Long-term Effects of a Peer-led Asthma Self-Management Program on Asthma Outcomes in Adolescent Peer Leaders

Abstract

Objective:

To examine the long-term effects of a peer-led asthma self-management program on urban adolescent peer leaders with asthma.

Methods:

This longitudinal study includes 51 adolescents (16–20 years) enrolled in an asthma self-management program implemented at a one-day camp as peer leaders. Study outcomes, including quality of life, asthma control, asthma knowledge, and attitudes toward asthma were collected for 15 months post-intervention. Mixed-effects models were conducted to estimate time effects, and effect sizes were calculated for each model.

Results:

Of 51 enrolled, 41 completed the training, of which 35 successfully participated in the camp program. A total of 17 peer leaders withdrew between enrollment and 15-months follow-up. Quality of life, asthma control, and knowledge significantly improved after peer leader training and remained elevated for 15 months, while significant improvement in attitudes emerged immediately after camp, in which they served as leaders, and sustained for 15 months.

Conclusion:

This study demonstrates the long-term positive effects of a peer-led program on a wide range of asthma outcomes in urban adolescent peer leaders.

Practice Implications:

A peer-led approach to asthma education providing peer leaders with intense training and leadership experience can be effective and sustainable in improving asthma outcomes among urban adolescents.

1. INTRODUCTION

Asthma is currently the leading pediatric chronic respiratory condition in the U.S., affecting approximately 5.5 million children in 2018, of which 2.4 million were youth between 12 and 17 years of age [1]. Of these, an estimated 60% had chronically uncontrolled symptoms [2]. While morbidity and mortality in adolescents is concerning in general, urban youth bear a disproportionate burden of disease and often have poorer outcomes and quality of life [3,4]. The increased burden of asthma in this population is associated with multiple environmental risk factors such as greater exposure to indoor and outdoor triggers, poverty, and chronic psychosocial stressors [5–7], as well as personal factors, such as suboptimal self-management [8–11] and poor knowledge about symptoms, triggers and medications [9–11]. This points to the need for an asthma self-management program specifically designed to ameliorate the burden of asthma among urban adolescents.

Program leaders play critical roles in the success of programs targeting adolescents. Specifically, program effects can be enhanced when delivered by program leaders who share similar characteristics with participants (i.e. peers) [12–14]. A peer-led program can maximize positive effects by capitalizing on interpersonal similarities that increase participants’ receptiveness to information [14,15], and facilitating vicarious learning [14]. Several school-based studies have utilized peer-led approaches in delivering asthma self-management programs and reported positive impact on participants’ asthma-related knowledge, asthma control, or quality of life [16–18]. Evidence suggests that peer-led health programs are beneficial to peer leaders as well, through a chance to develop and practice interpersonal and communication skills [19,20], and improve knowledge of disease management, and attitudes toward the health condition [12,19–27]. Nonetheless, little is known about the extent to which peer-led asthma programs targeting adolescents are beneficial to peer leaders.

Our earlier study evaluating the effects of a peer-led asthma self-management (PLASMA) program demonstrated beneficial effects on peer leaders’ asthma control and quality of life [28]. However, these findings were based on the small number of peer leaders (n=12) who were predominantly female, white suburban adolescents. The study was unable to determine whether the beneficial effects on peer leaders were due to intense training or leading educational sessions. Further investigation was needed to determine (1) whether the positive effects could be reproduced in a larger group of peer leaders who are primarily urban minority adolescents, (2) if the positive effects would sustain beyond 9 months, and (3) which program components are responsible for the positive changes.

Recently, we replicated the PLASMA program in a large multi-site study conducted in three urban communities where participants were followed for 15 months post-intervention. The purpose of this study is to examine the long-term effects of a peer-led program on peer leaders from urban communities. We hypothesize that urban adolescents who participated in an asthma education program as peer leaders would improve their asthma control, quality of life, and asthma-related knowledge and attitudes over the course of 15 months after the program implementation.

