Patient Safety Tip of the Week
August 24, 2010
The BP Oil Spill – Analogies in Healthcare
We often like to use lessons learned from root cause analyses (RCA’s) done in other industries to demonstrate by analogy what can go wrong in various healthcare settings (see, for example, our Patient Safety Tips of the Week for August 28, 2007 “Lessons Learned from Transportation Accidents”, October 2, 2007 “Taking Off From the Wrong Runway”, May 19, 2009 “Learning from Tragedies”, and May 26, 2009 “Learning from Tragedies. Part II”).
So we’ve been itching to do a column on lessons learned from the BP Gulf oil disaster because they will undoubtedly have resemblances to serious adverse events in healthcare. There are multiple investigations of this accident ongoing (BP itself, DOI MMS, Congress, special committee appointed by President Obama, DOJ, etc.). Unfortunately, the root cause analysis has not yet been completed (or, if it has, not all details have been released). We also find that interesting and ironic. In healthcare, almost every state mandates that a thorough and credible RCA be performed within a very short time frame (usually 30-45 days) and most healthcare organizations perform them much sooner. Admittedly, some aspects may take longer for proper understanding and corrective action but in most cases we are interested in taking necessary steps to help prevent similar untoward incidents from occurring. Quite frankly, given the lack of transparency surrounding the BP spill, we’ll be surprised if the results of a thorough and credible RCA are ever released publicly.
However, there have been a series of reports in The Wall Street Journal and other conventional media on testimony in some of the above hearings and we are beginning to see Wikipedia and multiple blog sites releasing plausible root causes for the BP spill even in the absence of complete objective facts about that event. Most of what is out there is about events leading up to the disaster. Little has yet been published on root causes in the response to the event. But there are probably enough lessons learned to date for us to apply to healthcare.
The TapRoot® website has an excellent blog on the BP gulf oil spill that includes many preliminary facts and findings from BP’s interim incident investigation and from an independent investigation done by Dr. Robert Bea at UC, Berkeley. Wikipedia does a nice job on the chronology and background, plus provides many of the facts available to date. An excellent blog from safety expert Tom Krause focuses on failures of leadership, at both the private corporate and governmental levels, primarily because virtually every major disaster tends to have such leadership failures as root causes.
Richard Posner’s blog on the Gulf oil leak draws many analogies to the recent collapse of the financial systems and our economy. One of the root causes he notes in both disasters is the “rapid and relentless advance of technology”. Regulators, whether overseeing financial organizations or oil companies, typically lag behind the advances in technology that move forward very rapidly.
Sound familiar? How many new drugs or technologies in healthcare get endorsed and heavily promoted based on preliminary studies, only to be associated with serious consequences when science catches up with hope and hype.
Posner also talks about some of the attributes of risk taking. When the probability of failure or disaster is perceived to be low, companies or organizations or individuals tend to take risks. No one gets credit for preventing a disaster that was considered unlikely in the first place. We reward the gambler and the swashbuckler. Also, the riskier the venture, the more likely the return on investment will be high. He also notes that the overcautious business will lose profits, investors and staff to bolder competitors.
The healthcare analogy: local medical “arms races” where hospitals throw caution to the wind because they fear the hospital down the street will put in that new program or new piece of equipment before they do. Community hospitals began bariatric surgery programs, which have steep learning curves, and mortality/morbidity rates soared. Hospitals raced to implement high dose radiation therapy programs or the newest CT technologies without adequate training of all staff, and patients suffered radiation overdoses.
When we talk about root cause analyses we always mention our “Big 3” that show up in almost every incident having a serious untoward outcome:
- Failure to heed alarms
- Failure to buck the authority gradient
- Failed communications
The BP Gulf oil disaster was no exclusion to that.
Warning signs were not heeded.
In the CNN interview workers noted that the well would frequently “kick”, meaning that gas was ascending up through the mud and one mechanic noted the pressures were much higher than he had seen in other drilling. In a letter to the Wall St. Journal (Barr 6/11/10) Terry Barr noted multiple red flags that went unheeded. He notes that the fact that the cementing job did not go completely as expected and should have been a red flag. A pressure test performed was a second red flag that should have pointed out the cement had failed to form a seal. But most importantly data from pressure monitoring as the mud was being replaced with lighter sea water should have led to the conclusion that hydrocarbons were flowing within the well and that the pressure was not being controlled because more fluid was being pushed out than was being pumped in. Barr asserts that at this point the BP supervisors should have gone into a well kill operation and begun pumping the heavier mud back in to control the pressure.
In healthcare, issues related to alarms or failure to heed alarms are a big problem (see our Patient Safety Tips of the Week March 5, 2007 “Disabled Alarms”, March 26, 2007 “Alarms Should Point to the Problem”, April 2, 2007 “More Alarm Issues”, June 19, 2007 “Unintended Consequences of Technological Solutons”, April 1, 2008 “Pennsylvania PSA’s FMEA on Telemetry Alarm Interventions”, February 23, 2010 “Alarm Issues in the News Again”, March 2, 2010 “Alarm Sensitivity: Early Detection vs. Alarm Fatigue”).
Failure to buck the authority gradient.
