On the Escape of Tigers & Other Public Health Hazards
Considering the Ebola Outbreak in Haddon’s Matrix and Reason’s Swiss Cheese Model
October 16, 2014
Preface: In 1970, Dr. William Haddon wrote a brilliant editorial that changed forever how we evaluate accidents and other failures in complex systems. The paper was titled, On the Escape of Tigers: An Ecological Note, and it looked at accidents through an etiological rather than descriptive approach. The work was immediately applied to automotive and aviation safety, beginning the enormous reduction in accidents in both fields that continues to this day. In 1990, Dr. James T. Reason published his first work on the role of barriers to sequential failure, and how they fail in truly catastrophic accidents. Together, their combined work forms the basis for much of the best practice for risk mitigation in the automotive, airline, oil & gas and healthcare industries. In this article, I take a look at the failed efforts to control the Ebola outbreak relative to the concepts of Drs. Haddon and Reason.
Haddon’s Strategies and Matrix
Haddon focused his original work on accident prevention, specifically focusing on the reduction in the amount and structural impact of energy released in a harmful way in an accident. He promulgated 10 strategies specifically targeted at this goal (see table below). The first four of these strategies were designed to be implemented before an accident or other untoward outcome (“pre-event”) to prevent events from happening or minimize the potential negative outcome (“effect”), the second four strategies are design to protect the party subject to the effect (“host”) in the course of the event, and the final two were designed to mitigate the negative consequences post event. Reducing these to a more generalized format we get:
|Strategy Number||Strategy||Examples(Ebola strategies bold)|
|1||Prevent creation of the hazard||Ban 3-wheel ATVs, restrict access to explosives, Design effective Ebola strategy|
|2||Reduce the amount of the hazard||Limit pills per container, decrease water temperature, strict reductions on travel to and from areas impacted by the virus|
|3||Prevent release of the hazard||Improved brakes, puncture resistant gas tanks, Quarantines, travel limitations|
|4||Alter release of the hazard||Blister packaging on pills, public health restrictions of travel for potentially infected patients|
|5||Separate person and hazard in time and space||Child restraints, Quarantines, travel restrictions|
|6||Place barrier between person and hazard||Bike helmets, pool fences, Protective clothing & gear for care givers|
|7||Modify basic qualities of the hazard||Breakaway light poles, Protective clothing & gear for potentially infected patients|
|8||Strengthen resistance to the hazard||Earthquake building codes, Special hospital units for Ebola patients|
|9||Detect quickly||Smoke detectors, Effective blood tests and screening based on physical characteristics|
|10||Repair the damage||EMS, treatment & rehabilitation, Effective and timely treatment for Ebola patients|
Adapted from slides by Carolyn Fowler, PhD
What makes Haddon’s strategies so compelling is their common sense simplicity, and that they are fundamentally hopeful; we can alter the frequency and impact of terrible events through simple and thoughtful actions. Haddon started a revolution in accident prevention by looking at the causes of accidents and the effects of accidents on victims. These techniques have been broadly adopted by regulatory authorities around the world, with a long term trend in the reduction in both number of accidents and resulting injuries and fatalities. It is a method for optimists, and it really works.
Reasons to be Optimistic about Airline Safety
Happy Motoring – The Result of Successful Accident Prevention and Mitigation Strategies
When later reduced to practice, these strategies were combined into a matrix format.
Adapted from slide presentation by Carolyn Fowler, PhD
It is instructive to look at the use of Haddon’s Matrix and related tools in the control of the SARS (Severe Acute Respiratory Syndrome) outbreak in Toronto in the time period of February to July of 2003. There are a couple of important differences between SARS and Ebola that are worth noting from the outset. On first blush, SARS is much more communicable (being capable of airborne transmission), and Ebola is much more fatal (50% fatality rates in Ebola vs. less than 10% for SARS). A slightly deeper examination of the facts suggests that in both cases, these are primarily nosocomial infections, and that the vast amount of risk of both infection and transmission lies at the patient/caregiver interface. This was clearly the case in SARS (although not immediately expected), and the early evidence in the US and Europe certainly points to this interface as the greatest point of risk for Ebola.
