Coaching to create a smoke-free home in a brief secondhand smoke intervention

Abstract

Few community interventions exist to reduce secondhand exposure to tobacco smoke in the home. This study presents the coaching process of a larger intervention to promote smoke-free homes across an efficacy and 2 effectiveness trials. It furthers assesses the coaching call’s reach and participants’ satisfaction with the call across three intervention sites. The sources of the data were from baseline and 3-month follow-up surveys, coaching forms or online tracking system and interviews with coaches. Reach for the coaching call across trials was fairly high from 72% to 92%. Overall, the majority of participants were highly satisfied with the intervention (M = 3.76) and found it useful in creating a smoke-free home (M = 3.63). Common goals set were changing the environment to support a home smoking ban (e.g. putting up signs, removing ashtrays) (82%) or picking a date (60%). Challenges to a smoke-free home were consistent with other literature on barriers related to household smoking restrictions, including need for assistance in quitting, outside weather and smokers who do not want to quit. Additional research is needed to explore differential reach and reactions to the coaching call as it is disseminated and the impact of coaching call on the outcome of a smoke-free home.

Introduction

The national prevalence of households with a smoke-free ban, defined as no smoking any place at any time, increased from 43.0% during 1992–1993 to 83.0% during 2010–2011 in the United States [1, 2]. An estimated 58 million non-smokers were exposed to secondhand smoke (SHS) from 2011 to 2012 and exposure is higher among Black and lower income households [2]. However, households with at least one smoker and lower income households still are less likely to have a smoke-free rule [2, 3]. Continued adoption of smoke-free home rules is critical to reducing health impacts of SHS such as cardiovascular disease, lung cancer and respiratory diseases [4].

Efforts to reduce SHS in the home have focused mostly on clinical settings with an emphasis on cessation or clinician counseling and national campaigns (e.g. Environmental Protection Agency’s Smokefree Home Pledge) [5–10]. Rosen et al. [11] found that interventions for SHS exposure reduced air nicotine in homes in a review of seven studies with families from medical settings or government assistance programs (i.e. Health Start); all but two were in clinical settings. Therefore, there are few community interventions to promote household smoking bans [12, 13].

Research on community interventions to reduce SHS is an understudied area [13]. In a recent systematic review, out of 57 studies interventions aimed to reduce SHS exposure of children, only 7 studies targeted at the community setting [14]. Few community interventions to promote smoke-free homes have been evaluated. Alwan et al. [15] found that a multicomponent intervention of school education, community events and brief intervention with a professional, for families of children aged 9–12 years, significantly increased homes with total bans from 35% at pretest to 68% at post-test. Other interventions found some positive effects on home air quality for an enhanced intervention with home visits with personalized home air quality feedback between earlier and later visits [16] and for a multicomponent intervention (personalized feedback on home air quality, behavioral support, and nicotine replacement therapy) over usual care [17]. Baxi et al. [14] reported on 11 studies that incorporated motivational interviewing (MI) or intensive counseling and 7 studies included stages of change as strategies to reduce SHS exposure. More research, to determine the effectiveness of these intervention strategies in different settings, is needed to build the evidence base for community programs to reduce SHS exposure [18].

To fill this research gap in community interventions for promoting smoke-free environments, we developed an intervention to create smoke-free households over a series of trials [19–21]. It aimed to increase household smoking bans and reduce non-smokers’ exposure to SHS and has four components (three mailings and one coaching call). Printed educational materials were designed around the theme of ‘Some Things are Better Outside’ and include a Five-Step Guide to a Smoke-Free Home, Challenges and Solutions booklet, pledge, information on secondhand and thirdhand smoke, and the benefits of smoke-free homes, a photo story depicting a family going smoke-free, stickers, a window cling and signs. The intervention was delivered at 2-week intervals beginning with a mailing, followed by a brief coaching call, and the remaining two mailings. The intervention materials were developed based on the Social Cognitive Theory, Transtheoretical Model and MI principles. We held two focus groups to elicit ideas about the intervention and to pilot materials and then implemented a pilot study with 40 participants [22]. We conducted a randomized controlled efficacy trial with 2-1-1 participants in Atlanta and then conducted two randomized controlled effectiveness trials in North Carolina and the Texas Gulf Coast region [19–21]. All three trials demonstrated positive results on home smoking bans 6 months after baseline.

