A New Framework for Serious Injury and Fatality

A New Framework for Serious Injury and Fatality Prevention Thomas R. Krause, Ph. D.

Occupational Fatalities and Non-fatalities

Industry Data

A Common Situation for Leading Companies Our recordable and lost time injury rate are both declining steadily, but our fatality rate is level or increasing. How can this be? HSE people have been telling us for years about the Safety Triangle, and the idea that smaller injuries predict larger ones is embedded in our culture. We have spent time and resources on improving safety, yet the most important safety events, serious injuries and fatalities, are not improving.

How Could This Happen? Four months ago we celebrated our safety success. Our recordable rates were lower than ever. Our leaders got bonuses. This quarter we had a fatal injury, and a severe burn. Our safety leaders are perplexed and our plant manager and his boss are angry.

What does it mean when serious injuries and fatalities (SIFs) increase while recordable injuries decrease? • Existing safety systems are not reducing SIF events. Why? • This is a specific question about safety systems, not a general one. • The question implicates the design and implementation of existing safety systems. • Unless we know the answer to this question we are not in a position to address the problem.

Serious Injury and Fatality Prevention Study Participants BST Mercer ORC Exxon. Mobil Corporation Archer Daniels Midland Company Shell BHP Billiton Petroleum Potash Corporation A. P. Moeller – Maersk Group Cargill Inc.

Objective: Develop a model to understand prevent the occurrence of Serious Injuries and Fatalities (SIF’s) Using data from participating companies, study these questions: 1) Is the traditional Safety Triangle accurate descriptively? 2) Is the traditional Safety Triangle predictive? 3) Is it possible to develop intervention principles, criteria and methods to address SIF events?

Definition of Serious Injury • Any injury or illness that resulted in: • 1. Life-threatening injury or illness: one that if not immediately addressed is likely to lead to the death of the affected individual, and will usually require the intervention and/or external emergency response personnel to provide life-sustaining support. Examples include, but are not limited to: a. Laceration or crushing injuries that results in significant blood loss; b. An injury involving damage to the brain or spinal cord; c. An event which requires the application of cardiopulmonary resuscitation or an external defibrillator; d. Chest or abdominal trauma affecting vital organs; e. Severe burns • • • 2. Life-altering injury or illness: one that results in permanent or longterm impairment or loss of use of an internal organ, body function, or body part. Examples include, but are not limited to: a. Significant head injuries b. Spinal cord injuries c. Paralysis d. Amputations e. Broken or fractured bones

Question 1 Is the Safety Triangle Accurate Descriptively?

• The traditional safety triangle is accurate descriptively 12791 **SIF, 293 2984

Implications of the descriptive validity of the Safety Triangle 1) It provides an accurate description of the quantitative nature of accidents and incidents. 2) It provides insight that informs prevention strategies. 3) It means that a single incident has significance.

Single incidents tell us about the system One serious incident occurs every x times when the potential for a serious incident exists. Drunk drivers don’t have serious accidents each time they drive. It takes a set of variables happening at the same time to produce a serious accident. The driver is drunk, the car brakes are faulty, the child runs into the street when visibility is low etc. So when an accident or incident happens, we are wrong-headed to think it is a one-time single event. The set of variables that led up to the event have happened many times previously with no incident. We know this from the Safety Triangle.

“This was a one-off event” “Sometimes you step off the curb and get hit by a bus” “This catastrophic event was not related to systematic factors in the industry” “We don’t see a pattern in the data”

Question 2 Is the Safety Triangle Accurate Predictively? A. Do less serious injuries have similar or different potential to be SIF’s? B. Do SIF’s have different kinds of characteristics and causes than less serious injuries?

The traditional safety triangle is not accurately predictive SIF Of the 300 sampled injuries, 64 had the potential to be SIFs Not all injuries have SIF potential. A reduction of injuries at the bottom of the triangle does not correspond to an equivalent reduction of SIFs 21% Potentially SIF

Results of Longitudinal Root Cause Analysis 71% of SIF’s (N=55) 17% of non SIF’s (N=35) are related to Safety Absolutes.

Results of Quantitative Analysis SIF’s are disproportionately related to certain types of activities and also to activities related to certain types of safety control. Type of Activity Out of 126 incidents that occurred in connection with these activities 114 were SIF’s. Operation of mobile equipment or watercraft, working under suspended loads. Type of Safety Control Out of 47 cases that occurred in connection with these types of safety control all 47 were SIF’s. Lock out tag out, machine guarding and barricades, confined space entry, use of hot work permits, equipment and pipe opening of hazardous chemicals.

Based on both kinds of analysis, can SIF precursors be identified that will inform intervention strategies?

