savvy contractors have significantly cut jobsite injuries in the past few decades by transforming safety practices into a science. Recent innovations include using more statistics as well as the latest technology. Some firms are taking a cue from psychology and employing behavior-modification techniques.

公司use statistical data from outside resources such as OSHA and the Construction Industry Institute (CII) at the University of Texas at Austin, but many also compile their own. They track recordable injuries, medical treatments, near misses and lost-time cases using tools such as job-safety analysis. Then they analyze data to help prevent future injuries.

丹佛的PCL Construction Enterprises Inc.健康，安全和环境总监Rich Baldwin说：“我们的公司正在使用更多的统计数据，随着可记录的事件率下降，我们必须联系并寻找更多信息以进行改进。”“我们正在更深入地挖掘。”鲍德温补充说，PCL的事故发生率“大大低于1”。相比之下，根据美国劳工统计局的数据，2011年最新数据的发病率是美国建筑中非致命伤害和疾病的发病率为3.8。发病率是每100名全职工人的伤害或疾病数量。

承包商过去主要依靠滞后指标，例如总可记录事件率（TRIR）和经验修改率（EMR），这些指标衡量了过去的安全性能，但对预防没有太大帮助。为了解决问题，现在越来越多的公司检查了主动的主要指标，例如工人的态度和行为以及工作场所条件。

“我们透过挡风玻璃,而不是the rear-view mirror," says Matt Ogle, safety manager at Kansas City, Mo.-based JE Dunn Construction Co., about the contractor's use of leading indicators.

Another effective safety tool used by contractors is behavior modification, long practiced by psychologists. Firms employ the "active caring" method rather than punishment to change unsafe behavior.

CII主任韦恩·克鲁（Wayne Crew）表示：“与工人脱颖而出的旧战略相反，承包商正在试图建立一种雇员完全从事安全工作的文化。”“他们正在培训员工如何告诉某人他们正在表现出高风险的行为。”

奥格尔说：“如今的安全钥匙包括积极的关怀。……如果我看到一个不戴安全眼镜的人，我不只是告诉他戴眼镜。”“我说我关心他的眼睛。这导致蒸发行为为零。”

PCL instituted a program whereby employees learn to be safety observers. If observers see a worker or a crew doing something unsafe, they address the problem with the workers. They can even stop work, if necessary.

"Ninety percent of all injuries are caused by behaviors," says PCL's Baldwin. "We are doing more behavioral safety training with supervisors and workers—things like keeping your mind on your work, following the rules, and why you should follow the rules."