was worn. The BSA protected varied

was worn. The BSA protected varied from 10% to 100%
for boys and 25% to 100% for girls.
Statistical Analysis
The primary outcome of interest was change in the
mean percent of BSA protected after 1 and 2 years of
adolescent exposure to the intervention package. Protection
levels of adolescents in intervention communities
were compared with adolescents of the same grades
in control communities at baseline (summer 2001) and
follow-up years (summers 2002–2004). All subjects
were considered baseline in 2001, as were subjects entering
the 6th grade in subsequent years. In 2002, adolescents
entering the 7th or 8th grade, and in years 2003
and 2004 students entering the 7th grade, were classified
at “1-year follow-up.” In 2003 and 2004, students entering
the 8th grade were classified at “2-year followup.”
Thus, students who received 1 or 2 years of intervention
exposure were compared with students of the
same grade in control communities in each data-collection
period.
Descriptive statistics (eg, means and SDs) were used
to compare the characteristics of the cross-sectional
samples and observation environment according to exposure/grade
group and intervention status. Specifically,
we compared environmental factors (temperature, UV
index) and adolescent characteristics (skin type, gender).
Items that differed between intervention and control
sites during any of the time periods were included as
adjustment factors in the multiple regression analysis.
The primary analysis consisted of a multiple linear
regression analysis that predicted the percent of BSA
protected for each adolescent. Covariates included calendar
year of assessment, intervention status, and the
interaction between these factors. Additional covariates
included adjustment factors as appropriate. A randomeffect
factor was included to control for community clustering.
Effect modification for gender and UV-exposure
level was evaluated including appropriate interaction
terms in the regression model. On the basis of the fitted
model, we computed the adjusted mean protection level
for the adolescents in the intervention and control
groups according to measurement time point. Subject
reports of sunscreen application and protection-advice
variables were analyzed by the 2 test for proportions
and the Wilcoxon rank-sum test for continuous variables.
RESULTS
Of the 13 middle schools approached, 10 principals
agreed to participate. Of the 14 primary care practices
approached in the intervention communities, 11 agreed
to participate in the intervention. This represents 80% of
the providers that school nurses identified as caring for
the majority of their middle school students. All athletic
and recreation programs in the intervention communities
agreed to participate. Overall, there were an average
of 30 observation days per year with an average of 10
that occurred at school-sponsored events. In the 5 intervention
schools, requests for the project’s curricular materials
ranged from 21% to 53% of the teachers. In the
following 2 years, 39% (132 of 323) of all teachers in
these schools reported integrating teaching about solar
radiation and skin cancer or sun protection into their
subjects.
Town populations varied from 6300 to 34 000, with
no significant difference in mean student-body size between
intervention/control-town pairs. Of the population
of middle school students, 94% were white. Information
about the school population was available from a
concurrent survey conducted at baseline in the 10
schools on 20% of the students (n
827). Seventy
percent of middle school students participated in spring
or fall athletics, and 36% reported a visit to their primary
care provider in the previous year. Adolescent refusal to
be interviewed at the beach/pool was rare (1% per
observation session). Characteristics of the adolescent
population observed and observation conditions during
each of observation time periods (baseline and at 1 and
2 years) are listed in Table 1 for the intervention and
control communities. There was considerable variation
in the UV index and temperature across years resulting
from weather-condition variation during New England
summers with high UV-exposure levels with moderate
temperatures.
The main effect examined in the study was whether
the intervention changed the decline in adolescent sun
protection known to occur during the middle school
years. Table 2 shows our full model of the independent
predictors of what percent of the body surface is protected.
Grade is not included as a predictor, because the
baseline and follow-up periods are closely linked to
grades 6, 7, and 8. Students who burned more easily had
higher levels of protection, and hotter weather resulted
in more unprotected skin.
A significant intervention effect was found after 2
years but not after 1 year of exposure to the intervention
(intervention status by time). In addition to this intervention
effect, the SunSafe intervention compared with
control was more effective in improv