On April 20 (the day the well erupted and the explosion occurred) there apparently was a disagreement (Casselman & Gold, WSJ 5/27/10) about the decision to replace the heavy drilling mud, which was keeping pressure down, with much lighter seawater. That was one of the most crucial decisions in the cascade of events leading to the disaster. Concerns of those workers on the rig were apparently overruled by BP. The culture aboard the drilling rig apparently was not one where contrary viewpoints were sought out. A CNN report on interviews with survivors of the blast talked about that argument that took place on the morning of the blast. The survivors said it was commonly perceived that you could get fired if you expressed concerns about safety that might delay drilling and that some workers had already been fired.
The healthcare analogy is obvious. In almost every serious event we see there was someone who knew that something was wrong but was afraid to speak up, usually because the culture of the organization was too hierarchical and did not foster open expression of opinions.
Failure of communications.
While we have not seen in the reports evidence of fumbled handoffs or similar communication breakdowns in the days proximate to the disaster, some of the testimony at the various hearings has focused on communication and chain of command issues (Weber & Plushnik-Masti 8/23/10). Investigators noted that the person in charge of keeping the crew and vessel safe and preventing pollution actually had little say and awareness of what was going on in terms of risk.
Workarounds, Cutting Corners, and
Efficiency-Thoroughness Tradoffs (ETTO’s)
A series of decisions were crucial in the chain of events leading up to the explosion. In testimony before a federal panel (Weber 8/24/10) a Haliburton official noted he had recommended BP use 21 “centralizers” to help create a good seal on the well. BP instead opted to use only 6, citing the extra 10 hours it would take to install the larger number. A decision to use seawater rather than much denser mud (Bea 5/20/10) removed a pressure barrier that had been keeping oil and gas from moving up the well. Another decision (King & Gold, WSJ 6/15/10) involved using a “long string” design in which a pipe runs all the way from the sea floor to the bottom of the well, allowing gas to potentially rise straight to the top. This was considerably less expensive and faster to implement than the alternative design which would added other barriers to prevent any gas from flowing unchecked to the surface. Another decision was made not to take 12 hours to completely circulate the heavy drilling fluid (helps check for leaking gas) and another to forgo a test to determine if the cement had properly bonded that would also have taken about another 12 hours. Congressmen Henry Waxman and Bart Stupak noted that all these decisions were trade-offs between cost and safety (see our September 15, 2009 Patient Safety Tip of the Week “ETTO’s: Efficiency-Thoroughness Trade-Offs” about Erik Hollnagel’s book on ETTO’s).
The healthcare analogy comes right from Erik Hollnagel’s book on ETTO’s: He describes a case in which both a patient and his oncologist each had reasons to initiate chemotherapy before the final pathology report on a gastric tumor was available. The patient died after 5 months, receiving the wrong chemotherapy (his tumor actually had a good prognosis had the correct chemotherapy been given).
There clearly were also time pressures. One interview by CNN’s Anderson Cooper (Bronstein & Drash, CNN 6/9/10) noted the project was costing $750,000 per day and was already 5 wks late.
Note that we cannot find any reference as to whether any of the decision makers had personal financial incentives to finish the job earlier or at lower cost. Such incentives, of course, would clearly tip the ETTO balance in favor of efficiency over safety and, if so, would clearly be an important root cause. In healthcare, where bonuses and incentives typically comprise part of executive salaries, one must be very careful to ensure that such incentives do not lead to projects being pushed to early completion at the expense of proper attention to safety issues.
Overreliance on the Blowout Preventer (BOP)
It also is quite apparent that there was an overreliance on the BOP (blowout preventer). Throughout, it seems that both those designing and those implementing the drilling technology were confident that if something went wrong the BOP would shut down the well and prevent a disaster. The healthcare analogy: overreliance on the computer and other high tech devices. We’ve written extensively on this site about the unintended consequences of technology. One of those consequences is putting too much trust into the computer system and feeling that the computer will capture any errors we make.
The BOP (blowout preventer) apparently had several potential means of activation (see Wikipedia article). One method of communication may have been lost due to the explosion. A second involved a “dead man switch” but examination of at least one of the units controlling that switch showed a dead battery. Also, a (hydraulic) leak in the blowout preventer had been noted about 2 weeks prior to the accident. Those were obviously latent errors in the system that contributed to the disaster once more immediate causes occurred. Also during the federal investigation issues were raised about the most recent safety inspection of the blowout preventer.
No one really knew what the risk of a blowout was (all assumed it to be low despite frequency of such in the oil industry). Tom Krause notes the influence of recency bias in that regard, noting that there had been no serious explosions in the Gulf since 1979 (in fact, that very bias may have played a role in the Obama administration’s decision to allow expansion of off-shore drilling just days before this disaster). That sort of bias has been seen in many previous disasters such as the space shuttle disaster due to damage to the thermal insulating tiles. NASA had over the years stretched its tolerance for complying with standards related to those tiles because the shuttle had flown so many times without problem. (You may recognize the latter phenomenon as “normalization of deviance” where an organization comes to accept a deviation from best practice as the new standard practice, i.e. a workaround becomes the norm). How many of you know of examples in your organizations where there is a workaround around a faulty piece of equipment or a faulty procedure that has now become accepted “standard” practice? If you don’t start looking today – you’ll find them!
