Unidad 1 Westmaas Et Al 2007 2004

52 History and Methods
3
J. Lee Westmaas, Virginia Gil-Rivas,
and Roxane Cohen Silver
Designing and Implementing

Interventions to Promote Health
and Prevent Illness
Warnings in the media are plentiful about the dan- mately 42% of the U.S. population smoked (CDC,
gers of potential threats to our health such as flu 2004). Through a combination of laws restricting
pandemics, mad cow disease, and excessive use of smoking in public places, bans on various forms
pesticides and antibiotics. Although efforts to pre- of advertisement, tobacco taxes, the availability of
vent such scenarios from becoming reality are well cognitive-behavioral programs for smoking cessa-
placed, many other health conditions in which in- tion, and advances in pharmacotherapies, the rate
dividuals can play a role in their prevention are al- of smoking in the United States in 2004 was ap-
ready taking the lives of millions of people. For proximately 21% (CDC, 2005), a 50% reduction in
example, the Centers for Disease Control and Pre- prevalence. In the early years of the AIDS epidemic,
vention (CDC) estimate that more than 440,000 the increase in safer-sex activities among gay men
smokers in the United States die prematurely every that accompanied messages about the dangers of
year from smoking-related diseases (CDC, 2002). In unprotected sex was also a remarkable example of
addition, although HIV infection has been known for the effectiveness of behavior change interventions
more than a decade to be mostly preventable by (Revenson & Schiaffino, 2000; Shilts, 1987). How-
behaviors such as using a condom, approximately ever, the recent increases in HIV infection rate
40,000 persons become infected with HIV each year among men who have sex with men (Elford & Hart,
(Glynn & Rhodes, 2005), and 4.3 million adults 2003) and the increases in smoking rates observed
worldwide were newly infected with HIV in 2004 among high school students in the 1990s (CDC,
(UNAIDS, 2004). These sobering statistics point out 1999), and among college students in the 2000s
the need to develop, and the challenge of develop- (Rigotti, Lee, & Wechsler, 2000; Wechsler, Kelley,
ing, effective interventions to promote health and Seibring, Juo, & Rigotti, 2001), demonstrate that
prevent illness. effective prevention interventions need to be at-
Although the task of persuading thousands or tuned to the dynamic, ongoing, and complex na-
millions of people to change their behaviors may ture of human behavior. This chapter presents
seem daunting, this is not an unrealistic goal. When important conceptual and practical issues in design-
the surgeon general announced that cigarette smoking and implementing behavioral and psychologi-
ing was a leading cause of cancer in 1964, approxi- cal interventions whose goal is to promote health
52
Designing and Implementing Interventions to Promote Health and Prevent Illness 53
and prevent illness. Our aim is not to present a entire communities that differ on sociodemographic
comprehensive review of each of these issues (read- and other dimensions. These interventions may use
ers will be provided with references to articles that the media and social organizations to educate and
provide more in-depth discussions) but to direct encourage people to adopt healthy behaviors and
attention to their importance and their implications discourage unhealthy ones. For example, advertise-
for conducting effective or informative prevention ments by the government of Canada encouraging
interventions. Examples that illustrate topics under physical activity in its populace, the ParticipAction
discussion will be taken from the smoking cessa- campaign in the 1970s, 1980s, and 1990s empha-
tion and HIV-prevention literatures, not only be- sized the positive health benefits of exercise and were
cause of the substantial morbidity and mortality expected to be viewed and acted upon regardless of
associated with smoking and HIV infection but also age, gender, or socioeconomic status (Canadian Pub-
because these topics have generated a substantial lic Health Association, 2004). Population-based ef-
amount of research illustrating the challenges of forts usually involve simple messages that can be
conducting effective prevention interventions. understood by a majority of a society’s members. On
their own, however, they can be less effective than
Primary, Secondary, other approaches in changing individual behavior.
and Tertiary Interventions Population approaches can sometimes be cost-
effective, however. If only a tiny fraction of the
Interventions can be identified by the point along population is motivated to change their behavior as
the health-illness continuum at which they occur. a result of the message, the cost savings resulting
Primary prevention focuses on changing behaviors from the prevention of illness among these indi-
to prevent illness from occurring. For example, a viduals can be significantly greater than the cost of
primary prevention program for HIV-negative in- the intervention (Thompson, Cornonado, Snipes,
dividuals would aim to prevent infection by pro- & Puschel, 2003).
moting the use of condoms and other safe-sex Population-wide interventions also include
strategies. Secondary prevention interventions are laws that mandate health-promoting behaviors, for
those that occur after the individual has been diag- example, seat belt use, the wearing of protective
nosed with a condition, disease, or illness and seek headgear for motorcyclists in some jurisdictions, or
to stop or reverse its progression. In the case of HIV, laws restricting smoking in the workplace. These
a secondary prevention intervention would focus interventions can lead to behavior change not only
on behavior change to prevent other strains of the by increasing levels of perceived threat but also by
virus from infecting those already infected. Current influencing individuals’ attitudes, beliefs, and ap-
health policy emphasizes secondary prevention, praisals. At the interpersonal level, these campaigns
although it has been argued that devoting more may result in changes in social attitudes and norms
resources to primary intervention might benefit that may further contribute to behavioral change.
population health more substantially (Kaplan, Action at the policy or population level can also
2000). Tertiary prevention interventions seek to provide additional motivation for behavior change
control the devastating complications of an illness among individuals contemplating action as a result
or negative health condition. An intervention to get of other prevention efforts. For example, county-
hospitalized cancer patients to give up smoking to or statewide restrictions on smoking in workplaces
promote recovery from their surgery is an example and eating establishments, which have already en-
of a tertiary prevention intervention. couraged thousands of smokers to attempt to quit
smoking (Chapman et al., 1999), might need to be
Levels of Intervention combined with steep tobacco taxes to encourage
some smokers to quit.
Interventions to promote health and prevent illness Less broad in their reach are community and
can also attempt to influence behavior at the indi- organizational activities that seek to promote healthy
vidual, organizational, community, or societal level. behavior in their members. Many community inter-
Action at the societal (population) level represents ventions have adopted a social ecological perspec-
the broadest level of influence; interventions fo- tive, recognizing that behavior change is a result of
cused on this level of influence seek to motivate social and environmental influences. The program
components of community-level interventions are individual-level approaches. For example, social

