ArticlePDF Available

Re: Hutchinson Smoking Prevention Project: Long-Term Randomized Trial in School-Based Tobacco Use Prevention--Results on Smoking

Re: Hutchinson Smoking
Prevention Project: Long-Term
Randomized Trial in School-
Based Tobacco Use Prevention—
Results on Smoking
The Hutchinson Smoking Prevention
Project (HSPP), whose findings were
published recently (1), was a well-
designed trial that was conducted from
1984 to 1999. The intervention tested
was social-influences based and was de-
livered to cohorts of youth from grades 3
to 10 in 20 of 40 randomly assigned
school districts (n 8388). No statisti-
cally significant differences were found
in prevalence of smoking between stu-
dents in program and control districts in
the 12th grade and 2 years after high
school. The editorial accompanying the
article (2) and recent media accounts
may lead people to conclude that the
social-influence approach is ineffective.
That conclusion is unwarranted in light
of the larger prevention literature. A re-
cent meta-analysis by Tobler et al. (3) of
207 universal school-based drug preven-
tion programs, including 138 prevention
programs that included social-influence
components, provides strong support for
the efficacy of social-influence program-
ming at a 1-year follow-up. The results
of HSPP must be interpreted within the
context of all of these other studies,
many of which were methodologically
rigorous, included large sample sizes,
and collectively provide compelling evi-
dence that some school-based preven-
tion approaches are effective.
To draw meaningful conclusions
about HSPP findings and their implica-
tions, we need much more information
about the study data and outcomes that
is not contained in the current article. In
the absence of such information, it is
worth considering possible alternative
interpretations of these data. For ex-
ample, it is possible that social-influence
approaches do not work equally well
with all youth (e.g., HSPP was con-
ducted at small schools in primarily ru-
ral settings with primarily white youth).
It is possible that youth did not like this
particular program, since no process
data are reported from youth. It is not
clear how representative the HSPP in-
tervention is of other prevention ap-
proaches that include a focus on pro-
smoking social influence. In particular,
there should have been immediate ef-
fects on variables specific to the social-
influence process, such as reductions in
tobacco prevalence overestimates or
perceived acceptability of tobacco use,
that are not discussed. Failure to affect
these variables would reflect program
failure, not the failure of the general so-
cial-influence approach.
Also, the HSPP curriculum did not
include several currently discussed ele-
ments of the social-influences approach
that might be important to the achieve-
ment of prevention effects, such as lis-
tening and communication skills, deci-
sion making, and making a commitment
(4). HSPP appears to be an example of
what Tobler et al. (3) label as “social-
influence” programming (minus a pub-
lic commitment component) rather than
more comprehensive approaches, such
as “comprehensive life skills program-
ming” [e.g., Life Skills Training or Proj-
ect Towards No Tobacco Use (4,5)]or
“system-wide programming” [e.g., Proj-
ect Northland (3)]. The effect sizes of
non-social-influence programs are 0.
While the effect sizes of programs that
contain social-influence components are
in the small-to-moderate range (mean
effect size .12 to .27; standard devia-
tion .25), comprehensive life skills
programming (i.e., social influence plus
life skills material) is 40% more effec-
tive than social influence-only program-
ming. Furthermore, system-wide pro-
gramming (i.e., social influence or
social influence plus life skills program-
ming, delivered through multiple com-
munity units) is 60% more effective
than comprehensive life skills program-
Most important, it is not clear wheth-
er the HSPP is the case of a program
whose effects later decay or if no effects
ever occurred. Maintenance of program
effects is a central issue in prevention
(5). The HSPP findings (1) provide con-
vincing evidence that this program’s ap-
proach was not effective in the long run.
However, given that more than 400 000
people die of smoking-related diseases
each year and 3000 youth begin smok-
ing each day, the results of one study
should not be allowed to deter efforts to
promote the use of smoking-prevention
approaches for which there is convinc-
ing evidence of effectiveness.
(1) Peterson AV, Kealey KA, Mann SL, Marek
PM, Sarason IG. Hutchinson Smoking Preven-
tion Project: long-term randomized trial in
school-based tobacco use prevention—results
on smoking. J Natl Cancer Inst 2000;92:
(2) Clayton RR, Scutchfield FD, Wyatt SW.
Hutchinson Smoking Prevention Project: a
new gold standard in prevention science re-
quires new transdisciplinary thinking [edito-
rial]. J Natl Cancer Inst 2000;92:1964–65.
(3) Tobler NS, Roona MR, Ochshorn P, Marshall
DG, Streke AV, Stackpole KM. School-based
adolescent drug prevention programs: 1998
meta-analysis. J Prim Prev 2000;20:275–336.
(4) Botvin GJ. Preventing drug abuse in schools:
social and competence enhancement ap-
proaches targeting individual-level etiologic
factors. Addict Behav 2000;25:887–97.
(5) Donaldson SI, Sussman S, MacKinnon DP,
Severson HH, Glynn T, Murray DM, Stone
EJ. Drug abuse prevention programming: do
we know what content works? Am Beh Sci-
entist 1996;39:868–83.
G. J. Botvin is the copyright holder of the Life
Skills Training program and has a publication
agreement with Princeton Health Press.
Affiliations of authors: S. Sussman, Institute
for Health Promotion and Disease Prevention
Research, University of Southern California, Los
Angeles; W. B. Hansen, Tanglewood Research,
Greensboro, NC; B. R. Flay, Health Research and
Policy Centers, University of Illinois at Chicago;
G. J. Botvin, Weill Medical College of Cornell
University, New York.
Correspondence to: Steve Sussman, Ph.D.,
Professor of Preventive Medicine and Psychology
and Institute for Health Promotion and Disease
Prevention Research, University of Southern
California/IPR, 1000 S. Fremont, Unit #8, Rm.
