Benefits of a School-Based Asthma Treatment Program in the Absence of Secondhand Smoke Exposure: Results of a Randomized Clinical Trial

Background  Daily maintenance medications are recommended for all children with mild persistent to severe persistent asthma; however, poor adherence to these medications is common.

Objective  To evaluate the impact of school-based provision of inhaled corticosteroids on asthma severity among urban children with mild persistent to severe persistent asthma.

Design  Children aged 3 to 7 years with mild persistent to severe persistent asthma were identified at the start of the 2000-2001 and 2001-2002 school years in Rochester. Children were assigned randomly to a school-based care group (daily inhaled corticosteroids provided through the school) or a usual-care group (inhaled corticosteroids not given through school).

Main Outcome Measure  Improvement in parent-reported symptom-free days.

Results  Of 242 eligible children, 184 were enrolled from 54 urban schools. Data for 180 children were available. Parents of children in the school-based care group had a greater improvement in quality of life compared with parents of children in the usual-care group (change score, 0.63 vs 0.24; P = .047); also, children in the school-based care group vs the usual-care group missed less school because of asthma (mean total days missed, 6.8 vs 8.8; P = .047) and experienced more symptom-free days during the early winter months (mean days per 2-week period, 9.2 vs 7.3; P = .02). A post hoc analysis revealed that all significant findings were produced by differences among children who were not exposed to secondhand smoke. Furthermore, among children not exposed to smoke, those in the school-based care group vs the usual-care group had more symptom-free days overall (11.5 vs 10.5; P = .046), had fewer days needing rescue medications (1.6 vs 2.3; P = .03), and were less likely to have had 3 or more acute visits for asthma (6 [13%] of 47 children vs 17 [31%] of 54 children; P = .03).

Conclusions  School-based provision of inhaled corticosteroids significantly improved symptoms, quality of life, and absenteeism among urban children with mild persistent to severe persistent asthma. This effect was seen only among children not exposed to secondhand smoke.

In the United States, poor and minority children disproportionately suffer from asthma.^1-6 While national guidelines recommend daily maintenance medications for all children with mild persistent to severe persistent asthma,^7,8 undertreatment and poor adherence to these medications are common, particularly for young urban children.^9-13 This may partially explain the discrepancy in asthma morbidity between poor and nonpoor children.

A few community-based studies have been aimed at improving the health of urban children with asthma. Most of these studies^14-20 have involved intensive case management and educational interventions, and their impact has varied. Others^21,22 have involved innovative multimedia programs for education and monitoring, again with mixed results. This project was designed to help urban children with asthma receive the medications they need through a simple strategy using the school system as the site for the provision and delivery of preventive medications. Part of the impetus for this study was the evidence in other therapeutic settings that treatment that is directly observed can be effective.^23-26 By providing daily preventive medications through the school system, adherence can be assured at least on the days when the child attends school.

The primary objective of this study was to evaluate the impact of school-based provision of inhaled corticosteroids as opposed to the standard of care on asthma symptoms among urban children with mild persistent to severe persistent asthma. Furthermore, because environmental tobacco smoke exposure is associated with increased morbidity among young children with asthma,^27-43 and because environmental tobacco smoke exposure may be significant for young urban children,^43-47 we performed a single post hoc analysis of the effectiveness of this intervention among children with and without exposure to secondhand smoke in the home.

This study was conducted within the Rochester City School District in Monroe County, New York. The population of Rochester is greater than 200 000, and the district consists primarily of minority and low-income children, 87% of whom are eligible for free or reduced-price lunches. All parents of children entering the Rochester City School District complete a school health and development form on enrollment. This school survey inquires about chronic diseases and includes several questions about asthma. At the start of the 2000-2001 and 2001-2002 school years, all children whose parents indicated either a diagnosis of asthma or a history of asthma symptoms on their school health forms were identified. These parents received a telephone survey from our research team (J.S.H., K.M.C., T.L.N., P.M.C., S.C.L.) to assess their child's eligibility for this study. The Institutional Review Board of the University of Rochester School of Medicine and Dentistry approved the study protocol.

To be eligible for participation, children were first required to have symptoms consistent with mild persistent or more severe asthma. We used the following criteria for symptom severity, adapted from the National Heart, Lung, and Blood Institute guidelines: (1) an average of more than 2 days per week with asthma symptoms,^7,8 (2) more than 2 days per month with nighttime symptoms,^7,8 (3) 3 or more acute visits for asthma during the past year, and (4) 1 or more hospitalizations for asthma during the past year.