often supported by the results of individual-level pressure by a spouse or family members to change
or clinic-based research. Indeed, it has been argued an individual’s behavior (i.e., social control) has
that the costs of community-wide programs to pro- been associated with the degree to which spouses
mote healthy behaviors are justifiable only if prior reduce their levels of smoking (Westmaas, Wild, &
research supports program components (Sorensen, Ferrence, 2002).
Emmons, Hunt, & Johnston, 1998). An example Advocates of ecological models of health pro-
of a community intervention to prevent the uptake motion interventions recommend an integration of
of smoking in youth was the Healthy for Life Project these various levels of influence in any effort to
conducted in the United States. Recognizing the change health behaviors (Stokols, Pelletier, & Field-
various social influences on smoking, the interven- ing, 1995). Nonetheless, the decision as to whether
tion targeted peers, schools, and parents, as well as an intervention should be at the individual, commu-
community agencies (Piper, 2000). A justification nity, or population level, and whether it should be
for the use of community prevention approaches primary, secondary, or tertiary, will be influenced by
is based on the concept of population-attributable a number of factors. These include the amount of
risk, which refers to how much risk produces a financial resources available, political considerations,
given amount of disease in a population (Rose, findings from prior prevention/intervention research,
1985). According to the epidemiologist Geoffrey the likelihood of community cooperation, and the
Rose (1992), changing the risk levels of a popula- believed causes of the behavior or illness in question.
tion to a small degree can impact public health more Of these, an important first step is to understand the
strongly than substantially changing the risk of a various influences on the illness- or health-promot-
smaller number of people. ing behavior and to use a theoretical model to guide
Some interventions are limited to specific in- the design of the intervention.
stitutions such as work sites or schools. Consid-
ering that many individuals spend a substantial
amount of time at places of employment or educa- Psychosocial and Other
tion, the proliferation of work site and school pro- Pathways to Disease
grams addressing a wide range of health issues such
as smoking, weight loss, and physical activity is not Research in the last three decades has provided con-
surprising. vincing evidence of the contribution of psychoso-
Individual-level interventions are characterized cial, biological, and behavioral factors in illness and
by higher levels of personal interaction between the health, and several theories to explain these as-
targets of the interventions and their providers and sociations have been elaborated (Schneiderman,
are more likely to be based on psychosocial or bio- 2004). Understanding how psychosocial, behav-
medical explanations for behavior. An example is ioral, and biological factors independently and/or
a program to reduce smoking prevalence by having interactively contribute to health is an important
physicians provide advice and support for smoking step in deciding when and how to intervene. Hav-
cessation (Goldstein et al., 1997). Evaluations of ing a theoretical model as a template from which
such interventions led to the recommendation in to understand the influence of these variables on
1996 by the United States Agency for Health Care health and behavior change is important in design-
Policy and Research (AHCPR) that physicians prac- ing a successful and cost-effective intervention.
tice the “Four Rs” during their patients’ checkups:
emphasize the risks of smoking, the rewards of
quitting, the relevance of the risks of smoking to Models of Health Behavior Change
the smoker (and the rewards of quitting), and rep-
etition of these messages (Fiore, Jorenby, & Baker, To date, prevention intervention efforts have been
1997). These recommendations for physician inter- largely guided by individual-level theories in which
ventions were found to be cost-effective compared social and cognitive variables play a central role
with other prevention interventions (Cromwell, (Kohler, Grimley, & Reynolds, 1999; Rutter &
Bartosch, Fiore, Hasselblad, & Baker, 1997). Fam- Quine, 2002). Chief among these are the health
ily members or friends can also become involved in belief model (HBM; Rosentock, Strecher, & Becker,
1994); the theories of reasoned action (Ajzen & Fish- reactions to one’s behavior are also important in-
bein, 1977, 1980) and planned behavior (Ajzen, fluences on health behavior. Barriers to the ini-
1991, 1998); social cognitive (learning) theory (Ban- tiation and maintenance of behavioral change may
dura, 1986, 1997); and the transtheoretical model of exist within the individual (such as whether he or
change (Prochaska & DiClemente, 1983). These she has the resources and skills needed), may be
models and theories overlap to a considerable extent, situational, or may be the result of larger social and
but each emphasizes key concepts that significantly structural factors. An extensive body of research has
influence health behavior change (Elder, Ayala, & documented the influence of self-efficacy beliefs on
Harris, 1999). individuals’ efforts to implement and maintain di-
The HBM proposes that behavior change will etary changes (McCann et al., 1995); physical ac-
occur if individuals perceive a threat to their well- tivity and exercise adherence (McAuley, Jerome,
being and believe that the benefits of engaging in Marquez, Elavsky, & Blissmer, 2003); smoking ces-
behavior change outweigh the barriers or costs as- sation (Shiffman et al., 2000); condom use (Baele,
sociated with that behavior. Cues to action (e.g., Dusseldorp, & Maes, 2001); alcohol use (Maisto,
education, symptoms) are viewed as prompting be- Connors, & Zywiak, 2000); and drug use (Reilly
havior change, particularly when levels of perceived et al., 1995). Prevention intervention programs based
threat are high (Rosentock, Strecher, & Becker, on social-cognitive theory include several compo-
1994). The HBM has been used to predict a variety nents, including an informational component to
of health behaviors such as breast self-examination increase perceptions of the risks and benefits asso-
(Champion, 1994), safe-sex practices (Zimmerman ciated with a particular behavior, teaching social
& Olson, 1994), and exercise (Corwyn & Benda, and cognitive skills that can be used to initiate be-
1999), among others. Although the HBM has been havior change, building self-efficacy to promote
widely used, the relationship between key elements behavior maintenance, and building social support
of the model and behavior change are rather small to sustain change (Kohler, Grimley, & Reynolds,
(Sheeran & Abraham, 1996), suggesting the need to 1999).
consider the influence of factors in addition to those The theories of reasoned action (Ajzen & Fish-
central to the model. More recently, principles of bein, 1977, 1980) and planned behavior (Ajzen,
social-cognitive theory have been incorporated in 1988, 1991) propose that for behavior change to
interventions guided by the HBM in an effort to in- occur, individuals must experience a strong inten-
crease the likelihood of health behavior change. tion to change. Behavioral intentions, in turn, are
Social-cognitive (learning) theory (Bandura, predicted by (a) expectancies that a behavior will
1997, 1998) posits that self-efficacy beliefs, goals, produce a particular outcome, (b) attitudes toward
outcome expectations, and perceived barriers or the behavior, (c) beliefs about what others think is
aids involved in enacting a behavior jointly inappropriate behavior (subjective norms) and mo-
fluence human motivation, action, and health tivation to comply with others’ opinions, (d) per-
(Bandura, 1998). Self-efficacy refers to one’s perceptions of control over one’s behavior, and (e)
ceived ability to take the action necessary to achieve other behavioral, normative, and control beliefs
the desired effects or outcomes. Self-efficacy beliefs (Albarracin, , Johnson, Fishbein, & Muellerleile,
are the result of direct and vicarious experience and 2001). The application of these theories to preven-
verbal persuasion. Bandura (1998) suggests that tion intervention efforts and prediction of behav-
personal self-efficacy beliefs play an influential role ior requires defining the targeted individuals’ key
in health in two ways: (a) by influencing biological beliefs, values, and attitudes and their levels of per-
pathways (i.e., sympathetic nervous system activa- ceived control (Ajzen & Fishbein, 1980). These
tion, immune functioning) involved in the relation- theories have been widely used to predict health
ship between stress and illness, and (b) by its impact behaviors and to develop prevention interven-
on individuals’ decisions to make behavioral changes, tions. The empirical evidence, however, suggests
their motivation to maintain these changes and their a weak to moderate association between key ele-
ability to resume those efforts when they face a set- ments of the theory and condom use (Albarracin
back. Outcome expectations regarding the physi- et al., 2001), contraceptive use (Adler, Kegeles,
cal effects of a health behavior (e.g., discomfort), Irwin, & Wibbelsman, 1990), physical activity
the social reactions it evokes, and self-evaluative (Blue, 1995), alcohol use (Johnston & White,
2003), and smoking (Higgins & Conner, 2003), efficacy in performing the behavior, outcome ex-
among others. pectancies) and have also examined how they in-
The transtheoretical model of change (TMC; fluence physical health or physiological outcomes.
Prochaska & DiClemente, 1983) proposes that be- Sobel (1995) argues that psychosocial variables
havior change is a process. Key elements include such as sense of control and optimism, in addition
stages of change, the process of change, decisional to self-efficacy, not only directly impact health be-
balance (pros and cons of change), and situational haviors but also have direct effects on physiological
self-efficacy. The stages of change are precontem- processes that in turn influence health. Interventions
plation (not ready to change within the next that attempt to increase levels of these “shared de-
6 months), contemplation (thinking about change terminants of health,” he believes, are important in
within the next 6 months), preparation (ready to changing any health-relevant behavior but have not
change in the next 30 days), action, and maintenance been given the attention they deserve. For example,
(more than 6 months of sustained action). The TMC although feelings of self-efficacy have been found
posits that tailoring interventions to individuals’ to be an important predictor of behavior change,
readiness to change based on their current stage will few interventions have been developed in which
be more likely to produce behavioral changes. This creating feelings of self-efficacy regarding the tar-
theory has been used to predict a wide range of geted behaviors is an important goal. The studies
health behaviors, including alcohol and drug use of Lorig and colleagues at the Stanford Arthritis
(Prochaska, DiClemente, & Norcross, 1992), physi- Center were offered by Sobel as an example in
cal activity (Marshall & Biddle, 2001), and sexual risk which the finding that improvement in symptoms
behaviors (Grimley, Prochaska, & Prochaska, 1993), (reduced pain) was predicted most strongly by an
but there have been null effects reported by some enhanced sense of control over symptoms led to a
interventions using this approach (Adams & White, change in intervention focus (Lorig & Fries, 1990;
2005). Other theoretical models guiding health pro- Lorig et al., 1989). The result was a restructuring
motion research include cognitive/information pro- of the intervention to focus on enhancing feelings
cessing (Joos & Hickam, 1990) and social support of self-efficacy based on achievable goals (e.g., walk-
theories (Gonzalez, Goeppinger, & Lorig, 1990). ing up two steps rather than a whole flight of stairs),
and which produced significant reductions in pain
Key Elements of Successful Interventions and subsequent physician visits.
The research and interventions of Kemeny with
Based on research demonstrating the value of key HIV-positive patients have also targeted health be-
concepts from the preceding models in predicting havior change variables such as outcome expectan-
behavior change, Elder et al. (1999) summarized cies (see Kemeny, 2003). Their research program
the important ingredients for successful health pro- found that in men diagnosed with AIDS, negative
motion and prevention programs. Specifically, for expectancies about their future health were the
a person to change, she or he must “(1) have a strongest predictor of accelerated time to death,
strong positive intention or predisposition to per- controlling for a variety of confounding factors such
form a behavior; (2) face a minimum of information as baseline health status or immune functioning.