4124, Bldg. A-4, Alhambra, CA 91803 (e-mail:
Peterson et al. (1) found no evidence
for the long-term effectiveness of a
school-based social-influences approach
to smoking prevention. Appropriately,
this finding received considerable media
attention. It will, and should, have an
impact on decision makers responsible
for directing smoking-prevention initia-
tives. As Clayton et al. (2) noted in an
accompanying editorial, the results from
the Hutchinson Smoking Prevention
Journal of the National Cancer Institute, Vol. 93, No. 16, August 15, 2001 CORRESPONDENCE 1267
by guest on January 5, 2012 from
Project (HSPP) are compelling because
of the quality of the design and imple-
Does the HSPP study mean that
school-based social-influences smoking
prevention programs should be aban-
doned? Or do programs of this type still
have a place in a comprehensive ap-
proach to smoking prevention? If so, un-
der what circumstances are they useful?
These are key questions for both re-
search and practice.
In their editorial, Clayton et al. (2)
noted that we must reorient the preven-
tion field away from asking the main
effects question (what works?) to the
moderated model question (what works,
for whom, under what conditions, how,
and why?). This call echoes the previ-
ously expressed concern (3) that out-
comes of prevention programs be exam-
ined not only as a function of program
content but also in relation to other vari-
ables that may affect outcomes (notably,
characteristics of program participants,
providers, and settings).
Setting factors may be very important
in determining the impact of an inter-
vention. For instance, in a recent study
of a school-based social-influences smok-
ing prevention program, we found no
significant differences in smoking rates
between students in the intervention and
comparison conditions at the end of
grade 8 if data were pooled across all
schools within conditions (4). But a more
fine-grained analysis revealed a differ-
ent picture. Smoking norms vary widely
across school settings, and to avoid con-
founding treatment conditions with
school-level risk, we had, therefore,
stratified schools on the basis of risk
(high-risk schools were characterized by
a relatively high prevalence of smoking
among grade-8 students when the inter-
vention cohort was in grade 6). This
stratification, which made it possible to
examine the interaction between inter-
vention and school risk, revealed that
the intervention had a substantial impact
in reducing smoking in high-risk schools
but not in other schools (4).
Our data thus tangibly illustrate the
importance of Clayton et al.s call for
more nuanced examination of interven-
tion effects. If the results of our study
(4) are robust, they suggest that selective
deployment of school-based programs
in high-risk settings may be a valuable
component of a comprehensive smok-
ing-prevention strategy.
Policy questions about tobacco pre-
vention are too important to base on any
single study, no matter how well de-
signed. We urgently need a careful, bal-
anced review of the evidence and prac-
tice recommendations.
(1) Peterson AV, Kealey KA, Mann SL, Marek
PM. Hutchinson Smoking Prevention Project:
long-term randomized trial in school-based
tobacco use preventionresults on smoking.
J Natl Cancer Inst 2000;92:197991.
(2) Clayton RR, Scutchfield FD, Wyatt SW.
Hutchinson Smoking Prevention Project:
a new gold standard in prevention science re-
quires transdisciplinary thinking [editorial].
J Natl Cancer Inst 2000;92:19645.
(3) Best JA, Thomson SJ, Santi SM, Smith EA,
Brown KS. Preventing cigarette smoking
among school children. Ann Rev Public
Health 1988:9;161201.
(4) Cameron R, Brown KS, Best JA, Pelkman CL,
Madill CL, Manske SR, et al. Effectiveness of
a social influences smoking prevention pro-
gram as a function of provider type, training
method, and school risk. Am J Public Health
Affiliations of authors: R. Cameron (Canadian
Cancer Society/National Cancer Institute of
Canada Centre for Behavioural Research and
Program Evaluation), K. S. Brown (Statistics and
Actuarial Science), University of Waterloo, ON,
Canada; J. A. Best, Centre for Clinical Epidemi-
ology and Evaluation, Vancouver Hospital and
Health Sciences Centre, BC, Canada.
Correspondence to: Roy Cameron, Ph.D., Ca-
nadian Cancer Society/National Cancer Institute
of Canada Centre for Behavioural Research and
Program Evaluation, Lyle Hallman Institute,
University of Waterloo, 200 University Ave., W.,
Waterloo, ON N2L 3G1, Canada (e-mail:
Peterson et al. (1) recently reported a
randomized controlled trial of a social-
influences approach to smoking preven-
tion administered in schools annually
from grades 3 to 10. They found no evi-
dence of reduced smoking prevalence,
either at grade 12 or 2 years after high
school. This result is in contrast to that
of Botvin et al. (2), who in a previous
randomized school-based trial of an in-
tervention that was administered inten-
sively in grades 79 and that also aimed
at teaching skills for resisting social in-
fluences to use tobacco and other sub-
stances, found a statistically significant
reduction in smoking at high school
graduation. Peterson et al. noted but did
not discuss the difference in outcome
between their study and the study by
Botvin et al. and, in fact, they concluded
that, consistent with previous trials,
their negative results showed that a
school-based social-influences approach
is ineffective in deterring smoking.
What could account for this differ-
ence in results? Although Peterson et al.
offered no explanation, Clayton et al.
(3), in their accompanying editorial, im-
plied that the higher attrition rate, 39%
in Botvin et al. versus 7% in Peterson et
al., made the results of Botvin et al.
problematic. However, Botvin et al. had
used statistical techniques that revealed
no differential effects related to attrition
in the study groups. Clayton et al. pro-
vided no other specific criticisms of
Botvin et al., concluding, like Peterson
et al., that the social-influences ap-
proach does not work.
