Trooper Scott Rendfrey uses a pupilometer while conducting a twelve-step DRE (Drug Recognition Expert) test on instructor Cpl. Michael Carr during a role playing exercise at the Colorado State Patrol Training Academy in Golden, Colo. on March 6, 2014. (Craig F. Walker, The Denver Post file)

The State Patrol arrested fewer people on allegations they were driving under the influence of marijuana last year than in the previous year, according a report released Thursday.

The report is the first glimpse at how the change in law is affecting highway safety, because the patrol did not keep statistics on the number of people accused of driving under the influence of marijuana prior to 2014, when recreational pot became legal. 
                                  
In 2015, the second year the patrol kept statistics, 4,546 citations were issued for driving under the influence of alcohol or drugs.

Of those, 665 people — or nearly 15 percent — had marijuana in their systems when they were charged, according to State Patrol statistics.

Overall, the number of people cited for marijuana-related driving under the influence of drugs dropped 1.3 percent between 2014 and 2015, the report said. In 2015, 665 impaired drivers had marijuana in their systems, compared with 674 in 2014.

The data do not include arrests made by local police and sheriff's departments. In Colorado, the legislature has decided that the legal limit for impairment by marijuana is 5 nanograms of THC in the blood. However, that is a presumption only and has been rejected in court.

Highway safety experts said they were not reading too much into the numbers. But others said it is a good indicator that Coloradans are getting the message about the dangers of stoned driving.

"It's not as much of a public danger as it was made out to be when recreational marijuana was first legalized," said Jay Tiftickjian, a Colorado DUI attorney who edited a textbook on marijuana and the law.

He believes government educational campaigns are working.

But Trooper Josh Lewis, a highway patrol spokesman, said his agency wants three to five years of data before drawing conclusions. One year of numbers is not enough to spot a trend.

"We'd certainly like to think education and enforcement action are making our roads safer, but until we have data for multiple years we simply don't know," Lewis said.

Colorado Department of Transportation spokesman Sam Cole said the report provides a valuable snapshot, but his agency also continues to evaluate data. So far, it doesn't have enough information to spot any trends related to drivers who are high.

For example, CDOT keeps statistics on drugged driving in fatal crashes. From 2013 to 2014, the number of drivers in fatalities who tested positive for cannabis rose to 83 from 44.

However, researchers do not know if the pot's psychoactive effects had worn off before the crash or whether the driver was impaired because of it, Cole said.

CDOT also has surveyed residents about their attitudes toward marijuana and driving.

The results indicate marijuana use is rising in Colorado, but about half of the people surveyed did not understand the risks of driving while high, including that they could be charged with driving under the influence.

"What we know is too many people are driving high, and they need to know about the danger and they need to know the legal consequences of doing so," Cole said.

Colorado was the first state to legalize recreational marijuana, and other states are looking to it for guidance as legalization becomes more widespread.

However, the state is playing catch-up when it comes to understanding the impact and what it means to be high while driving.

Last year, the State Patrol began testing five devices to help troopers detect whether drivers are too high to be behind the wheel.