Systematic Reviews of Asthma Self-Monitoring, Self-Management, or Patient Education

There have been numerous systematic reviews of asthma self-monitoring, self-management or patient education—many of which have been methodologically rigorous.^{10, 29}–⁴⁵ We direct interested readers to these reviews for detailed descriptions of their results. In general, the results of these systematic reviews highlight the heterogeneity of these literatures, including disparities in the relevant information reported about the interventions and the content and educational approaches evaluated. A common finding was that interventions directed at improving patient knowledge, typically through non-interactive formats (e.g., lecture, video, print) do not necessarily improve health outcomes. In contrast, interventions that focus on improving self-monitoring or self-management skills through behavior change techniques, often result in reductions in health services utilization and asthma symptoms and improve functional status. The authors of the systematic reviews described the common methodological weaknesses in the evidence base: small sample sizes, lack of long-term follow up, and use of interventions without strong theoretical or empirical foundations.

The research question addressed by many of these reviews was whether written action plans improve the outcomes of care for patients with asthma.^{10, 29, 30, 34, 35, 40} For example, the 2001 AHRQ funded Evidence Report entitled, “Management of Chronic Asthma”¹⁰ (which was used to inform the recommendations of the 2002 NHLBI asthma guideline) evaluated whether written asthma action plans improve asthma outcomes and, specifically, whether peak flow monitor-based plans are superior to symptom-based plans. They synthesized the evidence from 36 controlled trials of the efficacy of written asthma action plans to improve outcomes for pediatric and adult populations with asthma. The authors found that most study designs were confounded by multiple asthma management interventions—only nine studies with a total of 1501 patients evaluated self-management programs in isolation.¹⁰ Of the five trials comparing a peak flow-based action plan to no action plan, only one found a statistically significant result—a reduction in emergency department visits for the peak flow-based action plan group. None of these five trials reported any other statistically significant difference in any other measure of clinical status or health services utilization. The authors note that most studies were underpowered to find statistically significant results.¹⁰ They concluded that there was insufficient evidence to demonstrate an association between the use of written asthma action plans and improved asthma outcomes.¹⁰ Moreover, they found that there was insufficient evidence to support the hypothesis that peak flow monitoring-based plans were superior to symptom-based plans.¹⁰

The 2004 Cochrane review by Toelle and Ram was designed to assess whether the provision of a written individualized management plan increased adherence with self-monitoring or self-management behaviors.³⁰ Toelle and Ram concluded that there was insufficient evidence to recommend for or against the use of written management plans for asthma. Toelle and Ram commented that the failure to demonstrate any difference in health outcomes between written individualized plan groups and control groups may be because the provision of a written plan does not lead to any change in behavior or that enrollment in a study may lead to similar improvements in both control and intervention groups.³⁰

Additionally, although not systematic reviews, the Working Group Reports from the 1998 World Asthma Meeting are notable for providing succinct descriptions of key trials of a variety of self-management and patient education interventions and discussions of important gaps in the literature.^{46, 47} In particular, the article by Partridge and colleagues details the list of self-management skills (including self-monitoring) that are widely accepted as required for effective self care and reviews a number of interventions and international efforts to improve patient-provider communication, provider education, and policy-level interventions for asthma care.⁴⁷

In summary, prior systematic reviews found that non-interactive educational interventions were not effective, and that there was insufficient evidence to determine the value of written self-management plans.

To avoid duplicating the work done in prior reviews, we did not evaluate the comparative effectiveness of symptom-based versus peak flow-based self-monitoring or self-management interventions. Instead, our aim was to evaluate specific intervention characteristic (e.g., setting, teaching strategy, intensity of the intervention) and population characteristics (e.g., adolescents, country of residence) associated with improvements in outcomes of care for children and adults with asthma.