An alternative explanation for the
difference, however, might be that the
instructional materials and techniques of
Botvin et al. were more effective than
those of Peterson et al. in preventing
smoking. No easy method of testing this
possibility offers itself, but I believe we
should not yet discard the social-
influences approach to prevention of
smoking initiation in youth.
(1) Peterson AV Jr, Kealey KA, Mann SL, Marek
PM, Sarason IG. Hutchinson Smoking Preven-
tion Project: long-term randomized trial in
school-based tobacco use preventionresults
on smoking. J Natl Cancer Inst 2000;92:
(2) Botvin GJ, Baker E, Dusenbury L, Botvin
EM, Diaz T. Long-term follow-up results of
a randomized drug abuse prevention trial in a
white middle-class population. JAMA 1995;
(3) Clayton RR, Scutchfield FD, Wyatt SW.
Hutchinson smoking prevention project: a new
gold standard in prevention science requires
new transdisciplinary thinking [editorial].
J Natl Cancer Inst 2000;92:19645.
Correspondence to: Harry A. Bliss, M.D.,
Maine Medical Center, 22 Bramhall St., Portland,
ME 04102 (e-mail:
1268 CORRESPONDENCE Journal of the National Cancer Institute, Vol. 93, No. 16, August 15, 2001
by guest on January 5, 2012 from
The comments and questions regard-
ing the Hutchinson Smoking Preven-
tion Project (HSPP) trial raised by
Sussman et al., Cameron et al., and Bliss
address four main issues: 1) the scien-
tific question addressed by the HSPP
trial, 2) appropriate conclusions to be
drawn from the trial results, 3) com-
parison of HSPP results with those of
other studies, and 4) implications for
the future.
1) The HSPP scientific question.
The HSPP trial was designed to deter-
mine to what extent a comprehensive
grade 310 social-influences curriculum
could deter smoking in youth at the end
of the age period of child/adolescent
smoking acquisition. The HSPP results
pertain only to the intervention approach
tested in the HSPP trialthe social-
influences approach [e.g., (1)]and not
to the life skills training approach [e.g.,
(2)], brought up by Sussman et al. and
by Bliss, or to school-plus-community-
wide intervention (i.e., system-wide
programming), brought up by Sussman
et al.
In accordance with the trials goal,
its main endpoints were smoking at
12th grade and at 2 years after high
school. So, it is appropriate that the re-
port of trial results focused on these
long-term endpoints, on which the an-
swer to the scientific questionthe suc-
cess or failure of the intervention in
deterring smoking at the end of the
period of child/adolescent smoking ac-
quisitionrests, rather than on data
at younger ages, before the end of
the intervention period, for indication
of delayed onset (as suggested by
Sussman et al.). Analyses of the latter
data are planned as part of the trials
secondary analysis of smoking acquisi-
tion during the period from grades 3
to 12.
2) Appropriate conclusions to be
drawn from the HSPP trial. Because
of the high rates of implementation and
follow-up and other achievements re-
lated to scientific integrity, alternative
explanations for the findings were ruled
out. In response to Sussman et al.s
question about students interest in the
curriculum, both the teachers (via post-
implementation interviews and a self-
report questionnaire) and HSPP staff
(via classroom observations) reported
that the youth were engaged by and
interested in the classroom activities
[e.g., (3)].
We agree with Sussman et al. that,
because the HSPP trial was not per-
formed in urban areas or in predomi-
nately minority populations, the trial
itself can make no conclusion specifi-
cally about these populations. Neverthe-
less, the study population was demo-
graphically and geographically diverse,
covering 40 different rural and suburban
communities with a wide range of socio-
economic status, smoking prevalence, and
percent of minorities (4).
3) Comparison of HSPP results
with other studies. Sussman et al. are
concerned about how the HSPP trial re-
sults compare with those of other trials
reported in a recent meta-analysis by
Tobler and colleagues. But the appropri-
ate comparison for the HSPP trial is
with trials that investigated the same
scientific question. In contrast to the
many studies included in the meta-
analysis, which included a variety of dif-
ferent intervention approaches, a num-
ber of different targeted substances
(alcohol, illegal drugs, and cigarettes),
and studies with short-term endpoints,
only four trials [HSPP and those in
(5–7)] evaluated the long-term impact
of the social-influences approach for
the prevention of smoking. In none of
these trials was there any evidence of
long-term impact of the intervention.
Thus, the null results from the HSPP
trial are consistent with the relevant
4) Implications for the future. Our
findings have implications for both re-
search and public health practice. Con-
cerning implications for research, future
work is needed in the following areas:
(a) additional analyses of data from
long-term trials and longitudinal studies,
both to investigate why the social-
influences approach hasnt worked and
to obtain information on process and
risk factors to inform future intervention
development. In particular, related to
comments by Sussman et al., (i) prelimi-
nary analysis of variables targeted by
the HSPP intervention (e.g., knowledge,
beliefs, self-efficacy, and perception of
smoking norms) reveals that the HSPP
intervention did impact these variables
(as we reported at the March 2001 meet-
ing of the Society of Behavioral Medi-
cine) and (ii) analyses are planned to
evaluate various aspects of the process
of smoking acquisition among the HSPP
cohort, including comparisons of smok-
ing prevalence early in the smoking ac-
quisition process. Also needed are (b)
additional theory development for smok-
ing acquisition, informed and guided by
data, and (c) future trials, once promis-
ing new approaches have been identi-
fied. In this context, we agree with Cam-
eron et al. that one should not assume
that interventions will necessarily work
in all environments. As they point out,
for example, the characteristics of the
schools (e.g., high-risk versus low-risk)
and of youth (e.g., rebellious versus not
rebellious) would need to be considered,
first in theory development and then in
sharpened scientific questions for inter-
vention research.