Definition of Precursor A precursor is an unmitigated high risk situation which will result in a serious or fatal injury if allowed to continue.

Examples of Precursors Precursor: A In order to change the doctor roll on a paper re-roller two workers must stand beneath it and guide it. The doctor roll weighs about 6 tons. Precursor: A worker is working on the bottom of an elevated vessel. There is no approved place to secure the lanyard. Precursor: During an emergency shutdown workers are unable to follow the procedure because it is not understood and appears impractical.

Examples of Activities that May Have High Proportions of Precursor Events Mobil equipment (operation and interaction with pedestrians) Confined space entry Jobs that require lock-out tag-out Lifting operations Working at height Caustic liquor handling Manual handling

Examples of Situations that May Have High Proportions of Precursor Events Process instability Significant process upsets Unexpected maintenance Unexpected changes High energy potential jobs Emergency shutdown procedures

Should We Develop a New Paradigm?

Old Paradigm All injuries of low severity have the same potential for serious injury. Injuries of differing severity have the same underlying causes. One injury reduction strategy will reach all kinds of injuries equally. Reduce minor injuries by 20% and you will also reduce major injuries by 20%.

The Old Paradigm A valid generalization which obscures important relationships. Fatalities Lost-Time Injuries Recordable Injuries

Effects of the Old Paradigm This paradigm has been useful but it has also raised issues: Elevation of the trivial. Creative classification of injuries. Loss of credibility with labor organizations. Cynicism in the organizational culture. Lack of effectiveness in fatality prevention.

New Paradigm All minor injuries are not the same: a sub-set of low severity injuries are associated with precursors to serious injuries and fatalities. Injuries of differing severity have differing underlying causes. Reducing serious injuries requires a different strategy than reducing minor injuries. The strategy for reducing serious injuries should use precursor event data drawn from all available sources of data: accidents, injuries, near misses and exposures.

The Old Paradigm Revisited A focus on the potential for serious injury Serious Injuries Precursor Events: Unmitigated high risk situations

The New Paradigm Precursors Unmitigated High Risk Situations High-Risk Event Combinations High Potential Events Near Misses Accidents Injuries Exposures High Risk Activities Outcomes Serious Injuries and Fatalities

Effects of the New Paradigm More focused effort on the prevention of serious injuries. Stronger safety culture. More engagement of labor organizations in safety. Lower rates of serious injuries.

Intervention Plan for the Prevention of Serious Injuries and Fatalities Step 1. Educate the organization on the New Paradigm for Serious Injuries and Fatalities.

Intervention Plan for the Prevention of Serious Injuries and Fatalities Step 2. Institutionalize the SIF Rate • The SIF Rate is the number of serious and fatal injuries and recordable injuries with high potential divided by hours worked. • Data on the SIF Rate should be gathered for the past two-three years and from here forward monthly. • The SIF Rate should be given high visibility throughout the organization.

The Critical Importance of the SIF Potential Rate 1. Gives visibility to SIF performance-as a leading and lagging indicator. 2. Enables new research needed to develop intervention strategies. 3. Enables root cause analysis of large numbers of SIF events. 4. Sets the stage for Predictive Analytics

SIF Rate Example

Intervention Plan for the Prevention of Serious Injuries and Fatalities Step 3. Integrate findings from the SIF study with existing safety systems. Some examples: Incident investigation Observation and Feedback Pre-task risk assessment Data analysis systems

Intervention Plan for the Prevention of Serious Injuries and Fatalities Step 4. Develop mechanisms for the ongoing identification and remediation of SIF precursors. Some examples: Longitudinal analysis Predictive analytics Discovery conversations

Discovering SIF Prevention Opportunities 87% of SIF cases studied had underlying precursors/preconditions/root causes that were discoverable from interview-based observations N=55

Predictive Analytic Insights Analysis of 264 SIF’s in one 15, 000 worker business unit shows that 33% are related to a particular management system failure. Of these, 52% involve workers who have been on the job for 1 -3 years.

Intervention Plan for the Prevention of Serious Injuries and Fatalities Step 5. Develop and validate intervention strategy. (1) Identify intervention group. (2) Gather baseline SIF Rate data. (3) Design and implement intervention plan. (4) Track SIF Rate data to measure effectiveness.

New research questions We know that leadership and culture measures predict recordable incident rates. Do specific leadership behaviors predict SIFs? Do specific cultural attributes predict SIFs? What other variables are related to SIFs?

Implications of these new findings Work from the top down rather than the bottom up Basic safety systems are not fully in place Metrics

A New Framework for Serious Injury and Fatality Prevention Thomas R. Krause, Ph. D.