A child fulfilling at least one of the previously mentioned criteria was considered for participation. Participants were eligible for the study if the child met symptom criteria, was aged 3 to 7 years, and was enrolled in the Rochester City School District and if the family had access to a working telephone for monthly follow-up telephone calls. Children who were scheduled to move from the school district within fewer than 6 months were not eligible for enrollment. Because of a lack of available resources, Spanish-speaking families were enrolled in study year 2 only.

If a child was eligible and the family wanted to participate, they were asked to provide informed consent. The child's primary care provider was notified of the child's asthma severity and asked for permission for the child's enrollment in the program. Providers were asked to sign authorization only if they agreed with the child's need for daily inhaled corticosteroids. A child was not enrolled into the program without clinician authorization.

Parents of enrolled children completed a baseline assessment by telephone interview. Interviews addressed demographic information, asthma severity, medication use, environmental risk factors for asthma, and parent's quality of life.

Following completion of the baseline assessment, each child was randomized into either the school-based care group or the usual-care group. Randomization was stratified by current use of preventive medications and was blocked in groups of 6. Pairs of siblings were assigned randomly to the same group. Randomization cards were made from a table of random numbers and were kept in sealed, opaque, sequentially numbered envelopes until after the baseline assessment was completed. Following randomization, families and primary care providers were notified of the child's group allocation. Families were sent diaries for symptom tracking during the study.

For each child assigned to the school-based care group, 2 metered-dose inhalers of fluticasone propionate, 110 µg per puff (Flovent 110), along with spacers, were provided. The school nurse used one inhaler to administer one dose (2 puffs) each day that the child was in school (although fluticasone is labeled for twice-daily dosing, once-daily dosing is effective,⁴⁸ and was used in this study to allow for the administration of medication during school hours). The other inhaler was given to the family to use for medication administration at home on the days when the child did not attend school. The school nurse recorded each administration of medication that was delivered at school. Children who were already using fluticasone or a different inhaled corticosteroid at home were instructed to use the fluticasone provided through the study instead. Children who were using more than one preventive medication were instructed to continue with their other medications (in addition to the fluticasone given through school) at the discretion of their primary care provider. The primary care providers were able to "step up" the dose of preventive medications by increasing home doses, but the dose given at school was not altered throughout the study.

Primary care providers and parents of children in the usual-care group were notified of the child's asthma severity, and parents were advised to contact their provider for asthma care. While maintenance medications were recommended for these children, and primary care providers agreed with the use of preventive medications for each child before their authorization for the study, the children did not receive medications directly through the program. If children in the usual-care group were taking a daily preventive medication, parents were responsible for filling prescriptions from their physician and administering medications to their child.

All families were contacted monthly by telephone to assess the main outcome measure: the number of symptom-free days during the 2 weeks before the follow-up interview. They were asked to refer to their symptom diaries to respond to questions about their child's symptoms. Symptom-free days were described to parents as a 24-hour period that their child experienced no symptoms of asthma (including coughing, wheezing, or shortness of breath). Other measures of asthma severity included daytime asthma symptoms, nighttime asthma symptoms, and the need for rescue inhaler use (all measured as the number of days during the 2 weeks before the follow-up interview). To assess absenteeism, parents were asked how many days of school the child missed because of asthma since the last telephone call from the research group. Parents also were asked monthly about any visits to either the physician's office or the emergency department for asthma attacks and about any hospitalizations for asthma since the last follow-up telephone survey. Last, parents were asked monthly about any adverse effects their child was experiencing, and specifically about yeast infections of the mouth.

The quality of life of parents and caregivers was measured using the Paediatric Asthma Caregiver's Quality of Life Questionnaire.⁴⁹ The Paediatric Asthma Caregiver's Quality of Life Questionnaire is a standard instrument used to assess the quality of life of caretakers of children with asthma, and provides a quantitative score that reflects answers to questions about the burden of caring for a child with asthma. This scale was administered at baseline and at the end of the school year. A change in quality-of-life score was ascertained by calculating the improvement from baseline to final quality-of-life scores.

An overall asthma symptom score was calculated for each measure of severity (symptom-free days, daytime symptoms, nighttime symptoms, and rescue medication use) by computing the mean of all follow-up data obtained for each child. The total number of school days missed because of asthma was calculated for each child. Results also were analyzed by grouped months of follow-up to evaluate whether there were certain times during the year that the program was most beneficial.

