Document Detail

Title: RS321-1 - Precursors of High-Impact, Low-Frequency Events, Including Fatalities
Publication Date: 10/1/2016
Product Type: Research Summary
Status: Supporting Product
Pages: 33
Research conducted in other industries than construction has revealed that high-impact, low-frequency (HILF) events are preceded by events or conditions that can be detected and, if acted upon, can prevent the HILF incident. This summary describes the CII project that yielded the construction industry's first valid and reliable method for identifying the leading conditions that predict HILF events, and the predictive scorecard the team developed to implement those findings.
Order Now  


Annual total recordable injury rates have shown great improvement over the past few decades, yet the rate of fatalities appears to have plateaued. Studies of catastrophic incidents in other industries have revealed that high-impact, low-frequency (HILF) events are preceded by events or conditions that can be detected and, if acted upon, can prevent the HILF incident. Although the precursors are often unique to each industry, the methods implemented to identify and analyze precursors are consistent. Using existing precursor analysis programs as guidance, HILF events were defined as high-energy events that result in or have the potential to result in a fatality or life-altering injury or illnesses and precursors as reasonably detectable events, conditions, or actions that serve as warning signs of an event.
The ultimate goal of the Research Team 321 (RT-321) project was to conduct rigorous scientific research that yielded a precursor analysis protocol for construction that enables practitioners to pursue the following series of steps:
  1. Assess conditions in a leading fashion.
  2. Identify the presence of precursors, and quantify their extent, in a structured and methodical fashion.
  3. Predict and prevent the potential for a HILF event
In pursuit of this goal, RT-321 addressed the following essential questions:
  • Are there precursors to HILF construction events?
  • If so, what are they, and how can they be identified, analyzed, and used in a predictive fashion to prevent HILF events?
Addressing these research questions will help to address the fact that fatality rates have recently plateaued and even increased.
By using a combination of literature review, input from industry experts, empirical data collection, a series of randomized and blinded experiments, and objective multivariate statistical analyses, RT-321 was able to achieve the aforementioned goal and exceed original expectations. This project yielded the construction industry’s first valid and reliable method for identifying leading conditions that predict HILF events.
The RT-321 research process delivered the following key findings:
  1. The process for predicting a HILF event is far more difficult than conducting a retrospective root cause analysis.
  2. Precursors are different from leading indicators, and precursor analysis is different from monitoring and evaluating leading indicators.
  3. The quantity of that is energy present in a work operation or condition before an incident occurs is a direct predictor of the severity of an injury.
  4. Professionals are able to use the precursor analysis protocol developed in this research and their intuition to correctly predict the occurrence of HILF events with significantly better than random frequency.
  5. The errors made in prediction were most often conservative.
  6. Mathematical models provide a valid, reliable, and objective method for predicting the occurrence of HILF events.
  7. The precursors for fatalities and severe injuries are indistinguishable from those that were involved in high-energy near misses.
RT-321’s detailed research findings are presented in Chapter 4. The user-friendly predictive scorecard that RT-321 developed from these findings is the focus of Implementation Resource 321-2, Guide to Precursor Analysis for Construction Fatalities.