General Characteristics of Self-Monitoring, Self-Management, or Patient Education Interventions for Children With Asthma

Study Design Characteristics. Among the included articles, 69 had some component of self-monitoring, self-management, or education for children with asthma or their caregivers. The median duration of the follow up period was 12 months (S.D., 5.8 months). The median sample size of these interventions was 90 subjects (Interquartile range: 43, 181). Fifty-nine were RCTs, three were quasi-RCTs, and ten were controlled before-after trials. In univariate analysis, the longest studies were more likely to have the largest number of subjects (p=0.005). Eleven were primarily self-monitoring or self-management, 36 were primarily patient or caregiver education, and 13 included both.

Intervention Characteristics. The included interventions were highly heterogeneous in terms of educational materials provided, setting, frequency and duration of contact with asthma patients and their caregivers, among other key characteristics.

Forty-three (62%) specifically described an underlying conceptual framework or theoretical background as the basis for the intervention. We direct interested readers elsewhere for discussions of the theoretic foundations of asthma educational and self-monitoring or self-management interventions which include theories of empowerment, social ecology, self-regulation, and self-efficacy.⁴⁸–⁵⁰ In univariate analyses, these studies tended to be more likely to report statistically significant improvements in emergency department visits (36 studies reported both emergency department visits and described a conceptual framework; p=0.014) but not the other outcomes of interest.

Nineteen of the self-monitoring, self-management, or patient education interventions were performed in outpatient primary care and most were taught by either physicians or nurses/nurse practitioners (Figure 5).

Figure

Figure 5. Intervention characteristics: setting and educator type.

Thirty-two (46%) of the self-monitoring, self-management, or patient education interventions provided 2 to 5 educational sessions to the children or their caregivers. Twenty-six interventions principally provided individualized instruction, six principally used interactive group teaching sessions, and 23 used combinations of teaching modalities (Figure 6)

Figure

Figure 6. Number of educational sessions provided by educational strategy.

Eight studies used video games or Web-based programs as educational or self-monitoring or self-management tools for children with asthma.⁵¹–⁵⁸ For example, “Bronkie's Asthma Adventure” (Click Health, Inc, Mountain View, CA) is a Nintendo^®-based game that has been designed to teach children self-monitoring or self-management strategies and provide feedback (in English or Spanish language) on their performance.⁵⁸ Similarly, the “Asthma Control” video game features a superhero named “Spacer” whom the player has to lead through six game levels, accumulating points by avoiding both indoor and outdoor triggers/allergens and using controller medications.⁵⁴ The use of video games or Web-based teaching modalities was not associated consistently with statistically significant improvements in outcomes or processes of care for children with asthma.

Outcomes Reported. The 69 self-monitoring, self-management, or patient education studies reported on a variety of outcomes. The most frequently reported outcomes were asthma symptoms, days lost from school or work, urgent care or emergency department visits, and hospitalizations due to asthma. Relatively few studies reported on guideline adherence including adherence to self-monitoring or self-management plans. Table 10 presents a summary of the number of studies reporting each of the primary outcomes of interest. Thirty-one studies reported on asthma symptoms; however, these were measured in highly heterogeneous ways (e.g., self reported symptoms from children's diaries, physicians' ratings of asthma symptoms, multiple different asthma symptom questionnaires). Thirty studies reported on days lost from school or work due to asthma. Some authors adjusted days lost from school on the basis of expected seasonal variation in absenteeism, but most reported school days lost in the intervention group compared to the control group.

Table 10

Outcomes reported by the pediatric self-monitoring, self-management, or patient education studies.

Self-Monitoring or Self-Management Interventions for Children With Asthma

Background. The International Consensus Report on Asthma suggests that peak flows of 80 to 100% of the individual's best are satisfactory and necessitate only routine treatment.²⁸ Peak flows of 50 to 80% of personal best peak flow should stimulate a treatment change (e.g., increasing bronchodilators or anti-inflammatory medications).²⁸ Peak flows below 50% of personal best should lead the patient to use their urgent medications (e.g., start oral steroids) and seek medical attention.²⁸ Most of the self-monitoring or self-management interventions for children with asthma utilize a written, often color-coded plan in which instructions and medications are labeled (green) for routine care, (yellow) for caution/early treatment, and (red) for urgent treatment. The purpose of this section is to present the characteristics of the individual self-monitoring or self-management interventions.