The study continued for the duration of the school year. To ensure an unbiased assessment, an independent research group, blinded to each child's group allocation, conducted the follow-up interviews.

Independent variables included sex, race (white, black, or other), ethnicity (Hispanic or not Hispanic), maternal educational level (less than high school or high school degree and higher), maternal age (<30 or ≥30 years), insurance type (Medicaid, New York's State Child Health Insurance Program, private or other, or no insurance), and poverty status (public assistance or no public assistance). Smoking status was assessed by inquiring whether anyone in the home smoked cigarettes, cigars, or pipes at enrollment.

In accordance with the intention-to-treat principle all randomized subjects were kept in their originally assigned treatment groups for analysis. Subjects who were enrolled in the study but participated for less than 1 month were not included because of lack of data (4 in the school-based care group and 0 in the usual-care group). Analysis was performed using a commercially available software program (SPSS for Windows, version 10.0; SPSS Inc, Chicago, Ill). Bivariate analyses were used to compare the dependent variables between the 2 groups, including t tests for numeric data and χ² tests for categorical data. Nonnormal data were transformed using the square root function before the application of statistical testing (symptom days, symptom nights, rescue inhaler use, and school absences). A single post hoc analysis was done with data stratified by the presence or absence of smoking in the home. Multiple regression analyses were performed to predict continuous outcomes (symptom-free days, quality of life, and days absent from school), controlling for baseline asthma status (baseline symptom-free days and quality of life), demographic factors (race, sex, and ethnicity), parent factors (educational level, poverty status, and age), and secondhand smoke exposure (as an independent variable and as an interaction term with the treatment group). For dichotomous outcomes (acute visits for asthma), a similar logistic regression model was used.

In planning this study, the sample size calculation was based on the ability to detect a difference of 1 symptom day per 2-week period (assumed SD, 2.5 days¹⁹), which is considered to be a clinically significant difference.¹⁹ A total sample size of 196 patients was required (α = .05, β = .20, 2-sided test). P<.05 was considered statistically significant.

Three hundred eighty-seven children entering the school district had symptoms of asthma noted on their school health forms and were assessed for eligibility for the study. Of these children, 145 were ineligible and the remaining 242 had mild persistent to severe persistent asthma and were eligible to participate (Figure 1). Fifty-eight children were not enrolled because the parent or the physician did not want the child to participate or written parental consent was not obtained. One hundred eighty-four children from 54 different schools and preschools throughout Rochester were enrolled in the program (74% of those eligible). Of the children, 93 were allocated to the school-based care group and 91 to the usual-care group.

One hundred eighty children had data available for analysis (89 in the school-based care group and 91 in the usual-care group). The response rate for all of the follow-up interviews was 94%. Table 1 shows the demographic characteristics and outcome measures at baseline for the children in the study. Of the children, 54% were aged 3 to 5 years and 46% were aged 6 to 7 years. Most of the children were male. Most families received Medicaid insurance and public assistance, and 53% of parents had less than a high school education.

There were no differences in demographic characteristics, exposure to secondhand smoke, or prior use of maintenance medications between those children assigned to the school-based care group and those assigned to the usual-care group. Furthermore, there were no significant differences in baseline asthma severity between children in the school-based care group and children in the usual-care group.

Children assigned to the school-based care group actually received the intervention (medication delivered at school) an average of 84% of the days that school was in session. Of parents in the usual-care group, 63% (57/91) reported their child using a daily maintenance medication during the study. Table 2 shows the overall mean results of the intervention for the 180 children included in the analysis. The parents of children in the school-based care group experienced a greater improvement in quality of life compared with the parents of children in the usual-care group. A change score for quality of life of 0.5 or more is considered to be clinically significant.⁵⁰ Furthermore, children in the school-based care group missed fewer days of school because of asthma compared with children in the usual-care group. There were no differences in the mean number of symptom-free days, symptom days, symptom nights, and days with rescue medication use among the children in the 2 groups. Furthermore, there was no statistical difference in the percentage of children in the 2 groups with 3 or more acute office or emergency department visits for asthma or with 1 or more hospitalizations for asthma. There were a total of 144 acute visits and 3 hospitalizations among children in the school-based care group compared with 175 acute visits and 8 hospitalizations among children in the usual-care group.