The intervention coaching call used evidence-based components of MI [23–25], stages of change [26–27], goal-setting [28–29] and problem solving [30] and drew on the overarching theory of Social Cognitive Theory. Because change is often best achieved when these techniques are combined, this intervention utilized evidence-based counseling techniques drawn from the three frameworks (described below). We used the effective components of these frameworks to develop a protocol for the smoke-free homes brief ‘coaching call’. This study presents the smoke-free homes intervention coaching process and results related to the coaching call’s reach and participants’ satisfaction with the call across the three intervention sites, and it describes participants’ stage of change and major goals at the mid-point of the overall intervention for the Atlanta site. These findings offer information about the feasibility and acceptability of this approach as part of the larger intervention to promote smoke-free homes from an efficacy trial to effectiveness trials and ultimately, for broad scale dissemination.

Methods

Participants and procedures

Study sample and data are from the Atlanta trial (total n = 498, intervention group = 247) testing the efficacy of a smoke-free homes intervention to increase household smoking rules in Atlanta [19] and the two effectiveness trials to further test the effectiveness of the intervention in different locations and among different populations, with the coaching calls made by trained 2-1-1 information & referral (I&R) specialists [20, 21]. A year prior to the Atlanta trial, the proportion of people with total household restrictions in Georgia was 91.5 [1]. I&R specialists at the United Way of Greater Atlanta (UWGA) 2-1-1 social services information and referral system recruited and consented participants beginning June 2012 and completed study activities in December 2013 for the efficacy trial [19]. Effectiveness trials were conducted by 2-1-1’s in North Carolina and Texas Gulf Coast [20, 21]. 2-1-1 covers all 50 states with over 200 state and local call centers [31]. Callers to 2-1-1 are disproportionately low-income, unemployed, uninsured and have fewer years of education relative to the general population, and seek assistance with basic human needs such as finding shelter, heat, electricity and food [32]. At the time of the trials, the smoking rates were 20.4% in Atlanta (2012), 20.3% in NC (2013) and 14.5% in Texas (2014) [33].

Eligibility criteria were 18 years of age or older, speak and understand English, and a smoker or non-smoker living in a smoking discordant household. Consenting participants completed a baseline interview, were randomized to the intervention (n = 764 pooled) or control (n = 742) conditions, and interviewed by phone at 3 and 6 months post-baseline. Those in the intervention condition received the smoke-free homes intervention as described below. A customized, web-based tool was used to manage aspects of the study, including eligibility screening, interviews, intervention delivery and follow-up and was designed to guide staff through all tasks. The study and procedures were approved by the respective university Institutional Review Boards.

Coaching intervention theoretical frameworks and strategies

The call’s content and messages were based on the modalities of therapy described below as well as the Social Cognitive Theory [34], which posits that environmental change is directly linked with behavior and individual-level factors, in a reciprocally determined relationship, such that each exerts influence on the other. The call had a semi-structured, open-ended format to elicit responses about participants’ stage of change for making their homes smoke-free (i.e. no interest in making home smoke-free, thinking about, decided to or already have a smoke-free home). The participant’s stage determined the call flow, script, and stage-based messages to address barriers, reinforce efforts, and set goals for making and/or keeping a smoke-free home (action plan) (Fig. 1). Specific change strategies included persuasion, role modeling to boost self-efficacy, goal-setting, environmental cues and reinforcement [34]. These coaching strategies were selected to facilitate tailoring to characteristics of the participants, reinforcement and feedback for actions discussed, and motivating behavior change as recommended change strategies for individuals by intervention mapping [35].

Fig. 1.

SFH coaching call flow diagram.

MI is a counseling technique to strengthen personal motivation for behavioral change [23–24]. MI posits that behavior change takes place when a client feels supported and is allowed to develop and evaluate his own ambivalence towards change. Controlled studies have demonstrated the efficacy of MI in behavior change with a large evidence-base for its use in the treatment of substance use such as alcohol [23, 36], heroin [37] and cigarette smoking [38]. MI involves collaborative talking and listening by the counselor. This style was integrated throughout our intervention in the way in which counselors approached the protocol (e.g. using reflective listening, developing goals). In the treatment of behavioral disorders, MI is frequently integrated with the transtheoretical model (TTM) [39]. A tenet of TTM is that behavior change takes place over time in a series of steps in which a person progresses from thinking about a problem to taking action and then maintaining that change [37, 38]. While this change is taking place, processes of change guide the person through these steps. For those less committed to changing at the start of an intervention, MI is used to gauge what level of change an individual was ready for (e.g. pre-contemplation, contemplation) and then, the individual is matched with stage-appropriate strategies [40–42]. In this intervention, MI techniques were used to resolve ambivalence, support self-efficacy and encourage stage-appropriate goal-setting for making the home smoke-free.