Results. We found 35 interventions designed principally for the improvement of self-monitoring or self-management of asthma symptoms by children with asthma and their caregivers (24 of these studies are presented in Table 11). Nine of the pediatric self-monitoring or self-management studies compared two self-monitoring or self-management interventions without including a control group that did not also receive a self-monitoring or self-management intervention.^{55, 56, 59}–⁶⁵ Studies with this design are described in detail in a subsequent section of the report (Table 21).

Table 11

Summary of self-monitoring or self-management interventions directed at children with asthma.

Table 21

Trials comparing two or more intervention groups without a control group that did not also receive an intervention.

The studies of pediatric self-monitoring or self-management interventions are notable among all the types of studies synthesized for this report, in two ways. First, many of these studies reported statistically significant improvements in subjects receiving the intervention—specifically, 29 (83%) of the pediatric self-monitoring or self-management studies reported at least one statistically significant outcome for the recipients of the self-monitoring or self-management intervention compared with the control group. We explore the intervention factors associated with improvements in outcomes and processes of care for children with asthma at the end of this section (Table 13 through Table 16).

Table 13

Association of pediatric self-monitoring, self-management, or patient education intervention characteristics and significant clinical outcomes.

Table 16

Association of pediatric self-monitoring, self-management, or patient education intervention characteristics and reductions in asthma-related hospitalizations.

Second, among the included studies, the reports of self-monitoring or self-management interventions were most likely to have described their efforts to design interventions that were well-grounded in theoretical frameworks such as social learning theory, cognitive development, and behavior change. Twenty-two (63%) of the included pediatric self-monitoring or self-management studies described such a theoretical rationale for either the content of the program or the selection of teaching methods employed.⁵²–^{54, 61, 64, 66}–⁸¹ Typically, these interventions relied less on lecture-based or pamphlet-based teaching methods and utilized multiple educational modalities including role-playing, videotapes, and games to reinforce patient learning.

The 35 interventions were heterogeneous with respect to the specific content and delivery methods of the self-monitoring or self-management program. However, two programs, Open Airways and “Superstuff” are worthy of specific mention because they were the subject of multiple evaluations. Five studies evaluated a school-based self-monitoring or self-management program utilizing multiple educational components called Open Airways.⁶⁸–^{70, 72, 82} This program consists of six 40–60 minute group sessions for inner-city third to fifth graders to increase their ability to care for their asthma on a daily basis. It includes information about asthma pathophysiology, recognizing and responding to asthma symptoms, using asthma medications and deciding about when to seek care, staying active, identifying and controlling asthma triggers, and managing asthma-related school problems. Interactive teaching methods used in Open Airways include group discussions, storytelling, games, and role-playing. Each of these studies found important clinical improvements in asthma outcomes for the participants of this self-monitoring or self-management program (Table 11). The study by Ronchetti and colleagues of children receiving asthma care in 12 Italian centers compared the outcomes of three groups: subjects receiving Open Airways, subjects receiving another self-monitoring or self-management program called Living with Asthma, and control subjects receiving usual care.⁷² The content of the Living with Asthma program is similar to Open Airways and it also uses a group format; however, it makes more extensive use of written diaries for developing asthma management skills and does not rely as heavily as Open Airways on encouraging group members to share problems and develop solutions together. One year after enrollment, patients in the Open Airways but not the Living With Asthma groups had fewer emergency treatments for asthma than controls (p<0.03).⁷²

