Do no harm

A recent CPC in the New Englad Journal of Medicine (November 11, 2010) on wrong-site surgery (wrist instead of one finger) and wrong procedure (carpal tunnel release instead of trigger-finger release) has an excellent discussion on causes of human error, disclosure of errors and caring for the caregivers. True to the tradition of the series, there is an excellent list of references on these topics. I would like to draw attention to two of those articles.

The discussion in the NEJM points out how error can occur due to breakdown in skill-based behavior or rule-based behavior or knowledge-based behavior. Skill-based behaviors involve routine tasks. Rule-based behaviors involve algorithms and check-lists. Knowledge based behaviors require “conscious problem solving and application of standardized knowledge to novel and unexpected situations”. Obviously the causes will be different for each of these areas of skill.

The first reference article reviews these causes and models of human error (Reason J. Human error: models and management BMJ 2000; 320: 768-770). This excellent review points out how one of the commonest responses to an error is to blame the involved individual. This is the Person approach model of error of causation and we all know that “blaming individuals is emotionally more satisfying than targeting institutions”.

In the System Approach, we concede the fact that humans are fallible. “We cannot change the human condition” but “we can change the conditions under which humans work”. The emphasis is not who made the mistake but asking whether steps were taken to minimize errors and if so where and how they broke down.

We are told that in plane accidents, the pilot is not singled out except when he/she was under the influence of mind-altering substances or knowingly violated the safety protocol. Based on studies conducted in US Navy Nuclear aircraft carriers, Nuclear Power plants and air traffic control centers with heavy work-loads and demanding activities but with low adverse incident rates, the system approach lays emphasis on defenses, barriers and safeguards. Even with these safeguards, errors can and do occur because of unfortunate alignment of errors in different areas. This is called the “swiss cheese model” of harm creation.

The second, more theoretical article on “The role of error in organizing behavior” by Rasmussen appeared in Ergonomics in 1990 (Ergonomics 1990; 33: 1185-1199). I found it a difficult read. However, Table 1 is very useful to understand the concept of “interaction between different levels of cognitive control”, namely knowledge-based domain, rule-based domain and skill-based domain”.