Setting goals can help individuals strive towards behavior change and persist, even when difficult [43, 44]. Goal-setting is most effective when accomplished through a collaborative process with both the individual trying to change and the counselor [29]. Once a problem has been identified and a goal has been set, a person often needs to use problem solving skills to achieve that result [29–30]. Problem solving is an appropriate technique when the issue an individual is facing is a structural or interpersonal issue [30]. This technique helps individuals to have the capability of facing and coping with life stressors; it includes being able to anticipate a challenge, identify all possible solutions, and select an appropriate solution to that problem [30]. The integration of MI and problem solving therapies has been shown as an effective, brief intervention for substance use treatment in low resource settings [45].

Coaching intervention description

The coaching call was one of the four intervention components and followed an initial mailing, after which were two more mailings [19]. There were no major changes to the coaching call structure or content from the efficacy to the effectiveness trial. The initial mailing included the ‘Five-Step Guide to a Smoke-Free Home’ the five steps of which guided the order of the coaching script. The Atlanta/Emory research team designed the protocol for the coaches to deliver the 20 min-long coaching call based on formative research. Its primary purpose was to motivate the participant to take the recommended actions to establish and enforce home smoking rules.

Motivational interviewing elements

Coaches used non-judgmental and positive regard during the entire call. Participant responses were validated with simple reflections. Coaching calls began with a discussion of Mailing 1, by asking open-ended questions about the packet. This acted as an invitation for the participant to tell what was most helpful (e.g. information on health effects of secondhand smoke). The coach then assessed the importance and confidence in making or keeping the home smoke-free on a scale from 0 = not at all to 10 = extremely important. After providing a numeric response, coaches then asked participants ‘why a X and not a lower number?’ to elicit change talk. Responses were used to generate motivation and determine stage of change.

TTM stages of change

Based on participants’ responses to the importance and confidence questions, participants’ were placed in an appropriate stage. The stage (e.g. pre-contemplation, contemplation) determined the type of platform for appropriate messaging and tailored tips and goals to facilitate a total ban. Participants who indicated low levels of confidence were provided with more non-directional, MI to elicit reasons for change and enhance self-efficacy to implement this change.

Problem solving and goal setting

The next part of the coaching call focused on 5 steps to making a home smoke-free from Pre-Contemplation (Step 1-No Interest), Contemplation (Thinking about SFH), Preparation (Decided to have SFH), Action (Have SFH) to Maintenance (Step 5-Maintaining SFH). Each participant was prompted to set goals based on her stage of change within the process of making the home smoke-free. Those in pre-contemplation (e.g. thinking about a ban) or contemplation (e.g. deciding to go smoke-free) were prompted to engage in activities designed to elicit ambivalence and get them to start thinking about making a change (e.g. having a family talk). Participants in the preparation stage were encouraged to set a date to make their homes smoke-free. Those in maintenance stages (e.g. made home smoke free) were encouraged to celebrate their success. In some cases, participants raised a concern about a goal (e.g. having a family talk), particularly if they were not the smoker in the household. In these cases, counselors led individuals through problem-solving skills to help them brainstorm ways in which they could accomplish the goal. If a participant was unsure, the coach provided feedback in an MI style (e.g. ‘Would you like to hear what others have done?’)

Health coach training, qualifications, and assessment of fidelity of delivery

The coaches for the Atlanta trial were female and held bachelors-level degrees in anthropology and psychology, respectively. One had served in an intervention capacity in a previous trial; neither had received extensive training in counseling, however. The coaches were based at the Emory School of Public Health to ensure fidelity of implementation and quality control for the later trials. For the effectiveness trials, coaches were female 2-1-1 I&R specialists. The agents’ ages ranged from 30’s to 50’s and education level ranged from associate to master’s degrees. Of note, 2-1-1 I&R specialists were familiar with serving the population of callers as this was part of their job prior to and after the study.

The Atlanta coaches were trained in the intervention components in an intensive, 1-week instruction on basic MI techniques (e.g. use of empathy, reflective statements) and counseling. All calls were audiotaped and reviewed weekly during group supervision. A master’s-level clinician held weekly supervision in a group format, which was used to process any issues from sessions. The supervision meetings also included tape review, with directive feedback on ways to improve adherence to the treatment protocol and/or therapeutic style. At the North Carolina and Texas sites, the Atlanta team prepared I&R specialists as coaches through a 2-day in-person training. Each coach had to complete five consecutive, satisfactory recordings before they were approved for participation in the trial. Thereafter, the tracking tool prompted 10% of all calls to be recorded; recordings were reviewed for protocol adherence and feedback.