Three studies reported evaluations of self-monitoring or self-management programs that utilized the “Superstuff” pediatric self-monitoring or self-management kits produced by the American Lung Association (or materials developed from “Superstuff”).^{66, 83, 84} “Superstuff” kits include a Parent's Magazine containing 29 articles on asthma pathophysiology, triggers, relaxation techniques, and personal control and decision making. The Children's Kit includes riddles about asthma facts, the “Breathe Easy” board game, puzzles and dolls with self-care messages, a comic strip about relaxation exercises, a mystery house for games about discovering allergens, a phone book with advice about when to call the doctor and to record emergency numbers, and assorted asthma-related door signs, posters, stickers, records, and paper cut-outs. In the study by Rakos and colleagues, 20 children with moderate to severe asthma received the “Superstuff” kit in the mail with instructions on its self-administration while 23 children received usual care.⁸³ One year after receiving the intervention, parental reports suggested a statistically significant decrease in “interruptions due to asthma” (p<0.04) but there was no difference in school absenteeism between groups. The study by Whitman and colleagues⁸⁴ provided the “Superstuff” kit in addition to eight teaching sessions to 19 children and found that three months after the intervention, there was no difference in number of asthma episodes or days without asthma between recipients and controls (N=19); however, participant's knowledge (p=0.02) and asthma skills (p<0.01) improved compared to controls. Also, the intervention evaluated by Pérez and colleagues for Venezuelan children with asthma was based on a self-monitoring or self-management packaged adapted from Superstuff and Living with Asthma—they found that intervention subjects reported fewer asthma crises and their physicians reported less severe asthma than among control subjects (Table 11).⁶⁶

Of the other studies, numerous reported statistically significantly greater improvements in outcomes for the intervention group over the control group. However, often these statistically significant findings (e.g., improvement in percent predicted FEV₁ of less than five percent) are of only modest clinical significance.

Conclusions. The 35 self-monitoring or self-management interventions, although heterogeneous in terms of content delivered and method of instruction, tended to be associated with statistically significant improvements in outcomes for children with asthma. Additionally, they tended to be grounded in established theoretical or behavioral frameworks. The interventions with well-established theoretic foundations typically utilized multiple educational modalities including role-playing, videos, and games to reinforce patient learning. Overall, many of the reported improvements were of only modest clinical significance.

Patient Education Interventions for Children With Asthma

Background. In general, pediatric patient education strategies for asthma are based on imparting knowledge about asthma pathophysiology to patients and their parents or caregivers and encouraging the appropriate use of peak flow meters and inhaled medications.

Results. We found 54 evaluations of interventions designed primarily to educate children or parents or caregivers of children with asthma (Table 12). The included studies were highly heterogeneous with respect to the target of the intervention (e.g., patients versus their parents or caregivers), the setting of the intervention (e.g., home, school, clinic), and the information being provided (e.g., asthma pathophysiology, allergen reduction). Thirty-seven (69%) studies demonstrated at least one statistically significant improvement in clinical outcomes, functional status, health services utilization, or guideline adherence. The studies reporting decreases in health services utilization tended to have described a theoretical basis for their intervention (p=0.048). Half of the pediatric patient education interventions (N=27) were based on a conceptual or theoretical framework (compared to 63% of the self-monitoring or self-management interventions).

Table 12a

Summary of patient education interventions directed at the parents or caregivers of children with asthma.

In this section (Table 12), we describe the included studies in groups according to these key characteristics. (Note: The number of studies described in Table 12 is greater than 54 because there is overlap between the school-based interventions and other types of interventions so those studies are presented more than once). At the end of this section, we present the results of our synthesis of the association between intervention characteristics and likelihood of finding statistically significant improvements in the outcomes of interest for patient education or self-monitoring or self-management interventions.^‡

