I've Been "Invited" To A Root Cause Analysis, Now What?

Identifying the causes of errors can help prevent similar errors from happening in the future. The primary objective of the article is to give nurses some idea of what to expect in a Root Cause Analysis experience. Nurses Safety Article


This article was reviewed and fact-checked by our Editorial Team.
I've Been "Invited" To A Root Cause Analysis, Now What?

Dealing With Error

I have written about error in Why do we continue to harm patients? In that article, I give statistics about the number of people who are killed while in the hands of trusted healthcare professionals. The Joint Commission has mandated Root Cause Analysis (RCA) since 1996 for patient death, or serious harm from error (sentinel events).1 Some hospitals also voluntarily use RCA for errors that reach the patient, but don't cause harm; and even for near misses -- where error has the potential to occur. RCA is based on a tool used by the airline industry, nuclear power, and chemical engineering to reduce accidents and improve safety for consumers and workers.

What is Root Cause Analysis?

RCA is the retrospective analysis of errors; it happens after the fact, but is an incredible tool for future prevention. The goal of RCA is to identify all of the factors involved in an error, with the intent to create actions to prevent future error. Two important characteristics of RCA should emerge with its use; underlying causes are fixable and the problem is uncovered within a reasonable amount of time using a reasonable amount of resources. RCA involves creating three "discoverables" - things you have to turn in to The Joint Commission: a timeline, a causal tree, and an action plan. These items must be filed within 45 days of discovery of a sentinel event, so it's important to be prepared to step into action.

The first step in RCA is to gather a team, and investigate. Creating a team can be the most difficult part. The Director Of Nursing wants to be present, and the Chief Medical Officer likes to be there, and many times you need a pharmacist, and the charge nurse, and the nurses involved. Some RCA methodologies recommend having an RCA team already in place to expedite the process. I recommend using schedule a meeting scheduler like doodle2 to avoids long chains of emails and miscommunications.

Get Bedside Nurses to the Table

Getting the nurses involved in the meeting can be a big challenge, often due to the lack of information out there about the process and goals of RCA. So often they assume they are "in trouble" - and why not? The news is full of the typical response to nurse error: blame and fire the nurse. Nurses are often full of guilt over the error, and terrified of loss of employment and of being labeled incompetent. I once had a nurse tell me, "I was terrified I would enter the room and everyone would show me all the reasons why it was my fault."

Reduce fear and anxiety with information: RCA is about problems, not people (and have them read this article!) It may help to share that participation in RCA can have meaningful benefits. Here is a sample of what nurses have said to me after participating: "This was a transformative experience!", "I am so glad I came. It helped so much with my healing to be part of the process of making sure this never happens again", "I clearly saw my role in the error, but so many other things that went wrong to allow it to happen -- the things we always complain about -- I saw upper management listening and making plans to address those issues."

That is what makes RCA so special to me - that an interdisciplinary experience can be such a game changer. As nurses we beat ourselves up on a daily basis, chasing unattainable perfection. When we enter a room of peers that include pharmacists and physicians, and we hear them acknowledge the systems factors that contributed to the error, and then we see them take action to correct problematic factors, that can make all the difference. It can change a nurse from wanting to quit, to wanting to contribute.

RCA Process

When I ran RCA meetings, they took place in conference rooms with access to electronic records, since there was always something else we needed to look up. I sent out the timeline by email, and also handed it out at the meeting. After explaining the process of RCA, we would get into creating a causal tree. I always cautioned team members to resist jumping to obvious conclusions. We used something called the 5 whys to stimulate creativity and out of box thinking.3

We would cover a wall with sticky notes (they are great because you can move them around, and easily change them). The goal was to create a tree showing all the possible causes of an error. I am including one I created for an error that was published in 2002 - a woman died after experiencing unexpected hypoglycemia after her IV was flushed with insulin instead of heparin.4 For that RCA, it was important to choose a good starting point (called the problem statement) to maximize the number of possible causes. For example, in the causal tree below (which I created from information from the article), if the statement that "insulin was used to flush an arterial line" had been used to start the tree, factors concerning the actual medication event would be the only root causes. By using the statement "there was delayed discovery of unexpected hypoglycemia", additional factors can be brought to light, such as the timing of labs, and the protocol for hypoglycemia in the ICU.

Analysis tree chart for nurses dealing with hypoglycemia

Making a causal tree can take a few hours, and typically the construction of the action items is scheduled for a follow up meeting. Recent criticisms of RCA have focused on the creation of action items. Critics have argued that often the action items are weak, and little follow up occurs to determine outcomes. A good action item chart includes at least one, strong, actionable item for each branch of the causal tree. The chart should also include outcomes measures, a time frame for completion and an owner - the person responsible for reporting out about the item. Below are some action items I came up with from the data from the article. Strong items, like prohibiting multi-dose vials, are those that don't rely on individuals.

This table is better viewed on desktop devices...