Data collection

The principal data for this study were derived from three sources: a baseline survey, 3-month follow-up survey (conducted by research staff) and the coaching call documentation. At baseline (n = 764), participants provided demographic information (e.g. gender, race, income, education) and smoking related information (e.g. current smoker, ban status). During the coaching call (n = 624), coaches followed a scripted protocol and noted responses about importance of and reasons for having a smoke-free home, and challenges experienced in their stage of readiness. In the efficacy trial, coaches documented responses to structured questions about stage and goals and open-ended prompts on a paper-based form; coaches in the effectiveness trials documented responses using an online tracking tool. In the 3-month follow-up survey (n = 564), participants were asked their ban status and the usefulness, relevance and satisfaction with the intervention materials and coaching call (responses were 1 = not at all to 4 = very). Participant characteristics and ratings of the calls from all three sites were analyzed using SAS 9.3 [46].

In addition, at the end of the trial in Texas the project co-ordinator for the SFH trial interviewed the Houston-based 2-1-1 coaches to collect data on their expectations, satisfaction with the training, supervision and support, and their success in balancing usual 2-1-1 I&R specialist duties with their SFH intervention coaching responsibilities.

Results

Participant demographics (all sites)

In the efficacy trial, 92% of the intervention participants completed the coaching call. In the effectiveness trials, 81% completed the call in North Carolina and 72% in Texas. There were no statistically significant differences among those who received the coaching call compared to those who did not, except for the Texas trial. In Texas, men and those employed outside the home were less likely to have received the call (Table I). Across the sites, the sample receiving the call was largely female (85%), earning under $20 000 a year (77%) and African American (73%). Most of those who received the call had finished high school (59%) and were single (53%). Approximately 74% of the pooled sample was smokers, with the rest comprised of non-smokers living in households with smokers. Over 78% had a child under the age of 18 living in the home.

Table I.

Participant and household characteristics by receipt of coaching call across trials (Atlanta, NC and TX)

Site	Across trials					Atlanta					North Carolina					Texas
	Received		Not received			Received		Not received			Received		Not received			Received		Not received
	n	(%)	n	(%)	P-value	n	(%)	n	(%)	P-value	n	(%)	n	(%)	P-value	n	(%)	n	(%)	P-value
Total	621	81.3	143	18.7		225	91.5	21	8.5		211	81.2	49	18.8		185	71.7	73	28.3
Gender
Male	88	14.17	28	19.58		44	19.56	3	14.29		26	12.32	8	16.33		18	9.73	17	23.29
Female	533	85.83	115	80.42	0.10	181	80.44	18	85.71	0.56	185	87.68	41	83.67	0.45	167	90.27	56	76.71	0.004
Race/ethnicity
African American/Black	450	73.05	84	61.76		192	85.33	18	85.71		125	60.39	25	52.08		133	71.89	41	64.38
White	113	18.34	32	23.53		20	8.89	2	9.52		64	30.92	19	39.58	0.76	29	15.68	11	15.07
Other	53	8.60	20	14.71	0.12	12	5.33	1	4.76	0.93	18	8.69	4	8.34	0.25	23	12.43	15	20.55	0.44
Employment
Employed	68	32.23	22	44.90		51	22.67	4	19.05		68	32.23	22	44.90		42	22.70	27	36.99
Unemployed	57	27.01	13	26.53		105	46.67	13	61.90		57	27.01	13	26.53		58	31.35	24	32.88
Homemaker/retired/ disabled/other	86	40.76	14	28.57	0.18	69	30.67	4	19.05	0.38	86	40.76	14	28.57	0.18	85	45.95	22	30.14	0.03
Income
$10 000 or less	300	48.39	68	47.55		120	53.33	14	66.67		92	43.81	18	36.73		88	47.57	36	49.32
$10 001–$20 000	179	28.87	49	34.27		62	27.60	5	23.80		61	29.05	18	26.73		56	30.27	26	35.62
$20 001–$35 000	96	15.48	18	12.58		33	14.70	2	9.50		38	18.09	9	18.36		25	13.52	7	9.59
$35 001–$50 000	23	3.71	4	2.80		5	2.22	0	0.00		9	4.29	2	4.08		9	4.86	2	2.74
$50 001–$75 000	7	1.13	3	2.10	0.93	1	0.44	0	0.00	0.21	1	3.48	1	2.04	0.44	5	2.70	2	2.74	0.57
Education
Less than/some high school	135	21.74	31	21.68		55	24.50	5	23.81		44	20.85	8	16.33		36	19.46	18	24.66
High school graduate/GED	229	36.88	60	41.96		83	36.89	12	57.14		63	29.86	20	40.82		83	44.86	28	38.36
Vocational/technical school/some college	214	34.46	44	30.77		68	30.20	4	19.05		86	40.75	18	28.73		60	32.43	24	30.14
College graduate or higher	43	6.92	8	5.59	0.40	19	8.44	0	0.00	0.23	18	8.53	3	6.12	0.63	6	3.24	5	6.85	0.93
Marital status
Not married, living w/partner	172	27.70	37	25.87		61	27.11	5	23.81		66	31.28	14	28.57		45	24.32	18	24.66
Married	120	19.32	18	12.59		41	18.22	2	9.52		42	19.91	10	20.41		37	20.00	6	8.22
Single	329	52.98	88	61.54	0.099	123	54.67	14	66.67	0.50	103	48.82	25	51.02	0.93	103	55.68	49	67.12	0.06
Smoking status
Smoker	456	73.55	112	78.32		176	78.22	15	71.43		155	73.81	39	79.59		125	67.57	58	79.45
Non-smoker	164	26.45	31	21.68	0.24	49	21.78	6	28.57	0.47	55	26.19	10	20.41	0.40	60	32.43	15	20.55	0.06
Children under 18 in the home
Yes	483	77.78	108	75.52		181	80.44	18	85.71		169.00	80.09	41.00	83.67		133.00	71.89	49.00	67.12
No	138	22.22	35	24.48	0.56	44	19.56	3	14.29	0.56	42.00	19.91	8.00	16.33	0.57	52.00	28.11	24.00	32.88	0.45
	Mean	SD	Mean	SD		Mean	SD	Mean	SD		Mean	SD	Mean	SD		Mean	SD	Mean	SD
Age	40.3	11.40	39.4	12.30	0.39	40.2	11.08	37.8	9.82	0.32	39.98	10.77	39.67	13.38	0.88	40.66	12.49	39.60	12.30	0.54

