Reliability or Dependability
Webster’s dictionary uses the word dependable as the first definition of reliable, but the second definition as “giving the same result on successive trials.”  Dependable is defined as capable of being trusted or reliable. Perhaps that is why I have usually thought of dependable as applying to people rather than processes. It occurred to me I had seen many articles discussing reliability in healthcare, but not dependability, so I did a Google search. This turned up many references to the reports from the Institute for Healthcare Improvement (IHI), which has been discussing building highly reliable healthcare systems for many years.
In a 2018 summary their framework for safe and reliable care was described as consisting of an appropriate organizational culture, leadership, and a learning system. The four components of a learning system were reliability, transparency, continuous learning, improvement and measurement. Reliability was further defined as “applying best evidence and minimizing unnecessary variation with the goal of failure free operation over time.” I believe most healthcare professionals are motivated toward this goal and most of the arguments revolve around what evidence is most relevant (and thus reliable.) But the measurement component has been distorted by the government, which has devised a zero-sum payment scheme that does not reward measurable improvement, but only attainment relative to other players.
The same Google search turned up several references showing computer engineers talk about computer systems using reliability and dependability to mean somewhat different things. Dependable systems are “trustworthy” and reflect reliability, availability, safety, confidentiality, survivability, integrity, and maintainability. Reliability is defined as “the system behaves as expected with very few errors.” Integrity is defined as “system data should not be modified without intent and authorization.” The authors point out dependability must be measurable, which poses challenges even in computer systems. To computer system engineers, then, dependability is a more overarching concept which subsumes reliability. But the more significant distinction may be the word “trustworthy.”
Presently our culture demonstrates a low level of trust and many institutions are deemed untrustworthy. Where this will lead is unknown, but Brooks posits the younger generation will value “not liberation, but security; not freedom, but equality; not individualism, but the safety of the collective; not sink-or-swim meritocracy, but promotion on the basis of social justice.” Given these currents in the larger society, should healthcare organizations strive for reliability or dependability, assuming the latter implies some element of being trustworthy. Or is this a false dichotomy? I have had the privilege of working with many people who were dependable—I could trust them to show up, do their best, and help me take care of the patient. And I trust they found me dependable as well. Isn’t this what we all want from our care-givers? It really isn’t an either/or question, we need both. But I think there is an important distinction.
The question really is can we build reliable systems without dependable people? I think not. Further, can we retain dependable people if our systems are not dependable? Again, I think the answer is no. The surge in burnout and the high staff turnover may be signs that our systems are not dependable in the mind of those who work within them. Patients don’t really have much choice—it is sort of like commercial air travel. I really have no choice but to trust the mechanics, pilots, and engineers and others who work to make it safe.
In the early days of continuous quality improvement, I used to argue that we should do this because I wanted to provide excellent, not average (mediocre?) care. Then, realizing the nature of the reimbursement game, I began to argue the goal was indeed to be average, since no one could consistently beat the averages. Both doctors and health care executives are highly educated people, which means they have all had to compete for entry into their educational track by getting “A’s” on their required courses. Somehow, striving for a “C” seems counter-intuitive. But CQI requires a permanent dissatisfaction with the status quo, which is hard to maintain. In the current system, striving for excellence is both unrealistic and likely to produce many unintended consequences. On the other hand, reliability may be too restricted a goal. Clinicians are not robots nor are patients and delivery of medical care is more than the sum of its component industrial systems. Striving to be robotic and mechanical won’t work either. So, I propose we think in terms of being dependable, meaning both reliable and trustworthy.
How can we start? The challenge for all leaders is to understand while they can specify the boundaries of acceptable performance, they cannot micromanage without destroying trust. A leader has to trust the staff to do the right thing, but the staff have to trust the leader to support the staff even when doing the right thing requires exceptions to the rule. To become trustworthy to our public, we must first become trustworthy to each other. This means all parties must actively listen to each other, respectfully communicate, mean what we say and say what we mean, and prove that we can be dependable colleagues. Then maybe we can restore public trust in our institutions and achieve the goal of being dependable sources of health care. A tall order? Perhaps, but I remember days when this was a general goal. We did not agree about everything, of course, but we could stay in communication and disagree until we worked to resolve the issues at hand.
Maybe the real challenge is bigness. Trust is difficult to establish in big organizations which are, by their nature, impersonal. Working closely with a group of people, where you get to know each other, seems a necessary pre-condition to build trust. This logic suggests we need to deliberately break our big organizations into smaller, more collegial (and functional) subunits. All the trends presently seem to be going in the other direction.
6 December 2020
 https://www.merriam-webster.com/dictionary/reliable. Accessed 25 November 2020.
 https://www.merriam-webster.com/dictionary/dependable. Accessed 25 November 2020.
 http://www.ihi.org/resources/Pages/IHIWhitePapers/Framework-Safe-Reliable-Effective-Care.aspx. Accessed 25 November 2020.
 Frankel A, Leonard M. A Framework for High-Reliability Organizations in Healthcare. 30 November 2018. Accessed at https://healthcatalyst.com/insights/high-reliability-organizations-in-healthcare 25 November 2020.
 The other challenge is how to implement your learning system. The tendency is to set up bureaucratic regimes, which, as I have argued elsewhere, are one of the things making our system both less resilient and perhaps less reliable.
 Khan MF, Paul RA. Dependable and Secure Systems Engineering. Advances in Computers, 2012. Accessed at https://www.sciencedirect.com/topics/computer-science/dependable-system 25 November 2020.
 Brooks D. America is Having a Moral Convulsion. The Atlantic. 5 October 2020. Accessed 6 October 2020 at https://www.theatlantic.com/ideas/archive/2020/10/collasping-levels-trust.
Answering Strategic Questions
Answering strategic questions is hard work and also uncertain, but these are not reasons to avoid asking them.
Engaging Burned Out Physicians
Horizontal Violence and Nursing Staff Turnover
A recent study shows horizontal violence - conflict between nurses in a hospital - is common and a major cause of job dissatisfaction and intention to leave. What can be done about it?
More on Turnover from the Departing Staff Point of View
Building good teams is hard work. What does it take to be successful?