ผลลัพธ์ (ไทย) 1:

was worn. The BSA protected varied from 10% to 100%for boys and 25% to 100% for girls.Statistical AnalysisThe primary outcome of interest was change in themean percent of BSA protected after 1 and 2 years ofadolescent exposure to the intervention package. Protectionlevels of adolescents in intervention communitieswere compared with adolescents of the same gradesin control communities at baseline (summer 2001) andfollow-up years (summers 2002–2004). All subjectswere considered baseline in 2001, as were subjects enteringthe 6th grade in subsequent years. In 2002, adolescentsentering the 7th or 8th grade, and in years 2003and 2004 students entering the 7th grade, were classifiedat “1-year follow-up.” In 2003 and 2004, students enteringthe 8th grade were classified at “2-year followup.”Thus, students who received 1 or 2 years of interventionexposure were compared with students of thesame grade in control communities in each data-collectionperiod.Descriptive statistics (eg, means and SDs) were usedto compare the characteristics of the cross-sectionalsamples and observation environment according to exposure/gradegroup and intervention status. Specifically,we compared environmental factors (temperature, UVindex) and adolescent characteristics (skin type, gender).Items that differed between intervention and controlsites during any of the time periods were included asadjustment factors in the multiple regression analysis.การวิเคราะห์หลักประกอบด้วยหลายแบบเชิงเส้นวิเคราะห์การถดถอยที่ทำนายเปอร์เซ็นต์ของบีเอสเอป้องกันสำหรับวัยรุ่นแต่ละ Covariates รวมปฏิทินปีที่ประเมิน สถานะการแทรกแซง และปฏิสัมพันธ์ระหว่างปัจจัยเหล่านี้ Covariates เพิ่มเติมรวมปัจจัยการปรับปรุงตามความเหมาะสม Randomeffect การปัจจัยรวมอยู่ในการควบคุมคลัสเตอร์ชุมชนผลการเปลี่ยนแปลงเพศและแสง UVระดับที่ประเมินรวมทั้งการโต้ตอบที่เหมาะสมเงื่อนไขในแบบจำลองถดถอย โดยการจัดรูปแบบ ที่เราคำนวณระดับการป้องกันหมายถึงปรับปรุงสำหรับวัยรุ่นในการแทรกแซงและควบคุมกลุ่มตามจุดเวลาวัด ชื่อเรื่องรายงานการใช้ครีมกันแดดและคำแนะนำการป้องกันตัวแปรที่วิเคราะห์ โดยการทดสอบสัดส่วน 2และทดสอบ Wilcoxon ลำดับผลรวมสำหรับตัวแปรต่อเนื่องผลลัพธ์โรงเรียนกลาง 13 ประดับ แบบที่ 10ตกลงที่จะเข้าร่วม แนวทางปฏิบัติหลัก 14ประดับในชุมชนแทรกแซง 11 ตกลงการมีส่วนร่วมในการแทรกแซง นี้แสดงถึง 80% ของผู้ให้บริการที่โรงเรียนพยาบาลเป็นการดูแลส่วนใหญ่เด็กนักเรียนมัธยม แอธเลติกทั้งหมดและโปรแกรมนันทนาการในชุมชนแทรกแซงตกลงที่จะเข้าร่วม โดยรวม มีค่าเฉลี่ยเก็บข้อมูล 30 วันต่อปีโดยเฉลี่ย 10ที่เกิดขึ้นในโรงเรียนสนับสนุนกิจกรรม ในการแทรกแซง 5schools, requests for the project’s curricular materialsranged from 21% to 53% of the teachers. In thefollowing 2 years, 39% (132 of 323) of all teachers inthese schools reported integrating teaching about solarradiation and skin cancer or sun protection into theirsubjects.Town populations varied from 6300 to 34 000, withno significant difference in mean student-body size betweenintervention/control-town pairs. Of the populationof middle school students, 94% were white. Informationabout the school population was available from aconcurrent survey conducted at baseline in the 10schools on 20% of the students (n
827). Seventypercent of middle school students participated in springor fall athletics, and 36% reported a visit to their primarycare provider in the previous year. Adolescent refusal tobe interviewed at the beach/pool was rare (1% perobservation session). Characteristics of the adolescentpopulation observed and observation conditions duringeach of observation time periods (baseline and at 1 and2 years) are listed in Table 1 for the intervention andcontrol communities. There was considerable variationin the UV index and temperature across years resultingfrom weather-condition variation during New Englandsummers with high UV-exposure levels with moderatetemperatures.The main effect examined in the study was whetherthe intervention changed the decline in adolescent sunprotection known to occur during the middle schoolyears. Table 2 shows our full model of the independentpredictors of what percent of the body surface is protected.Grade is not included as a predictor, because thebaseline and follow-up periods are closely linked togrades 6, 7, and 8. Students who burned more easily hadhigher levels of protection, and hotter weather resultedin more unprotected skin.A significant intervention effect was found after 2years but not after 1 year of exposure to the intervention(intervention status by time). In addition to this interventioneffect, the SunSafe intervention compared withcontrol was more effective in improv