The High Cost of (Not) Failing

Earlier this year a National Post article written by Sarah Boesveld on the role of failure in our children’s education served as “grist for the mill” during one of our early morning “Elgin Street Diner” think tank discussions.  Entitled “In Praise of Failure”, Ms. Boesveld’s opinion piece makes the case that we should view failure not as a roadblock or a blemish on a child’s record but, rather, as a key ingredient to success.  Perhaps, we wondered, a similar perspective applies to one of our favorite topics – innovation in health IT.   Is failure a necessary ingredient to successful innovation?

If asked to put your hand up to indicate whether you had implemented a wildly successful health IT project, would you?  We bet that if this question was asked at one of the e-Health conferences that few people, if they were honest, would be waving their hands.  Health IT projects, like any large-scale IT project, have an unfortunate propensity to disappoint.

Consider, for example, the National Health Service’s National Programme for Information Technology (NPfIT) in the U.K.  Established in 2002 to deliver electronic health records across the NHS, the NPfIT was axed in 2011 because, according to the Department of Health, it “has not and cannot deliver to its original intent.”   The final cost of the program is expected to exceed £10B while the benefits, based on Department of Health estimates, are projected to be £10.1B or roughly what was spent to implement the program.  Even this rather dismal result is questionable with the U.K. National Audit Office stating that there is “very considerable uncertainty about whether the forecast benefits will be realised.”

Closer to home the Canadian Firearms Registry is another example of the failure of a major IT project to deliver the expected benefits.  When the registry was first established in 1995, the net cost to the taxpayer was estimated to be $2M once registration fees were subtracted from the $119M cost to implement.  By 2005, the Auditor General estimated the cost to the taxpayer to be on the order $1B.

Large scale IT projects are not only difficult but take a long time to bring to fruition and, in most cases, are not very user friendly. Equally troubling, senior management often has an aversion to failure which compels them to put in place onerous processes in an attempt to gain control.  These processes prolong the procurement cycle, complicate specifications and lead to long implementation time lines.  The result is IT solutions that are two to three generations out of date by the time they are brought on line.

Consider the changes that we witnessed during the period the Canadian Firearms Registry was developed.  When it was first conceived neither Rogers nor Bell were offering Internet services.  By the time the system was in production,  Internet use was increasing at a rapid pace and both LinkedIn and Facebook had been established.

The less than stellar track record of large IT projects is driving a culture of timidness in our digitization agenda in Canada, one in which a Minister once advised us to “play error free ball” while, paradoxically, extolling the need for innovation.  Similarly, Deputy Ministers clamor about the need for innovation in our health care system.  The Council of the Federation even established an Innovation agenda!   Yet, despite their pleas that innovation will drive change, the actual implementation of the Innovation agenda has been an overly cautious endeavor.

We believe this aversion to failure is misplaced.  Without failure we relegate ourselves to middling solutions that fall far short of the needs of the end users.  Failure, according to Edward D. Hess, Professor of Business Administration and Batten Executive-in-Residence at the Darden Graduate School of Business, is a “necessary part of the innovation process.”  In a June 2012 guest post on the Forbes web site, Dr. Hess asserts that failure leads to “learning, iteration, adaptation, and the building of new conceptual and physical models through an iterative learning process” and claims that “almost all innovations are the result of prior learning from failures.”

The blame cannot be placed on the vendors. They are responding to what government and regional authorities ask for in their RFPs. It is a process that squeezes out any innovative dimension of a project by asking for solutions that are already in use at other organizations.  As John Halamka, physician, CIO and highly regarded healthcare IT thought leader so succinctly summarizes this situation, “Traditional procurement approaches are likely to acquire technology at the end of its lifecycle.”

James Dyson, celebrated entrepreneur and inventor of the radically different Dyson vacuum cleaner, claims “You can never learn from success, but you do learn from failure.”  Perhaps, instead of avoiding failure and sweeping it under the rug, we need to celebrate it.  This idea was explored at last May’s CHIEF event (a gathering of CIOs from across Canada) in the form of a Failures Conference suggested by one attendee.  The proposed conference would only be open to those people willing to admit to having an unsuccessful project implementation.  The agenda would be structured so that lessons learned from these unsuccessful projects could be shared.

While the idea of a Failures Conference was made somewhat tongue in cheek at the CHIEF event, similar ideas of have been floated at other events.   During a recent gathering of community leaders and entrepreneurs in York region outside Toronto, one participant suggested an awards banquet to celebrate failures.

A panelist at this same event noted that one of the best attended sessions at an Innovation conference he had attended earlier in the year involved three successful entrepreneurs sharing their most spectacular failures and the lessons that they learned from them.  Perhaps there is merit in a Failures Conference or, at the very least, a Failures track at the next eHealth conference.

Health care is perhaps the most complicated business sector in which to develop IT-based solutions.  In an Information Week article on the Clinical Query solution that he championed at Beth Israel Deaconess Medical Center, John Halamka noted that while there are only six data elements in an average ATM transaction, the average patient record contains about 65,000 data elements. To further complicate matters, individual clinicians may define each element differently.

This level of complexity does not lend itself to a well defined, detailed specification process. Instead, small scale projects that explore different approaches and concepts are better suited to this complex environment.  This approach involves deliberately taking risks and accepting that version 1 will probably be a failure. Iterative refinement and learning from failure will lead to solutions that best address end user needs.

Dr. Rick Irving recently commented in an article for Canadian Healthcare Technology that major innovations have been the result of bricolage – that is constant experimentation (sic failure). We like the term because it echoes advice from many commentators creating an eco-system that encourages innovation in the health care sector; keep it small, focused, short turn around time and as a constant learning process.

When asked about the apparent lack of results from his many failed experiments Thomas Edison famously replied, “Results! Why, man, I have gotten a lot of results! I know several thousand things that won’t work.”  While failure won’t always lead to innovation, creating an environment that embraces the results from failed endeavors and that celebrates those who make the attempt is a necessary condition for innovation.  Without failure there can be no innovation.  The cost of failure is really an investment in innovation.

share this article...
Share on FacebookShare on Google+Tweet about this on TwitterShare on LinkedIn