The Haddon Matrix & SARS Hospital Infection Control
|Human||Agent & Carrier (Vector)||Physical and Social|
|Pre-Event||1) Staff Training in infection control, 2) Case mix of patients, 3) Surveillance for SARS by health care providers, 4) Public health risk communication||1) Level of contagiousness, 2) Incubation period, 3) Subclinical infection potential, 4) Lethality, 5) Potential modes of transmission||1) Employee awareness of daily infection control practice, 2) Staff adherence to infection directives and protocols, 3) Risk communication to staff, 4) Budget allocations 5) Plans for surge requirements, 6) Proximity to borders, airports and access points, 7) Availability of Protective gear, 8) Infection control checklists and forms, 9) Hospital infection control infrastructure, 10) Lab facilities|
|Event||1) Mental health for staff during event, 2) Staff adherence to infection controls, 3) Isolation and quarantine implementation, 4) Risk communication to staff and patients.||1) Modes of dissemination of virus during actual outbreak.||1) Hospital surge capacity, 2) Trust in administration crisis management performance, 3) Availability of designated SARS hospitals, 4) Budget, 5) Communication network and capacity, 6) Effective incident command system, 7) Crisis command center, 8) Efficiency of medication and equipment delivery, 9) Positive bias toward healthcare providers and media accuracy, 10) Accurate and appropriate messages to staff and patients, 11) Moral support to healthcare community, 12) Patient and family compliance.|
|Post Event||1) Risk communication, 2) Post-mortem management, 3) Psychology of post-event reaction, 4) Surveillance||1) persistence of agent in the environment||1) Decontamination, 2) messaging, 3) Government financial support, 4) Restoration of stockpiled medication and equipment, 5) Mental health support, 6) economic impact on affected community.|
Source: Environmental health Perspectives, May 2005; 113(5): 561-566
The current Ebola outbreak began in March of this year and now totals 8,997 cases in 7 countries (WHO bulletin 10/15/2014), dwarfing the next largest (425 cases in Uganda in 2000), of the 22 outbreaks since 1976. This is fundamentally a different scale and risk profile than the world has witnessed since the SARS outbreak in 2003. It is unclear if the US, the EU or any member of the G20 has made any special preparations to deal with the Ebola outbreak in the 7 months since it began.
The control of SARS in Toronto ultimately required a ratio of 100 persons quarantined for every confirmed case. In Canada alone, this was a total of 23,000 people in quarantine during the course of the crisis. The premature declaration of victory in the SARS case, resulted in what was in reality back-to-back pandemics, with a doubling of the cost of illness, lives and economic loss.
Based on the SARS Matrix as an example, and the facts and circumstances of the current Ebola outbreak, it is worth taking a minute to sketch out an ideal response matrix, and then compare it to the actual response – at least as far as it has been made public.
Ebola Pandemic – an Ideal Haddon Matrix
|Human||Agent & Carrier (Vector)||Physical and Social|
|Pre-Event||1) Direct and permanent intervention in West Africa, 2) Extensive drug research and testing, 3) Education regarding risks in animal transmission.||1) Ebola is moderately contagious, with a long and variable incubation period, is highly lethal and is easily confused with other diseases – active testing and public health are a requirement, 2) Every attempt must be made to support curative therapies and vaccines.||1) Education is key to public health in the endemic countries, 2) Mass elimination of other mammalian hosts is a necessity, 3) Alternative foods must be available, 4) Development of multinational response teams, 5) Stockpiling of equipment and supplies, 6) Checklists and extensive “in-country” training, 7) Rapid response for burial training and practice in-country.|
|Event||1) Mental health for staff during event, 2) Staff adherence to infection controls, 3) Isolation and quarantine implementation, 4) Risk communication to staff and patients.||1) Modes of dissemination of virus during actual outbreak must be weakened and eliminated, 2) Every effort must be made to limit dispersion of virus, 3) Healthcare workers are particularly vulnerable to contagion and need special prophylactic therapy where available.||1) Hospital surge capacity, 2) Trust in administration crisis management performance, 3) Availability of designated Ebola hospitals, 4) Budget, 5) Communication network and capacity, 6) Effective incident command system, 7) Crisis command center, 8) Efficiency of medication and equipment delivery, 9) Positive bias toward healthcare providers and media accuracy, 10) Accurate and appropriate messages to staff and patients, 11) Moral support to healthcare community, 12) Patient and family compliance.|
|Post Event||1) Risk communication, 2) Post-mortem management, 3) Psychology of post-event reaction, 4) Surveillance, 5) rapid return to pre-event priorities||1) Persistence of agent in the environment must be eliminated wherever possible to prevent the double pandemic seen in SARS.||1) Decontamination, 2) messaging, 3) Government financial support, 4) Restoration of stockpiled medication and equipment, 5) Mental health support, 6) economic impact on affected community.|
Ebola Pandemic – Actual Haddon Matrix to date
|Human||Agent & Carrier (Vector)||Physical and Social|
|Pre-Event||1) No meaningful or successful strategy or intervention, 2) Limited drug research and testing, 3) Limited education regarding risks in animal transmission.||1) Ebola is moderately contagious, with a long and variable incubation period, is highly lethal and is easily confused with other diseases – despite this, no testing and public health took place, 2) No meaningful attempt was made to support curative therapies and vaccines.||1) No Education in the endemic countries, 2) No systematic elimination of other mammalian hosts, 3) No alternative foods available, 4) No development of multinational response teams, 5) Limited stockpiling of equipment and supplies, 6) No plans or “in-country” training, 7) No cultural changes.|
|Event||1) Mass confusion and in action at the government level. Mental health for staff during event ignored, 2) Infection controls incorrect or not specified, 3) Isolation and quarantine implementation late, ineffective or non-existent, 4) No risk communication to staff and patients.||1) Modes of dissemination of virus during actual outbreak neither weakened nor eliminated, 2) No effort made to limit dispersion of virus, 3) Healthcare workers are particularly vulnerable to contagion – grossly inadequate training and protective gear.||1) No hospital surge capacity, 2) No trust in administration crisis management performance, 3) No designated Ebola hospitals, 4) Budget?, 5) Communication network and capacity wholly inadequate, 6) Ineffective incident command system, 7) No crisis command center, 8) Efficiency of medication and equipment delivery?, 9) Positive bias toward healthcare providers and media accuracy?, 10) Inaccurate and inappropriate messages to staff and patients, 11) Limited moral support to healthcare community, 12) No patient and family compliance, 13) Command and control turned over to untrained political representative, 14) Legal and financial protection afforded all quarantined persons at Federal, State and local levels.|
|Post Event||?||1) Persistence of agent in the environment must be eliminated wherever possible to prevent the double pandemic seen in SARS.||?|
The entire response US to the Ebola crisis has been done without an articulated strategy, seemingly responding to political imperatives rather than public health best practice. None of the obvious strategies or methods that would fit Haddon’s 10 strategic categories has been employed, and the discourse has taken an alarming political tone. The appointment of Ron Klain, a political operative and lawyer as administrative chief of the effort begs the question of credentials, and is certain to further politicize this public health fiasco.
