The 90% syndrome in construction safety: distributional concentration of construction accidents in the final project phase

Do construction accidents peak near the end of the project?

This analysis of >33k Korean construction accidents examined when accidents occurred during the project phases.

They describe the ‘90% Syndrome’, borrowing from the 90-90 rule in software engineering: the first 90% of code accounts for the first 90% of development time, while the remaining 10% accounts for the other 90%.

Similarly, the near completion construction phase may account for a disproportionate complexity, hazard accumulation or injury toll.

PS. Check out my YouTube and recent Critical Control / barrier vid: https://youtube.com/@safe_as_pod?si=iUaDPJynPemQRZhY

Shout a coffee: https://buymeacoffee.com/benhutchinson

Extracts:

·      “findings suggest that reported accident records are non-uniformly distributed across the project lifecycle and exhibit a systematic concentration in the final phase”

·      “The 90% mark doesn’t create problems. It reveals them”

·      “Most of the late-stage problems I diagnose were created in the first half of the project”

·      “You can post all the rules you want. The physical reality makes it very hard to control”

·      “What’s needed is not more forms — it’s physically reorganizing how the work is sequenced and separated in space”

·      “Schedule compression in the early and middle phases doesn’t disappear — it transfers forward”

·      “The schedule problem becomes a safety problem by the time you reach 90%”

·      “No safety budget solves a fundamentally compressed schedule”

·      “The ≥90% interval accounts for 14.7% of reported accidents and 16.6% of fatalities”

·      “Compared with the preceding 80–89% interval, the ≥90% interval shows a 56.3% increase in reported accident counts”

·      “Sites with both DfS and a Safety Management Plan…show a ≥90% proportion of 11.5%, compared to 23.1% for sites with neither”

·      “Without DfS, you’re managing risks that were never anticipated. Without the plan, the DfS analysis sits in a drawer”

·      ““Safety management approaches focused exclusively on terminal-phase controls may have limited effectiveness if upstream schedule and coordination conditions remain insufficiently managed”

·      “The ≥90% interval accounted for 14.7% of reported accidents (OR 5 1.552, 95% CI: 1.482–1.626) and 16.6% of fatalities (OR 5 1.793, 95% CI: 1.441–2.230)”

·      “Findings inform recommendations … [like]  formalizing the ≥90% phase as a heightened oversight period”

Study: https://doi.org/10.1108/ECAM-03-2026-0497

Leave a comment