March / April 2008
Fix Root Causes with Closed-Loop Corrective Action
Dekker and Laursen (2007) show that many healthcare organizations still regard mistakes and near misses as "violations" or "errors" by personnel. When so-called corrective action consists of disciplinary action, retraining, and admonitions to "be more careful," it is hardly surprising that healthcare personnel do everything possible to avoid reporting problems and that the problems' root causes therefore go uncorrected. If management finds someone to blame for a system-related problem, its underlying cause still goes uncorrected. Responsible organizations, in contrast, encourage workers to report mistakes and near misses. That results in closed-loop corrective action (CLCA), which prevents the problem from recurring.
Bad Systems Beat Good Employees
A focus on disciplinary action and retraining demonstrates a lack of familiarity with W. Edwards Deming's 85/15 rule: 85% of the trouble in any work environment comes from the system in which people must work, while 15% is due to carelessness, negligence, and similar employee deficiencies. Juran and Gryna (1988, 17.4-17.5) quote an 80:20 ratio, with 80% of the errors being "management controllable" and only 20% being "worker controllable." Disciplinary action and retraining might therefore be appropriate in one out of five cases at most. In the others, management must look to its own errors or omissions for the problem's real source.
Andell (2007), while discussing excessive reliance on computerized data, states that "a bad system beats a good employee every time." Furthermore, cultures of blame almost universally discourage the reporting of bad news. In contrast, Japanese workers initiate closed-loop corrective action (CLCA) through a hiyari or "scare report" that describes a quality or worker safety incident that could have happened. Based on the hiyari, countermeasures are put into place to prevent it. The Inoue Hospital in Osaka, Japan, implemented similar procedures more than 20 years ago (Imai, 1997, 273-276).
Closed-Loop Corrective Action
CLCA begins when (1) a quality problem like a medical error takes place or (2) a near miss is forestalled only by human vigilance. The hiyari or "scare report" is the means for reporting near misses.
TAM [Toyota Astra Motors] has two special programs directed at anticipating problems in advance. One is called the hiyari (scare) report, and the other is called the quality hiyari, or kiken-yochi training anticipating danger in advance, report. The hiyari (scare) report points out unsafe conditions or actions that could eventually lead to problems in the workplace...
...management rewards the efforts of employees to anticipate or detect problems in advance and solve them before they become a reality. Management regards such an approach to problem solving as more valuable than addressing the problem after it has become a reality (Imai, 1997, 142-143).
Suppose that a nurse receives a verbal instruction from a doctor to administer what sounds like an overdose of a medication, and that only the nurse's knowledge of the usual dosage range saves the patient from death or injury. In a poorly managed hospital, the doctor's and nurse's inclination would be to keep the problem quiet because administrative action would focus on whom to blame, and it would probably not even address the underlying system deficiency that made the error possible. In a well-managed hospital, neither the nurse nor the doctor would hesitate to file a hiyari. This would lead to identification of what to blame (such as medications with similar-sounding names, lack of a verbal order read back procedure, and so on) followed by permanent countermeasures to prevent anyone from making the same mistake in the future.
As another example, suppose that healthcare workers are not cleaning their hands between patients. Willful or negligent failure to comply with an established procedure is technically cause for disciplinary action. On the other hand, noncompliance might easily be symptomatic of an impediment to job performance, such as the need to walk to a hand-cleaning station. This impediment's removal is management's responsibility. The Ford Motor Company discovered in the 1920s that workers would not throw trash on the floor if a waste container was within seven steps of any position in the factory. Hospitals are now discovering that doctors and nurses will clean their hands routinely if cleaning stations are readily available.
All CLCA procedures are variants of the basic PDCA (Plan, Do, Check, Act) quality improvement cycle. It does not really matter which one an organization uses, as long as it uses it diligently and effectively. Six Sigma's DMAIC (Define, Measure, Analyze, Improve, Control) is, however, more oriented toward numerical measurements and statistics. Ford Motor Company's TOPS-8D (Team Oriented Problem Solving, Eight Disciplines) is a very simple CLCA method, while the Automotive Industry Action Group's (2006) Effective Problem Solving Process is another. I recommend one of the latter for organizations that are not already committed to DMAIC.
8D Problem-Solving Process
I summarize the eight steps or disciplines as follows:
1. Form an appropriate cross-functional team with a champion (the person with the resources and authority to implement the selected solution); team leader; subject matter experts like doctors, nurses, laboratory personnel, and pharmacists; and a facilitator. Don't forget the patient, who can provide the customer's first-hand perspective.
