6 ways to improve your root cause analysis

Risk managers routinely use a root cause analysis (RCA) to determine the true source of an adverse outcome or other event, but are your RCAs as good as they could be? There is no single way to conduct an RCA, but one experienced investigator explains how you can improve your results by following a few tips and avoiding the most common mistakes.

No two RCAs are ever the same, says Kathryn Schulke, RN, principal with the consulting firm Booz Allen Hamilton in Baltimore, MD. Each one must be tailored to fit the circumstances of the incident being investigated and the people involved, Schulke says.

“Most risk managers could improve their RCAs,” she says. “I think a lot of people know that, but are unsure what might be lacking or how they can make their RCAs better.”

There is no one way to conduct an RCA, Schulke says. However, some RCA best practices apply to all situations, so Schulke offers these tips for improving your RCAs:

• Do not include in your RCA the parties who were involved in the adverse event. This is one of the most common mistakes that Schulke sees with an RCA. A better process is to interview those individuals one-on-one and capture the information they can provide, but do not include them in the RCA exercise or meeting. Instead, bring in people from their service line and other relevant areas.

“If that person is from dietary or pharmacy, for instance, you make sure you have people from those departments involved in the exercise, but not the person who was directly involved with the error,” Schulke explains. “People can’t help but be defensive and try to speak for themselves when they were part of the error, so it makes the exercise more objective when they are not involved in the actual RCA.”

• Frame your questions on facts, not a hypothesis. This distinction is important, Schulke says. Do not ask, “Was the lack of proper equipment a contributing factor to this error?” for example, because that is a hypothesis, and you could lead the person into answering yes. Instead, ask factual questions about what equipment was available, what was needed, and what was used. Those participating in the RCA, particularly those from risk management or quality improvement who might take the lead in the investigation, should be properly trained on this point.

• Involve physicians and hospital leadership. Too often, Schulke says, the RCA is conducted by other parties, and then the results are presented to the physicians and hospital leaders as a final conclusion. Involving those parties in the RCA will result in more substantial and sustained process improvements stemming from the RCA results, she says.

“The physician community is key to looking at some of these sentinel events and the serious harm that has occurred,” Schulke says. “Getting them involved early on, along with the C-suite at the hospital, is a superior way to conduct a root cause analysis.”

• Focus on the cause, not the problem. The cause relates to the steps that were taken, not taken, or missed, whereas the problem is the resulting adverse outcome. With a patient fall, for example, the problem is that the patient fell, but the cause is that the floor was wet.

• Use a multidisciplinary team for the RCA. Schulke cautions that many hospitals will build a multidisciplinary team related to the event in question, but they stop there. A pharmacy error, for example, might prompt a multidisciplinary team that includes representation from pharmacy, physicians, and nursing, because those are the fields involved in that incident.

“You really should involve people from areas that were not necessarily involved in that error, but who might have a perspective or point of view that is useful,” Schulke says. “Everyone brings a different perspective and a different history to the analysis, and that’s what you’re looking for in an RCA: insight that is not immediately apparent to those involved.”

• Build your timeline from documentation, not personal recollections. The medical record should be considered the reliable source of factual data such as time points. The statements of people involved most likely will contradict because memories are imperfect and sometimes because the parties involved have a vested interest in skewing information such as the timeline to their advantage.

“If it is in the medical record, you should be able to rely on that as the accurate source,” she says.


  • Kathryn Schulke, RN, Principal, Booz Allen Hamilton, Baltimore, MD. Telephone: (301) 825-7104. Email: schulke_kathryn@bah.com.