Feasibility and acceptability of coaching calls (all sites)

For the Atlanta efficacy trial, it took an average of four attempts (M = 4.24, SD = 4.60, range 1–27) to reach participants; the average coaching call lasted approximately 18 min (M = 18: 21, SD = 5: 47, range 7: 00–40: 34). Overall, the mean satisfaction with the calls was 3.76 (SD = 0.57), with a range from 3.67–3.80 across sites, on a scale from 1 to 4 (4 = 'very satisfied’) (Table II). Generally, the mean ratings of relevance were the same across sites (M = 3.24–3.56) and did not vary significantly between smokers and non-smokers. The mean usefulness of the call and information provided by the coach varied from 3.57 to 3.69; there was no difference in reported usefulness of the call across the sites. Smokers and non-smokers’ ratings did not differ on satisfaction, relevance, or usefulness, except for the North Carolina trial that reported a difference in usefulness (P = 0.03).

Table II.

Reactions to coaching call by smoking status and study site (Atlanta, NC, TX)

	Atlanta				NC				TX				Total
	n	Mean	SD	P-value	n	Mean	SD	P-value	n	Mean	SD	P-value	n	Mean	SD	P-value
Relevance, total	147	3.56	0.70		147	3.46	0.77		122	3.24	0.87		416	3.43	0.79
Smoker	108	3.60	0.63		97	0.34	0.75		80	3.33	0.79		285	3.47	0.72
Non-smoker	38	3.45	0.89	0.33	49	3.49	0.82	0.73	42	3.07	1.00	0.13	129	3.34	0.91	0.16
Usefulness, total	147	3.69	0.62		147	3.61	0.70		122	3.57	0.74		416	3.63	0.68
Smoker	108	3.72	0.54		96	3.51	0.74		80	3.56	0.73		284	3.61	0.71
Non-smoker	38	3.58	0.79	0.31	50	3.78	0.58	0.03	42	3.60	0.77	0.82	130	3.61	0.67	0.44
Satisfaction, total	147	3.77	0.55		148	3.24	0.87		122	3.79	0.56		417	3.76	0.57
Smoker	108	3.78	0.54		97	3.67	0.62		80	3.80	0.54		285	3.75	0.57
Non-smoker	38	3.74	0.60	0.70	50	3.80	0.53	0.21	42	3.76	0.62	0.72	130	3.77	0.58	0.72

Rating was from 1 = not at all to 4 = very. The smoking status is at 3 months follow-up.