Parent or caregiver education programs. Whereas most school-age children are typically considered to be sufficiently mature to benefit from asthma education offered outside the context of their families, preschool children learn new skills best within the context of their families, and parents are the primary target of the education of the youngest children with asthma (less than seven years).⁹⁰ Among the included articles, we found 21 that included educational interventions directed at parents or caregivers (Table 12a)—13 (62%) of which found statistically significant improvements in processes and outcomes of care for children with asthma. In particular, this type of study was likely to report improvements in clinical outcomes (13 studies found improvements in asthma symptoms and other clinical outcomes among the 15 studies reporting these types of outcomes). Four studies compared two or more parent or caregiver education programs with each other but did not include a control group that did not also receive an educational intervention—these are described in Table 21.^{60, 62, 64, 65}

Notable for its size and methodological rigor, the National Cooperative Inner-City Asthma Study (NCICAS) evaluated the effectiveness of a multifaceted, home-based intervention for 1,033 inner city children aged 5 to 11 years with asthma from seven U.S. cities living in census tracts in which at least 20% of households had income levels below the federal poverty level.⁷³–⁷⁵ The NCICAS evaluated an educational intervention designed to teach caregivers about asthma management and to reduce those environmental asthma triggers to which their children had positive skin tests.⁷³ Intervention families were given training on asthma triggers, environmental controls, and asthma physiology by social workers and were given tools to reduce environmental allergens such as vacuum cleaners, pillow covers, and air filters. Two years after enrollment, intervention children had more symptom free days (565.1 versus 538.5 days), fewer asthma symptoms (p<0.001), days lost from school (0.54 versus 0.71 days per two weeks, p<0.009), and lower allergen levels.⁷³ There were no differences in hospitalization rates, physician visits, or emergency department between intervention and control groups.^{74, 75} The cost of the intervention was $337 per child for 2 years resulting in an estimated incremental cost-effectiveness ratio of $9.20 per symptom-free day gained (95% CI: -$12.56 to $55.29 per symptom free day gained).⁷⁵ We cannot assess the extent to which it was the educational component or the reduction in environmental allergens that resulted in improvements for intervention subjects.

The study by Toelle and colleagues⁸⁸ differed from the others in this section in that it was an educational program directed broadly at children with asthma and all the adults who influence their care including parents, physicians, teachers, pharmacists, community nurses, and school teachers. Six months after the intervention, pulmonary function (FEV₁(L) was 2.13 at six months vs. 1.78 at baseline) improved in the intervention group but not the control group (p<0.001). The number of children with wheeze and symptoms that limited their activity did not change but night cough decreased a statistically significant amount in the intervention group (37.3% at six months vs. 68.3% at baseline; p<0.001). There was no statistically significant difference in physician or emergency department visits or days absent from school.

School-based education programs. Sixteen of the studies of pediatric QI studies delivered some portion of their intervention in schools—13 of these were patient education programs (Table 12b). Eleven school-based programs (69%) reported statistically significant improvements in processes and outcomes of care for children with asthma. Although these interventions all occurred in schools, they were highly heterogeneous in terms of the curriculum delivered, training of the person(s) delivering the curriculum, intensity of the program, and target audience. Two studies used the same 6 session educational program called “The Roaring Adventures of Puff.” McGhan and colleagues⁹¹ studied the effects of this program in 7 to 12 year old children in Edmonton and Cicutto and colleagues evaluated this program in 6 to 11 year olds in Toronto.⁷⁷ In both studies, this program—which is based on asthma practice guidelines, social cognitive theory, and self-regulation theory and utilizes numerous teaching modalities including puppetry, games, role playing, model building, discussions, and asthma diary recordings—found reductions in asthma-related limitations in activity (McGhan et al. reported that in the intervention group, 41.5% of children at baseline vs. 29.2% children post-intervention were limited in their kinds of play; Cicutto et al. reported 6.2 versus 9.1 days of limited activity in the intervention compared to the control group).^{77, 91}

Table 12b

Summary of school-based patient education interventions directed at children with asthma.