2. Define the problem. This problem definition becomes part of the project charter.
3. Contain the problem to protect patients from its consequences.<
- Containment addresses only the problem's symptoms. It does not remove the underlying cause. When organizations stop with containment, e.g. by making things right with the affected customer, they leave the door open to the problem's recurrence.
4. Identify the problem's root cause. The traditional cause and effect diagram (fishbone diagram, Ishikawa diagram) has six categories of potential root causes. The acronym 6M, or "5Ms and E," makes the categories easy to remember.
- Machine (equipment). Lack of preventive maintenance, unsuitable equipment, and defective equipment falls into this category. In a hospital, equipment that is hard to disinfect or sterilize would be a Machine problem. The same goes for equipment that requires a nurse's vigilance to ensure that a malfunction does not harm a patient. Jidoka (autonomation) means that equipment can detect and react to abnormal conditions, e.g. by sounding an alarm.
- Method (the procedure for doing the job). Deficient procedures are the root causes of many medical errors and malpractice cases.
- Manpower (personnel; the gender-specific term permits the acronym "5Ms and E" or "6M"). Inadequate training, or assignment of personnel to jobs for which they are not adequately trained, has been cited as the cause of numerous medical tragedies. The ISO 9001:2000 standard for quality management systems asks specifically how the organization ensures that people are qualified for the tasks to which they are assigned.
- Measurements. This includes gauges or, in a hospital setting, tests, and diagnostic equipment.
- Materials. In a hospital, this category includes consumable items like medications, dressings, antiseptics, and blood products.
- Environment (or Medium). This refers to the surroundings in which the work takes place, and temperature and humidity are usually key considerations. In healthcare, bacteria are yet another issue. The new Royal Liverpool Hospital will reputedly have silver (probably silver plated) or copper door handles (Bartlett, 2007). Silver forms a thin oxide layer that is fatal to bacteria, thus preventing transmission from one healthcare worker's hands to another's.
5. Select a permanent corrective action.
6. Implement the proposed solution, and verify its effectiveness.
7. Standardize the change.
- Standardization means an update in written procedures to make sure that everybody follows the new procedure. ISO 9001:2000 requires a controlled documentation system for doing this, as well as closed-loop confirmation that everybody who does the job in question is aware of the changes.
- Share the improvement with related activities. This is best practice deployment; the concept has been around for more than eight decades. There is a very strong case for some kind of keyword searchable national database in which healthcare providers can share CLCA projects with each other. In addition, the Agency for Healthcare Research and Quality, Leapfrog Group, Institute for Safe Medication Practices, and Institute for Healthcare Improvement (IHI) have best-practice recommendations that any hospital can implement.
8. Recognize the team's achievement.
The AIAG's Effective Problem Solving Process is similar, and it uses the same kinds of problem solving techniques. Furthermore, many vendors offer web-based software that both facilitates and documents the CLCA process. The key point is, however, that every medical error and near-miss (as reported in a hiyari) requires closed-loop corrective action to prevent recurrence.
Every medical error and near miss should result in not only containment (action to protect patients from the consequences) but also closed-loop corrective action to remove the root causes that allowed that error to occur. The hiyari or "scare report" is an established means for reporting near misses that were averted only through worker vigilance. A culture of blame contravenes these quality improvement best practices, and often ensures that the real cause of the trouble will go uncorrected.
William Levinson is the principal of Levinson Productivity Systems, P.C., in Wilkes-Barre, Pennsylvania. He is an ASQ Certified quality engineer, quality auditor, quality manager, reliability engineer, and Six Sigma Black Belt. Levinson may be contacted at firstname.lastname@example.org.
Andell, J. L. (2007). Data-driven decision-making and organizational excellence. American Society for Quality, World Conference on Quality Improvement. Orlando, April-May 2007.
Automotive Industry Action Group. 2006. CQI10, effective problem solving: A guideline. Southfield, MI: Author.
Bartlett, D. (2007, July 24). Blueprint for new Royal Liverpool Hospital unveiled. Liverpool Daily Post. Available at, http://www.liverpooldailypost.co.uk/liverpool-news/regional-news/2007/
Dekker, S., & Laursen, T. (2007). From punitive action to confidential reporting. Patient Safety & Quality Healthcare, 4(5).
Imai, M. (1997). Gemba kaizen: A commonsense, low-cost approach to management. New York: McGraw-Hill
Juran, J., & Gryna, F. (1988). Juran's quality control handbook (4th ed.). New York: McGraw-Hill, 17.4-17.5