Readiness to change (Atlanta only)

When asked about the importance of having a smoke-free home, the majority ranked it highly on a 0–10 scale (M = 9.24, SD = 1.47). Most also reported confidence in their ability to make their homes smoke-free on a 0–10 scale (M = 7.83, SD = 2.20). Stage-of-change reported by most participants reflected both the importance and confidence to make their homes smoke-free. Most participants said they were either in the Contemplation (33.3%) or Preparation stage (57.8%). By the time of the coaching call, a small portion had already made their homes smoke-free (16.7%) (Maintenance stage). Of the last group, 18 self-identified as having already made their home smoke-free whereas the remaining 20 participants indicated they had decided to make it smoke-free (Preparation stage), but later revealed that the home was already smoke-free.

Challenges to going smoke-free and goal setting (Atlanta only)

During the coaching call, participants were invited to comment on the challenges to going smoke-free (Table III). For those in the Contemplation stage, the top challenges were that the smoking habit is difficult to break (70.7%), the smoker wants to quit and needs help (56.0%), the smoker does not want to quit right now (38.7%) and the smoker was worried about weather for smoking outside (38.7%) (Table III). For those in the Action stage, the top two challenges were that the smoker wants to quit and needs help (64.6%) and smoking habit is difficult to break (37.7%). In the Maintenance stage, barriers were the smoker wants to quit and needs help (61.1%) and having friends, family or visitors who want to smoke inside (38.9%). The three goals most frequently selected from a predefined list were getting ready to go smoke-free (e.g. pick place outside) (82.2%), pick a date and/or sign the pledge (60.0%) and do a 1-week SFH trial (Table IV). Goals suggested by participants were to go outside to smoke (36.8%), doing another behavior instead of smoking (12.4%) and smoke less (12.4%).

Table III.

Challenges to going smoke-free by participant stage for the efficacy trial (Atlanta)

	n	%
No interest in making your home smoke-free (pre-contemplation) (n = 2)
You or the another smoker does not want to quit smoking now	1	50.0
You are not head of the household and you can’t make the rules for your home	1	50.0
You or the smoker think it’s okay to smoke in a separate room with the door closed	2	100.0
Smoking inside is a habit that’s hard to break	1	50.0
Thinking about making your home smoke-free (contemplation) (n = 75)
Smoking inside is a habit that’s hard to break	53	70.7
The smoker wants to quit smoking an needs help	42	56.0
You or the another smoker does not want to quit smoking now	29	38.7
You are worried it will be hard to smoke outside in bad weather or at night	29	38.7
You or the smoker think it’s okay to smoke in a separate room with the door closed	27	36.0
You or the smoker are smoking more because you are stressed or worried	20	26.7
You have friends, family, or visitors who want to smoke inside	14	18.7
You are not head of the household and you can’t make the rules for your home	12	16.0
You or the smoker are not ready to make your home smoke-free right now (e.g. no motivation, resistant smoker, busy/no time, etc.)	9	12.0
Smoking outside is not convenient or comfortable (e.g. bugs, physical limitations, safety, etc.)	9	12.0
A smoker in the home doesn’t like the rule and sometimes ignores it	7	9.3
You live in an apartment and there is no porch or yard to use as a smoking area	2	2.7
You can’t go outside to smoke because you have young children	1	1.3
A smoker ‘forgets’ the rule	1	1.3
other	5	6.7
Decided to make your home smoke-free (preparation) (n = 130)
The smoker wants to quit smoking an needs help	84	64.6
Smoking inside is a habit that’s hard to break	61	47.0
You or the smoker think it’s okay to smoke in a separate room with the door closed	49	37.7
You are worried it will be hard to smoke outside in bad weather or at night	32	24.6
You or the smoker are smoking more because you are stressed or worried	30	23.1
You have friends, family, or visitors who want to smoke inside	22	17.0
A smoker in the home doesn’t like the rule and sometimes ignores it	17	13.1
You are not head of the household and you can’t make the rules for your home	13	10.0
You or the smoker are not ready to make your home smoke-free right now (e.g. no motivation, resistant smoker, busy/no time, etc.)	14	10.8
Smoking outside is not convenient or comfortable (e.g. bugs, physical limitations, safety, etc.)	8	6.2
A smoker ‘forgets’ the rule	7	5.4
You or the smoker are having trouble dealing with your cravings to smoke	3	2.3
You can’t go outside to smoke because you have young children	2	1.5
You live in an apartment and there is no porch or yard to use as a smoking area	1	0.8
Other	5	3.8
Already have a home smoke-free (Action) (n = 18)
The smoker wants to quit smoking an needs help	11	61.1
You have friends, family, or visitors who want to smoke inside	7	38.9
Smoking inside is a habit that’s hard to break	5	27.8
You are worried it will be hard to smoke outside in bad weather or at night	5	27.8
You or the another smoker does not want to quit smoking now	2	11.1
You or the smoker are smoking more because you are stressed or worried	2	11.1
You or the smoker think it’s okay to smoke in a separate room with the door closed	1	5.6
You or the smoker are having trouble dealing with your cravings to smoke	1	5.6