processing and physical, logistical, and social envi- Other important psychosocial predictors were nega-
ronmental barriers to performing the behavior; (3) tive appraisals of characteristics or the self and re-
perceive her/himself as having the requisite skills for jection sensitivity. Rejection sensitivity about one’s
the behavior; (4) believe that material, social, or other homosexuality was significantly related to the rate
reinforcement will follow the behavior; (5) believe of CD4 decline and to faster progression to AIDS
that there is normative pressure to perform and none and mortality (Cole, Kemeny, & Taylor, 1997).
sanctioning the behavior; (6) believe that the behav- Interventions to alter cognitive appraisals of the
ior is consistent with the person’s self-image; (7) have disease process among these men, in addition to
a positive affect regarding the behavior; and (8) en- cognitive-behavioral stress management, have been
counter cues or enablers to engage in the behavior found to produce significant changes in physiologi-
at the appropriate time and place” (p. 276). cal parameters relevant to HIV, such as CD4 T cells
Some interventions have targeted one or more and viral load (Schneiderman, Antoni, & Ironson,
of these requirements for behavior change (e.g., self- 2003).
Social Influences and Health fect and cravings, responses that in prior research
Behavior Change predict the likelihood of lapsing (Kassel, Stroud, &
Paronis, 2003). However, they hypothesized that
One psychosocial variable that may be a valuable the gender of the support provider and recipient
component of interventions to change health be- would moderate the effects of support in reducing
haviors is providing support for achieving the de- negative affect and cravings. Prior research on gen-
sired goal. Many interventions have included social der and social support suggested to them that
support in an effort to delay illness or prolong among men, emotional support during quitting
life (e.g., Zabalegui, Sanchez, Sanchez, & Juando, should originate from a romantic partner, whereas
2005). For example, interventions to help smokers among women, effective sources of support could
quit have included strategies to elicit social support include same-sex friends or strangers. They found
from others or have assigned smokers to buddies that, indeed, women smokers’ negative affect and
who provide support during situations with a high withdrawal symptoms were minimized during a
risk for relapse (May & West, 2000; Park, Schultz, stress task if a female stranger provided support,
Tudiver, Campbell, & Becker, 2004). whereas men smokers’ negative affect and with-
In many cases, however, the promise of social drawal symptoms increased if the support provider
support has not lived up to expectations. For ex- was a female confederate. These results indicated
ample, some studies have found null or adverse that smoking cessation interventions may need to
effects of critical incident stress debriefing (CISD), take into account theory and findings on gender in
a form of immediate social support provided to constructing interventions that include supportive
survivors of acute trauma, on the incidence of post- components.
traumatic stress disorder (McNally, Bryant, & In other areas of health psychology research,
Ehlers, 2003). In addition, a recent Cochrane meta- gender, age, and sociocultural factors may be im-
analysis of psychosocial interventions for women portant factors in whether the provision of social
with metastatic breast cancer found no evidence of support is an effective component of interventions.
long-term effects, although methodological features Attention to how social support is defined (struc-
of the trials reviewed, as well as insufficient power tural, emotional, social pressure, etc.) and other
to detect effects, may have precluded finding effects. methodological factors such as the use of standard-
Personality factors may also moderate the extent to ized measures and process evaluation, are also im-
which social support is beneficial, but few studies portant in evaluating the effects of social support
have examined the role of personality factors or components. These methodological factors are dis-
other individual differences as they interact with cussed in subsequent sections in more detail.
support provision. Possible candidates are hostil-
ity (Lepore, 1995) and defensiveness (Strickland &
A Social Ecological Approach
Crowne, 1963; Westmaas & Jamner, in press); ex-
to Health Behavior Change
perimental studies have found these dispositional
qualities to moderate the extent to which social Social ecological models of behavior change address
support is beneficial in reducing subjective and the multiple sources of influence on health-relevant
physiological reactions to stressors. behaviors. In a model described by Sorensen and
In the smoking cessation literature, some inter- colleagues (1998), these sources of influence are
vention studies that have sought to increase the explained in terms of lenses through which various
amount of social support for smokers likewise have disciplines view the behavior or illness:
proved to be ineffective (May & West, 2000). How-
ever, data suggest that attention to potential mod- At the micro level, the biomedical lens focuses
erators such as gender and the use of theoretical on biophysiological theories of disease
models to guide research may be valuable in un- causation. . . . The psychosocial lens maintains
derstanding how social support can be used effec- a primary focus on the individual, investigating
tively in interventions. For example, Westmaas and questions about individual and social behav-
Billings (2005) hypothesized that social support iors such as personality structures, a sense of
might facilitate smoking cessation by reducing sub- control, and self-efficacy. . . . The epidemio-
jective responses to stressors such as negative af- logical lens examines disease patterns within
populations and aims to understand differen- are metabolized, and by implication the likelihood
tial risk factors, including biological predispo- of lung cancer development (Harrison, Cantlay, Rae,
sitions as well as behavioral and environmental Lamb, & Smith, 1997; Nyberg, Hou, Hemminki,
exposures. By contrast, the society-and-health Lambert, & Pershagen, 1998; Jourenkova-Mironova
lens brings to the foreground cultural, social, et al., 1998). These and other advancements, such
economic, and political processes and aims to as the development of a vaccine to prevent nicotine
understand the ways in which these social from reaching the brain (Shine, 2000), offer the pos-
structures influence differential risks. The sibility of future biologically based interventions to
social ecological model cuts across these prevent the development of lung cancer among
disciplinary lenses and offers a theoretical smokers (secondary prevention).
framework that integrates multiple perspec- Psychosocial factors are also associated with
tives and theories. This framework recognizes smoking initiation, such as parental or sibling smok-
that behavior is affected by multiple levels of ing, and perceived norms about the acceptability of
influence, including intrapersonal factors, smoking. School-based primary interventions have
interpersonal processes, institutional factors, addressed these psychosocial factors. Recent meta-
community factors, and public policy. (p. 390) analyses of school-based interventions found that
the most effective approaches were those that in-
Social ecological models to promote healthful cluded a focus on social reinforcement for not
behaviors can also address the influence of physi- smoking, whereas the least effective were those that
cal environments. According to Stokols (1992), in sought only to increase awareness of the dangers
a social ecological approach, “the healthfulness of of starting to smoke (Levinthal, 2005).
a situation and the well-being of its participants are In addition to psychosocial factors as contribu-
assumed to be influenced by multiple facets of both tors to smoking behavior, societal-level factors are
the physical environment (e.g., geography, archi- implicated, such as the price of cigarettes and the
tecture, and technology) and the social environment portrayal of smoking among actors in movies (Ander-
(e.g., culture, economics, and politics)” (p. 7). Of son & Hughes, 2000). Community approaches to
five health-related functions of the sociophysical the prevention of smoking have recognized that in
environment noted by Stokols, one is environment addition to psychosocial factors, these societal-level
as “an enabler of health behavior exemplified by the factors are also important.
installation of safety devices in buildings and ve- Sociodemographic and cultural variables, such
hicles, geographic proximity to health care facili- as age, gender, ethnicity, and/or socioeconomic sta-
ties, and exposure to interpersonal modeling or tus, may moderate the impact of biological, psycho-
cultural practices that foster health-promotive be- social, and societal influences on smoking. These
havior” (pp. 13–14). variables have taken on increased importance in the
design of interventions to reduce or prevent smok-
Multiple Influences on Health Behavior ing because of recent evidence that smoking initia-
Change: The Case of Smoking tion is also now occurring at a later age through
cigarette promotion activities in bars that cater to
A good example of the multiple levels of influence college students (Rigotti, Moran, & Wechsler,
that comprise a social ecological approach to behav- 2005). This has occurred as laws curbing advertis-
ior change is the case of smoking. Smoking is im- ing directed at youth have been enacted (e.g., the
plicated in many illnesses, and the ability to quit Joe Camel campaign). Such recent developments
appears to be a function of societal, psychosocial, suggest that to appropriately evaluate population-
and biological variables. Cigarette smoking is be- level interventions such as those limiting the ad-
lieved to account for approximately 90% of all vertising or price and availability of cigarettes, age,
lung cancer cases (Siemiatycki, Krewski, Franco, or educational level will need to be considered as
& Kaiserman, 1995), but most smokers will not de- possible moderators of the effectiveness of these ac-
velop lung cancer. Genetic polymorphisms in glu- tivities. Gender or ethnic differences in smoking
tathione s-transferase enzyme activity may influence initiation, in reasons for smoking, and in smok-
the degree to which carcinogens in cigarette smoke ing prevalence or ability to quit have also been dem-
onstrated (Mermelstein, 1999; Perkins, 2001; Perkins, exist. For example, cultural beliefs may play a sig-
Donny, & Caggiula, 1999), suggesting their poten- nificant role in the adoption of the behavior, and
tial role as moderators of intervention effectiveness. culture-appropriate materials may be needed to de-
Although a social-ecological perspective in deliver the intervention. In an intervention to reduce
signing health promotion interventions has been the likelihood of HIV infection among migrant farm-
promulgated and extensively implemented (see re- workers in California who have sex with other men,
views by Merzel & D’Afflitti, 2003; Sorensen et al., but who would not self-identify as gay, Conner and
1998), a single unifying framework that integrates colleagues distributed a novella (ongoing sagas pre-
multiple levels of influence on a particular behavior, sented in comic book format) to promote condom
and that acknowledges possible interactions among use (Mishra, Sanudo, & Conner, 2004). In this par-
them, has been absent (Merzel, & D’Afflitti, 2003). ticular socioethnic group, such an approach was seen
Goodman and colleagues have also argued that with- as a legitimate source of information compared with
out the specificity of an integrative social-ecological other possible options.
model with which to test hypotheses, it is difficult In addition to cultural factors, specialized groups
to properly evaluate the effects of community-level may also differ on other sociodemographic charac-
prevention interventions (Goodman, Liburd, & teristics, such as literacy or socioeconomic status,
Green-Phillips, 2001). which will dictate the methods used to deliver the
intervention. If English is not the native language of
the intended recipients, the intervention will of course
Designing Interventions need to be presented in their language. If reading
ability is limited, print media are not appropriate.
An initial step in designing prevention interventions The sociodemographic and cultural character-
is deciding who should be the target of the inter- istics of a targeted population can also influence
vention. This decision should be partly related to whether an intervention should be undertaken at
whether the focus of the intervention is primary, the individual, community, or societal level. If the
secondary, or tertiary prevention. Many illness-pro- targeted population is difficult to recruit for face-
moting behaviors begin during youth or adoles- to-face interactions, then community- or society-
cence, and so primary interventions will often need level interventions, rather than individual-level