Adolescent education programs. During adolescence, developmental behavioral changes can have adverse effects on asthma management if medication adherence declines or medical supervision becomes less consistent.⁹² We found six studies of educational interventions that exclusively targeting adolescents with asthma (Table 12c). None of these six adolescent-targeted interventions, even those that relied on peer teachers or intensive educational programs from physicians or nurse educators, resulted in statistically significant durable improvements in inhaled bronchodilator use, asthma symptoms control, or health services utilization. We note that five of these were relatively short with interventions lasting eight months or less.

Table 12c

Summary of patient education interventions directed at adolescents with asthma.

We found 24 studies of self-monitoring, self-management, or patient education of general pediatric populations with asthma that included adolescents. Of these, 18 (75%) reported at least one positive outcome. Similarly, adolescents were included in 26 of the interventions directed at general populations with asthma—18 (70%) of these reported at least one positive outcome. Among these 50 studies that included adolescents as part of their study population but were not exclusively targeting adolescents, 14 (28%) reported statistically significant improvements in asthma symptoms or disease severity compared to controls and 12 (24%) reported significantly fewer emergency department/urgent care visits compared to controls. Because none of these studies reported outcomes by age strata, the extent to which these improvements were found among adolescents is unclear.

Outpatient education programs. We found 15 additional studies of educational interventions designed for children with asthma that were coordinated from the outpatient setting, sometimes including home visits and calls (Table 12d). Several of the included studies that directed the educational intervention at children but did not include parental involvement (particularly for young children) did not find statistically significant improvements in the processes or outcomes of care for the asthmatic patients. Several authors noted that even among families with older children, parents did not always respond to the suggestions raised by their children after participation in the educational intervention.⁵⁴ Sometimes this was because parents could not afford or were unable to remove environmental factors such as wall-to-wall carpeting in their public housing residences.⁵⁴ Others who were capable of removing some allergens or irritants sometimes denied that a particular environmental exposure, such as their smoking, was harmful to the child. These observations strongly suggest that parallel educational activities focusing on parents as well as their children may be needed.

Table 12d

Summary of outpatient patient education interventions directed at children with asthma.

Conclusions. Patient education interventions can be effective for improving the processes and outcomes of care for children with asthma. In particular, school-based programs and those directed at parents or caregivers of young children with asthma (even among lower socio-economic groups) tend to be associated with the greatest improvements in asthma outcomes. However, the effect sizes were often only of borderline clinical significance. Additionally, the few QI strategies that exclusively targeted adolescents have not resulted in much success—emphasizing the need for additional study of this key asthma population.

Synthesis of Evidence From Self-Monitoring, Self-Management, or Patient Education Interventions for Children With Asthma

In this section we present the results of our evaluation of the association of study design characteristics (e.g., whether the authors specified a theoretical/conceptual framework, year of study, sample size, country), intervention characteristics (e.g., whether parents or caregivers were a target of the intervention, the number of educational sessions provided to students, the setting in which the intervention took place), and improvements in the outcomes of interest for the self-monitoring, self-management, or patient education interventions. We present our analyses according to each of the primary outcome types.

Clinical Outcomes. The 48 studies of pediatric or caregiver education interventions that evaluated clinical outcomes were highly heterogeneous with respect to their reporting of asthma symptoms, spirometric measures, and asthma attacks. (For example, asthma symptoms were reported varyingly as percentage of days with wheeze, amount of daytime wheezing, self-reported severity, and change in physician reported asthma symptoms, among others.) Similarly, we were interested in the effect of these interventions on the amount of rescue medications used in contrast to routine inhaled corticosteroids; however, the heterogeneity of reporting of these outcomes limited our ability to combine them quantitatively. Thus, we evaluated the study design and intervention characteristics associated with the finding of one or more statistically significant clinical outcomes using logistic regression (Table 13). We found that among all studies of pediatric asthma evaluating self-monitoring, self-management, or patient education interventions, those directed at parents or caregivers were most likely to be associated with a statistically significant improvement in clinical outcomes (p=0.02). However, there is insufficient evidence to assess the effect of other key intervention factors such as the intensity of the intervention (i.e., number of times the educator met with the child, parents, or caregivers), the type of provider of the education (e.g., nurse, physician), setting of the educational intervention (e.g., home, school), or the type of educational materials on clinical outcomes.