Table IV.

Goals setting from coaching call for efficacy trial (Atlanta)

	n	%
Goals (n = 225)
Get ready for the change by setting up a place outside, putting up the sign, and taking away the ashtrays	185	82.2
Pick a date to go smoke-free and/or sign the pledge (make it official)	135	60.0
1-week SFH trial	96	42.7
Call for tips and information on quitting smoking	71	31.6
Talk to a member of your household about going smoke-free	59	26.2
Post your pledge	56	24.9
Celebrate your success in making your home smoke-free by planning a special dinner, playing a game, watching a favorite show together or going to a movie	33	14.7
Tell friends and visitors that your home is smoke-free	29	12.9
Share guide and print materials with other household members	30	13.3
Remind/encourage smoker to go outside politely	25	11.1
Set an example and go outside to smoke	12	5.3
Suggested new goals (n = 225)
Go outside to smoke/make home smoke-free	83	36.9
Try replacing smoking with another behavior/staying busy	28	12.4
Smoke less/buy fewer cigarettes	28	12.4
Attempt to quit or quit smoking altogether	19	8.4
Find ways to spend more time outside	3	1.3

Coaches’ perceptions of training, supervision, and support (NC and TX sites)

To assess thoughts about coaching delivery and feedback about coaching delivery, we interviewed coaches in NC and TX. In Houston, the 2-1-1 manager recommended the I&R specialists who were trained as coaches. After the training, one coach declined the role and decided to participate as a recruiter. Two remaining specialists liked the training although they mentioned that it was very intensive. Suggestions for improvement included: (i) fewer attempts to reach participants, (ii) telling future coaches not to become discouraged with participants who may not be willing to change immediately and (iii) freedom to make the calls at all times of the day. They also said the participants who really wanted to make their homes smoke-free made them feel ‘it was all worth it’. In the NC trial, UNC strongly suggested that the call center separate the roles of recruiters and coaches among different staff members, but call center management requested shared roles. After the trial, the coaches reported that it was challenging to maintain the dual roles. Overall, both call center staff and management reported that participating in the intervention improved their overall skills as I&R specialists, and that they professionally improved based on their participation in the intervention.

Discussion

This article presents the coaching call protocol used in the Smoke-Free Homes: Some Things are Better Outside program and related process evaluation results. Similar to many coaching interventions in health promotion, aspects of MI were used in combination with TTM and other therapeutic modalities [38–40]. Results showed that 2 weeks into the intervention and after receipt of the first mailing, the majority were already in the preparation stage. Simply consenting to participate in a Smoke-Free Homes intervention may have shifted some to preparation stage. This finding is consistent with research that shows committing to a change and/or behavioral contracting alone can elicit behavior change [47]. Alternatively, the first intervention mailing (including a contract) may have been effective in shifting participants along the stage of change continuum. We did not assess stage of change for creating a smoke-free home at baseline so we are not able to disentangle when in the change process participants increased their readiness to create a smoke-free home.

The coaches elicited barriers to creating a smoke-free home after assessing stage of change. The most common barrier reported across stages was the difficulty of smoking cessation. Participants reporting interest in wanting to quit were then offered Quitline information. Coaches typically gave participants the Quitline number and explained that participants were not being asked to quit as part of this intervention, rather simply to smoke outside. Among the very few in pre-contemplation, a belief that smoking in a room with the door closed provided adequate protection was a barrier to change. Environmental monitoring studies have shown, however, that ventilation systems are unable able to offer protection from SHS [4] and that smoke from cigarettes can linger for hours [48]. Not being head of the household and therefore not having authority to establish rules was also mentioned. This barrier has been identified in other studies, including our own, and coaches were ready to offer a solution that usually included protection of children, pets or others in the home [49, 50]. In some cases, role play of the conversation was also a strategy that was used to increase participant self-efficacy in having the conversation with the head of household.