to target individuals in these age groups. Good ex- prevention approaches, may be more appropriate.
amples are school-based interventions to prevent If individuals from some communities may have
initiation of smoking, to prevent obesity, or to pre- to endure a long commute to attend a cognitive-
vent pregnancies and sexually transmitted diseases. behavioral smoking cessation clinic provided by a
However, primary interventions can also target hospital, or may not have the financial resources to
older individuals, such as older HIV-negative men make the trip, public service messages on radio sta-
who have sex with men, or college students who tions or campaigns delivered through church groups
are in danger of starting to smoke or binge drink. may have a better chance at reaching them. For ex-
Age and gender differences are important consid- ample, Brandon and colleagues recruited recently
erations because they are related to maturational quit smokers for a relapse-prevention intervention
or sociocultural factors that are likely to play a role through newspaper, radio, and media advertise-
in the factors influencing health behaviors and ments that provided them with a toll-free number
whether the behavior is adopted. The use of explicit to call to register for the program (Brandon et al.,
sexual language in print messages urging gay men 2004). The intervention consisted of brochures that
to use condoms to prevent HIV infection, for ex- were mailed directly to participants. This technique
ample, while believed to have been effective for this was effective in decreasing relapse rates.
population, could be offensive if applied to young Sometimes an intervention seeks to target those
women who are also in danger of becoming infected. who are most at risk. If reducing risk in the most
If the target population for an intervention con- at-risk individuals is the overarching goal, it might
sists of specialized groups such as ethnic and ra- be assumed that increasing the intensity of the in-
cial minorities, immigrant populations, children, tervention will be needed to change behavior. How-
the elderly, or the physically ill, other challenges ever, even moderately intense prevention efforts
could increase the attrition rate in such a popula- trol groups represents the best strategy for ruling out
tion. Equally important, the intervention will be less alternative explanations, but participant responses
likely to show positive effects among those most at to and reactions against the randomization process
risk. For example, the aim of the Community In- must be attended to and minimized, if possible
tervention Trial for Smoking Cessation (COMMIT) (Wortman, Hendricks, & Hillis, 1976). In addition,
was to increase cessation rates among heavy smok- other designs, including longitudinal research, can
ers (i.e., those smoking more than 25 cigarettes a be used to support causal inferences. However, lon-
day). However, postintervention analyses indicated gitudinal designs are based on the assumption that
that COMMIT succeeded in increasing quit rates certain parameters do not change over time, and
among light and moderate smokers (i.e., those smok- there is still the possibility of spuriousness that
ing less than 25 cigarettes a day) but not among needs to be accounted for in order to make causal
heavy smokers (Fisher, 1995). inferences (Kenny, 1979). One threat to the abil-
Another important consideration in selecting ity to make causal inferences is the problem of se-
participants is how generalizable the results of the lection bias. Selection bias occurs if the units
intervention are intended to be, which in turn will making up the intervention and control groups
influence whether the intervention is likely to be differ before the intervention is even implemented
adopted by others. The RE-AIM framework, devel- (Larzelere, Kuhn, & Johnson, 2004). These biases
oped by Glasgow and colleagues, is a system of can lead to both overestimating and underestimat-
evaluating health promotion interventions that in- ing the effects of interventions. In some interven-
cludes an assessment of the representativeness of tions, selection biases operate so that the sickest
participants, and the settings in which the interven- or riskiest groups are targeted for behavior change.
tion was conducted (Glasgow, Bull, Gillette, Klesges, Regression to the mean by these individuals, de-
& Dzewaltowski, 2002). Among the components fined as the tendency over time to approach mean
of RE-AIM (Reach, Effectiveness, Adoption, Imple- levels of a behavior (Cook & Campbell, 1979), can
mentation, Maintenance), Reach refers to “the per- give the appearance that the intervention pro-
centage of potential participants who will take part duced positive effects. Without randomization
in an intervention, and to how representative they to intervention and control groups of the most at-
are of the population from which they are drawn” risk individuals, regression to the mean as a plau-
(p. 63). Glasgow and colleagues evaluated health sible alternative interpretation of results cannot be
promotion interventions conducted between 1996 eliminated.
and 2000 that attempted to change dietary, smok- Well-designed randomized clinical trials repre-
ing, and physical activity behaviors, and that in- sent one of the most powerful means of assessing
cluded a comparison or control group. They found health behavior theories and the effectiveness of
that among 36 studies, although a majority reported interventions. However, as Helgeson and Lepore
on the percentage of eligible patients who partici- (1997) note, designing a randomized clinical inter-
pated, few studies reported whether participants, vention will often require balancing “the needs of
compared with those who declined, differed on the individual patient with the requirements of the
sociodemographic or medical variables. Knowing research protocol” and “the practical or logistical is-
who declined, why they did so, and whether they sues in conducting an intervention with the theoreti-
differed from participants on sociodemographic cal and experimental issues.” Helgeson and Lepore
variables such as sex, age, and socioeconomic status further note that this balancing act will sometimes
can help in the design and revision of recruitment require unforeseen modifications to the research
strategies so that generalizability of an intervention protocol in order to ensure patient recruitment and
is enhanced. retention and/or the cooperation of clinic staff. As
an example they mention the occasional cancer
Selecting the Appropriate Design patient who is dismayed by his or her assignment
to the control group and asks to be put in the in-
To be able to conclude that an intervention is ef- tervention group. The authors resolved this issue
fective, plausible alternative explanations must be in their own research by favoring patients’ well-
ruled out. Randomization of individuals, schools, being (e.g., providing them with referrals to other
work sites, or communities to intervention and con- support groups in the community; see Hohmann
& Shear, 2002, for another discussion of these designing interventions at the community level, “an
problems). expanded range of research methodologies is re-
When there is nonrandomization of units to quired to address the diverse needs for scientific
intervention and control groups, assessing possible rigor, appropriateness to research questions, and
preexisting differences between intervention and feasibility in terms of cost and setting” (p. 401).
control groups on variables that may influence the They describe other designs that could supplement
targeted behavior becomes paramount. These prin- the randomized control trial in answering questions
ciples have not always been followed in commu- about the effectiveness of community interventions,
nity interventions to prevent the uptake of smoking including observational studies, qualitative research
in young people, however. According to a recent methods, and action research methods. For example,
Cochrane database review, among the 17 commu- qualitative research methods would be appropriate
nity intervention studies designed to prevent youth for understanding community needs, priorities, and
smoking that included control groups and assessed resources before an intervention is designed.
baseline characteristics (their criteria for inclusion With the increased popularity of the Internet,
in the review), in 8 studies the allocation of com- a number of Web-based interventions have also
munities or schools to the intervention or control been developed. Web-based interventions offer the
groups was nonrandom, and some studies did not advantages of accessibility, low cost, data complete-
account for baseline differences in smoking in their ness, standardization, personalization or tailoring
follow-up analyses (Sowden, Arblaster, & Stead, of information, and potentially greater accuracy of
2005). reporting symptoms or illegal or stigmatizing be-
In the absence of randomization to intervention haviors. Subjects can also participate in program
and control groups, matching individuals or com- elements in the privacy of their own homes and at
munities from intervention and comparison groups their own convenience, and the degree of program
on variables associated with the targeted behavior participation can be easily assessed. Currently, there
is appropriate. In a review of 32 community in- is a paucity of well-controlled research on the effi-
terventions to reduce smoking in adults that in- cacy of Web-based interventions to promote health
cluded a control group, however, only 5 studies behaviors, but there are promising signs. For ex-
demonstrated that the intervention and control ample, current Web-based interventions address-
communities were comparable on demographic ing smoking cessation, substance use, depression,
variables at baseline (Secker-Walker, Gnich, Platt, and post-traumatic stress disorders have demon-
& Lancaster, 2005). strated positive treatment effects compared with
In randomized controlled interventions con- control groups (Barr Taylor & Luce, 2003; Bock
ducted at the community level, the need to main- et al., 2004; Copeland & Martin, 2004).
tain scientific rigor through standardization and
control can sometimes conflict with community Power Analyses
goals and priorities. A certain amount of flexibility
in accommodating the needs of participating com- For any research endeavor, conducting power analy-
munity organizations is important for ensuring in- ses is an important means of determining the num-
tervention integrity and can ultimately influence the ber of units to be assigned to experimental and
effectiveness of the intervention. Involving commu- control groups in order to answer questions about
nities in the design and implementation process will the intervention’s effectiveness. If a proposed study
help both researchers and communities understand is not adequately powered, the absence of reliable
each other’s perspectives and can ensure that the (significant) differences between groups could be
goals and priorities of both parties are met. attributed to lack of power. Power analyses can also
The expense of randomized controlled trials at determine whether there are sufficient data points
the community level, in which the unit of alloca- to adequately evaluate if intervention effects on out-
tion to experimental and comparison groups is the come variables are moderated by other variables (e.g.,
community or organization, can be a motivating motivation to quit smoking or gender). Such analy-
factor in considering alternative designs, especially if ses represent the testing of “group X moderating
required levels of statistical power are to be achieved. variable” interactions and provide valuable informa-
Sorensen and colleagues (1998) have stated that in tion, especially if no main effects are obtained.
In individual-level interventions, the unit of be important in maintaining their interest and co-
allocation is the participant, with power analyses operation. Maintaining contact with staff, particu-
indicating the number of participants that should larly nurses, who are often vital to the successful
be recruited in each group in order to detect sig- implementation of the project, rewarding them for
nificant main or interaction effects at predetermined their cooperation, and providing updates and evi-
levels of power (usually 80%). In community in- dence of the intervention’s value will also help to
terventions, the unit of allocation is the work site, achieve this goal (Helgeson & Lepore, 1997, Grady
school, hospital, city, or town. Power analysis to & Wallston, 1988).
determine adequate sample sizes in community For interventions at the community level, the
interventions need to account for statistical depen- skills and priorities of the individuals, agencies, and
dencies of responses within each unit or cluster institutions participating in the intervention are more
(Donner & Klar, 1996; Koepsell et al., 1992). When varied. Altman (1995a) summarized four recommen-
only one community receives the intervention, with dations for improving community-level interven-
