WASHINGTON Auditors for the U.S. Department of Transportation say the Federal Motor Carriers Safety Administration SafeStat system still is not ready for prime time.

In fact, say the auditors, before the public can view SafeStat scores the FMCSA needs to make additional improvements.

SafeStat, which is officially known as the Motor Carrier Safety Status Measurement System and is used by the FMCSA to identify high-risk truck and bus operators, was the subject of an extensive review by the USDOT Office of Inspector General.

Three years ago, the FMCSA removed from its public Websites both the overall SafeStat scores and all SafeStat accident data, calling the move temporary and saying it was necessary to make sure the accident data was accurate, complete and timely.

U.S. Rep. Thomas Petri, R-Wis., former chairman of the House Highways, Transit and Pipelines Subcommittee, requested the review by the USDOT Inspector General.

Although the FMCSA has improved SafeStat, problems still exist with states reporting crash data to the agency, the report said. While states are reporting more truck and bus crashes, anomalies call into question the completeness of the non-fatal crash reporting, the report said.

'We recommended that FMCSA implement a new, more reliable estimate that would allow evaluation of non-fatal crash reporting, both nationally and state by state, before the department makes all SafeStat scores available to the public,' the report said.

The FMCSA is working on a new, more accurate national estimate of how many non-fatal crash reports are absent from its database and how many are missing from each state, with a target date of October. The OIG recommended SafeStat scores not be available to the public until this step is complete.

In June, the Government Accountability Office, the investigative arm of Congress, reported SafeStat is better than random selection at identifying high-risk carriers but said a statistical approach would be better. (See July 1 Bus & Motorcoach News.)

Late-reported crash data, cited as a potential statistical problem by the USDOT, in fact has little effect on SafeStats usefulness, the GAO reported.