The habitual nature of smoking in the home was a prominent barrier across stages and has also been highlighted in the literature [49–53]. Other major barriers were consistent with those identified in other studies, including friends, family members and guests who smoke, concerns about the weather and going outside at night [50, 52–56].

We also examined goals set through the coaching call. The most common was environmental: removing ashtrays, putting up signs and setting up a place to smoke outdoors. Altering environments and related stimuli is a longstanding behavior change strategy [57–60]. A 1-week trial was also common, an approach that could be used to build self-efficacy for a longer-term change by providing a sense of mastery [61]. Picking a date was also a common goal, borrowed from smoking cessation interventions [62].

Reach varied across the sites, with fewer calls completed in NC compared to Atlanta, and dropping to 72% completed in Texas. Although the tracking tool data could not easily explain the differences, it appears that the Atlanta calls were made in evenings and on weekends as well as daytime hours, whereas in Texas, coaches were restricted to set daytime hours as the result of organizational structure and 2-1-1 management decisions. This scheduling difference is consistent with reaching many fewer men and fewer employed participants in Texas. Also in Texas, 30% had telephone numbers not in service only 2 weeks after the baseline interview. Although some of these numbers would be reinstated later, this was outside the window of allowable time for the coaching call to ensure fidelity of implementation. Reactions to the coaching call were positive from both smokers and non-smokers, which suggests no changes are needed to the protocol to better address the needs of smokers, unless cessation issues are addressed more directly. Across both smokers and non-smokers, the coaching call was viewed as relevant and useful, with high levels of satisfaction.

This study has a few limitations. The stages of change of the participant were assessed only at the coaching call and follow-up; therefore, we cannot assess change shift except for the creation of a smoking ban at the follow-up. In addition, frequency of behaviors related to making a smoke-free home were based on self-report and not observed. However, in the Atlanta trial, we did have nicotine monitors placed at a proportion of homes to validate a smoke-free home. This process evaluation provides support for the feasibility, reach and satisfaction with the health coaching process. The assessment of the coaches’ facilitation of the goal setting and problem solving process was limited to the data provided by the Atlanta trial; however, Atlanta research assistants who served as independent evaluators, not coaches, delivered the call. We did not examine Texas or North Carolina 2-1-1 health coaches’ facilitation of these processes to confirm these findings, although impact of the intervention on smoke-free homes was similar across the sites. Another limitation was that we evaluated delivery of the intervention in a context in which university-based project co-ordinators provided oversight of implementation using structured project quality control protocols to ensure that coaches followed the detailed protocols and scripts for intervention delivery. Findings based on ongoing dissemination studies will provide insight into factors influencing implementation feasibility, reach, satisfaction and fidelity in a more ‘real world’ context, where 2-1-1 managers provide oversight for all SFH activities.

There are several areas of future research. Further examination of the differences reported by smokers and non-smokers regarding program usefulness may inform future adaptations to tailor messages for these different groups of participants based on their own unique challenges. We also could explore the steps to creating a smoke-free home in more depth using qualitative methods. We believe the family talk is a critical component of the intervention. It was a commonly reported step in creating a SFH and it would be helpful to know more about the conversation and its context [i.e. how it was brought up and to whom, the activities at the time (e.g. dinner), and what points were used]. Prior studies have examined facilitators and barriers to smoke-free homes, but have typically not focused on the change process [49, 50]. The prominence of cessation concerns is also noteworthy. Future studies could compare associated outcomes when cessation is addressed more directly as in some similar studies [6] in comparison to our approach which steered participants to a focus on reducing SHS exposure by smoking outside. Finally, conducting moderator analyses of the impact of the coaching call on household ban status across the effectiveness trials could provide evidence of its impact as an intervention component.

This research demonstrates the feasibility of the coaching component of a Smokefree Homes intervention in terms of its reach and acceptability, to augment the effectiveness results for the brief intervention for encouraging families to make their homes smoke-free to enhance scalability [63]. To our knowledge, this is the first study to test an easily disseminable intervention to reduce secondhand smoke in low-income, urban households in three different states. Our use of already existing infrastructure in United Way 2-1-1 I&R systems as well as demonstrated efficacy, feasibility and acceptability, indicate that this intervention has the potential to be adopted by other 2-1-1 systems, increasing the impact and reach of the current study.