Concerning implications for public
health practice, as stated eloquently and
accurately by Sussman et al., without
enlightened and effective action now
our nations youth will continue to take
up smoking and die in disheartening
numbers. The conclusion from the
HSPP trial and the other three trials that
evaluated the long-term impact of the
social-influences approach for school-
based smoking prevention seems clear:
The social influences approach has
worked no better than the usual health
curricula already in the schools in long-
term deterrence of youth smoking. In
our study, one quarter of 12th grade
youth smoked daily, regardless of the
presence of a comprehensive social-
influences intervention.
So, what to do now to help our na-
tions youth? Unfortunately, the HSPP
trial does not give the answer to this
question. The answer must come from
positive experiences from other inter-
vention approaches. For example, the
experience and results from statewide
initiatives in California and other states
indicate that a comprehensive statewide
anti-tobacco program may be effective.
Also, it is clear that parents smoking
is an important risk factor for their chil-
dren becoming smokers. Accordingly,
action to inform and change the behav-
ior of parents who smoke is worthy of
Noteworthy from the HSPP trial was
the tremendous cooperation received
from youth, parents, and teachers, which
helped to provide the excellent data par-
ticipation and implementation rates that
contributed to the trials high scientific
integrity. For the nation, such coopera-
tion is a promising sign that the citizenry
Journal of the National Cancer Institute, Vol. 93, No. 16, August 15, 2001 CORRESPONDENCE 1269
by guest on January 5, 2012 from
may cooperate in new initiatives to help
our nations youth avoid tobacco. Our
responsibility is to ensure that such ini-
tiatives are based on the best evidence
available. In particular, from the null
results of the HSPP trial and three
other trials that investigated long-
term impact of the social-influences
approach, there is strong evidence
that a school-based, social-influences
approach alone is unlikely to be the
(1) Flay BR. Psychosocial approaches to smoking
prevention: a review of findings. Health Psy-
chol 1985;4:44988.
(2) Botvin GJ, Eng A, Williams CL. Preventing
the onset of cigarette smoking through life
skills training. Prev Med 1980;9:13543.
(3) Kealey KA, Peterson AV Jr, Gaul MA, Dinh
KT. Teacher training as a behavior change
process: principles and results from a longitu-
dinal study. Health Educ Behav 2000;27:
(4) Mann SL, Peterson AV Jr, Marek PM, Kealey
KA. The Hutchinson Smoking Prevention
Project trial: design and baseline characteris-
tics. Prev Med 2000;30:48595.
(5) Ellickson PL, Bell RM, McGuigan K. Pre-
venting adolescent drug use: long-term results
of a junior high program. Am J Public Health
(6) Flay BR, Koepke D, Thomson SJ, Santi S,
Best JA, Brown KS. Six-year follow-up
of the first Waterloo school smoking preven-
tion trial. Am J Public Health 1989;79:
(7) Murray DM, Pirie P, Leupker RV, Pallonen U.
Five- and six-year follow-up results from four
seventh-grade smoking prevention strategies.
J Behav Med 1989;12:20718.
Affiliations of authors: A. V. Peterson, Jr., Can-
cer Prevention Research Program, Division of
Public Health Sciences, Fred Hutchinson Cancer
Research Center, Seattle, WA, and Department of
Biostatistics, University of Washington, Seattle;
K. A. Kealey, S. L. Mann, P. M. Marek, Cancer
Prevention Research Program, Division of Public
Health Sciences, Fred Hutchinson Cancer Re-
search Center; I. G. Sarason, Cancer Prevention
Research Program, Division of Public Health Sci-
ences, Fred Hutchinson Cancer Research Center,
and Department of Psychology, University of
Correspondence to: Arthur V. Peterson, Jr.,
Ph.D., Division of Public Health Sciences,
Fred Hutchinson Cancer Research Center, MP-
603, 1100 Fairview Ave., N., P.O. Box 19024,
Seattle, WA 981091024 (e-mail: avpeters@
Bliss interprets our editorial as saying
that the higher attrition rate in the study
by Botvin et al. (1) [39%, versus 7%
in the study by Peterson et al. (2)]
makes Botvin et al.s results question-
able. Although he does note, correctly,
that Botvin et al.s statistical analysis
showed that there were no differential
effects related to attrition in the study
groups, he missed our point. Our point
was that the Peterson et al. study repre-
sents a new gold standard in prevention
science with regard to the implementa-
tion of a randomized trial, one index
of which is the difference in attrition
Bliss then offers another explanation
as to why Botvin et al. (1) found statis-
tically significant reductions in several
measures of smokingi.e., that their in-
structional materials and techniques
were more effective than those used by
Peterson et al. (2). Although this is cer-
tainly a plausible hypothesis, it seems
unlikely, for several reasons. First, the
Hutchinson Smoking Prevention Project
(HSPP) curriculum started with children
when they were younger [3
versus 7
grade in Botvin et al. (1)].
Second, it embodied in each year all
15 essential elements of school-based,
curriculum-driven smoking prevention
recommended by the National Cancer
Institute/National Centers for Disease
Control and Prevention. Third, because
the HSPP program was delivered from
the 3rd through the 10th grades, it had
considerably more opportunities for re-
inforcement of lesson materials than
the curriculum used by Botvin et al.,
which had lessons in the 7th through the
9th grades only. Finally, prevention sci-
entists generally agree that prevention
curriculum materials need to be deliv-
ered interactively rather than in a didac-
tic fashion, and both programs were in-
teractive. So, there must be other
explanations for the different findings of
the two studies.
One possible explanation was the ba-
sis of the comment by Cameron et al.
(3), who call attention to results from a
study they conducted. A first look at the
data from a main-effects perspective re-
vealed no statistically significant differ-
ences between intervention and com-
parison conditions at the end of grade 8,
after pooling data within conditions.
