A systematic review of parent based programs to prevent or reduce alcohol consumption in adolescents

Peer-reviewed literature was systematically searched to identify relevant studies published between database inception and August 2019 from eight databases including; PubMed, EBSCO, Emerald, ProQuest, Ovid, ScienceDirect, Taylor & Francis and Web of Science. A copy of the review protocol was not prospectively registered, however, the systematic search of the literature followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [54]. Due to the heterogeneity of the identified programs in regards to study populations and outcome measures, meta-analysis was not possible [55, 56]. Eight databases were searched using the keywords alcohol*, parent* and school* in combination with randomized controlled trial, intervention, program, and evaluation. See Table 1 for a detailed list of search terms.

A total of 6837 records were retrieved as summarized in Table 2. Each record was downloaded to Endnote and 2549 duplicates were removed. Inclusion and exclusion criteria were applied to the remaining 4288 unique records to ensure they were accurate representations of studies evaluating parent alcohol programs.

Studies published in English and meeting the following criteria were included in the review: 1) detailed parent reported outcome measures on any factors related to general parenting or substance-related parenting, 2) published between database inception and August 2019, 3) reported results of an alcohol prevention program that involved an element of parent training, 4) included a control or comparison group, and 5) included parents of adolescents’ aged 10–18 years old as substance use typically first occurs in adolescence.

Studies were excluded if; 1) they did not evaluate a universal program (i.e. studies targeting specific populations such as immigrant families), 2) there was no program implemented, 3) they did not assess parent reported outcome measures associated with general or alcohol specific parenting behaviours and, 4) they were not a journal article. The flowchart based on the PRISMA guidelines [54] is shown in Fig. 1. In total 17 studies detailing 13 individual parent programs were identified prior to forward and backwards searching. To allow for accurate reporting of program development, implementation and evaluation, forward and backward searching was undertaken. Specifically, searches using authors’ names and program names were undertaken in Google Scholar. A further 25 relevant studies detailing further information on the included programs were located. In total 42 studies were included in the analysis of 13 different programs. The full list of the 42 studies for the 13 programs can be found in Table 4 in Appendix.

Two independent reviewers screened titles, abstracts, and full texts of potential articles. A third reviewer resolved disagreements regarding inclusion of a study. Following the application of the exclusion criteria, 17 studies evaluating 13 universal parent alcohol programs on parent outcome measures were identified.

The included studies were analyzed in terms of; 1) program effectiveness on parent reported outcome measures, 2) the level of stakeholder engagement in program design, 3) the level of theory use and, 4) the quality of program design and delivery.

In total, 13 unique programs reported in 42 studies qualified for inclusion. Two programs were culturally adapted. Project Northland was culturally adapted for Russia [55], and Orebro (later named Effekt) was culturally adapted for Estonia [68]. The Orebro program was trialed on two separate occasions in Sweden [69, 70]. The majority of programs identified were conducted in the USA (46%, n = 6). The sample size of parents ranged from 64 [71] to 2048 [72]. Parent based programs are typically delivered through means of; workshops [73‐75]; posting materials via mail [69, 72]; and take home materials from school [55, 76, 77]. The majority of programs focused on reducing parents permissive norms towards underage drinking [68, 69, 72‐74]; encouraged parents to set clear alcohol specific rules [71, 72, 74, 75, 77] and aimed to increase parent-child communication [55, 72, 74, 78]. The follow-up assessment period ranged from immediately post program [79] to 36-months [69]. Standard health education was administered to the majority of the control conditions in each trial, however, informational booklets on different aspects of adolescent development were mailed to control parents in the Preparing for the Drug Free Years (PDFY) trial [75]. An additional file provides an overview of the characteristics of the studies included in the review (see Additional file 1).

The majority of the identified parent alcohol programs reported limited levels of stakeholder engagement during program design. Specifically, over half of the programs (n = 8, 61%) reported methods used to inform stakeholders [68, 71, 73, 74, 79, 81, 82, 84]. For example, studies reporting on the Orebro [84] and PDFY [82] programs do not mention any form of stakeholder input during program design, reporting instead that an extensive review of the literature was conducted to inform the development of the respective programs. Three programs (23%) reported methods used to consult with stakeholders [55, 68, 76]. When designing the Self-help home ATOD communication program, formative research was conducted with parents involving a self-complete questionnaire and structured small group discussions [76]. Parents provided insights into their specific needs in terms of communicating with their adolescent (e.g. what topics to talk about) and their preferences towards the nature and delivery of the program (e.g. interactive, home-based, easy to read) [76].

Additionally, two programs utilized methods to involve, working with multiple stakeholders and developing alternative parent ideas based on stakeholder input [72, 77]. Formative research was conducted with parents representing the target audience of the Family Matters program [85]. Informal discussion with parents led to the development of a pilot test which was administered to the parents. Program materials and procedures were refined based on parent feedback [85]. Throughout this process advice was solicited from a variety of experts such as a clinical psychologist, researchers in the field of adolescent alcohol use and developers of prior family based prevention programs [77]. No programs reported use of methods to collaborate or empower stakeholders during program design.

