Prevent accidents from happening again
WC trust fund mandates analysis to cut injuries
"What could have been done to prevent this accident?" That’s a standard question posed after an employee suffers an injury. But at the 35 hospitals of the Palmetto Hospital Trust (PHT) in Columbia, SC, workers’ compensation self-insurance program, there’s a new focus on finding the answer — the "root cause" of serious accidents.
PHT is requiring member hospitals to conduct a root-cause analysis of "serious debilitating injuries," including occupational exposure to airborne infections such as tuberculosis or severe acute respiratory syndrome (SARS) that results in transmission, blood or body fluid exposure that leads to seroconversion, or injuries related to violence that require police intervention. PHT also is strongly advising hospitals to conduct a root-cause analysis of leading injuries as a way to reduce injury rates.
Hospitals typically have conducted such analyses related to patient safety, notes Larry Gray, AIC, senior vice president of PHT Services Ltd., the program administrator for PHT.
It only makes sense to extend the activity to worker safety, he says.
For PHT, the new requirement evolved after a health care worker became ill following exposure to a patient with meningitis. The board of directors — CEOs and senior managers of the member hospitals — agreed to add new minimum requirements. "If we truly want to reduce the frequency and severity of the injuries and correct the problems, for the financial security of the Trust and make a safe work environment for employees, why shouldn’t we do this?" Gray asks.
Looking beyond the blame game
Root-cause analysis means looking at a chain of events and digging beyond the obvious error that triggered an accident or near miss. It does not, as some employees fear, involve seeking someone to blame.
"Discipline is totally separate from root-cause analysis," emphasizes Craig Clapper, PE, CQM, MBA, partner and chief operating officer of Performance Improvement International in San Clemente, CA, who provided training for PHT member hospitals.
"Root-cause analysis is all about the system," he says. "The system led to the error; let’s change the system."
Here’s an example: A technologist in the nuclear medicine department is mounting a collimator on a camera to perform a procedure. She has tightened only two of the four bolts when the phone rings. Distracted, she forgets to bolt the other two. When she turns the camera, the collimator crashes, crushing her fingers and barely missing the patient’s head.
The obvious error is a lack of attention. But the overriding error, Clapper says, is the lack of an interlocking mechanism that would have prevented the operation of the camera if the bolts weren’t fully latched.
The hospital could work with the manufacturer to add a safety feature or could add a process of checking the bolts before activating the camera, he says.
"Most events aren’t caused by simple human errors," he says. "They’re caused by errors that normally occur in the system, but the system then allows them to become an event."
Clapper recommends setting up a multidisciplinary team to look for the root cause of accidents. That is the approach taken by Spartanburg (SC) Regional Health Care System. The team includes employee health, risk management, industrial hygiene/safety, infection control, and other managers. The health system has looked at needlestick incidents to improve both safety devices and work practices.
"Everybody has to have a buy-in on safety," says Denise Hollis, RN, director of employee occupational health. "Everybody has to want to do it. No one person can do it alone. No one department can do it alone."
Steps toward finding the root cause
Clapper identifies seven basic steps in root-cause analysis:
1. Define the problem.
Look not just for the human error, but the underlying causal factors, he says. It could be a process, a piece of equipment, an organizational element, or a leadership issue, he says.
Don’t settle for the easy answer — too little money or too few staff, Clapper says. "The one thing you’ll never get is more people or more money, so you might as well quit complaining about those two," he advises.
2. Collect data and process them into facts.
Brainstorming alone probably will not help you find solutions, he says. "We ask people to be less problem solving-based and more investigation-based, where they have actual facts from the case and they’re looking at lists of possible causes. That gives a better result," he says. You will look at physical evidence, such as the scene of the accident. You will interview employees and supervisors and review policies and other documents.
3. Determine failure modes.
Look for areas of weakness that involve individuals, processes, organizational structure, and management. Ask the question, "Why did this happen now and not before? What are the causal factors that led to the injury event?"
4. Construct failure scenarios.
Determine the sequence of events that led to the accident. You may want to diagram what happened, with chains linking the causal factors that worked together to lead to the injury. You should still be asking the question, "Why?" to make sure you are capturing all causal factors.
5. Validate findings through benchmarking.
Look for data or literature related to your injury. In some cases, you may find that another facility has addressed an identical problem and identified solutions.
6. Develop "root solutions."
Corrective action is the key to this process. "You’re really not in the root-cause analysis business. You’re in the root-solution business," Clapper says.
7. Monitor for effectiveness.
Root-cause analysis is a version of quality improvement. You should use the plan-do-check-act cycle. In other words, you may want to implement changes (root solutions) as a pilot project in one unit then verify that corrective actions are working. For example, if you are seeking a behavior change, such as checking that the camera bolts are tightened before starting a procedure, then you monitor whether employees are taking that extra step.
"Most corrective actions don’t work as well as people thought when they recommended them," cautions Clapper. "If it was important enough to do root cause, it’s important enough to check and see if it worked."
On a long-term basis, you should monitor your injury rates to make sure that your root-cause program is having an impact.
"You have to view root cause as part of a balanced diet in improvement," he says.
[Editor’s note: For more information on root-cause analysis, contact Amber Driscoll, Practice Coordinator of The Greeley Co., Marblehead, MA. Telephone: (781) 639-8030.]
Identify Root Causes
1. If the identified condition is corrected or removed, would the event of interest be prevented/avoided?
2. Could the identified condition be corrected/ removed without creating another undesirable situation?
3. Is the identified condition able to be corrected in a cost-effective manner?
4. Is the identified condition within management’s control to correct?
5. Is the identified condition or performance substandard when compared to similar conditions in other locations/organizations? In other words, is the identified condition or barrier designed to prevent the event of interest with a high degree of reliability?
If all five conditions are satisfied (i.e., YES answers), then the identified condition should be considered a root cause. If any answer is NO, then additional analysis is needed to identify the true root cause(s).
Source: The Greeley Co. and Performance Improvement International, Marblehead, MA. Copyright 2003.