Table 2 Studies included based on eligibility criteria: studies with increases in enforcement activities but without statistical models to account for those increases

Fell 2008 [23]

Objective: Evaluate the impaired-driving demonstration projects conduced in 7 states.

Indicators relative to surrounding states

Compared to surrounding states, Georgia, Tennessee, Indiana, Michigan had statistically significant decreases in the Ratio, whereas in some States (Louisiana, Texas), there were increases in the Ratio.

12–18 months

 CITS

Ratio: ratio of drinking drivers (BAC > =0.01) to nondrinking drivers (BAC = 0.00) in fatal crashes

 2000–2003

 7 selected states in the US (Georgia, Louisiana, Pennsylvania, Tennessee, Texas, Indiana, and Michigan)

Use of paid media (+/− earned media):

Georgia, Louisiana, Tennessee, Texas, Indiana, Michigan

VMT: alcohol-related fatalities (driver or pedestrian total BAC > 0.01) per 100 million

Sobriety checkpoints:

VMT

 Comparator (within-state comparison: Georgia, Tennessee, Michigan; neighboring states: selected nearby states, pooled; the rest of the nation, pooled).

Georgia, Louisiana, Pennsylvania, Tennessee, Indiana

Georgia:

Saturation patrols:

Louisiana, Tennessee, Indiana, Michigan

Community education/partnership:

Pennsylvania, Michigan

Ratio: −14 % (p < 0.05), VMT −5 %

Louisiana:

Ratio: 1 %, VMT15% (p < 0.05)

Pennsylvania:

Ratio: −9 %, VMT: −2 %

Tennessee:

Ratio: −11 %(p < 0.035), VMT: 1 %

Indiana:

Ratio: −13 %(p < 0.018), VMT: −20 % (p < 0.002)

Michigan:

Ratio: −14 % (p < 0.07), VMT −18 % (p < 0.003)

Texas:

Ratio: 3 %, VMT: 5 %

Zwicker 2007a [24]

Objective: Evaluate the effect of the National Highway Traffic Safety Administration impaired driving high-visibility enforcement model in 2002 in West Virginia

Alcohol-related fatalities in targeted counties: reduction of 0.99 lives each month. - 24 % (p = 0.01)

Alcohol-related fatalities in targeted counties: − 24 % (p = 0.012)

18 months

 CITS

 2000–2004, monthly

Alcohol-related fatalities in targeted countries for men 21–34yo: reduction of 0.09 lives per month (p = 0.79)

 Intervention period: July 2003- Dec 2004

 Comparison period: Jan 2000- June 2003

Mass Media:

Paid media (TV)

 Intervention site: 6 counties in West Virginia

Enforcement:

Statewide alcohol-related fatality trend: reduction of 1.6 fatalities per month (p = 0.20)

Sobriety checkpoints, saturation patrols

 Comparator site: 49 non-targeted counties

Zwicker 2007b [25]

Objective: Evaluate Connecticut’s statewide impaired-driving publicity and enforcement campaign

The overall alcohol-related fatality trend for the State:

Net change in alcohol-related fatalities in the state: −36.4 %

18 months

 CITS

 Jan 2000- Dec 2004, monthly

Estimated reduction of 2.604 lives each month (p = 0.01) for the 18 mo. following the beginning of the campaign (Net change: lives saved during 18 mo., 36.4 % decrease)

Net change in alcohol-related fatalities among men 21–34 years old: −29.7 %

 Intervention phase: July 2003- Dec 2004, Comparison phase: Jan 2000- June 2003

Mass media:

Paid and earned media targeting men 18–34 years old

Enforcement:

 Intervention site: Connecticut, USA

Sobriety checkpoint

The alcohol-related fatality trend for fatalities involving men 21 to 34 years old:

Estimated reduction in the number of fatalities by 1.568 lives each month for the 18 mo. following the beginning of the campaign (p < 0.03) compared to 0.16 lives per month saved in contiguous counties (Net change: 25 lives saved during 18months, 29.7 %)

 Comparator site: 3 neighboring states

Lacey 2008 [26]

Objective: Evaluate NHTSA Checkpoint Strikeforce program done July-December of each year, 2002–2004.

Alcohol-related fatal crashes in the intervention sites: −7.1 % relative to the nation as a whole (p = 0.119). In one State, West Virginia, the reduction was 16.7 % (p = 0.02) when compared to the Nation as a whole.

Alcohol-related fatal crashes: −7.1 %

3 years

 CITS

 1991–2004, annually

 Intervention sites: Delaware, Maryland, Pennsylvania, Virginia, West Virginia, District of Columbia

Mass Media:

Paid and earned media. “Checkpoint Strikeforce. You Drink & Drive. You Lose.”

 Comparator: entire nation

Enforcement:

Checkpoints. BAC measurements (Maryland, Delaware, and Virginia)

Agent 2002 [27]

Objective: Document

Number in 2002 compared to the average of the previous three years

Alcohol/drug related crashes: −9 %

4 years (13 days per year)

 ITS

the results of the “You Drink& Drive.

 Before intervention: 13 days around Labor day in1999–2001

You Lose” campaign.