2. MATERIAL AND METHODS

2.1. Study Design Overview

This study reports the analyses of longitudinal data collected from 51 adolescents with asthma who served as peer leaders in a randomized controlled trial (RCT) including 320 participants. The purpose of the RCT was to assess the effects of the PLASMA intervention that was implemented at camp settings in three urban sites including Buffalo NY, Baltimore MD, and Memphis TN between 2015 and 2018. We have already published details about the intervention protocol including peer leader training and program content [29], intervention format at camp [30] and intervention fidelity [31]. Peer leaders completed data at baseline, after training, at camp, and every 3 months thereafter for 15 months.

2.2. Study Sample

Peer leaders were recruited from clinics, schools, and the community through clinician and school referrals and word of mouth. Peer leader eligibility criteria included ages 16–20 years, urban dwelling, English proficiency, having had physician diagnosed asthma for at least 1 year, being on a controller medication, and either preventive or acute healthcare service use for asthma care within 12 months prior to recruitment. The age range was chosen for peer leaders to ensure cognitive maturity to process the program content and communicate the information to others. Those 19–20 years old were considered adolescents because they were within the age range defining adolescence by Sawyer et al. [32]. To further ensure the quality as a leader, a written adult nomination supporting the candidate’s strengths in asthma self-care (medication adherence), leadership skills, and/or emotional intelligence was required to be considered. Those with other chronic conditions requiring daily medication or mental health conditions were excluded due to the potential confounding effects on self-management routines and quality of life. We also excluded teens with behavioral issues reported by parents to ensure peer leaders’ qualification as role models and for the safety of program participants.

2.3. Peer Leader Study Activities

Peer leaders attended an intense training program consisting of three 4-hour sessions, covering basics of asthma, symptom/trigger monitoring, symptom prevention/ management and asthma-related psychosocial/communication issues. The program module for each session comprised content review, case study, learning strategies, and role play. The trainers evaluated and provided feedback on peer leader’s mastery of content and leading skills, and additional training was offered if necessary. Details of the training content and format are provided elsewhere [29]. Trained peer leaders were paired and led the intervention program for a small group of 4–6 adolescents of ages 12–17 years in the camp setting. Peer leaders made brief follow-up contacts (~10 minutes) with their group members every 2 months after camp for 12 months via phone calls or texting, during which any asthma-related issues or questions were discussed and good self-management was encouraged. We described details about the bimonthly contacts in another publication [31]. Peer leaders attended a 4-hour booster session at 6 months to refresh education content and their commitment to bimonthly contacts. Most peer leaders completed a booster group training session.

2.4. Data Collection and Measures

The study protocol was reviewed and approved by four academic institutional review boards at the study coordinating center and three study sites. Prior to data collection, written informed assent and consent were obtained from younger peer leaders (16–17 years old) and their parents or guardians. Older peer leaders (18–20 years old) provided written consent for themselves. Sociodemographic data including age, gender, race, health insurance type, and age at asthma diagnosis were obtained at enrollment. Study data were collected at 8 time points including enrollment, following training, at camp, and 3-, 6-, 9-, 12-, and 15-months post-intervention.

Quality of Life:

The Pediatric Asthma Quality of Life Questionnaire (PAQLQ) consists of 3 subscales that measure symptoms (10 items; e.g. ‘how bothered have you been during the last week by wheezing?’), activity limitations (5 items; e.g. ‘activities with friends and family’), and emotional functioning (8 items; e.g., ‘feel frustrated because of your asthma’) in the past week. Each item was measured on a 7-point scale [33]. Mean scores were computed for total and individual subscales, with higher scores indicating better quality of life. Cronbach alpha ranged from .96 to .98 across all time points in this study. Subscale reliability was mostly above α=.90.