These authors report that a more fine-
grained analysis looking for moderator
effects revealed wide variation in smok-
ing norms across school settings. By ex-
amining the interaction between inter-
vention and school risk, the authors
found that the intervention had a sub-
stantial impact in reducing smoking in
high-risk schools but not in other
schools. Students are nested in the envi-
ronmental context of the school they at-
tend, and there is good reason to believe
that schools differ from each other on a
number of dimensions, most of which
have been largely ignored in the school-
based curriculum-driven part of preven-
tion science.
In our editorial, we noted that one
of the most interesting findings from
the HSPP was the substantial difference
in the prevalence of daily smoking
within conditions. For example, among
-grade females in the 20 school
districts in the control condition, daily
smoking ranged from 0% to 41.9%.
If there is such heterogeneity across
districts even within one condition, het-
erogeneity must be even greater across
the 72 schools in the HSSP study and
the 56 schools in the study by Botvin
et al. (1). Heterogeneity at the macro
level, like individual differences, is
the norm, not the exception, and must be
taken into account. We believe the
explanation offered by Cameron et al.
provides an important clue for future
prevention science and programming
Finally, we agree with both Bliss and
Cameron et al. that policy questions
about prevention are too important to be
based or changed on the results of any
one study. The findings from the HSPP
study should, however, be seen as an
opportunity to begin addressing major
public policy questions. For example,
are the marginal effects of school-based,
curriculum-driven prevention program-
ming large enough to justify taking that
many hours away from more traditional
academic programming? What and how
strong are the effects of prevention pro-
gramming on academic achievement?
We would argue that, if prevention pro-
grams do not have positive effects on the
1270 CORRESPONDENCE Journal of the National Cancer Institute, Vol. 93, No. 16, August 15, 2001
by guest on January 5, 2012 from
academic purpose of the schools, there
is reason to question their place in
(1) Botvin GJ, Baker E, Dusenbury L, Botvin
EM, Diaz T. Long-term follow-up results of
a randomized drug abuse prevention trial in a
white middle-class population. JAMA 1995;
(2) Peterson AV, Kealey KA, Mann SI, Marek
PM. Hutchinson Smoking Prevention Project:
long-term randomized trial in school-based
tobacco use preventionresults on smoking.
J Natl Cancer Inst 2000;92:197991
(3) Cameron R, Brown KS, Best AJ, Pelkman CL,
Madill CL, Manske SR, et al. Effectiveness
of a social influences smoking prevention
program as a function of provider type, train-
ing method, and school risk. Am J Public
Health 1999;89:182731.
Affiliations of authors: Kentucky Prevention
Research Center and Kentucky School of Public
Health, University of Kentucky, Lexington.
Correspondence to: Richard R. Clayton, Ph.D.,
Kentucky School of Public Health, University of
Kentucky, 2365 Harrodsburg Rd., Suite B100,
Lexington, KY 40504 (e-mail: clayton@pop.
Journal of the National Cancer Institute, Vol. 93, No. 16, August 15, 2001 CORRESPONDENCE 1271
by guest on January 5, 2012 from
... Peterson [31] J Natl Cancer Inst, 2000 Hutchinson Smoking Prevention Project, WA 1984–1999 Shean [32] Aust J Public Health, 1994; Armstrong et al., Med J Aust, 1990 52 University of Minnesota (social consequences curriculum), Western Australia 1981–1988 Shope [33] ...
... and individual subjects [18] [22] [40] were also used as the unit of assignment. Two quasi-experimental studies selected community as the assignment unit [25] [30], and the Hutchinson Smoking Prevention Project [31] randomly assigned entire school districts to treatment conditions. The latter study specifically indicated that it sought to minimize social mixing of subjects between treatment groups by its assignment protocol. ...
... Of these 24 studies, cigarette smoking was typically assessed by questions relating to the frequency of lifetime, monthly, weekly, or daily use. In the vast majority of studies, each frequency measure was recoded into a dichotomous outcome (yes or no), with lifetime (ever, any, etc.) cigarette use [16] [18] [20] [21] [23] [24] [27] [31] [32] [35] [38] being the smoking category most often evaluated in this fashion, followed by monthly (30-day use) [18] [19] [22] [23] [26] [29] [31] [37] [38] and weekly [19] [23] [25] [28] [29] [31] [35] [37] [38], and then daily use [18 –20,23,25,28,38]. Some studies evaluated smoking by Table 3 Methodological design Investigators Smoking categories assessed (in treatment vs control group comparisons) Biochemical validation Abernathy [16] Never vs any NR Aveyard [17] Regular weekly (weighted average of one or more cigarettes per week and at least one cigarette per day)—yes/no NR Bergamaschi [18] Non smoker—yes/no; experimenter—yes/no; occasional (within past month)—yes/no; current (daily)— yes/no NR Botvin [19] Monthly—yes/no; weekly—yes/no; " heavy " cigarette smoking (a pack or more a day)—yes/no; tobacco and alcohol monthly—yes/no; tobacco and alcohol weekly—yes/no; tobacco and marijuana weekly—yes/no; tobacco, alcohol and marijuana monthly—yes/no; tobacco, alcohol and marijuana weekly—yes/no. ...