another community serving as the control group, tions, at least two of which refer to the importance
conducting power analysis is difficult if results are of community cooperation. The four recommen-
to be analyzed at the cluster level. Indeed, it has dations include “(i) integrate interventions into the
been argued that the modest or nonsignificant ef- community infrastructure, (ii) use comprehensive,
fects of several community interventions to pro- multi-level intervention approaches, (iii) facilitate
mote health and prevent illness may have been due community participation and promote community
to insufficient power to detect positive effects, even capacity-building, and (iv) conduct thorough needs
small ones (Secker-Walker et al., 2005). assessment/social reconnaissance in order to tailor
interventions to the community context.” Spend-
Enlisting Cooperation for Interventions ing the time to understand the priorities of com-
munity organizations, whose assistance is required
In clinic-based interventions, the goal is often to for the intervention to be implemented, and incor-
evaluate the efficacy of a specific treatment on a spe- porating their needs into the intervention goals will
cific outcome. An example is determining whether help to sustain their cooperation during the re-
a cognitive-behavioral intervention for smoking ces- search phase. At the same time, demonstrating how
sation is effective in getting hospitalized patients the intervention goals can benefit the community
to quit. Clinic-based interventions usually involve and obtaining consensus for their importance will
nurses, doctors, or other health care professionals help ensure that the needs of all parties are ade-
(e.g., therapists, psychologists, and psychiatrists). quately met. However, some flexibility is still re-
Helgeson and Lepore (1997) provide several guide- quired on the part of the academically oriented
lines and comments that are useful in enlisting the research team so that the intervention is tailored to
cooperation of medical personnel. For example, the community context.
they note that to gain access to a medical popula-
tion, the first step is to identify a physician who Process Evaluation
values research and can be convinced that the re-
sults of the intervention will translate to benefits for Designing, planning, and executing an intervention,
the patient and the medical community. Including especially one that requires the cooperation of re-
physicians in designing the intervention itself may searchers, community agencies, workplaces, and
be challenging, given the time constraints that many media, involves a tremendous amount of effort. To
have, but nurses, whose training emphasizes the be able to determine whether these considerable
psychosocial needs of the patient, can often provide efforts are effective in producing the intended
valuable information about patients and the opera- changes and are cost-effective, a rigorous evaluation
tion of the institution (Holman, 1997). This in- of intervention delivery is required. Without exten-
formation can be especially valuable in the design sive process evaluation, interpreting a lack of dif-
phase of the intervention. Because of the multitude ferences between intervention and control groups
of demands faced by clinic staff, minimizing the is particularly problematic. For example, if teach-
amount of work required of them (e.g., in devel- ers implementing a smoking resistance program
oping a list of eligible patients for the study) will among middle schoolers deviate significantly from
activities geared to changing norms about smoking level, this is likely to occur if the media extensively
and resisting offers to smoke, then nonsignificant cover intervention activities. Comparison commu-
effects of the intervention could be attributed to the nities may also independently conduct their own
intervention group having received a weaker dose health fairs, enact legislation, or provide media
of the treatment. Ongoing evaluation of program messages that produce effects similar to those of the
activities and delivery can also be used to appro- intervention. For example, in the Alliance of Black
priately modify program components once the in- Churches Project to reduce smoking through coun-
tervention is under way. seling by church members, the difference between
A model example of rigorous process evalua- the intervention group and the comparison group
tion occurred in the COMMIT trial, a five-year in whether they received information about smok-
community intervention trial to decrease smoking ing from a church member was 29% versus 20%,
prevalence among heavy smokers (Fisher, 1995). respectively (Schorling et al., 1997). In the COM-
Monitoring of program delivery was extensive, with MIT trial, differences in indices of penetration for
logs completed by staff and volunteers and com- the intervention and control communities were also
puterized record keeping of intervention activities. relatively small.
Process evaluation in other community interven- Secular trends in awareness of and engagement
tions has included surveys completed by the tar- in behaviors that promote health and prevent ill-
geted population and deliverers of the intervention, ness may also lead to behavioral changes in the
either by phone or through the mail, focus groups control group that are comparable to those in the
and semistructured interviews, and tracking and intervention group. For example, the decreased
documenting program activities. In some trials, social acceptability of smoking, facilitated first by
these functions were performed by computerized the surgeon general’s report and subsequently by
systems (Secker-Walker et al., 2005). tobacco taxes, public health campaigns, and the
Process evaluation in community interventions Master Tobacco Settlement Agreement, may be
can also include a determination of its reach and contributing to a decline in smoking rates in many
penetration. Reach refers to how aware members of geographic areas. The secular trend of reducing
the target population are of program activities such smoking levels has been cited for the observation
as radio or television advertisements, newspaper of a greater decrease in smoking prevalence in the
articles, health fairs, workplace programs, treatment control compared with the intervention community
clinics, self-help kits, and so on. Penetration refers in the Pawtucket Heart Health program (Carleton,
to the extent to which the targeted population par- Lasater, Assaf, Feldman, & McKinlay, 1995). Secu-
ticipated in these activities. Polling of represen- lar trends in smoking reduction may have also pre-
tative samples of individuals from the targeted cluded finding stronger effects in the COMMIT trial
population can determine reach and penetration, (Bauman, Suchindran, & Murray, 1999). A greater
which can be presented as the number or propor- intensity of intervention dosage may be needed to
tion of individuals who partook of intervention overcome secular trends observed in comparison
activities. This information should be an important communities. In addition, to better assess interven-
part of the dissemination of trial results because low tion effectiveness, investigators should determine
rates of awareness and/or penetration could account the extent to which secular trends in behavior
for nonsignificant differences between intervention change are occurring prior to the implementation
and comparison groups on key outcomes. Penetra- of the intervention (Secker-Walker et al., 2005).
tion (intervention dose) can also be used to deter-
mine dose-response relationships, a measure of the Outcome Evaluation
effect of the intervention.
A case can be made that assessing the reach and For any health promotion or disease prevention
penetration in comparison groups or communities intervention, what the outcome variables should be,
for activities that are similar to intervention com- by what means they should be assessed (question-
ponents should also be performed. There may be naires, interviews, etc.), and how often and when
diffusion or “spillover” of the treatment from the they should be measured need to be determined. In
experimental group or community into the com- community-level interventions, outcome variables
parison group or community. At the community should be relevant or salient to the individuals from
the communities or populations being studied (Hoh- The reliability of participants’ responses can also
mann & Shear, 2002). Hohmann and Shear (2002) be assessed through the use of collateral reports from
suggest that while symptoms and diagnostic catego- subjects’ romantic partners, family, and/or friends.
ries may be important for comparative purposes, for Convergence of evidence from these sources pro-
participants the important or expected outcomes vides a greater degree of confidence about the re-
may be different (e.g., an increase in level of daily liability of responses. In addition to self-report or
functioning). The use of standardized measures interview format, observational or archival mea-
with demonstrated reliability and validity, or measures can be useful indicators of intervention effec-
sures that have been used in prior research, should tiveness. An example would be documentation of
be encouraged because they allow for easier com- the number of teenage pregnancies before, during,
parison with results of prior research. and after an intervention promoting condom use in
Who the intervention is targeting, as discussed adolescents.
previously, may play a role in determining which These methods of assessment differ in conve-
methods should be used to assess outcome variables. nience and amount of resources required. As noted
Self-report measures are often the most convenient earlier, Web-based questionnaire assessments have
and may be the best option if highly personal infor- been used with increasing success in health psy-
mation, such as sexual practices or illegal activity, chology research because of their convenience and
is sought. Structured interviews conducted over the low cost, and it is likely that they will be seriously
telephone or computer-assisted interviews also pro- considered for use in future health-promoting in-
vide some degree of anonymity and are probably terventions. (See also Schlenger & Silver, 2006, for
less subject to the problem of missing data than are additional information on the pros and cons of the
questionnaires. Both questionnaire and interview use of the Web for data collection.)
formats allow some degree of control of the testing The assessment of outcomes, and of potential
context and data quality by researchers, but the mediating or moderating variables, should be con-
degree of control needed will depend on the ques- ducted before, during, and after the intervention.
tion being asked. For example, in a study investi- Assessing outcomes sometime after the intervention
gating why women stay in abusive relationships, has ended can answer important questions about
Herbert and colleagues used a strategy that ensured its long-term efficacy. For example, follow-up of
that women who completed the sensitive question- cognitive-behavioral smoking cessation programs
naires could do so without the knowledge of their has found impressive quit rates soon after the inter-
abusive spouses (Herbert, Silver, & Ellard, 1991). vention ended, but a substantial number of smok-
For interventions conducted at the individual ers relapse within the subsequent year (USDHHS,
level, demand characteristics are more likely to be 2000). This has led to a focus on devising relapse
a problem. For example, smokers who undergo a prevention programs for smokers, some of which
cognitive-behavioral intervention to quit smoking have been successful (e.g., Brandon et al., 2004).
may be motivated to misrepresent their actual lev- Many community-level prevention efforts have also
els of smoking at the end of the intervention be- led to short-lived behavioral changes. However,
cause of the expectation by clinic staff that they some of the health effects of interventions may take
should have quit or reduced their smoking. This years to be realized. This has stimulated efforts to
consideration has led many smoking cessation pro- encourage the sustainability or maintenance of in-
grams to include biochemical validation of smok- terventions after researchers have collected and
ing status. In general, however, the more limited the published their data.
contact between participants and clinic staff, the less
likely smokers are to misrepresent their actual lev-
els of smoking (Velicer, Prochaska, Rossi, & Snow, Sustaining Interventions
1992). In addition, findings from randomized ex- at the Community Level
periments document that when research participants
respond to questions without interacting directly If intervention activities are to be transferred to
with an interviewer, they are more likely to reveal community organizations, their leaders should be
sensitive and/or personal information (Lau, Thomas, involved in the planning and implementation of the
& Liu, 2000; Turner et al., 1998). intervention, which will need to be sensitive to the
priorities and limitations of community resources Ethical Issues