Asthma Control:

The Asthma Control Questionnaire (ACQ) contains 6 items measured on a 7-point scale regarding youth’s activity, asthma symptoms, and controller medication use in the past week [34,35]. A sample question includes ‘In general, during the past week, how much of the time did you wheeze?’ Mean scores were computed, with higher scores indicating worse asthma control. Cronbach alpha in this study ranged from .85 to .91 across time points.

Asthma Knowledge:

The Asthma Knowledge Questionnaire (30 items) [28] was used to measure knowledge related to asthma and its management. Items address multiple areas of asthma knowledge including symptoms, triggers, symptom prevention and management, with two response options, True or False. A sample question includes ‘Is it true that an asthma episode can be caused by stress?’ Knowledge was assessed at baseline, immediately following the intervention, and 6-, 12-, and 15- months post intervention. A total score was the sum of correctly responded items, with higher scores representing greater asthma knowledge. Kuder-Richardson 20 reliability of this scale ranged from .49 to .58 across time points.

Attitudes Toward Asthma:

The scale includes 13 items assessing adolescents’ feelings about having asthma and their perceptions of the degree to which asthma impacts their life [36]. Each item is measured on a 5-point scale. Sample question includes ‘How fair is it that you have asthma?’ Mean scores were computed with higher scores indicating more positive attitudes. Cronbach alpha ranged from .86 and .90 across time points.

2.5. Data Analysis

Of various sociodemographic and asthma-related factors, only sex, season, and site were significantly associated with at least one of the outcome measures, thus adjusted for as covariates in the final models. We fit linear mixed-effects models for each continuous outcome separately to estimate the time effects. In each of these mixed-effects models, significance was calculated by the chi-square statistic for the change in explained deviance.

For each model, partial eta squared (η²) was reported as the effect size for the overall change over time. Partial η² is the proportion of variance explained by a given variable out of the variance remaining after excluding that explained by other predictors. Partial η²>0.14 indicates large effects, while η²=0.06 to 0.14 are medium and <0.06 small effects [37]. Cohen’s d was calculated between enrollment and the subsequent time points. For all models, estimates and 95% confidence intervals for the time point effect were calculated along with p-values for the strength of the time effect. Residual analysis was examined to look for outliers that were investigated, but none were removed. No substantial departures from the model assumptions were found. We conducted intention to treat (ITT) analyses including all enrollment peer leaders and compared the results with those of sensitivity analyses that included only those who completed training. Because we found no differences in the significant findings, this paper reports results from the ITT analyses.

3. RESULTS

3.1. Sample Characteristics

Fig. 1 shows the detailed flow of peer leaders from screening to 15-months post-intervention. A total of 51 adolescents from Buffalo NY (n=23), Baltimore MD (n=14), and Memphis TN (n=14) were enrolled as peer leaders. Of those, 11 were withdrawn by the investigators before camp due to their insufficient mastery of program content and leadership skills. Six peer leaders dropped out at various points of assessment after camp primarily due to time conflicts. There were no significant differences between those who withdrew and remained regarding sociodemographic characteristics or study variables. All 51 participants were included in the ITT analyses.

Fig.1.

Flow of study participants from screening to analysis.

Recruitment from schools, teacher referrals or school events, yielded over 35% of the group, followed by word of mouth (23.5%) and clinician referrals (21.6%). A small number of peer leaders were recruited through previous study registry (10%), community events (6%) , and study flyers (4%). Nomination was obtained primarily from teachers or guidance counselors (n=37), followed by healthcare providers (n=8), community leaders (n=4), or employers (n=2). Table 1 summarizes the demographic and asthma characteristics of peer leaders at enrollment. Buffalo was the only site with white peer leaders (48%). The mean age of peer leaders in Memphis was slightly younger than the other two sites. Otherwise, sites were comparable in demographic and study variables.