Although the initial effectiveness of psychosocial strategies programming in preventing smoking and other drug abuse among adolescents has been well established through literature reviews and meta-analyses, much less evidence exists for the long-term follow-up success of these interventions. The primary goal of this paper, therefore, is to summarize the effectiveness of published program evaluation studies that have followed adolescents across the transitional period between junior high and high school for a period of at least 2 years. Studies for inclusion in this review were accessed primarily through a computerized search of Medline, Healthstar, and PsychINFO databases. Intervention studies that met five core criteria were retained for review. Two authors independently abstracted data on study characteristics, methodology, and program outcomes. Search results yielded 25 studies suitable for examination. The majority of these studies reported significant program effects for long-term smoking, alcohol, and marijuana outcomes, while indicating a fairly consistent magnitude of program effects. This review provides long-term empirical evidence of the effectiveness of social influences programs in preventing or reducing substance use for up to 15 years after completion of programming. However, this conclusion is still somewhat tenuous given the lack of significant program effects reported in several studies and the great variability that existed in the level of internal and external validity across all studies.
... Kürzlich ist die Effektivität dieser präventiven Programme kritisch hinterfragt worden. Das "Hutchinson Smoking Prevention Project", eine methodisch aufwändige Längsschnittstudie mit mehr als 8000 Schülern, von denen die Hälfte eingebunden war in ein kontinuierliches Interventionsprogramm, das auch Elemente des Standfestigkeitstrainings beinhaltete und welches von der dritten bis zur zwölften Klassenstufe durchgeführt wurde, fand keine Unterschiede zwischen der Experimentalgruppe und der "unbehandelten" Kontrollgruppe in der Klassenstufe 12 und zwei Jahre später in Bezug auf das tägliche Rauchen [32][33][34][35]. Ferner konnten Hansen und Graham [36] Der Wettbewerb wird auf Klassenebene durchgeführt. ...
Hintergrund: Diese Studie beschreibt die Konzeption und Evaluation eines schulischen Wettbewerbs zum Nichtrauchen. Die teilnehmenden Klassen verpflichten sich, für ein halbes Jahr nicht zu rauchen. Klassen, die dieses Ziel erreichen, können eine Reihe attraktiver Preise gewinnen. Ziel der Studie war zu überprüfen, ob das Programm eine geeignete Maßnahme darstellt, den Einstieg in das Rauchen zu verzögern und rauchende Jugendliche zur Raucherentwöhnung zu motivieren. Methoden: Im Schuljahr 1998/1999 wurde eine Kontroll-Gruppen-Studie mit Messwiederholung durchgeführt. An der Studie nahmen 2142 Schüler mit einem durchschnittlichen Alter von 12,9 (SD = 0,98) Jahren teil, die zu zwei Messzeitpunkten über einen Zeitraum von 12 Monaten a) vor der Intervention und b) 6 Monate nach Ende der Intervention zu ihrem Rauchverhalten befragt wurden. Ergebnisse: Während sich die Experimentalgruppe und die Kontrollgruppe zur Baselinemessung nicht in ihrem Rauchverhalten unterschieden (15,2 % vs. 18,5 % Raucheranteil), rauchten zum Katamnesezeitpunkt signifikant mehr Schüler der Kontrollgruppe (32,9 % vs. 25,5 %). Bedeutsam mehr Jugendliche der Experimentalgruppe (82,5 % vs. 77,6 %) blieben auch in der Katamnese Nichtraucher. Statistisch bedeutsame sekundärpräventive Effekte des Programms konnten nicht festgestellt werden. Schlussfolgerungen: Die Ergebnisse legen die Schlussfolgerung nahe, dass der Wettbewerb geeignet ist, im Sinne der Primärprävention des Rauchens den Einstieg in das Rauchen zu verzögern bzw. zu verhindern. Der Wettbewerb scheint weniger geeignet als Raucherentwöhnungsprogramm für jugendliche Raucher.
... The recent failure of school curriculum-based smoking education interventions (Peterson et al. 2000), along with the decay of promising short-term drug intervention results (Flay et al. 1989;Bell et al. 1993), has led to some scepticism of school-based approaches. While critics have been quick to identify methodological flaws in Peterson et al. 's (2000) Hutchinson study in particular (Sussman et al. 2001), few have considered the possibility that the underlying abstinence messages may be inconsistent with students' lives and experiences (Resnicow & Botvin 1993;Kay 1994). While abstinence messages may be appropriate for younger adolescents to assist in preventing the onset of alcohol use, as use becomes normative among peers during mid-adolescence this message may lack relevance. ...
... Norum [30] describes how some actions, that are otherwise difficult to illustrate, can be effectively shown to an entire class using video. Most single-component programs [31] [32] [33] [34] do not appear to be effective. Thus, using a video education tool in conjunction with a story book is likely to produce greater effects, especially when delivered by health care providers in the school-based setting. ...
Full-text available
Introduction. Cigarette smoking contributes to the deaths of more than 400,000 Americans annually. Each day >3,000 children and adolescents become regular smokers. This paper details a new antitobacco educational program titled “AntE Tobacco” Method. Children in grades 1–3 were administered a 10-item questionnaire to ascertain their baseline knowledge about the ill effects of smoking, shown an educational cartoon video depicting the ill effects of tobacco, and given a story book based on the video. At the end of video, children were administered a questionnaire to determine short-term recall of the antitobacco educational objectives of the program. Four to 6 weeks later, the children were then administered a follow-up survey to determine long-term retention of the anti tobacco educational program. Result. Eighty two percent of the children answered the outcome questions correctly immediately following the video. At follow-up, 4–6 weeks later, 83% of children answered all questions correctly. Conclusion. The anti tobacco education program used in this study effectively conveyed most of the educational objectives. The results of this study indicate that a multimedia (i.e., video and book) educational program can be used to educate and reinforce anti tobacco messages. This program may be very useful as a part of a comprehensive anti tobacco curriculum in school systems.