(Altman, 1995b). According to Altman, questions
of ownership and control of community programs Any research whose goal is to change behaviors,
should be addressed prior to implementation, and even health-promoting ones, must attend to ethi-
any conflicts resolved, so that there is “broad-based cal issues involving the use of human subjects. Be-
support from a cross-section of community con- cause of concerns regarding the use of community
stituencies” (p. 529) both during the intervention samples in intervention research, nonprofit agen-
and after the research phase has ended. An ex- cies, health care settings, schools, and work sites
change of skills, through education and training, may serve as gatekeepers to block access to poten-
between researchers and community staff and fos- tial research subjects. Convincing these gatekeepers
tering a sense of empowerment in communities to of the value of one’s research often requires demon-
obtain resources for programs also contribute to strating sensitivity to the ethics of human subject
sustainability. For example, educating community experimentation (Sieber, 1998). Moreover, conduct-
leaders on effective performance evaluation should ing interventions at any level of analysis must also
foster a sense of empowerment and produce skills involve the review of research plans by institutional
that leaders can use to design and implement review boards housed in academic institutions, as
unique interventions that address the same health well as in individual research settings (e.g., indi-
behaviors. vidual hospitals). The provision of an inherently
more appealing treatment may require eventually
Disseminating Research Findings offering it as an option to the control group at a later
date (e.g., designing a “waiting list” control group).
Although expenditures on health promotive re- Designing research on specialized populations,
search have been substantial, knowledge of how to such as geriatric, pediatric, or medically ill samples,
disseminate findings from research to primary-care requires special attention to issues of informed con-
clinics and community agencies that deliver health sent, avoiding coercive procedures, and providing
services is only now emerging as a research en- ample opportunity for refusal and termination of
deavor in its own right (Kerner, Rimer, & Emmons, participation in the research effort over time. It is
2005). Much work remains, however, for dissemi- important that the individual researcher conduct a
nation research to advance as a science. As Kerner careful cost-benefit analysis, weighing the personal
et al. note, the best research methods for evaluat- rights of individual participants against the potential
ing dissemination strategies have not yet been es- benefits for society of the research. While ethical is-
tablished, which limits the amount of guidance sues surrounding the design and implementation of
available for researchers who want to conduct dis- community-based intervention research may be chal-
semination research. These authors also recom- lenging and will undoubtedly require creativity and
mend more infrastructure support from private persistence, conducting methodologically rigorous
and public sectors and greater consensus among research on human participants is required for the
journal reviewers and editors on how to evaluate science of health promotion and illness prevention
dissemination research projects. Nonetheless, trans- to advance.
lating the empirical results of scientific investiga-
tions into practical recommendations for health
care professionals, schools, work sites, and com- Conclusion
munity organizations is critical. Working effec-
tively with the media and others to take research The value of health promotion and intervention
findings to the public—to ensure that they are programs to improve health and reduce illness has
effectively applied to both policy and practice— been amply demonstrated in several domains of
should be an important goal of health psychology behavior. Indeed, it can be argued that intervention
researchers. With increasing attention to these programs are victims of their own success when
issues, the efforts devoted to designing effective individuals in control groups show the same im-
health-promoting interventions will hopefully trans- provements in health behaviors as those in inter-
late to evidence-based practices that improve the vention groups. However, for progress to continue
public health. in our dynamic environment, where new threats to
mental and physical health emerge (e.g., bioter- researchers and communities. Health Psychology,
rorism) and old ones adopt new faces (e.g., water 14, 526–536.
pipes for smoking), lessons learned from prior in- Anderson, P., & Hughes, J. R. (2000). Policy
tervention research must be considered as well as interventions to reduce the harm from smoking.
Addiction, 95(Suppl), S9–S11.
new approaches. Attention to such principles as
Baele, J., Dusseldorp, E., & Maes, S. (2001). Condom
the use of theoretical models to guide research,
use self-efficacy: Effect on intended and actual
consideration of individual, cultural, and socio- condom use in adolescents. Journal of Adolescent
demographic differences and their moderating ef- Health, 28, 421–431.
fects on treatment outcomes, the equivalence of Bandura, A. (1986). Social foundations of thought and
intervention and control groups, and the appropri- action. Englewood Cliffs, NJ: Prentice-Hall.
ate use of statistical analyses and methods should Bandura, A. (1997). Self-efficacy: The exercise of
provide the foundation for health promotion and control. New York: Freeman.
intervention research. However, to improve the Bandura, A. (1998). Health promotion from the
health of the greatest number of individuals, the perspective of social cognitive theory. Psychology
expertise of others who are invested in advancing and Health, 13, 623–649.
Barr Taylor, C., & Luce, K. H. (2003). Computer- and
our ability to promote health and prevent illness is
Internet-based psychotherapy interventions.
needed. Their involvement, as well as that of the
Current Directions in Psychological Science, 12, 18–
targets of our research, will necessarily contribute 22.
to our understanding the most effective ways to Bauman, K. E., Suchindran, C. M., & Murray, D. M.
initiate and sustain health behavior change. (1999). The paucity of effects in community
trials: Is secular trend the culprit? Preventive
Medicine, 28, 426–429.
References Blue, C. L. (1995). The predictive capacity of the
theory of reasoned action and the theory of
Adams, J., & White, M. (2005). Why don’t stage- planned behavior in exercise research: An
based activity promotion interventions work? integrated literature review. Research in Nursing
Health Education Research, 20, 237–243. & Health, 18, 105–121.
Adler, N. E., Kegeles, S. M., Irwin, C. E., & Wibbels- Bock, B. D., Graham, A. L., Sciamanna, C. N.,
man, C. (1990). Adolescent contraceptive Krishnamoorthy, J., Whiteley, J., Carmona-
behavior: An assessment of decision processes. Barros, R. N. et al. (2004). Smoking cessation
Journal of Pediatrics, 116, 463–471. treatment on the Internet: Content, quality, and
Ajzen, I. (1988). Attitudes, personality and behavior. usability. Nicotine and Tobacco Research, 6, 207–
Milton Keynes, UK: Open University. 219.
Ajzen, I. (1991). The theory of planned behavior. Brandon, T. H., Meade, C. D., Herzog, T. A., Chirikos,
Organizational Behavior and Human Decision T. N., Webb, M. S., & Cantor, A. B. (2004).
Processes, 50, 179–211. Efficacy and cost-effectiveness of a minimal
Ajzen, I., & Fishbein, M. (1977). Attitude-behavior intervention to prevent smoking relapse:
relations: A theoretical analysis and review of Dismantling the effects of amount of content
empirical research. Psychological Bulletin, 84, 888– versus contact. Journal of Consulting and Clinical
918. Psychology, 72, 797–808.
Ajzen, I., & Fishbein, M. (1980). Understanding Canadian Public Health Association. (2004).
attitudes and predicting social behavior. Englewood ParticipAction—The Mouse That Roared: A
Cliffs, NJ: Prentice Hall. Marketing and Health Communications Success
Albarracin, D., Johnson, B. T., Fishbein, M., & Story. Canadian Journal of Public Health, 95(Suppl),
Muellerleile, P. A. (2001). Theories of reasoned 1–46.
action and planned behavior as models of Carleton, R. A., Lasater, T. M., Assaf, A. R., Feldman,
condom use: A meta-analysis. Psychological H. A., & McKinlay, S.(1995). The Pawtucket
Bulletin, 127, 142–161. Heart Health Program: Community changes in
Altman, D. (1995a). Strategies for community health cardiovascular risk factors and projected disease
intervention: Promises, paradoxes, pitfalls. risk. American Journal of Public Health, 85, 777–
Psychosomatic Medicine, 57, 226–233. 785.
Altman, D. G. (1995b). Sustaining interventions in Centers for Disease Control and Prevention. (1999).
community systems: On the relationship between Cigarette smoking among high school students:
11 States, 1991–1997. Morbidity and Mortality behaviour increasing among men who have sex
