International Journal for Quality in Health Care 16:263 (2004)
International Journal for Quality in Health Care vol. 16 no. 3 © International Society for Quality in Health Care and Oxford University Press 2004; all rights reserved
Letter to the Editor |
There is nothing wrong with the concept of a root cause
To the Editor: Dr Donald Berwick, president and CEO of the Institute of Healthcare Improvement, recently discussed in the Washington Post the failure of hospitals to reduce medical errors. There is much to agree with in this editorial [1]. However, one cause offered by Dr Berwick as to why hospital personnel fail to reduce medical errors was:‘They believe that analyzing errors will allow them to find a single root cause, even though in fact the very idea of a "root cause" is misleading. Most system failures come from complex interactions between latent failures (the little things that go wrong all the time) and specific actions. Saying that one factor is a "root cause" is usually an illusion created by hindsight bias. It is Monday morning quarterbacking.’
This view of error analysis is misleading. Techniques such as FRACAS (Failure Review and Corrective Action System) and RCA (Root Cause Analysis) have been used successfully for years in reliability engineering to improve quality [2,3]. The notion of a "root cause" is central to the FRACAS or RCA concept. There is no requirement that a root cause be either single or simple. Moreover, since these tools are used to analyze observed events, by definition they involve retrospective analysis, which is a less pejorative term than "Monday morning quarterbacking".
FRACAS is a closed loop process that entails: (i) establishing a quality goal; (ii) measuring the error rate; (iii) analyzing errors to determine their root cause; and (iv) proposing and implementing corrective actions.
Errors do not have to cause harm. For example, in air safety the Federal Aviation Administration distinguishes between incidents and accidents [4]. With modeling in FRACAS [5], "the little things that go wrong all the time" count as error events and are weighted in the model. Ranking of events allows an efficient use of limited resources to reduce the error rate.
Berwick argues in favor of quality improvement demonstration projects, yet one wonders how effective they have been. For example, there have been demonstration-type projects for years for using electronic medical records, but progress (e.g. implementation) has been slow in this area. An example of dramatic quality improvement is the original laboratory assay automation introduced by the Auto Analyzer from Technicon Instruments (now Bayer) in the 1960s and "70s. New technology is often fueled by prospects of financial success; quality improvement is often only a by-product.
In between new technology breakthroughs, quality improvement tools such as FRACAS and RCA analysis are valuable. A poorly executed FRACAS is of course of no value; yet there is nothing wrong with the concept of a root cause.
Krouwer Consulting, Sherborn, MA, USA
References
- Berwick, DM. Invisible injuries. We need a better system for tracking and preventing medical errors [Editorial]. Washington Post, 29 July 2003: http://www.ihi.org Accessed 23 September 2003.
- O’Connor, PD. Test Engineering: A Concise Guide to Cost-effective Design, Development and Manufacture, chapter 12. New York: Wiley, 2001.
- Krouwer, JS. Assay Development and Evaluation. A Manufacturer’s Perspective. Washington, DC: AACC Press, 2002: pp. 59–67.
- Federal, Aviation Administration. Aircraft Accident and Incident Notification, Investigation, and Reporting. http://www2.faa.gov/avr/aai/Chap1.htm Accessed 2 October 2003.
- Krouwer, JS. Using a learning curve approach to reduce laboratory error. Accred Qual Assur 2002; 7: 461–467.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||