Studies have shown that the effectiveness of programs or curricula may depend in part on who delivers the material. In adolescent health education programs, peer leaders are often recruited to implement programs because they are more persuasive to other adolescents than adults. Teachers also systematically vary how groups are constructed in school-based health education programs. This study compared the effects of three leader and group selection methods within the context of two tobacco prevention programs. Eight schools received a social influences program (Chips) and eight received a program with a multicultural emphasis (Flavor). Within these 16 schools 84 classrooms consisting of 1486 students were randomly assigned to one of three leader and group creation conditions: (i) leaders defined as those who received the most nominations by students and groups created randomly (random group), (ii) same as (i) but groups created by assigning students to the leaders they nominated (network), and (iii) leaders and groups created by teachers (teacher). One year follow-up data showed that main effects of the curriculum and network assignments were non-significant on smoking initiation when entered alone. Interaction terms of curriculum and assignment methods, however, were significant such that the network and teacher conditions were less effective than the random group condition with Chips, and more effective than random group condition with Flavor. These data show that school-based prevention programs should be evaluated in light of who implements the program. Even a peer-led program will be differentially effective based on how leaders are selected and how groups are formed, and this effect may be curriculum dependent.
This paper critically reviews the published evidence pertaining to Project Towards No Drug Abuse (Project TND). Publications from seven evaluation studies of Project TND are reviewed, and the results from these are discussed as related to the following outcomes: main effects on the use of cigarettes, alcohol and marijuana; main effects on the use of "hard drugs," defined in the evaluations as cocaine, hallucinogens, stimulants, inhalants, ecstasy and other drugs (e.g., depressants, PCP, steroids and heroin); subgroup and interaction analyses of drug use; and violence-related behaviors. Very few main effects have been found for cigarette, alcohol and marijuana use in the Project TND evaluations. While studies do report main effects for hard drug use, these findings are subject to numerous threats to validity and may be attributable to the data analyses employed. Similarly, while isolated subgroup and interaction effects were found for alcohol use among baseline nonusers and some violence-related behaviors in the early Project TND evaluations, these findings have not been replicated in more recent studies and may result from multiple comparisons between study conditions. In conclusion, there is little evidence to support the assertion that Project TND is an effective drug or violence prevention program. The broader implications of these findings for prevention science are discussed and suggestions are made as to how the quality of research in the field might be improved.
Full-text available
The current emphasis on best practices for school-based health and mental health programs brings with it the demand for evaluation efforts in schools. This article describes the challenges of launching a successful school program and evaluation, with lessons learned from three projects that focus on intimate partner violence. The authors discuss issues related to constraints on the research design in schools, the recruitment of schools and participants within schools, program and evaluation implementation issues, the iterative implementation-evaluation cycle, and the dissemination of programs and study findings. The authors emphasize the need for flexibility and cultural awareness during all stages of the process.
Full-text available
This paper reports six-year follow-up data from the first large-scale randomized trial of the social influences approach to smoking prevention. In 1979, 22 schools were randomly assigned to program or control conditions. Students in program schools received a social influences curriculum in six core and two maintenance sessions in grade 6, two booster sessions in grade 7, and one booster session in grade 8. All students were assessed at pretest (T1), immediate posttest (T2), end of grade 6 (T3), beginning and end of grade 7 (T4 and T5), end of grade 8 (T6), and grades 11 and 12 (T7 and T8). Ninety percent of study students were relocated and data obtained from over 80 percent of them at T8. Program effects on experimental smoking observed in grades 7 and 8 had completely decayed by T8, six years after the beginning of the program. Grade 6 smoking experience and social risk were each strong predictors of T8 smoking behavior. Subjects who had left school were smoking at more than twice the rate of subjects still in high school (grade 12) at T8. We discuss implications of the results.
Full-text available
Twenty-seven school-based studies of psychosocial approaches to smoking prevention are reviewed. Two major approaches are represented: the "social influences" approach and the broader "life/social skills" approaches. The research studies are considered in four "generations": the seminal work by Richard Evans and colleagues at the University of Houston; seven "pilot" studies of improved programs at Stanford, Minnesota, New York, and Washington, with one school or classroom per experimental condition; twelve improved "prototype" studies by these four groups and others, with two or three units randomly assigned to conditions; and six studies in which maximizing internal validity was of prime concern. Reported results were fairly consistent, with each tested program seeming to reduce smoking onset by about 50%. However, none of the pilot or prototype studies considered alone provided easily interpreted results. The major contributions were improved programs and methods. The findings from the fourth generation of studies were more easily interpreted, though only two of them were interpreted with high confidence. It seems that psychosocial approaches to smoking prevention, particularly the social influences approach--fourth generation tests of the broader life/social skills approaches have yet to be reported--are effective, but at this time we know very little about why, for whom, or under what conditions. Suggestions are provided for improved future research.
Full-text available
To evaluate the long-term efficacy of a school-based approach to drug abuse prevention. Randomized trial involving 56 public schools that received the prevention program with annual provider training workshops and ongoing consultation, the prevention program with videotaped training and no consultation, or "treatment as usual" (ie, controls). Follow-up data were collected 6 years after baseline using school, telephone, and mailed surveys. A total of 3597 predominantly white, 12th-grade students who represented 60.41% of the initial seventh-grade sample. Consisted of 15 classes in seventh grade, 10 booster sessions in eighth grade, and five booster sessions in ninth grade, and taught general "life skills" and skills for resisting social influences to use drugs. Six tobacco, alcohol, and marijuana use self-report scales were recorded to create nine dichotomous drug use outcome variables and eight polydrug use variables. Significant reductions in both drug and polydrug use were found for the two groups that received the prevention program relative to controls. The strongest effects were produced for individuals who received a reasonably complete version of the intervention--there were up to 44% fewer drug users and 66% fewer polydrug (tobacco, alcohol, and marijuana) users. Drug abuse prevention programs conducted during junior high school can produce meaningful and durable reductions in tobacco, alcohol, and marijuana use if they (1) teach a combination of social resistance skills and general life skills, (2) are properly implemented, and (3) include at least 2 years of booster sessions.