Table 1:

Comparisons of peer leader characteristics and study variables at baseline by site (N=51)

	Total (N=51)	Buffalo (n= 23)	Baltimore (n= 14)	Memphis (n= 14)	p
Age, Mean (SD), Range	17.22 (1.27), 16–20	17.48 (1.34), 16–20	17.43 (1.39), 16–20	16.57 (0.76), 16–18	.045
Sex, % (n) male	35.3 (18)	34.8 (8)	28.6 (4)	42.9 (6)	.686
Race, % (n) white	21.6 (11)	47.8 (11)	0 (0)	0 (0)	<.001
Health insurance, %(n) public insurance	54.9 (28)	56.5 (13)	71.4 (10)	35.7 (5)	.086
Age at diagnosis, Mean (SD), Range	4.33 (4.04), 0–14	3.83 (4.45), 0–14	5.64 (3.87), 1–12	3.86 (3.44), 0–10	.838
Family history of asthma, % (n) yes	60.8 (31)	52.2 (12)	50.0 (7)	85.7 (12)	.062
Quality of life, Mean (SD), Range
Total	5.73 (1.24), 1.96–7	5.89 (1.26), 1.96–7	5.67 (1.16), 3.57–7	5.52 (1.32), 3.22–6.96	.377
Symptoms	5.56 (1.36), 1.8–7	5.73 (1.33), 1.8–7	5.47 (1.44), 2–7	5.39 (1.41), 3–7	.458
Emotional function	5.99 (1.23), 2–7	6.15 (1.27), 2–7	6.01 (0.98), 4.75–7	5.71 (1.43), 2.38–7	.305
Activity limitation	5.64 (1.22), 2.2–7	5.8 (1.25), 2.2–7	5.53 (1.24), 2.8–7	5.49 (1.2), 3.8–7	.426
Asthma control, Mean (SD), Range	1.06 (1.05), 0–3.83	0.87 (0.97), 0–3.83	1.06 (1.04), 0–3.33	1.36 (1.17), 0–3.33	.175
Asthma knowledge, Mean (SD), Range	23.67 (3.43), 14–29	23.48 (4.04), 14–29	23.50 (2.95), 18–27	24.14 (2.91), 19–28	.595
Attitudes toward asthma, Mean (SD), Range	47.74 (8.99), 25–65	48.36 (9.12), 25–65	48.93 (9.50), 34–65	45.57 (8.55), 29–61	.411

The following describes the results of longitudinal analyses on various outcomes. Table A.1 displays the means and standard deviations of outcome measures for each time point.

3.2. Effects on Quality of Life (QOL) and Asthma Control

Table 2 summarizes the results of longitudinal effects on outcome measures in comparison to baseline. QOL significantly increased from baseline to post-training and further improved over time. Asthma control scores were significantly lower after training for 15 months, indicating better controlled asthma over time. Fig.2 illustrates the longitudinal trends of predicted scores for three subscales of QOL and asthma control. The partial η² values indicate 16.5%−24.5% of the between-subjects variance in QOL is accounted for by the change over time. Cohen’s ds compared to enrollment were between 0.33 and 0.95 (Table A.2), indicating moderate to large improvements in QOL at all time points after training. The largest d between sequential times was between baseline and post-training. The partial η² for asthma control was also significant at 0.12 suggesting clinically improved asthma control over time.

Table 2:

Longitudinal effects of the intervention on outcome measures in peer leaders in comparison to baseline. These models are all adjusted for teen sex, camp season, and site.

