Root Cause Analysis Under Pressure

Added on Nov 6, 2015

A recent Washington Post article brought back memories of a fratricide investigation I conducted during Operation Iraqi Freedom. Through the parallels I see some learning points reinforcing our HPI Root Cause Analysis training and the ever challenging environment to sustain HRO culture. For those who don’t know me, I am a retired Air Force Major General, now with HPI applying my experience with high reliability military flying operations to our cherished healthcare partners.

The mentioned article addresses last year’s combat fatality of US Army Maj Gen Greene at the hands of an Afghan sniper, and the outbrief of the investigation to his family. Spoiler alert: the investigation found no one at fault, and that nothing could’ve been done to prevent the episode, where the family was expecting someone be held accountable. As an experienced military leader and trained aircraft mishap board president, I can try to understand specifics of the investigation from the article, but will assure you a disciplined process was followed that examined multiple aspects of the environment and chain of events…like our Swiss cheese model. The similarities with my 2003 experience start with a recognition that “expected” outcomes of investigations such as this color every element of its conduct, as well as how results are viewed by “invested” receivers.

The 2015 article: http://www.washingtonpost.com/sf/local/2015/08/19/completely-betrayed/

Patient Safety

My fratricide investigation was detailed in a couple of nationally recognized publications, mostly focusing on the absence of holding someone accountable for 3 friendly fatalities resulting from a US Air Force fighter dropping a bomb on a misidentified US Army rocket company, and the fact that I declined interviews (as I was told to do) in support of the stories. The frustration I encountered in my investigation, however, isn’t with the absence of the expected accountability outcome, but with predetermined conclusions and expectations that weighed heavily on the process…I am convinced similar pressures were at play in the Maj Gen Greene investigation. In my 2003 effort, as I attempted to research the Army’s policies, tactics, and parallel investigation efforts, I hit a brick wall of resistance from the Pentagon’s service leadership…this incident was to be coded as a combat-fatality, so no further Army investigation would be conducted. They could not stop an Air Force effort, and would offer minimum (if any) cooperation. All of my efforts to collaborate, or get their concurrence on my draft report, were thwarted.

I served much of my career as a leader in joint jobs, holding the highest respect for (and fairly deep understanding of) my sister services. I understand the volume of casualties the Army safety office deals with, and the status/respect a combat death earns as opposed to friendly fire (not to mention benefits afforded to the surviving family). But, as we in HPI see frequently in our clients’ approaches to Root Cause Analyses, if you don’t discipline yourselves to ask ALL the RIGHT questions, with a deliberate effort to avoid supporting predetermined conclusions, you may never uncover those key preventative elements to avoid reoccurrence.

I made reference above to predetermining results…in the Maj Gen Greene case, as is noted in the “We Were Exposed” section of the WP article, Danish colleagues had building security responsibility, but we didn’t want to “blame” a NATO ally from a diplomatic angle, nor would we blame the Afghan Army we were there protecting, even though it was their rogue-member who was the sniper. For my task, no matter what proof I offered the Army as evidence their soldiers were killed by an unintended friendly bomb from a US jet, the only “acceptable” label they would allow was “combat-fatality”.

How does this relate to us and our RCAs? When system or board level leadership, or executive sponsors of RCAs, interject influence to “protect” parties involved in the incident, we NEVER get the complete and honest story of what occurred, and more importantly we don’t get what can be done to prevent a similar event occurring in the future. As highly reliable as our military services are, I would contend these 2 investigations present evidence of a system focus on avoiding the assignment of blame, rather than determination of factors contributing to the incident. Equally guilty (granted, a harsh term for this group) are the Greene family members who wanted someone held accountable, as well as the journalists reviewing my event repeatedly quoting my report as recommending no one be disciplined.

Let’s get this to a list of reinforcing points for how Root Cause Analyses are/should be conducted in healthcare.

Hospital Safety

  • 1.) HROs conduct thorough, disciplined analyses of events in order to fully understand the total sequence of contributing factors

  • 2.) HROs need to continually reinforce that investigating a mishap has NO inherent intent to find fault. Almost contrarily, the more contributing factors that are uncovered in fact show how most times the person at the final action point (surgeon, nurse, pilot) was enabled by system gaps, or other knowledgeable professionals who could have prevented the event.

  • 3.) It is natural for environmental pressures to predetermine or attempt to influence the thorough examination of root cause. Peer groups, unions, supervisors, investors, lawyers, diplomats, loved ones…you name it, anyone who identifies with a group interested in conclusions regarding an event, either expect a certain conclusion, or feel it necessary to monitor/steer the process to ensure their expectations are met.

  • 4.) As all articles regarding these events credibly present, there are ALWAYS loved ones (or patients) who are devastated by the results of a preventable harm error…leadership owes them honest feedback about what occurred, and how this incident will be used to prevent something similar from ever happening again.

  • 5.) The press is a factor that cannot be ignored, both in military tragedy, and hospital error events. Avoiding engagement rarely works in the favor of leadership. Within privacy protections, recommend that leadership engage, releasing as much factual evidence as is practical. Always predict and acknowledge alternate viewpoints with some sound logic or factual rebuttal, but demonstrate that alternate conclusions were examined in depth.

  • 6.) When clear gaps exist in the obvious expectation of accountability, nothing positive comes from failure to address or dismissing the issue. This is what I see in play in the Maj Gen Greene investigation…leadership will not acknowledge who was responsible for security, nor admit that actions could have taken place that would have changed the tragic outcome.

  • 7.) Just imagine if Pentagon leadership would diplomatically acknowledge that we were relying on our Danish allies for security at that point, but we don’t find fault with their procedures and will work more closely in future environments to prevent the unexpected as best possible. AND if they acknowledge that our entire presence in Afghanistan is to afford an opportunity for these allies to live in peace with a legitimate representative government, but our risk in a land of different culture is that pockets of hatred continue to exist…we must stay vigilant to protect ourselves at all times against the potential for an attack like this. No “blame”, but clearly stating we know where/how this should be preventable in the future. Do you think the family would have more satisfaction or closure? Can our clients approach the aftermath of harm events with the same spirit?