Full-text available
Although several studies have reported short-term gains for drug-use prevention programs targeted at young adolescents, few have assessed the long-term effects of such programs. Such information is essential for judging how long prevention benefits last. This paper reports results over a 6-year period for a multisite randomized trial that achieved reductions in drug use during the junior high school years. The 11-lesson curriculum, which was tested in 30 schools in eight highly diverse West Coast communities, focused on helping 7th and 8th grade students develop the motivation and skills to resist drugs. Schools were randomly assigned to treatment and control conditions. About 4000 students were assessed in grade 7 and six times thereafter through grade 12. Program effects were adjusted for pretest covariates and school effects. Once the lessons stopped, the program's effects on drug use stopped. Effects on cognitive risk factors persisted for a longer time (many through grade 10), but were not sufficient to produce corresponding reductions in use. It is unlikely that early prevention gains can be maintained without additional prevention efforts during high school. Future research is needed to develop and test such efforts.
Full-text available
This study determined the effect of provider (nurse or teacher) and training method (workshop or self-preparation) on outcomes of a social influences smoking prevention program. One hundred elementary schools were stratified by school risk score (high risk = high smoking rate among senior students) and assigned randomly to conditions: (1) teacher/self-preparation, (2) teacher/workshop, (3) nurse/self-preparation, (4) nurse/workshop, and (5) control. Intervention occurred in grades 6 to 8. Smoking status at the end of grade 8 was the primary endpoint variable. Intervention reduced grade 8 smoking rates in high-risk schools (smoking rates of 26.9% in control vs 16.0% in intervention schools) but not in low-risk schools. There were no significant differences in outcome as a function of training method and no significant differences in outcome between teacher-provided and nurse-provided interventions in high- and medium-risk schools. Although nurses achieved better outcomes than did teachers in low-risk schools, neither provider type achieved outcomes superior to the control condition in those schools. Workshop training did not affect outcomes. Teachers and nurses were equally effective providers. Results suggest that programming should target high-risk schools.
This paper reports on a meta-analysis of 207 universal school-based drug prevention programs that compared the self-reported drug use of treatment to control or comparison youth. Programs are classified into Interactive and Non-Interactive groups based on a combination of content and delivery method. Weighted categorical and weighted regression methods have been used to determine the attributes that most effectively reduce, delay, or prevent drug use, including program size, type of control group and leader, attrition, target drug, intensity, grade, special population and level of drug use. Program type and size are found to be significant predictors of effectiveness. Non-interactive lecture-oriented prevention programs that stress drug knowledge or affective development show small effects. Interactive programs that foster development of interpersonal skills show significantly greater effects that decrease with large-scale implementations.
Drug abuse continues to be an important public health problem throughout the world. Although considerable progress has been made in identifying effective prevention approaches, there is a large gap between what research has shown to be effective and the methods generally used in most schools. The most promising prevention approaches target individuals during the beginning of adolescence and teach drug resistance skills and norm setting either alone or in combination with general personal and social skills. Evaluation studies testing these approaches show that they can significantly reduce adolescent tobacco, alcohol, and marijuana use. While some studies show that these effects may decrease over time, booster interventions have been found to maintain and in some instances even enhance prevention effects. The results of one large-scale evaluation study shows that it is possible to produce reductions in drug use that last until the end of high school. Available evidence suggests that these approaches may be effective when taught by different kinds of teachers and with different populations. The current paper provides a brief review of school-based prevention approaches targeting individual-level etiologic factors, evidence supporting their effectiveness, and a discussion of potential mediating mechanisms.
Seven thousand one hundred twenty-four members of the Classes of 1985 and 1986 who had participated as seventh graders in one of several smoking prevention programs were tracked and surveyed for smoking habits at 5- and 6-year follow-up: participation exceeded 90% in both cohorts. These data indicated that participants who received seventh-grade interventions based on the social influences model had similar smoking patterns compared to participants in other conditions. This finding supports the call for booster sessions after the initial seventh-grade intervention program. Future follow-up studies will assess whether the earlier benefits associated with the social influences model will translate into measurable differences in adult smoking patterns.
Research to develop and ensure diffusion of smoking prevention programs must (a) be based on an appreciation of the social, psychological, and biological determinants at each stage in the onset process, (b) disentangle major interactions between program content, participant, provider, and setting factors as they determine impact, and (c) ensure both that diffusion is based on empirically grounded principles and that the process is monitored and its effectiveness evaluated. Sufficient evidence supports the tentative conclusion that social influence curricula can be efficacious--at least with some youth. However, we lack key information for diffusion, in particular concerning provider and setting factors. Thus, a cautious advance to diffusion research is recommended, noting that there is much we do not know, and that the public health need for applications must be balanced with continuing research to clarify for whom and under what circumstances current curricula work. At the same time, there should be strong continuing research to improve current interventions, especially for high risk populations.
The effectiveness of a 10-session social psychological approach to the prevention of cigarette smoking was tested on 8th, 9th, and 10th graders (N = 281) in suburban New York. The program was designed to address both the social and psychological factors promoting the onset of smoking by attempting to: (a) increase students' ability to cope with direct pressures to smoke, (b) decrease their susceptibility to indirect pro-smoking social influences, and (c) improve their ability to cope with anxiety. The program was conducted by allied health professionals and utilized group discussion, modeling, and behavior rehearsal. Results indicate significant differences between the experimental and control groups in terms of the proportion of new “experimental smokers” (P < 0.01). Furthermore, there were significantly greater pretest—post-test changes for the experimental group on several of the knowledge and psychological measures, suggesting that the decrease in the onset of smoking behavior among the experimental subjects did occur for the hypothesized reasons.