Of the programs identified eight (61%) reported the use of a behavior change theory. Of those that specified a theoretical framework, three used one theory [71, 76, 79], three used two theories [55, 72, 82] and two used three or more theories [74, 77]. The most frequently identified theories were social cognitive theory [72, 74, 76, 78, 82] social learning theory [74, 77, 82] and ecological frameworks [55, 71, 72]. In terms of theory utilization level, three programs applied theory [55, 71, 79]. For example, The Substance Use Prevention Promoted by Eating family meals Regularly (SUPPER) program was developed using eco-developmental theory as a guiding framework. Key theoretical constructs, namely family bonding and social interactions were clearly identified in study reporting [71]. Five programs tested theory [72, 74, 76, 77, 82]. For example, Beatty et al., [76] report that the Self-help home ATOD Communication program incorporated key elements of SCT (e.g. improving parents’ self-efficacy to discuss alcohol related topics with their adolescent). The study used and measured several key theoretical constructs in both program implementation and evaluation (e.g. self-efficacy, knowledge and outcome expectations) [76]. No programs built theory.

A quality assessment of the identified programs was conducted using the EPHPP tool (see Table 3). Of the 13 identified programs, ten were assessed as weak in the global rating, three were assessed as moderate and none were assessed as strong. Selection bias was likely in many studies. Only one third of the programs reported representative sampling methods [68, 72, 76, 78]. However, as one of these programs reported low participation levels, only two programs were assessed as strong in regards to selection bias. Only three of the included studies described how randomisation sequences were generated [68, 69, 78] and therefore these were assessed as strong. In terms of confounders, the majority of programs (n = 10, 77%) reported either no baseline differences between groups or studies controlled for at least 80% of relevant confounders resulting in a strong rating. The rest of the programs did not report potential confounders or account for them in analysis and were therefore assessed as weak [69, 74, 77]. In all programs, blinding for both assessors and participants was not reported. In terms of data collection methods, almost half (n = 6, 46%) [55, 73‐75, 78, 80] of the included studies provided evidence of the validity and reliability of the reported outcomes measures and were therefore assessed as strong. The remaining programs were assessed as either moderate (n = 3) [69, 76, 79] for reporting validity only or weak (n = 4) [68, 71, 72, 77] as they did not report validity. Regarding the retention rates of parents, only two programs were assessed as strong with more than 80% of parents completing the program.

Each of the identified programs focused on influencing specific parenting factors associated with preventing or reducing alcohol use among adolescents including; parents’ restrictive attitudes, parent-child communication, alcohol-specific rule setting and parental monitoring. These forms of parental involvement have been described in previous studies as important protective factors for risky drinking behaviors in adolescents [17, 86].

Parental attitudes towards underage drinking changed as a result of program participation with seven of the nine programs evaluating attitudes demonstrating significant positive results. Parents restrictive attitudes towards underage drinking acts as one of the strongest protective factor for risky drinking behaviours in adolescents [87]. A change in parental attitudes is an important step towards ultimately influencing parenting specific behaviors as influencing attitudes increases the likelihood of changing behavioral intentions [88]. Furthermore, significant program effects were reported in four of the five programs measuring parent-child communication. Frequent and open alcohol specific communication between parents and adolescents can reduce adolescents’ alcohol usage, while also increasing their perceptions of the negative consequences associated with alcohol use [89].

The results showed that parent alcohol programs are effective in increasing parents’ restrictive rules concerning underage drinking. Adolescents whose parents enforce strict alcohol specific rules are less likely to engage in risky drinking behaviors [8, 90]. Furthermore, the enforcement of strict rules can still exert a protective effect from early adolescence until early adulthood [91]. Concerningly, in each of the studies assessing parental monitoring practices no program effects were observed. Operating as one of the strongest protective factors for adolescent alcohol use [6, 13, 92], parental monitoring has been found to minimize underage alcohol use [4, 93] improve adolescents’ self-efficacy to refuse alcohol [86] and improve family closeness [6].

In line with previous systematic reviews, this study provides evidence to support the efficacy of parent alcohol programs in preventing or reducing adolescent alcohol consumption [1, 13, 14] by focusing on parent specific outcome measures. However, improvements in study quality would extend confidence in reported findings. Extensive literature lends support to the notion that parenting specific behaviors play an important role in influencing and predicting adolescents’ drinking behaviors [17, 86, 93]. The results highlight emerging evidence that parent alcohol programs can achieve dual aims, providing a protective effect on adolescents in addition to changing parenting practices and attitudes. Further investigation is needed before definitive conclusions can be drawn giving heterogeneity in outcome variables and the lack of strong quality studies.