Figure 2 shows the mean number of symptom-free days for children in the 2 groups by months of follow-up. All of the children became healthier over time, with a mean of approximately 11.5 symptom-free days at the end of the school year for both groups of children. In the early winter months (November and December), children in the school-based care group on average had more symptom-free days in the 2 weeks before each follow-up interview compared with children in the usual-care group (mean [SD], 9.2 [4.4] vs 7.3 [4.7] days; P = .02).

Table 3 shows the results of the intervention for the 180 children included in the study, stratified by exposure to secondhand smoke in the home. The data from the 79 children exposed to smoke in the home showed no differences between the school-based care and the usual-care groups. Both of the significant findings from the overall analysis (improved quality of life and fewer absences from school) are produced by differences among the 101 children in the school-based care and usual-care groups who were not exposed to secondhand smoke in the home. Furthermore, among children not exposed to smoke, 3 additional significant differences were found in the stratified analysis. Those in the school-based care group had significantly more symptom-free days compared with those in the usual-care group (P = .046). Also, those in the school-based care group vs the usual-care group had fewer days needing rescue medications (P = .03) and were less likely to have had 3 or more acute office or emergency department visits for asthma (P = .03). To obviate any potential effect of the cost of medications in the analysis, these data also were analyzed encompassing only those subjects whose medications were available at no cost (children with Medicaid insurance or those enrolled in New York's State Child Health Insurance Program—82% of the children). The significant results previously reported also were significant (P<.05) in this additional analysis.

All of the significant findings from the overall analysis remained statistically significant in a multivariate regression analysis controlling for baseline asthma severity (symptom-free days and quality of life), baseline use of maintenance medications, child demographic factors (sex, race, and ethnicity), parent factors (age, educational level, and poverty status), exposure to secondhand smoke, and smoke exposure × treatment group interaction. Therefore, only the unadjusted analyses are presented. The smoke exposure × treatment group interaction was statistically significant in the analysis predicting symptom-free days (β = .45, P = .048).

There were 8 sibling pairs included in the study. To account for the nonindependence of sibling data, we repeated our analysis with only one randomly selected member of each sibling pair included. All of the significant results remained the same in this analysis. There were no differences between study groups in the report of any adverse events during the study (report of yeast infection of the mouth, 7% [6/89] vs 8% [7/91] for the school-based care group vs the usual-care group; P>.99).

This study demonstrates 2 new and potentially important findings about the management of urban children with asthma: the utility of a system change involving school-based provision of inhaled corticosteroids and the ineffectiveness of this program among those children with secondhand smoke in their homes. The children who received prophylactic inhaled corticosteroids from the schools experienced more symptom-free days during the winter, missed fewer days of school because of asthma, and had a greater improvement in parental quality of life compared with the children who did not receive this intervention.

A small study⁵¹ in Dallas, Tex, also evaluated school-based delivery of preventive medications for older inner-city children. The researchers found historic improvements in measures of asthma severity for the children who received the intervention, but no comparison group was examined. Another community-based study¹⁹ through the National Collaborative Inner-City Asthma Study used master's-level social workers to deliver an intensive home-based educational intervention to poor urban families. These researchers found a moderate improvement in symptoms for the children receiving the intervention compared with the children in the control group. Compared with other resource-intensive interventions, the school-based intervention that we assessed is conceptually simple and inexpensive and had significant beneficial effects.

This study was designed as a health services trial to test the impact of a system change on the effectiveness of care. This precluded altering the system of care for the children in the usual-care group to avoid the intrinsic bias of adopting a new routine process. Thus, there were by necessity inherent differences between the usual-care and school-based care groups beyond the location (school vs home) of drug administration. In the usual-care group, these include the requirement to obtain prescriptions from the physician, fill the prescription at the pharmacy, and administer the medication to the child at home. We cannot tell from this study which specific component of the system change was most beneficial to the children (and their families) in the school-based care group. What we have shown is that the system of care involving the schools for the provision of maintenance medications can be effective for young urban children with asthma.

Payment for the medications was not a seminal difference between the usual-care and the school-based care groups. Full prescription coverage was available to 82% of the children through Medicaid or New York's State Child Health Insurance Program. To obviate the potential effect of cost of the medications among the remaining 18% of children for whom medications would require payment, the data were analyzed encompassing only those subjects whose medications were available at no cost, and the results were unchanged.

