, degree of change to establish procedures, access)
• problems originating with the patient or family (e.g. ignorance, unwillingness or inability to follow a regimen, lack of engagement in their own health management)

The next step is to identify the best available clinical evidence from the firehose of research, reports, trials and other data. To do that, they’ve created an organization called Guidelines Advisory Committee to review everything written about the areas where the gap was identified as being greatest, and assess, endorse and summarize Guidelines based on research and other knowledge (‘evidence’) in those areas. You can see what they’ve done on the GAC Canada website (take a look, for example, at their review of this Guideline on how to treat endometriosis). These reviews are governed by a rigorous system of evidence assessment called the AGREE system, and are just one of the mechanisms that the GAC is sponsoring to improve practices and policies informed by evidence. They are hoping to extend their reach beyond direct-to-practitioner actions, to include medical faculty development and curriculum reform, and to help nurses and pharmacists, and eventually patients as well (though the Guidelines are carefully written to be understandable and useful to the public, and they are available to everyone on the GAC website).

And then, they look at what Davis calls the ‘barriers’ (yes, that’s complex adaptive systems language) to effective use of best available evidence — i.e. knowledge transfer. In other words, why are perfectly intelligent clinicians not already using this best available evidence? Some reasons:

• too much information to keep up with (solution: distill it into endorsed Guideline summaries)
• delivered just-in-case instead of just-in-time (solution: embed it in tools used at point-of-care e.g. anaesthesia nozzles that are different sizes so you can’t accidentally connect the wrong gas to the patient’s mask)
• not clearly communicated (solution: more effective education programs, multiple communication media)
• not consistently or completely delivered or implemented (solution: coordinated delivery programs)

Davis summarizes all this with his Seven Steps to Better Care:

1. Collect information and gain deep understanding of where the gaps are, what the possible causes are, and why they are occurring despite the best intentions of those in the system i.e. know what is happening today and why
2. Identify and collect the best available evidence relevant to each identified gap
3. Conduct an analysis of the barriers that preclude this evidence from being effectively used
4. Identify interventions, tools, methods and strategies to get around these barriers
5. Use a combination of methods and media to communicate and implement these interventions, tools, methods and strategies
6. Create better linkages between the stakeholders in each process, to enable reinforcement, feedback and evolution of the interventions and capture additional evidence
7. Create continuous measures of effectiveness of these interventions

These seven steps won’t work in every industry or environment, for reasons I’ve written about elsewhere (best available evidence, like ‘best practices’, only applies in situations where many people are doing, at least some of the time, very similar activities, like diagnosing or treating specific diseases). And Davis is pragmatic — he sees the value of intuition and personal judgement sometimes overriding what best available evidence might suggest is appropriate, in specific situations, as long as the best available evidence has at least been considered.

I can see this approach working in quite a few areas, at least by analogy, and I’m already at work seeing if it will apply in the context of my current work project. If it could save some of the victims of knowledge management failure, people like Vanessa Young, it deserves serious study and consideration. In one hour Dr Davis managed to change my perceptions about what KM can and cannotachieve. Very impressive stuff.