1. The number of crashes in which alcohol and/or drugs were listed as a contributing factor or the driver was noted to be suspected of drinking: −9 % (not statistically significant)

Number of injuries and fatalities resulting from alcohol/drug related crashes: −5 %

Enforcement:

Checkpoints and saturated enforcement activity

 Intervention: 13 days around Labor day in 2002

Mass media:

Paid media: broadcast and cable television, radio (from 15 to 30 Aug, 2002), and outdoor billboards (15 Aug–15 Sep, 2002)

 Intervention site: Kentucky, USA

2. The number of injuries and fatalities resulting from these crashes: −5 %

 Comparator: none

Solomon 2008 [28]

Objective: Evaluate the effect of the National 2006 Labor Day holiday campaign, “Drunk Driving. Over the Limit. Under Arrest.” Targeting age group 18 to 34 years old

1. The total number of alcohol-related fatalities: 17,602 in 2006 compared to 17,590 in 2005 (0.07 %).

The total number of alcohol-related fatalities: 0.07 % increase (2005–2006)

4 months (Sep-Dec 2006)

 ITS

 Intervention period: 3 weekends leading up to and around the Labor Day holiday period in 2006

2. The number of motor vehicle fatalities for male drivers (BAC 0.01 or higher) age 18 to 34: decreased from 5782 to 5654 (−2.21 %)

Mass Media:

1. Earned media (Aug 7- Sep 10)

 Intervention: 2006

2. Paid media (Aug 16–20; 23–27; Aug 30- Sep 3)

3. The number of motor vehicle fatalities for male drivers (BAC 0.08 or higher) age 18 to 34: decreased from 4996 to 4872 (−2.48 %)

 Comparaison: 2005

Enforcement:

 Intervention site: USA (nationwide)

Sobriety checkpoints, saturation patrols

 Comparator site: none

Beck 2009 [31]

Objective: Evaluate

Net change in three-year averages before and during campaign in Maryland:

Alcohol-related total

6 years

crashes: 2.2 %

 CITS

the effect of the Checkpoint Strikeforce campaign

Total alcohol-related fatalities: 14.7 %

 Before intervention: 1999–2001, Intervention: 2002–2004

Mass Media:

● Alcohol-related total crashes: 2.2 %

Alcohol fatalities as a percentage of total fatalities: Net change 3 %

Paid and earned media

● Alcohol-related injury crashes: −4.7 %

Enforcement:

● Alcohol-related fatality crashes: −2.7 %

 Intervention site: Maryland (Pennsylvania, Delaware, West Virginia, Virginia, District of Columbia)

Sobriety checkpoints

● Total alcohol-related fatalities: 14.7 %

● Alcohol-related injured drivers: −3.8 %

 Comparator sites: Minnesota, Oregon, and Washington

Miller 2004 [30]

Objective: Evaluation of three incremental CBT program approaches

Mass media is estimated to have decreased in nighttime fatal or serious crashes decreased by 13.9 % (90 % CI = −26.1 to −0.1) nationally

Nighttime fatal or serious crashes: −13.9 %

10 years

 ITS

 Intervention (CBT): 1993~

 Intervention (media): 1995~

Mass Media: National anti-drunk-driving campaign with hard-hitting messages

 Intervention (CBT enhancement): 1996~

 Intervention sites: New Zealand (CBT enhancement in Northern Police Region)

Enforcement: CBT checkpoints, (Northern Region) highly visible CBT through booze busses

NHTSA 2007 [32]

Objective: Evaluation of the effect on the

Total declines in yearly average of fatal crashes for alcohol-impaired drivers from 2002 to 2005 were slightly greater for the non-SES, as compared with the SES (a 5 % drop in non-SES compared to a 2 % decline in SES, net decline: 3 %).

Net decline in yearly average of fatal crashes for alcohol-impaired drivers: −3 %

5 years

 CITS

National Impaired Driving Crackdown Campaign targeting men 21 to 34 years old

 Before: 2001 and 2002

 After: 2004 and 2005

 Intervention sites: 13 Strategic Evaluation States (SES) (Alaska, Arizona, California, Florida, Georgia, Louisiana, Mississippi, Montana, New Mexico, Ohio, Pennsylvania, Texas, West Virginia)

Mass Media: paid and earned media (done nationwide). Additional advertising done in

In the target group of 18–34 year-old-male drivers, the decline was greater in non-SES compared to SES (8.7 % in non-SES and 3.8 % in SES).

SES.

 Comparator sites: non-SES

Enforcement: Sobriety checkpoints or saturation patrols in SES

Suriyawongpaisal 2002 [33]

Objective: Evaluate the campaign against drink-driving and enforcement efforts

Percentage of the traffic injury victims who were drivers with illegal BAC (0.05 or more): 14.6 % increase in 9 months (p = 0.20)

Percentage of the traffic injury victims who were drivers with illegal BAC (0.05 or more): 14.6 % increase

9 months (assessed in alternating months)

 ITS

 March-Nov 2002, alternating months

Mass Media:

 Intervention sites: 4 of the 21 public hospitals in Bangkok, Thailand

 Comparator site: None

Active public education program at national scale (roadside posters; bumper; radio and TV programs or spots; public announcements; press reports), 1997

Enforcements:

Highly visible sobriety check points, 1999