The motivation for sending 3,000 US troops to aid in the Ebola fight is equally mystifying. Our military is just exiting two decade long wars, and is now being drawn back into further conflict with ISIS. The putative purpose of this deployment is to build treatment facilities and operate mobile labs. Surely there are contractors more capable and with fewer agenda conflicts than the 101 Airborne. Sending untrained and poorly equipped US troops in to fight Ebola may be politically expedient, but I struggle to understand how it fits into any of Haddon’s strategic framework.
It is interesting to note that Nigeria, a poor West African neighbor to the primarily impacted countries (Guinea, Sierra Leone and Liberia) successfully turned back Ebola using a well-articulated and systematically applied strategy that did not include armed combatants. Nigeria’s Emergency Operations Center sat down with representatives from the World Health Organization, UNICEF, CDC, Doctors Without Borders and the Red Cross and designed a plan focused on four teams:
- A point-of-entry team to monitor and screen passengers entering and departing the country,
- An information dissemination team,
- A case management team, and
- A contact-tracing group to track down, monitor and quarantine people who had been in contact with patients suffering with active Ebola.
And it did involve closing the borders.
Ebola virus – the new target of the 101st Airborne
Reason’s Swiss Cheese Model and a Just Culture
James T. Reason is the father of Swiss Cheese Human Error model. This model is based on the concept that 90% of errors are caused by systemic (latent) rather than simple human factors (active). Models based on blame, normally miss the fundamental element of passive failure of accident prevention barriers – much like lining up the holes in Swiss cheese.
Combining these work of Haddon and Reason, we get Haddon’s strategies as the slices of cheese (barriers to failure), and failure occurring when the holes line up, due to chance, poor planning or systematic strategic failure. Reason proposed a concept of a “Just Culture” as an integral part of the model of Human Error. Reason’s human Error Model postulates that the vast majority of errors happen despite the good intentions of the person “causing” the error. In Reason’s just culture, failed systems rather than failed people are normally the cause of errors, and we need to investigate how to improve systems rather than blame and punish people if we want to improve outcomes.
Considerations of the US Response to Ebola
Relative to Haddon’s strategies and matrix, the US failed to have any viable strategy for dealing with Ebola either before or after the outbreak. Prior to the outbreak, we took no steps to combat the virus through vaccines or drugs, and did nothing to reduce the contact between the virus and human hosts. This is consistent with the cold blooded calculus that this is a strictly African problem and of no direct concern to the US, but the spread of the arrival of Ebola in the Dallas proves this to be short sighted and stupid (as well as callous).
Relative to Reason’s Swiss Cheese model, the barriers to the spread of Ebola that existed prior to the outbreak were few and full of holes. No vaccine, and no effective drug therapies were developed, and no coordinated response was initiated when the disease first took root in March. When the disease finally arrived in the US, we had no training and no protective clothing for our front line medical staffs.
Our Nation’s response to date gives not even the slightest thought to the plight of people forced to stop their lives and go into isolation or quarantine because of innocent exposure to the virus. The forced isolation of quarantine can result in loss of jobs and income, and there is not a single discussion on any front regarding protecting these people – which will only influence them to avoid quarantine. This is a national disgrace, and the antithesis of just culture.
The arrival of Ebola into the heartland of America did nothing to stimulate a viable strategic response. The response of sending armed troops to battle a virus in Africa seems silly, except when considered as a political rather than a public health action. The best research that I have found indicates that cutting off flights from the afflicted nations would prevent the spread of the virus by up to 8 weeks, but when considered politically it is apparently untenable.
The intra-outbreak excellence shown in Toronto by quarantine of patients (as well as potentially impacted people), has been totally ignored in our country; where we put our potentially infected people on airplanes and fly them around the country – approved by the CDC. Finally, when the public has finally had enough, we put a political operative in charge – a man with no background in public health.
My Grandmother used to say that God watches out for fools and drunks – by her criteria we should be able to depend on divine intervention in this crisis – it’s as good as any strategy that we’ve had to date.