We found a differential effect of this intervention for children with and without exposure to secondhand smoke in the home. The children who did not live with smokers at home experienced significant benefits from the intervention, including more symptom-free days overall and fewer acute visits for asthma exacerbations. The intervention had no effect on those children who dwelled in homes with environmental tobacco smoke regardless of how the asthma severity was assessed. The relatively few subjects in these subgroups may have prevented our detection of a small benefit of the intervention for those children in homes with smokers. Smoke may have adversely affected the airways of these children, because environmental tobacco smoke has been associated with worsening asthma symptoms and decreased pulmonary function in young children with asthma.^27-43 It also is possible that some environmental factor associated with smoke exposure in the home, such as allergens in the home, may have inhibited the intervention's effectiveness.

Studies^52,53 of adult patients with asthma indicate that inhaled corticosteroids are not effective in decreasing inflammation among those patients who smoke. Similar studies of the effectiveness of inhaled corticosteroids among children with and without exposure to smoke are not available. However, these data are consistent with the observations among adults with asthma, and raise the possibility that providing inhaled corticosteroids to children who are exposed to smoke may have limited effectiveness in improving their pulmonary status.

Potential limitations of this study include parental reporting as the sole measure of asthma symptom burden. We did not obtain physiologic measures of asthma severity, such as peak flow readings or spirometric measurements. However, the National Heart, Lung, and Blood Institute criteria^7,8 for defining asthma severity depend on parent report of symptoms and, therefore, these data are consistent with that standard. Furthermore, although environmental tobacco smoke exposure also was determined by parent report, studies^32,54,55 have shown that parent report of smoke exposure is reasonably accurate.

The usual-care group for this study received a modified intervention simply by participating in the study. Their primary care provider was required to authorize participation in the study and, thus, was alerted to the child's asthma severity and reminded of the guidelines for asthma care before the child's randomization. Furthermore, parents of children in both groups were called monthly and asked to recall the child's asthma symptoms. It is plausible that enhanced provider awareness of illness and more parental attention to ongoing symptoms led to improved outcomes among children in the usual-care group. This would introduce a conservative bias, making differences between the school-based care and usual-care groups more difficult to detect.

Blinding of parents, children, and primary care providers to the allocated treatment group was not possible for this study. Therefore, symptom responses and quality-of-life reports could be biased by parents assuming that the treatment was beneficial. However, highly reliable observations, such as absenteeism and acute visits, are less likely to be biased.

Last, only young urban children from Rochester were included in this study. Older children with asthma and children from rural or suburban localities might have a different response to this type of intervention.

This study demonstrated reduced morbidity among children receiving inhaled corticosteroids through a school-based program. The benefits included fewer symptoms, decreased absenteeism, and improved parental quality of life. These findings translate into reduced suffering for the children and their families, and increased productivity for the parents or caregivers who can remain at work rather than stay home with an ill child. The program only seemed to benefit children who were not exposed to smoke in the home, suggesting a need for further study of the environmental effects on asthma therapy.

Most children included in this study were poor and from minority populations, and represent the group with the greatest need for assistance with this common chronic illness. If these findings are replicated in other settings, this school-based system of preventive care could become standard for the management of childhood asthma in underserved communities.

Article

National guidelines recommend preventive medications for all children with mild persistent to severe persistent asthma. Despite these recommendations, adherence to medications remains suboptimal and morbidity and mortality continue to increase.

This study shows that school administration of preventive asthma medications improved outcomes for urban children with mild persistent to severe persistent asthma. However, the program only benefited children who were not exposed to smoke in the home. These findings suggest that school-based preventive care could help many children with asthma. Further investigations of the effects of secondhand smoke on asthma therapy are warranted.

Corresponding author: Jill S. Halterman, MD, MPH, Department of Pediatrics, University of Rochester School of Medicine and Dentistry, 601 Elmwood Ave, PO Box 777, Rochester, NY 14642 (e-mail: jill_halterman@urmc.rochester.edu).

Accepted for publication October 30, 2003.

This study was supported by a grant from the Halcyon Hill Foundation, Webster, NY; and by the Robert Wood Johnson Foundation's Generalist Physician Faculty Scholars Program.

We thank GlaxoSmithKline, Research Triangle Park, NC, for the donation of the fluticasone propionate and spacers used in this study; Andrew MacGowan, of the Rochester City School District, Sue Peer, RN, the school nurse supervisors, the school nurses, and the health aides for their support throughout this study; A. Dirk Hightower, PhD, and Guillermo Montes, PhD, for their assistance in the planning phase of this study; George B. Segel, MD, and Michael Weitzman, MD, for their insightful review of the manuscript; Michael McDermott, PhD, for his statistical consultation; and the children and caregivers for their participation.