Variables	Partial Eta^2		Overall time effect p-value		Post-training (n=37) B (95%CI)	Post-camp (n=35) B (95%CI)		3 mo, (n=27) B (95%CI)		6 mo,(n=28) B (95%CI)		9 mo, (n=29) B (95%CI)		12 mo, (n=27) B (95%CI)			15 mo, (n=27) B (95%CI)
Quality of life
Total	0.213	<0.001		0.592 (0.31, 0.88)			0.609 (0.31, 0.90)		0.761 (0.45, 1.07)		0.864 (0.55, 1.18)		0.863 (0.55, 1.17)		0.969 (0.65, 1.28)	0.916 (0.59, 1.23)
Symptoms	0.194	<0.001		0.715 (0.38, 1.04)			0.723 (0.38, 1.06)		0.812 (0.45, 1.17)		0.933 (0.57, 1.30)		0.853 (0.49, 1.21)		1.058 (0.69, 1.42)	1.063 (0.69, 1.43)
Activity limitation	0.245	<0.001		0.532 (0.26, 0.80)			0.709 (0.43, 0.99)		0.816 (0.52, 1.11)		0.929 (0.63, 1.23)		0.927 (0.63, 1.22)		0.995 (0.69, 1.29)	0.825 (0.521, 1.13)
Emotional Function	0.165	<0.001		0.477 (0.18, 0.78)			0.404 (0.09, 0.71)		0.657 (0.33, 0.98)		0.729 (0.40, 1.06)		0.826 (0.50, 1.15)		0.833 (0.50, 1.16)	0.778 (0.44, 1.11)
Asthma Control	0.119	<0.001		−0.464 (−0.75, −0.17)			−0.411 (−0.71, 0.11)		−0.525 (−0.85, −0.20)		−0.592 (−0.91, −0.28)		−0.539 (−0.85, −0.23)		−0.694 (−1.01, −0.37)	−0.694 (−1.02, −0.37)
Asthma know ledge	0.147	<0.001		2.684 (1.60, 3.77)			1.804 (0.68, 2.93)		--		2.134 (0.93, 3.33)		--		1.851 (0.64, 3.06)	2.152 (0.94, 3.37)
Attitudes toward asthma	0.151	<0.001		0.131 (−0.03, 0.29)			0.298 (0.13, 0.46)		--		0.339 (0.16, 0.51)		--		0.359 (0.18, 0.54)	0.301 (0.12, 0.48)

Fig. 2.

Asthma related quality of life (three scales) and asthma control for each of the time points. Each * indicates a time point significantly different from Enrollment. These models are adjusted for teen sex, study site and camp season. Enr Enrollment; Trn peer leader training; Cmp Camp intervention.

3.3. Effects on Attitudes toward Asthma and Asthma Knowledge

Attitudes significantly improved post-camp and remained high for 15 months post-intervention (Table 2). A significant increase in knowledge was found post-training and remained high over time. Long-term trends of the predicted scores for these measures are illustrated in Fig.3. The partial η² values indicate 14.7% and 15.1% of the between subjects variance in asthma knowledge and attitudes, respectively, is accounted for by the change over time. Cohen’s ds compared to enrollment were between 0.25 and 0.83, suggesting moderate to large improvements in attitudes and knowledge at all of the post-training time points (Table A.2). The largest d between sequential times was between baseline and post-training for knowledge and attitudes.

Fig. 3.

Asthma knowledge and attitude toward asthma for each of the time points measured. Each * indicates a time point significantly different from Enrollment. These models are adjusted for teen sex, study site and camp season. Enr Enrollment; Trn peer leader Training; Cmp Camp intervention

4. DISCUSSION AND CONCLUSION

4.1. Discussion

This study examined the long-term effects of a peer-led asthma self-management program (PLASMA) on urban adolescent peer leaders who were predominantly African American. In support of the hypothesis, we demonstrated sustained positive effects of the program on asthma outcomes among peer leaders over the course of 15 months following intervention. Importantly, this is the first study reporting the benefits of a peer-led asthma program lasting longer than a year for peer leaders themselves. Prior studies evaluating peer-led asthma education [16,17,38] have not assessed peer-leader asthma outcomes as peer-leaders were not required to have asthma. In contrast, our studies employed only peer leaders with asthma to maximize similarities to learners, which afforded a unique opportunity to investigate the effects of the intervention on their asthma outcomes.