Representation and involvement of a broad range of stakeholders in the design stage of the identified parent alcohol programs was limited. The results of this study support an earlier review which identified that programs aiming to minimize harm from alcohol lacked stakeholder engagement, with a restricted focus on the people whose behavior needed to change (e.g. problem alcohol drinkers) [28]. While the importance of collaborating with key stakeholders during program design has been evidenced [21, 31, 35], this review highlights a lack of stakeholder consideration and inclusion during program design stages. Involving stakeholders during program design not only creates a sense of ownership among participants, through increased participation and empowerment [30] but may enhance effectiveness through greater acceptance and adoption of the program [21].

Over half of the identified parent alcohol programs reported methods used to inform participants, involving limited one way communications. This limited stakeholders, giving them no voice or power to influence decisions on the programs to be implemented. Collaborations with stakeholders allow program developers to tap in to the unique perspectives and insights held by various stakeholders [26] whose interests are varied. Discrepancies between expert and user views exist in the context of parent alcohol programs [94], suggesting expert designed programs may be failing to meet the unique needs of parents. Future program efforts should seek to include multiple stakeholder perspectives beyond the end user group to identify and acknowledge multiple views and resolve possible conflicts and discrepancies. For example, to assess needs and guide program development Project Northland developers consulted experts, conducted focus groups with parents and interviewed community leaders including mayors, police chiefs, school principals and local council members [95]. This allowed program developers to gain a comprehensive understanding of community expectations and views on adolescent alcohol consumption.

Stakeholder empowerment can improve innovation [36] reducing resistance towards desired change [37]. Therefore, program design processes need to evolve in order to achieve higher levels of empowerment and advance the sustainable development of programs. Thus, it is suggested that researchers consider novel methods that allow stakeholders to actively contribute during the design process as opposed to being passive participants. For example, empowering stakeholders through active collaborations such as co-design methods that provide stakeholder with the tools and a voice to design behaviour change programs of value to them [94].

Theories are used in the development of programs to effect better outcome change [38], through influencing constructs that are known to cause specific behavior [39]. Reporting of theory use in current parent alcohol programs is lacking with only eight (61%) of the 13 programs explicitly mentioning and utilizing theory. These findings are consistent with other reviews that indicate a lack of theory use in behavior change programs [47, 96]. Theories are commonly used for audience research and segmentation [97], program development [98], message formation, promotion [98] and evaluation [46, 99]. During the formative research phase theories may be useful to assist in setting explicit aims and objectives, segmenting target audiences [100], identifying messages that resonate with the target audience and identifying important barriers and benefits to focus on [101].

Social cognitive theory is one promising theory for use in parent alcohol programs. Bandura [102] highlights the importance of SCT in behaviour change programs, emphasizing the usefulness of the theory’s constructs in influencing behaviour change. From this theoretical perspective behaviour in influenced environmental and personal factors as well as behaviour [42]. As one of the most commonly used theory’s in health behaviour research [103], it has been suggested that alcohol education programs may benefit from the application of SCT [104]. For example, the Self-help ATOD Communication program was developed based on SCT and successfully improved parent-child communication regarding alcohol and other drugs [76]. Program materials focused on key constructs of the theory such as improving parent self-efficacy to communicate to the adolescents about alcohol and increasing their knowledge of the risks associated with underage drinking [76]. However, while SCT provides health promotion researchers with one possible theoretical lens through which to examine parent alcohol program, the current findings highlight the need for more thorough application, testing and reporting of theories in behavior change programs.

Using the EPHPP quality assessment tool the methodological quality of the included programs were assessed. Ten programs were rated as weak and three as moderate. None were rated as strong. Overall the methodological quality of the included programs was low and conclusions should be interpreted with caution. Notable, methodological problems included selection biases and lack of assessor blinding. Only two programs selected participants representative of the population and achieved greater than 80% initial participation. In the included studies selection biases arose around practical issues related to the recruitment of parents. Parent were largely self-referred through convenience sampling methods such as mailouts or approaching parents at school pick up zones [71, 80], and control groups often consisted of parents who elected not to participate in program delivery [73, 79]. These findings highlight the need for increased resources dedicated to the evaluation of parent alcohol programs that permit for large scale systematic recruitment procedures.

Finally, in the current review no programs reported blinding of participants or assessors. Prior criticism has pointed to the lack of external evaluations in school based alcohol prevention programs [105]. While double blinding may not be feasible in parent based alcohol programs due to the nature of the trials, more external evaluations of parent alcohol programs are needed. Parent based alcohol program evaluations are largely evaluated by program developers who may have a vested interest in program success, thereby influencing the interpretation and reporting of results [105].

To enhance empirical evidence, future research should aim to address these issues and improve the methodological quality of parent alcohol programs. Due to the inconsistent evaluation methods and outcomes measures it was not possible to directly compare programs and make meaningful comparison of program components. Therefore, to move beyond a narrative description of programs and provide evidence towards the effectiveness of program components, more systematic reporting and evaluations of parents programs are needed. In doing so, researchers and practitioners will be provided will a more comprehensive picture of what does and does not make a program successful.