No Key issues to address (root causes on a causal tree) Recommendation
1 Timely diagnosis of hypoglycemia crucial Standard tools for differential diagnosis
2 Lab test took 30 minutes Bedside glucose test (not acted upon - issues with reliability discovered with quality control tests, too expensive)
3 Hypoglycemia not a first suspect in emergent ICU situations Checklists for emergencies
4 Routine practice in ICU to use multi-dose vials, insulin and heparin vials look similar Use of multi-dose vials of insulin and heparin prohibited
5 Insulin usually kept in the refrigerator Insulin added to an automated dispensing device
6 Insulin often left on med cart Staff educated to keep medications in authorized places
7 Standard practice to leave heparin on med cart Nurses reminded to keep med carts locked
8 Entering locked medication cart multiple times daily is inconvenient Prepare insulin in pharmacy
9 Busy unit with frequent emergencies, high acuity patients, certain medication need to be on hand Interdisciplinary team to examine how to expedite delivery to medications to patients
10 Standard is to flush occluded line with heparin from multi-dose vial Use of saline flushes to restore patency of arterial lines required
11 Nurse didn't check vial label with med sheet Barcoding
12 No system of second checks in place Institute independent second checks

There are multiple resources for how to do an RCA. The VA has used RCA the most consistently and has records for over 4000 RCAs (if only I could get access to that dat base!)5. The Agency for Healthcare Quality and Research6 recently published revised guidelines for RCA, called RCA2 - these new guidelines are controversial because they recommend not including the people involved in the error in the actual RCA.

Does RCA Work?

Criticisms of RCA include weak leadership and poor training.7 If those on the team are not well trained and not invested in the process, the likelihood for a positive outcome is low. There are several important elements for an effective RCA: a) strong leadership and facilitation; b) interdisciplinary approach; c) those who participated get to tell their stories; d) frontline workers are educated on the process.

A huge criticism of RCA is the lack of data supporting that it actually works.7 There are two reasons for this: a silo mentality and documentation. Healthcare organizations don't routinely share data with one another, especially data surrounding sentinel events. Within healthcare organizations, we see this same problem: adult day-care, outpatient services, ambulatory clinics, and inpatient often don't share the results of RCA activities.

How often have you as a working nurse been told the results of an RCA? Why don't we routinely share these results with frontline workers, other units, and other organizations? Learning from other the mistakes of other healthcare professionals should be part of our daily communication, instead of something that is hushed up and hidden.

Other barriers include time and money. It's so important to do an RCA for each and every event, but healthcare organizations often don't see the financial case for spending money up front, even if it means saving money later on lawsuits that didn't happen due to patient harm that didn't occur.

Please advocate for the use of RCA in your place of work. Consider bringing in a highly trained professional to teach staff how to do RCA, using consistent methodologies for each case. Collect and aggregate data from the RCA, and make a point of publishing follow-up to frontline staff. Celebrate success and positive change. Reward those who are brave enough to step forward and report an error with the incredible feeling of knowing that their report made a difference - that their report may have saved a life.


1. Joint Commission on Accreditation of Healthcare Organizations. (2009). Sentinel event alert: Leadership committed to safety. Retrieved from http://www.jointcommission.org/assets/1/18/SEA_43.PDF

2. Doodle

3. 5 Whys - Wikipedia

4. Bates, D. W. (2002). Unexpected hypoglycemia in a critically ill patient. Annals of Internal Medicine, 137(2), 110-115.

5. The Department of Veterans Affairs (VA) National Center for Patient Safety (2015). Root Cause Analysis. Retrieved from Root Cause Analysis - VA National Center for Patient Safety

6. Agency for Healthcare Research and Quality. (2014). Root Cause Analysis. Retrieved from Root Cause Analysis | AHRQ Patient Safety Network analysis

7. Peerally, M., Carr, S., Waring, J. & Dixon-Woods, M. (2016). The problem with root cause analysis. BMJ Quality & Safety, Retrieved from: The problem with root cause analysis | BMJ Quality & Safety



Patient Safety Columnist / Educator

Dr. Kristi Miller works in Oncology, Home Health, and Patient Safety.

60 Articles   529 Posts

Share this post

Share on other sites
Specializes in Psych, Peds, Education, Infection Control.

Great article, Safety Nurse! I've been to a RCA or two in my time (comes with my position), and they can be incredibly informative. Yes, they can absolutely go awry if an administration is looking to blame vs educate, but in most cases, it's a really excellent learning tool. Obviously, there are some situations that are sheer gross negligence, but I find most RCAs discover that people are doing the best they can with the resources they have and it's a process issue. It's definitely a great chance to see what went wrong AND what went right.

Specializes in Education, Informatics, Patient Safety.
Great article, Safety Nurse! I've been to a RCA or two in my time (comes with my position), and they can be incredibly informative. Yes, they can absolutely go awry if an administration is looking to blame vs educate, but in most cases, it's a really excellent learning tool. Obviously, there are some situations that are sheer gross negligence, but I find most RCAs discover that people are doing the best they can with the resources they have and it's a process issue. It's definitely a great chance to see what went wrong AND what went right.

I am so glad you had a positive experience - thank you so much for sharing and for reading.