Functional Status Outcomes. Twenty-seven studies reported the mean (or median) days lost from school in a manner that could be use in a summary analysis (as opposed to providing a statement indicating that there was no difference between treatment and control groups for school absenteeism or a ratio of days in attendance to days absent). For those studies that did not specifically report the timeframe over which the school absenteeism was recorded (e.g., days absent per month versus per school year); we assumed that the reported days of missed school were for the entire study duration. When interpreting these data, we note that not all authors specifically identified these data as mean days lost from school and some likely represent other parameters such as total days lost from school for the whole cohort.

Figure 8 (next page) presents the individual study and summary standardized mean difference (between intervention and control groups) for school absenteeism per month due to asthma. The weighted mean reduction in days absent was 0.11 days/month (95% CI: 0.05, 0.17; p=0.0004). This corresponds to a standardized mean reduction in days absent of 0.53 (95% CI: 0.06, 0.99; p=0.03). However, the studies included in this analysis were highly statistically heterogeneous (Figure 8). Additionally, visual inspection of the funnel plot of these data (Figure 7) suggests that there may be publication bias affecting this result because most of the published studies demonstrated less school absenteeism in the intervention group. Thus, given this heterogeneity and possibility of publication bias, we cautiously conclude that among the self-monitoring, self-management, or patient education studies reporting mean school absenteeism, there may be an overall effect toward reducing asthma-related school absenteeism. Moreover, the magnitude of the effect is relatively small.

Figure

Figure 8. Self-monitoring, self-management, or patient education studies reporting mean number of school days missed due to asthma per month. Notes: The effect sizes presented here are the random effects standardized (more...)

Figure

Figure 7. Funnel plot: school absenteeism due to asthma.

We sought study design and intervention characteristics associated with the greatest reductions in school absenteeism. For this analysis, we used the standardized mean difference in school absenteeism as the dependent variable in a weighted least squares regression. We found that the longer the study duration, the greater the expected reduction in asthma-related school absenteeism (p<0.0001) (Table 14).

Table 14

Association of pediatric self-monitoring, self-management, or patient education intervention characteristics and reductions in school absenteeism.

Health Services Utilization Outcomes. Among the studies reporting measures of health services utilization, 17 studies reported the mean number of asthma-related emergency department or urgent care visits and 24 studies reported mean number of hospital days in a manner that could be use in summary analyses. For those studies that did not specifically report the timeframe over which the emergency department/urgent care visits or hospitalizations were recorded; we assumed that they were for the entire study duration. For some studies, the authors did not specify whether the reported number of emergency room visits was the total for their cohort or the mean visits per group. Because we calculated standardized mean differences (as opposed to weighted mean differences which are not normalized by the variance in the reported outcome), we included both of these types of data in our analyses. Some studies only reported the percentage of patients in each group who had at least one visit to an urgent care or hospital. We did not include these data in our standardized mean difference calculations. Physician office visits were reported too heterogeneously to be combined (e.g., emergency visits to physician office for asthma, regular physician visits for asthma, total outpatient physician visits).

Asthma-related emergency department or urgent care visits. Figure 11 presents the individual study and summary standardized mean difference (between intervention and control groups) for mean asthma-related emergency department/urgent care visits per month. The weighted mean reduction was 0.01 visits per month (95% CI: 0.00, 0.02; p=0.17). The standardized mean difference was -0.12 (95% CI: -0.29, 0.05; p=0.16). Additionally, the studies included in this analysis were highly heterogeneous—which can readily be seen from the wide range of reported mean visits across the studies (Figure 11-see below). Visual inspection of the funnel plot of these data (Figure 9) does not suggest significant publication bias (there are studies demonstrating a range of effect size). We conclude that among the self-monitoring, self-management, or patient education studies reporting mean emergency visits due to asthma, we did not find a statistically significant overall effect toward reducing asthma-related visits.