Weekly Report, 48, 686–692. with men? AIDS Education Prevention, 15, 294–
Centers for Disease Control and Prevention. (2002). 308.
Cigarette smoking—attributable mortality and Fiore, M. C., Bailey, W. C., Cohen, S. J., Dorfman,
years of potential life lost—United States. S. F., Goldstein, M. G., Gritz, E. R. et al. (2000).
Morbidity and Mortality Weekly Report, 42, 645– Treating tobacco use and dependence. Clinical
649. practice guideline. Rockville, MD: U.S. Department
Centers for Disease Control and Prevention. (2004). of Health and Human Services, Public Health
The health consequences of smoking: A report of the Service.
Surgeon General. Rockville, MD: U.S. Department Fiore, M. C., Jorenby, D. E., & Baker, T. B. (1997).
of Health and Human Services, Public Health Smoking cessation: Principles and practice based
Service. upon the AHCPR Guideline, 1996. Agency for
Center for Disease Control and Prevention. (2005). Health Care Policy and Research. Annals of
Cigarette smoking among adults—United States, Behavioral Medicine, 19, 213–219.
2004. MMWR Highlights, 54(44), 1121–1124. Fisher, E. B. (1995). The results of the COMMIT trial.
Champion V. L. (1994). Strategies to increase American Journal of Public Health, 85, 159–160.
mammography utilization. Medical Care, 32, 118– Glasgow, R. E., Bull, S. S., Gillette, C., Klesges, L. M.,
129. & Dzewaltowski, D. A. (2002). Behavior change
Chapman, S., Borland, R., Scollo, M., Brownson, intervention research in health care settings: A
R. C., Dominello, A., & Woodward, S. (1999). review of recent reports with emphases on
The impact of smoke-free workplaces on external validity. American Journal of Preventive
declining cigarette consumption in Australia and Medicine, 23, 62–69.
the United States. American Journal of Public Glynn, M., & Rhodes, P. (2005, June). Estimated HIV
Health, 89, 1018–1023. prevalence in the United States at the end of 2003.
Cole, S. W., Kemeny, M. E., & Taylor, S. E. (1997). Poster presented at the National HIV Prevention
Social identity and physical health: Accelerated Conference, Atlanta, GA. Abstract 595.
HIV progression in rejection-sensitive gay men. Goldstein, M. G., Niaura, R., Willey-Lessne, C.,
Journal of Personality and Social Psychology, 72, DePue, J., Eaton, C., Rakowski, W., et al. (1997).
320–335. Physicians counseling smokers: A population-
Cook, T., & Campbell, D. (1979). Quasi-experimental based survey of patients’ perceptions of health
design. Chicago: Rand McNally. care provider–delivered smoking cessation
Copeland, J., & Martin, G. (2004). Web-based interventions. Archives of Internal Medicine, 157,
interventions for substance use disorders: A 1313–1317.
qualitative review. Journal of Substance Abuse Gonzalez, V., Goeppinger, J., & Lorig, K. (1990).
Treatment, 26, 109–116. Four psychosocial theories and their application
Corwyn, R. F., & Benda, B. B. (1999). Examination of to patient education and clinical practice.
an integrated theoretical model of exercise Arthritis, Care, and Research, 3, 132–143.
behavior. American Journal of Health Behavior, 23, Goodman, R. M., Liburd, L. C., & Green-Phillips A.
381–392. (2001). The formation of a complex community
Cromwell, J., Bartosch, W. J., Fiore, M. C., Hassel- program for diabetes control: Lessons learned
blad, V., & Baker, T. (1997). Cost-effectiveness of from a case study of Project DIRECT. Journal of
the clinical practice recommendations in the Public Health Management and Practice, 7, 19–29.
AHCPR guideline for smoking cessation. Agency Grady, K. E., & Wallston, B. S. (1988). Research in
for Health Care Policy and Research. Journal of the health care settings. Newbury Park, CA: Sage.
American Medical Association, 278, 1759–1766. Grimley, D. M., Prochaska, G. E., & Prochaska, J. O.
Donner, A., & Klar, N. (1996). Statistical consider- (1993). Condom use assertiveness and the stages
ations in the design and analysis of community of change with main and other partners. Journal of
intervention trials. Journal of Clinical Epidemiology, Applied Biobehavioral Research, 12, 152–173.
49, 435–439. Harrison, D. J., Cantlay, A. M., Rae, F., Lamb, D., &
Elder, J. P., Ayala, G. X., & Harris, S. (1999). Theories Smith, C. A. (1997). Frequency of glutathione
and intervention approaches to health-behavior S-transferase M1 deletion in smokers with
change in primary care. American Journal of emphysema and lung cancer. Human Experimental
Preventive Medicine, 17, 275–284. Toxicology, 16, 356–360.
Elford, J., & Hart, G. J. (2003). If HIV prevention Helgeson, V. S., & Lepore, S. J. (1997). The hurdles
works, why are rates of high risk sexual involved in conducting a randomized clinical
intervention. Health Psychologist, 18(4), 4–5, 14– and disease prevention programs. Annual Review
16. of Public Health, 13, 31–57.
Herbert, T. B., Silver, R. C., & Ellard, J. H. (1991). Kohler, C. L., Grimley, D., & Reynolds, K. (1999).
Coping with an abusive relationship: 1. How and Theoretical approaches guiding the development
why do women stay? Journal of Marriage and the and implementation of health promotion
Family, 53, 311–325. programs. In J. M. Raczynski & R. J. DiClemente
Higgins, A., & Conner, M. (2003). Understanding (Eds.), Handbook of health promotion and disease
adolescent smoking: The role of the Theory of prevention (pp. 23–46). New York: Kluwer/
Planned Behavior and implementation intentions. Plenum.
Psychology, Health, and Medicine, 8(2), 173–186. Larzelere, R. E., Kuhn, B. R., & Johnson, B. (2004).
Hohmann, A. A., & Shear, M. K. (2002). Community- The intervention selection bias: An under-
based intervention research: Coping with the recognized confound in intervention research.
“noise” of real life in study design. American Psychological Bulletin, 130, 289–303.
Journal of Psychiatry, 159, 201–207. Lau, J. T. F., Thomas, J., & Liu, J. L. Y. (2000).
Holman, E. A. (1997). The nursing profession’s role in Mobile phone and interactive computer inter-
health psychology research: A reply to Helgeson viewing to measure HIV-related risk behaviours:
and Lepore. Health Psychologist, 19(2), 8. The impacts of data collection methods on
Johnston, K. L., & White, K. M. (2003). Binge- research results. AIDS, 14, 1277–1278.
drinking: A test of the role of group norms in the Lepore, S. J. (1995). Cynicism, social support, and
Theory of Planned Behavior. Psychology and cardiovascular reactivity. Health Psychology, 14,
Health, 18, 63–77. 210–216.
Joos, S., & Hickam, D. (1990). How health profes- Levinthal, C. F. (2005). Drugs, behavior, and modern
sionals influence health behavior: Patient provider society. Boston: Allyn and Bacon.
interaction and health care outcomes. In Lorig, K., & Fries, J. F. (1990). The arthritis helpbook:
K. Glanz, F. Lewis, & B. Rimer (Eds.), Health A tested self-management program for coping with
behavior and health education: Theory, research and your arthritis. Reading, MA: Addison-Wesley.
practice (pp. 216–241). San Francisco: Jossey Lorig, K., Seleznick, M., Lubeck, D., Ung, E.,
Bass. Chastain, R. L., & Holman, H. R. (1989). The
Jourenkova-Mironova, N., Wikman, H., Bouchardy, beneficial outcomes of the arthritis self-manage-
C., Voho, A., Dayer, P., Benhamou, S., et al. ment course are not adequately explained by
(1998). Role of glutathione S-transferase GSTM1, behavior change. Arthritis and Rheumatism, 31,
GSTM3, GSTP1 and GSTT1 genotypes in 91–95.
modulating susceptibility to smoking-related lung Maisto, S. A., Connors, G. J., & Zywiak, W. H.
cancer. Pharmacogenetics, 8, 495–502. (2000). Alcohol treatment, changes in coping
Kaplan, R. M. (2000). Two pathways to prevention. skills, self-efficacy, and levels of alcohol use and
American Psychologist, 55, 382–396. related problems 1 year following treatment
Kassel, J. D., Stroud, L. R., & Paronis, C. A. (2003). initiation. Psychology of Addictive Behaviors, 14,
Smoking, stress, and negative affect: Correlation, 257–266.
causation, and context across stages of smoking. Marshall, S. J., & Biddle, S. J. H. (2001). The
Psychological Bulletin, 129, 270–304. Transtheoretical Model of Behavior Change: A
Kemeny, M. E. (2003). An interdisciplinary research meta-analysis of applications to physical activity
model to investigate psychosocial cofactors in and exercise. Annals of Behavioral Medicine, 23,
disease: Application to HIV-1 pathogenesis. Brain, 229–246.
Behavior, and Immunity, 17(Supplement), S62– May, S., & West, R. (2000). Do social support
S72. interventions (“buddy systems”) aid smoking
Kenny, D. A. (1979). Correlation and causality. New cessation? A review. Tobacco Control, 9, 415–422.
York: Wiley. McAuley, E., Jerome, G. J., Marquez, D. X., Elavsky,
Kerner, J., Rimer, B., & Emmons, K. (2005). S., & Blissmer, B. (2003). Exercise self-efficacy in
Introduction to the Special Section on Dissemina- older adults: Social, affective, and behavioral
tion. Dissemination research and research influences. Annals of Behavioral Medicine, 25, 1–7.
dissemination: How can we close the gap? Health McCann, J. M., Bovbjerg, V. E., Brief, D. J., Turner,