Although the present study is a larger-scale replication of our earlier study conducted over a decade ago [28], this study is unique in two major ways. First, the sample size and demographic characteristics of peer leaders are different. The current study involved a larger number of peer leaders, predominantly African American, including increased male representation, and selected from three urban locations geographically remote from each other, whereas the earlier study involved only 12, mostly white, females. Despite the differences, similar positive program effects between the two studies provide compelling evidence supporting the generalizability of the program effects on peer leaders regardless of geographic locations, race, or gender. Second, this study used a longer follow-up period, 15 months vs. 9 months. Similar to our earlier study, up to 9-months of sustained positive effects on peer leaders’ health and behavior has also been reported in a study of a peer-led program for HIV prevention [24]. The extended observation period in this study provides compelling evidence of the long-term health benefits of a peer-led program for peer leaders with asthma. Consequently, the current study offers a unique opportunity to validate, expand, and complement our earlier findings of significant health benefit accruing to peer leaders in the PLASMA program.

As in our earlier study [28], we demonstrated positive changes over time in asthma control and psychosocial outcomes including quality of life, knowledge, and attitudes toward asthma among our peer leaders. Particularly, the sustained improvements in asthma control are clinically important and worth noting, as control typically declines with age in older adolescents [39,40]. In general, peer-led programs’ positive impact on adolescent peer leaders’ psychosocial well-being and targeted health-related knowledge and behavior have been well documented for shorter periods [22–27,41]. The medium to large effect sizes reported in this study were relatively stable and present over a longer period of time, which has not been previously reported. Notably, while positive changes in most outcomes emerged immediately following peer-leader training, full and sustained improvement in attitudes emerged after the camp. This suggests that positive attitudes may develop as peer leaders actively exercise leadership roles [23] and communicate learned information to their peers. In addition, peer leaders’ pre-existing aptitudes may have enhanced the program effects. To maximize the effects of a peer-led program, peer leader selection is essential. Our peer leaders were nominated by adults who vouched for their nominees’ qualification as a leader and/or role model in asthma management. Such underlying positive aptitudes might have made them more receptive of learned information, ultimately contributing to the positive asthma outcomes. Further research is warranted to assess the degree to which the program effects are attributable to adolescents’ aptitudes.

Our previous study [28] could not determine whether the positive effects were attributable to the intense peer-leader training or experience as a leader at camp, because of the lack of data collected after training. The current study elucidated peer-leader training as a decisive condition responsible for the first appearance of positive changes in almost all outcome measures, which endured for 15-months. Likewise, others studies attributed any positive changes in peer leaders’ health behaviors and outcomes to training that peer leader received [24,25]. Intense training for peer leaders can afford several inherent benefits including: exposure to educational content [25], opportunity to practice activities that are meaningful to them [24], and a chance to internalize gained knowledge through repetition and systematic evaluation and feedback provided during the training. Nonetheless, it is unclear whether the training is the sole cause of the positive effects sustained for 15-months. Our peer leaders actively engaged in multiple components of the peer-led program including completing training sessions, leading asthma education sessions, and maintaining periodic interactions with campers for 12 months. Given the multifaceted nature of the program the peer leaders had been exposed to throughout study participation, it is impossible to know with certainty which element(s) of the program were responsible for the sustained effects over 15 months. Likely every component of the program played a role in the observed long-term effects. The training may have provided peer leaders with asthma-related knowledge and skills necessary for positive changes, while performing as a camp leader offered them rewarding experience. The leadership experiences may have contributed to improved attitudes, which could have further enhanced and sustained positive effects [42–44] that emerged after the training. In addition, ongoing interaction with peers during the follow-up period may have been a constant reminder for them to adhere to optimum asthma routines. Therefore, it could be the program in its entirety rather than its separate parts that may account for the long-term positive changes in asthma outcomes.