Where was the FMEA?
One of the apparent key issues in the BP oil spill was the lack of an effective plan to mitigate the disaster once it happened. In all industries we pride ourselves in doing FMEA’s (failure mode and effects analyses). In these we flowchart all the steps in a process and try to anticipate what could go wrong at each step. We then develop plans to prevent such things from going wrong or to mitigate the effects of that step going wrong. That includes planning for how to mitigate when the ultimate step goes wrong. We assume that BP and the other contractors must have performed a FMEA prior to beginning their Gulf operations. Surely, the regulators would require submission of a FMEA before granting approval of a drilling operation, wouldn’t they? But we have not yet seen any publication of that FMEA.
Even if the statistical likelihood of an untoward outcome like a blowout is low, one still needs a plan of action should that outcome actually occur. In healthcare, we have discussed that in our April 6, 2010 Patient Safety Tip of the Week “Cancer Chemotherapy Accidents”.
In healthcare, we do FMEA’s both for processes we perform frequently and for those issues for which we think untoward outcomes are unlikely but would be disastrous if they occurred. For example, we recommend those organizations doing obstetrics do a FMEA on the potential of switched babies (see our Patient Safety Tips of the Week for November 17, 2009 “Switched Babies”).
The Public Relations Debacle
Lastly, the public relations response to the disaster by all involved parties leaves a lot to be desired. Had BP and others just admitted upfront that mistakes were made and apologized and not tried to downplay the magnitude of the disaster and focused on learning for the entire industry, the public might have been more understanding. Quite frankly, one suspects that this sort of disaster could have happened on any of the deep sea oil drilling rigs run by any number of companies. But the way the events were handled by BP focused the wrath of the public upon them. In healthcare we have learned that disclosure and apology are the right thing to do (see our June 22, 2010 Patient Safety Tip of the Week “Disclosure and Apology: How to Do It”).
Summary
In many respects the BP Gulf oil spill illustrates multiple points that one often sees in healthcare cases with adverse outcomes (cascade of errors, latent errors, violations, unsafe workarounds, communication breakdowns, failure to heed alarms, multiple design flaws, safety “culture” issues, time pressures, workflow issues, and technological advances with unintended consequences) similar to many of the healthcare cases we have discussed in the past (see, for example our April 2, 2007 Patient Safety Tip of the Week “More Alarm Issues”).
References:
TapRoot®. Root Cause Analysis Blog.
BP’s Deepwater Horizon interim incident investigation. May 24, 2010. Downloaded from TapRoot® website.
http://www.taproot.com/wordpress/wp-content/uploads/2010/07/BPInvestigationPresentation2.pdf
Bea R. Deepwater Horizon Study Group (Center for Catastrophic Risk Management; University of California, Berkeley). Failures of the Deepwater Horizon Semi-Submersible Drilling Unit. May 20, 2010. Downloaded from TapRoot® website.
http://www.taproot.com/wordpress/wp-content/uploads/2010/07/BobBeaPreliminaryAnalyses.pdf
Wikipedia. Deepwater Horizon oil spill.
http://en.wikipedia.org/wiki/Deepwater_Horizon_oil_spill
Krause TR. What Caused the Gulf Oil Spill?. Corporate Social Responsibility Newswire August 21, 2010
http://www.csrwire.com/press_releases/30321-What-Caused-the-Gulf-Oil-Spill-
The Becker-Posner Blog.
http://www.becker-posner-blog.com/2010/07/the-gulf-oil-leakposner.html
Bronstein S, Drash W. Rig survivors: BP ordered shortcut on day of blast.
CNN June 9, 2010
Barr T. The Oil Disaster Is About Human, Not System, Failure (letter). The Wall Street Journal. June 11, 2010
http://online.wsj.com/article/SB10001424052748703303904575293270746496824.html?mod=googlenews_wsj
Casselman B, Gold R. BP Decisions Set Stage for Disaster. The Wall Street Journal. May 27, 2010
http://online.wsj.com/article/SB10001424052748704026204575266560930780190.html?KEYWORDS=BP
Weber HR, Plushnick-Masti R. Associated Press. Oil spill probe focuses on communication among key players. Buffalo News August 23, 2010
http://www.buffalonews.com/world-nation/article170013.ece
Weber HR, Plushnick-Masti R. Associated Press. Oil spill panel hears about Halliburton warning. Buffalo News August 24, 2010
http://www.buffalonews.com/business/24-hour-business-news/article170895.ece
King N, Gold R. BP Crew Focused on Costs: Congress. The Wall Street Journal. June 15, 2010
http://online.wsj.com/article/SB10001424052748704324304575306800201158346.html
Wilson AB. BP's Disaster: No Surprise to Folks in the Know.
CBS News June 22, 2010
http://www.cbsnews.com/stories/2010/06/22/opinion/main6605248.shtml
http://www.patientsafetysolutions.com
Patient
Safety Tip of the Week Archive
What’s New in the Patient Safety World Archive