Child secondhand smoke exposure (SHSe) is linked to asthma, respiratory illness, otitis, hospitalization rates, headaches, sudden infant death, and behavior problems
[1–4]. Because parental smoking is a primary source of SHSe in children, the American Academy of Pediatrics developed pediatric tobacco-intervention guidelines
 that address parental smoking. Pediatric healthcare visits provide a teachable moment to increase parents’ awareness of the adverse effects of SHSe on children and motivate them to protect children from SHSe
. However, there are barriers to effective pediatric-clinic interventions. For example, competing demands, time limitations, and systems barriers may prohibit offering medications to address parents’ nicotine addiction or providing in-depth counseling and training in self-regulatory skills that can alter smoking behavior
[7, 8]. The Kids Safe and Smokefree (KiSS) program addresses these limitations via a multilevel intervention model that integrates a pediatric clinic-level intervention with more intensive individual-level behavioral counseling and navigation to community-level services for nicotine dependence.
The KiSS clinic-level intervention focuses on improving the quality of clinic-delivered tobacco-related messages for parents. It emphasizes three elements, known as “Ask, Advise, and Refer” (AAR): Ask about child SHSe, Advise about the harms of SHSe and benefits of reducing SHSe, and Refer to cessation resources. The clinic AAR intervention can motivate and assist parents to take the initial actions to protect their children from SHSe. The KiSS behavioral health intervention then provides more intensive intervention that may be necessary to promote smoking behavior change. The behavioral intervention combines personalized, intensive family-centered counseling, smoking urge management and coping skills training and social support with community-level systems navigation to facilitate access to and effective use of no-cost nicotine replacement therapy and reimbursable cessation medication. The behavioral intervention includes a home visit to introduce intervention concepts and initiate skills training around reducing children’s SHSe in the home and car as a primary step toward preparing to quit smoking. The home is a critical target of intervention because restrictions on smoking across the U.S. do not typically extend to private homes and cars - contexts in which child SHSe is greatest. Weekly telephone counseling following the home visit initially emphasizes reducing child SHSe and then progresses to address smoking cessation and relapse prevention.
The design and sequencing of procedures in our trial were influenced in part by a pivotal review of smoking cessation interventions in medical practice. In that review, Kottke
 concluded that information to a smoker from one type of personnel (e.g., clinician) may potentiate information from another type of personnel (e.g., health counselor), and the number and duration of reinforcing sessions are related to cessation success. Kottke’s review focused on cessation, but we believe the same effects of multiple message sources and repeated “doses” of advice can be harnessed for SHSe reduction using an integrated multilevel, multimodal intervention. We will evaluate efficacy of the intervention in predominantly low-income, urban and minority communities with excess SHSe-related morbidity and mortality risk.
To date, most pediatric SHSe interventions adopt what Anderson
 has described as a single level of analysis--focusing either on environmental factors (e.g., smoke-free policies), social factors (e.g., pediatrician recommendation), or individual factors (e.g., motivation to change). The failure to develop a multilevel approach has impeded progress in the field because it is well established that smoking is multidetermined, and when an intervention targets a single level of a multidetermined behavior, that intervention provides insufficient elements to maintain healthy behavior change. For example, interventions that target one particular cause (e.g., limited knowledge about harm) might not address other relevant causes (e.g., nicotine addiction). Hence, a multilevel approach is likely to be more effective than a single-level approach.
The multilevel KiSS model follows recent recommendations to advance the science of health behavior change by testing a multilevel approach that addresses individual, group, and environmental influences simultaneously and over time. The model minimizes burden on clinicians, but the clinician still acts as a credible gateway to the more intensive intervention for smoking parents. The specific components of the clinic- and individual-level interventions are informed by the literature, including our preliminary studies and theory
[11–13]. For example, research links social support
, coping skills
, and self-efficacy
 to smoking behavior change. The KiSS interventions have elements shown to be associated with improved social support
, coping skills
, and self-efficacy
. Other non-program factors, such as psychological symptoms
, nicotine dependence
, and presence of other smokers in home
, are known predictors of smoking outcomes and might moderate intervention efficacy.
Aims and hypotheses
Aim 1: Test the primary hypothesis that an intervention integrating pediatric clinic-level quality improvement with individual-level behavioral counseling (AAR + BC) will be more effective in reducing children’s SHSe than a clinic-level quality improvement plus attention control intervention (AAR + AC). Hypothesis: Compared with children in the AAR + AC condition, those in the AAR + BC condition will have significantly greater reductions in SHSe from baseline to 3- and 12-month follow-up.
Aim 2: Test the secondary hypothesis that AAR + BC will be more effective in increasing parental quit rates than AAR + AC. Hypothesis: Compared with parents in the AAR + AC condition, parents in the AAR + BC condition will have a significantly greater smoking abstinence at 3- and 12-month follow-up.
Aim 3: Test hypotheses that social cognitive variables (social support, urge management coping skills, self-efficacy) will mediate effects of the AAR + BC intervention on outcomes. Hypothesis: Compared with parents in the AAR + AC condition, parents in the AAR + BC condition will report greater social support, coping skills, and self-efficacy related to smoking cessation and SHSe reduction from baseline to 3- and 12-month follow-up. In turn, these changes will account for between-group differences in child SHSe and parent cessation outcomes.
Aim 4: Explore factors that may affect outcomes and moderate intervention effects, including presence of other smokers at home, level of nicotine dependence, and depressive symptoms.