The long-term sustainability of positive effects on peer leaders might have been due in part to a booster session we offered at 6-months after camp. A brief booster session was added to this study to prevent positive effects from waning over time, which we observed occurring at 6- or 9-months after camp in our previous study [28]. We found no substantial changes in the patterns of outcomes, which suggests the booster as a maintainer of intervention effects over time, rather than as an extra treatment with added effects. A peer-leader booster session in a longitudinal study can be offered to refresh education content and help renew excitement for and commitment to the project [28]. This is also an opportune time for peer leaders to connect with an adult supervisor and fellow peer leaders in a structured setting. During the session, investigators can provide peer leaders with timely and needed support, and identify and address challenges facing peer leaders [25,45]. Interactions with other peer leaders during the booster session can also facilitate peer support and promote a sense of comradery, leading to higher retention. Given the booster is a sustainer of beneficial effects for peer leaders, yet is simple to administer, this strategy should be considered as part of a peer-led health program implemented for an extended period.

There are several limitations warranting caution. First, in this multisite project, peer leaders were managed by local research staff. Although peer leaders were trained and led the education program using a standard curriculum with reasonable fidelity [31], the levels of support and supervision they received from local staff throughout the study varied, which necessitated controlling for the sites as a covariate in the analysis. Second, this study has a relatively high peer leader withdrawal (33%, n=17). The majority were withdrawn by the investigators due to their unsatisfactory competence in program content and leadership skills, while some dropped out primarily due to time conflicts. Challenges to retaining adolescent peer leaders in health programs have been recognized due to other competing priorities or waning interest and commitment over time [24,46]. Despite the use of intention to treat analysis and there being no significant differences between dropouts and the remaining peer leaders in most demographics and asthma-related factors, the high withdrawal may have affected the long-term trajectories of asthma outcomes. Third, effect sizes over time for some outcomes were somewhat modest, raising a question about the extent of their clinical relevance. Fourth, this study lacks a control group against which the long-term effects of the program on peer leaders could be compared and verified. To affirm the program’s positive effects on asthma outcomes in peer leaders, an RCT including a well-designed control group is essential. Fifth, although improved self-management due to the program is suggested to contribute to the improvement in self-reported asthma control, self-management actions were not directly measured in this study. Future research is needed to assess the causal links between self-management behavior measured objectively (e.g. digitally tracked medication adherence) and asthma control. Finally, generalizability is limited given the purposeful sampling (to ensure the qualification as peer leaders), the exclusion of those with co-morbid health/mental conditions or behavioral issues, small sample size, and the sample being primarily urban dwelling African American adolescents. However, the comparable findings between this study and the earlier study including predominantly white peer leaders greatly strengthen the overall external validity of the findings.

4.2. Conclusions

This study demonstrates the long-term positive effects of a peer-led asthma management program on a wide range of asthma outcomes, including quality of life, asthma control, asthma-related knowledge and attitudes in adolescent peer leaders. Although positive changes in asthma outcomes emerged immediately following peer leader training, the extent to which training was responsible for the sustained effects for 15 months is unclear because of peer leaders’ subsequent exposure to other components of the program including the camp program and bimonthly contacts. Nonetheless, an asthma management program adopting a peer leadership approach can be a viable and effective means of improving asthma outcomes in peer leaders for an extended period.

4.3. Practice Implications

A peer-led approach to asthma education for adolescents can be effective and sustainable for improving asthma outcomes for adolescents. This study used a multi-faceted program that allows adolescents to participate not only as recipients of delivered information but as sharers of the information while engaging in ongoing peer interactions. Any sustainable solution to asthma morbidity in urban adolescents requires a multi-dimensional approach not only addressing information needs but also providing positive experiences to promote positive attitudes and interactions with peers that can maximize the program effects. Thus, a peer-led intervention could arguably have more “bang-for-buck” than traditional adult led interventions, in that it can benefit both the mentor and the mentee.