Psychology, 24, 443–446. C., Follette, W. C., Fitzpatrick, V., et al. (1995).
Koepsell, T. D., Wagner, E. H., Cheadle, A. C., Relationship of self-efficacy to cholesterol
Patrick, D. L., Martin, D. C., Diehr, P. H., et al. lowering and dietary change in hyperlipidemia.
(1992). Selected methodological issues in Annals of Behavioral Medicine, 17, 221–226.
evaluating community-based health promotion McNally, R. J., Bryant, R. A., & Ehlers, A. (2003).
Does early psychological intervention promote Rigotti, N. A., Lee, J. E., & Wechsler, H. (2000). US
recovery from post-traumatic stress? Psychological college students’ use of tobacco products: Results
Science in the Public Interest, 4, 45–79. of a national survey. Journal of the American
Mermelstein, R. (1999). Ethnicity, gender and risk Medical Association, 9, 699–705.
factors for smoking initiation: An overview. Rigotti, N. A., Moran, S. E., & Wechsler, H. (2005).
Nicotine and Tobacco Research, 1(Suppl 2), S69– US college students’ exposure to tobacco
S70. promotions: Prevalence and association with
Merzel, C., & D’Afflitti, J. (2003). Reconsidering tobacco use. American Journal of Public Health, 95,
community-based health promotion: Promise, 138–144.
performance, and potential. American Journal of Rose, G. (1985). Sick individuals and sick popula-
Public Health, 93, 557–574. tions. International Journal of Epidemiology, 14,
Mishra, S. I., Sanudo, F., & Conner, R. F. (2004). 32–38.
Collaborative research toward HIV prevention Rose, G. (1992). The strategy of preventive medicine.
among migrant farmworkers. In B. P. Bowser & New York: Oxford University Press.
S. I. Mishra et al. (Eds.), Preventing AIDS: Rosentock, I. M., Strecher, V. J., & Becker, M. H.
Community-science collaborations (pp. 69–95). (1994). The health belief model and HIV risk
New York: Haworth. behavior change. In R. J. DiClemente & J. L.
Nyberg, F., Hou, S. M., Hemminki, K., Lambert, B., & Peterson (Eds.), Preventing AIDS: Theories and
Pershagen, G. (1998). Glutathione S-transferase methods of behavioral interventions (pp. 5–24).
mu1 and N-acetyltransferase 2 genetic polymor- New York: Plenum.
phisms and exposure to tobacco smoke in Rutter, D., & Quine, L. (2002). Changing health
nonsmoking and smoking lung cancer patients behavior: Intervention and research with social
and population controls. Cancer Epidemiology cognition models. Buckingham, PA: Open
Biomarkers and Prevention, 7, 875–883. University Press .
Park, E., Schultz, J. K., Tudiver, F., Campbell, T., & Schlenger, W. E., & Silver, R. C. (2006). Web-based
Becker, L. (2004). Enhancing partner support to methods in terrorism and disaster research.
improve smoking cessation. Cochrane Database of Journal of Traumatic Stress, 19(2).
Systematic Review, 1, CD002928. Schneiderman, N. (2004). Psychosocial, behavioral,
Perkins, K. A. (2001). Smoking cessation in women: and biological aspects of chronic disease. Current
Special considerations. CNS Drugs, 15, 391–411. Directions in Psychological Science, 13, 247–251.
Perkins, K. A., Donny, E., & Caggiula, A. R. (1999). Schneiderman, N., Antoni, M. H., & Ironson, G.
Sex differences in nicotine effects and self- (2003). Behavioral medicine in HIV infection.
administration: Review of human and animal Psychotherapeutics, 48, 342–347.
evidence. Nicotine and Tobacco Research, 1, 301– Schorling, J. B., Roach, J., Siegel, M., Baturka, N.
315. Hunt, D. E., Guterbock, T. M., et al. (1997). A
Piper, D. L. (2000). The Health for Life Project: trial of church-based smoking cessation interven-
Behavioral outcomes. Journal of Primary Preven- tions for rural African Americans. Preventive
tion, 21, 47–73. Medicine, 26, 92–101.
Prochaska, J. O., & DiClemente, C. C. (1983). Stages Secker-Walker, R. H., Gnich, W., Platt, S., &
and processes of self-change of smoking: Toward Lancaster, T. (2005). Community interventions
an integrative model of change. Journal of for reducing smoking among adults. Cochrane
Consulting and Clinical Psychology, 51, 390–395. Database of Systematic Review, 1, CD001745.
Prochaska, J. O., DiClemente, C. C., & Norcross, J. C. Sheeran, P., & Abraham, C. (1996). The Health Belief
(1992). In search of how people change: Model. In M. Conner & P. Norman (Eds.),
Applications to addictive behaviors. American Predicting health behaviour: Research and practice
Psychologist, 47, 1102–1114. with social cognition models (pp. 23–61). Buck-
Reilly, P. M., Sees, K. L., Shopshire, M. S., Hall, S. M., ingham, England: Open University Press.
Delucchi, K. L., Tusel, D. J., et al. (1995). Self- Shiffman, S., Balabanis, M. H., Paty, J. A., Engberg, J.,
efficacy and illicit opioid use in a 180-day Gwaltney, C. J., Liu, K. S., et al. (2000). Dynamic
detoxification treatment. Journal of Consulting and effects of self-efficacy on smoking lapse and
Clinical Psychology, 63, 158–162. relapse. Health Psychology, 19, 315–323.
Revenson, T. A., & Schiaffino, K. M. (2000). Shilts, R. (1987). And the band played on: Politics,
Community-based health interventions. In people, and the AIDS epidemic. New York:
J. Rappaport & E. Seidman (Eds.), Handbook of St. Martin’s.
community psychology (pp. 471–493). New York: Shine, B. (2000). Nicotine vaccine moves toward clinical
Kluwer Academic/Plenum. trials. NIDA Notes: Nicotine Research, 15, 1–2.
Sieber, J. E. (1998). Planning ethically responsible Pleck, J. H., & Sonenstein, F. L. (1998).
research. In L. Bickman & D. J. Rog (Eds.), Adolescent sexual behavior, drug use, and
Handbook of applied social research methods violence: Increased reporting with computer
(pp. 127–156). Thousand Oaks, CA: Sage. survey technology. Science, 280, 867–873.
Siemiatycki, J., Krewski, D., Franco, E., & Kaiserman, UNAIDS, Joint United Nations Programme on HIV/
M. (1995). Associations between cigarette AIDS (2004). Global summary of the AIDS
smoking and each of 21 types of cancer: A multi- epidemic. Aids Epidemic Update—2004.
site case-control study. International Journal of Velicer, W. F., Prochaska, J. O., Rossi, J. S., & Snow,
Epidemiology, 24, 504–514. M. G. (1992). Assessing outcome in smoking
Sobel, D. S. (1995). Rethinking medicine: Improving cessation studies. Psychological Bulletin, 111, 23–41.
health outcomes with cost-effective psychosocial Wechsler, H., Kelley, K., Seibring, M., Juo, M., &
interventions. Psychosomatic Medicine, 57, 234– Rigotti, N. A. (2001). College smoking policies
244. and smoking cessation programs: Results of a
Sorensen, G., Emmons, K., Hunt, M. K., & Johnston, survey of college health center directors. Journal
D. (1998). Implications of the results of commu- of American College of Health, 50, 141–142.
nity intervention trials. Annual Review of Public Westmaas, J. L., & Billings, M. (2005). Gender, social
Health, 19, 379–416. support, and stress reactions in smokers. Unpub-
Sowden, A., Arblaster, L., & Stead, L. (2005). lished manuscript.
Community interventions for preventing smoking Westmaas, J. L., & Jamner, L. D. (in press). Paradoxi-
in young people. Cochrane Database of Systematic cal effects of social support on blood pressure
Review, 1, CD001291. reactivity among defensive individuals. Annals of
Stokols, D. (1992). Establishing and maintaining Behavioral Medicine.
healthy environments: Toward a social ecology of Westmaas, J. L., Wild, C., & Ferrence, R. (2002).
health promotion. American Psychologist, 47, 6–22. Effects of gender in social control of smoking.
Stokols, D., Pelletier, K. R., & Fielding, J. E. (1995). Health Psychology, 21, 368–376.
Integration of medical care and worksite health Wortman, C. B., Hendricks, M., & Hillis, J. W.
promotion. Journal of the American Medical (1976). Factors affecting participant reactions to
Association, 273, 1136–1142. random assignment in ameliorative social
Strickland, B. R., & Crowne, D. P. (1963). Need for programs. Journal of Personality and Social
approval and the premature termination of Psychology, 33, 256–266.
psychotherapy. Journal of Consulting and Clinical Zabalegui, A., Sanchez, S., Sanchez, P. D., & Juando,
Psychology, 27, 95–101. C. (2005). Nursing and cancer support groups.
Thompson, B., Coronado, G., Snipes, S. A., & Journal of Advanced Nursing, 51, 369–381.
Puschel, K. (2003). Methodological advances and Zimmerman, R. S., & Olson, K. (1994). AIDS-related
ongoing challenges in designing community- risk behavior and behavior change in a sexually
based health promotion programs. Annual Review active, heterosexual sample: A test of three
of Public Health, 24, 315–340. models of prevention. AIDS Education and
Turner, C. F., Ku, L., Rogers, S. M., Lindberg, L. D., Prevention, 6, 189–205.

Unidad 1 Westmaas Et Al 2007 2004

Uploaded by

Copyright:

Available Formats

Unidad 1 Westmaas Et Al 2007 2004

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Unidad 1 Westmaas Et Al 2007 2004

Uploaded by

Copyright:

Available Formats

52 History and Methods

Designing and Implementing

components of community-level interventions are individual-level approaches. For example, social

priorities and limitations of community resources Ethical Issues

You might also like