Table 1.

Aims and Methods for Studies of WSB Impact (Category 1)

Study	Study Length	N (I/C)	Study Aim	Method of Evaluation	Robustness of Evidence
Heelan et al3	2 years	324 (201/123)	To evaluate school‐wide prevalence of walking to school and to compare activity levels among 2 WSB intervention schools and 1 control school.	A quasi‐experimental design was used with 2 intervention schools and 1 control school in Nebraska, USA. WSBs had designated stops within a 1‐mile radius of schools and adult leaders. This study evaluated school‐wide prevalence of walking to school by self‐report at 6 time points. This study also compared objective physical activity levels among a subsample of research participants (201 intervention children and 123 control children) by having participants wear an accelerometer during 4 time periods to assess daily physical activity. They were also measured for body mass index (BMI) and body fat 2 times a year.	Prevalence of walking to school was measured by self‐report which could have limited validity of the study. However, authors measured objective physical activity using accelerometer but only among a subsample of children at 4 time periods.
Mendoza et al8	1 year	820 (347/293)	To evaluate the impact of a WSB program on student transport in a low‐income, urban neighborhood, specifically, would a WSB program increase the proportion of students walking to school and decrease the proportion being driven in short‐ and long‐term.	Controlled, quasi‐experimental trial with cross‐sectional assessments. There was 1 intervention school with a WSB program with parent volunteers and a part‐time coordinator and 2 control schools. All schools were urban, socioeconomically disadvantaged, elementary schools in Seattle, WA, USA. This study assessed students' method of transportation to school by a classroom survey at baseline and 1‐year follow‐up. McNemar's test was used to examine the change from baseline to 12‐month follow‐up for walking versus all other forms of school transport at the intervention or control schools.	This study assessed travel to school method self‐report publicly in classrooms which could limit validity. Also schools were not randomized but authors propose that control schools were comparable to intervention school (in that they all served predominantly disadvantaged, minority populations) which should minimize threats to internal validity.
Mendoza et al9	5 weeks	149 (70/79)	To evaluate the impact of a WSB program on children's rates of active commuting to school and on daily moderate‐to‐vigorous physical activity (MVPA).	Pilot cluster randomized control trial with 4 intervention and 4 control schools (all low‐income public elementary schools in Houston, TX, USA). Intervention schools had 1‐3 WSBs with trained staff, to and from schools, 5 days per week. Outcomes were measured the week before intervention (time 1) and during weeks 4 and 5 of intervention (time 2). The main outcome (percentage of trips made by active commuting) was assessed every school day for 1 week during times 1 and 2 using a questionnaire which authors report had a high test‐retest reliability and convergent validity with parental report. This asked children how they got to school. Children had to select 1 answer (from: school bus, carpool, car, metro bus, walked with an adult, walked without an adult, or biked). Walking or biking was considered active commuting. The secondary outcome (MVPA minutes per day) was measured using GT1M accelerometers worn by students for 7 days at times 1 and 2. These measured acceleration in the vertical plane and intensity every minute. A valid day was defined as 10 hours (600 minutes) of accelerometer wear. Participants who had at least 1 valid day were included in analyses.	A brief intervention period, therefore, limited generalizability. There were also baseline differences in percentage of active commuting (intervention 23.8%, control children 40.2%). Al though they used self‐report to measure active commuting, authors say study had good convergent validity as this was checked with parental report.
Mendoza et al10	5 weeks	149 (70/79)	To test feasibility of evaluating changes to pedestrian safety behaviors (PSB) during a WSB program. Outcomes measures were pedestrian safety behaviors related to crossing a street at a school level prior to WSB and during WSB trial.	This observational study was a pilot study of children's pedestrian safety behaviors and was conducted as part of the cluster RCT of WSB program (Mendoza et al9). Children's pedestrian safety behaviors related to crossing a street were observed by trained research assistants using a previously validated observational instrument. Children of all grades were unobtrusively observed at major intersections. Each of these behaviors ([1] crossed at a corner or crosswalk, [2] crossed with an adult or safety patrol, [3] stopped at the curb, [4] looked left‐ right‐left, and [5] walked and did not run across the street) were scored (yes = 1; no = 0) and a composite score (the sum from 0 to 5 of the 5 behaviors) were determined for each school.	This study used cross‐sectional school‐level data of pedestrian behavior of children of any grade and not longitudinal data on just the fourth graders who were in the WSB study. This could have diluted the impact of the intervention because many of the observed children at intervention schools did not use the WSB program. However, this enabled “real world” observations.
Moodie et al11	1 year	7840 (modeled)	To assess from a societal perspective the effects on BMI and disability adjusted life years (DALYs) of the Victorian WSB program if applied throughout Australia. Cost offsets and DALY benefits were modeled until the eligible cohort reached 100 years of age or death.	This study used a logic pathway to model the effects on BMI and DALYs of the Victorian WSB program if applied throughout Australia. Cost offsets and DALY benefits were modeled until the eligible cohort reached 100 years of age or death. The reference year was 2001. Second stage filter criteria (“equity,” “strength of evidence,” “acceptability,” “feasibility,” “sustainability,” and “side‐effects”) were assessed to incorporate additional factors that impact on resource allocation decisions.	Reports of missing data for 7 of the 33 local governments with WSB Programs therefore program activity may have been underestimated. Evaluation assumed the intervention to be in steady state (ie, implemented and working at its full effectiveness), but authors found unutilized capacity. The authors report that the number of participating local governments, schools, WSBs per school, and children per WSB could all be increased without expansion of infrastructure capacity at national, state, local government, or school level.
Sirard et al12	1 week	11 (5/6)	To (1) test the feasibility of a WSB as an intervention strategy to increase children's physical activity and (2) to identify any changes in physical activity from walking to school.	A randomized, controlled trial design. All participants attended an elementary school in Menlo Park, CA, USA and completed 1 week of baseline automobile commuting. After randomization, control group continued to be driven to school and intervention group used the WSB (route = 1.1 km). Inclusion criteria were that participants had to be in third‐fifth grade and were driven to school 4 or more days per week. Twelve students were randomized to control (N = 6) or intervention (N = 6) group but 1 had monitor malfunction, leaving only 5 in the intervention group. Students were instructed to wear Acti‐Graph monitors for 14 consecutive days except when swimming, bathing, or sleeping and to maintain normal activities during baseline week. Accelerometer data were summarized for total week and weekdays. Each weekday was also divided into 4 time blocks (before, during, afterschool, evening). Average monitor counts per minute and the average percentage of time spent in moderate‐to vigorous physical activity (%MVPA) were calculated for each student.	This was a very small study with only 5 students in the intervention group and 6 in the control group. Also this study was carried out over a very short time span.

N, number of children; WSB, walking school bus; RCT, randomized controlled trial.