Ultimately, changing asthma outcomes for youth may require a drastic change in the way that we deliver care, as current care approaches have not achieved desired improvements in outcomes particularly in poor minority adolescents. Implementing a peer-led program in a non-research setting would be one such possible change, and one that has proven evidence of efficacy. Given that asthma has become a dominant youth health issue with broad-spectrum negative impact on psychosocial and physical health outcomes, and given a growing recognition that families in vulnerable circumstances are often challenged to adequately manage asthma at home, it may be time to consider novel approaches to care, such as PLASMA. In doing so, new payer models (i.e. policy changes at insurer, state, and federal levels) must be considered that would enable programs such as this to be funded or reimbursed, and integrated into general pediatric asthma care.

Highlights.

• A peer-led program improved asthma control and quality of life in urban peer leaders.

• A peer-led program improved asthma knowledge and attitudes in urban peer leaders.

• Positive effects began to emerge after training and sustained for 15 months.

• A peer-led program can be an effective education model for urban teens with asthma.

Table A.1.

Mean and standard deviation (SD) of each study variable by time point

	Baseline Mean, SD (n)	Training Mean, SD (n)	Camp Mean, SD (n)	3 Month Mean, SD (n)	6 Months Mean, SD (n)	9 Months Mean, SD (n)	12 Months Mean, SD (n)	15 Months Mean, SD (n)
Quality of Life	5.73, 1.24 (51)	6.29, 0.89 (39)	6.32, 0.95 (35)	6.4, 0.83 (29)	6.47, 0.65 (28)	6.5, 0.59 (29)	6.57, 0.57 (28)	6.49, 0.85 (27)
Symptoms	5.56, 1.36 (51)	6.25, 0.88 (39)	6.26, 0.96 (35)	6.27, 0.95 (29)	6.36, 0.68 (28)	6.31, 0.8 (29)	6.49, 0.66 (28)	6.46, 0.88 (27)
Activity Limitations	5.64, 1.22 (51)	6.15, 1.02 (39)	6.35, 0.83 (35)	6.43, 0.8 (29)	6.5, 0.65 (28)	6.52, 0.58 (29)	6.56, 0.68 (28)	6.36, 0.96 (27)
Emotional Function	5.99, 1.23 (51)	6.43, 0.91 (39)	6.37, 1.09 (35)	6.55, 0.78 (29)	6.58, 0.67 (28)	6.72, 0.51 (29)	6.68, 0.54 (28)	6.61, 0.88 (27)
Asthma Control	1.06, 1.05 (51)	0.57, 0.75 (38)	0.63, 0.94 (35)	0.6, 0.81 (27)	0.52, 0.66 (28)	0.54, 0.55 (29)	0.42, 0.73 (27)	0.44, 0.64 (27)
Asthma Knowledge	23.67, 3.43 (51)	26.33, 3.18 (39)	25.49, 3.59 (35)	-	25.86, 2.58 (28)	-	25.48, 3.58 (27)	25.85, 3.22 (27)
Attitude	3.67, 0.69 (50)	3.9, 0.63 (38)	4.03, 0.71 (35)	-	3.99, 0.54 (28)	-	4, 0.69 (27)	3.95, 0.74 (27)

Table A.2.

Cohen’s D for longitudinal effects of the intervention on outcome measures in peer leaders in comparison to value at baseline. These models are all adjusted for teen sex, camp season, and site.

Variables	Partial Eta^2	p-value	Training	Camp	3 months	6 months	9 months	12 months	15 months
Quality of Life	0.213	<0.001	0.526	0.532	0.641	0.744	0.789	0.864	0.711
Symptoms	0.194	<0.001	0.598	0.581	0.606	0.732	0.670	0.851	0.773
Activity Limitations	0.245	<0.001	0.460	0.700	0.776	0.898	0.944	0.952	0.659
Emotional Function	0.165	<0.001	0.417	0.336	0.547	0.598	0.758	0.716	0.578
Asthma Control	0.119	<0.001	−0.488	−0.396	−0.451	−0.563	−0.566	−0.657	−0.653
Asthma Knowledge	0.154	<0.001	0.836	0.501	-	0.754	-	0.532	0.679
Attitudes	0.153	<0.001	0.255	0.432	-	0.408	-	0.389	0.308