Figure

Figure 11. Self-monitoring, self-management, or patient education studies reporting mean number of emergency department/urgent care visits due to asthma per month. Notes: The effect sizes presented here are the random (more...)

Figure

Figure 9. Funnel plot: asthma-related ED/urgent care visits.

We sought study design and intervention characteristics associated with the greatest reductions in asthma-related emergency department/urgent care visits. For this analysis, we used the standardized mean difference in visits per month as the dependent variable in a weighted least squares regression. Our analysis was limited by the small number of studies reporting this outcome. We did not find any statistically significant predictors of reductions in emergency department/urgent care visits (Table 15).

Table 15

Association of pediatric self-monitoring, self-management, or patient education intervention characteristics and reductions in asthma-related emergency department/urgent care visits.

Asthma-related hospitalizations. Figure 12 (see below) presents the individual study and summary standardized mean difference (between intervention and control groups) for mean hospital days per month due to asthma. The weighted mean difference was a reduction of 0.09 hospital days per month (95% CI: -0.26, +0.08; p=0.3). The standardized mean difference was -0.09 (95% CI: -0.26, 0.08; p=0.3). Additionally, the studies included in this analysis were highly heterogeneous—which can readily be seen from the wide range of reported mean hospital days across the studies (Figure 12-see below). Visual inspection of the funnel plot of these data (Figure 10) also suggests that there may be publication bias affecting this result (most studies report greater reduction in mean hospital days in the intervention group). We conclude that among the self-monitoring, self-management, or patient education studies reporting mean hospital days due to asthma, there is no overall effect toward reducing asthma-related hospitalizations.

Figure

Figure 12. Self-monitoring, self-management, or patient education studies reporting mean number of hospital days due to asthma per month. Notes: The effect sizes presented here are the random effects standardized mean (more...)

Figure

Figure 10. Funnel plot: hospitalizations due to asthma.

We sought study design and intervention characteristics associated with the greatest reductions in asthma-related hospitalizations. For this analysis, we used the standardized mean difference in hospitalizations as the dependent variable in a weighted least squares regression. We found that the greater the number of educational sessions provided, the greater the expected reduction in asthma-related hospitalizations (p=0.01) (Table 16). However, in none of our exploratory analyses was the intensity of the intervention, the type of provider of the education, the setting of the educational intervention, the type of educational materials used, or the number of QI strategies used in the intervention associated with reductions in school absenteeism (not all data shown).

Guideline Adherence Outcomes. The 27 studies reporting guideline adherence outcomes were highly heterogeneous. We performed logistic regression, using any statistically significant guideline adherence outcome as the dependent variable and study/intervention characteristics as the independent variables. In none of our analyses was the study design, the use of a theoretical/conceptual framework, the duration of the intervention, the frequency of interactions with the patient, the type of provider of the education (e.g., nurse, physician), setting of the educational intervention (e.g., home, school), the type of educational materials used associated with improved outcomes (data not shown).

Conclusions. We conclude that self-monitoring, self-management, or patient education interventions may result in improvements in clinical outcomes for children with asthma—particularly interventions that include a component directed at parents or caregivers. Additionally, longer intervention duration and greater numbers of educational sessions may be associated with improvements in functional status. We note that our regression analyses were limited by the small number of studies consistently reporting similar outcomes and the significant heterogeneity among the studies (which likely reflects, in part, the heterogeneity in study populations including key characteristics such as asthma severity and socio-economic status). This relatively low statistical power limited our ability to corroborate some of the univariate associations described in the preceding sections.

General Characteristics of Self-Monitoring, Self-Management, or Patient Education Interventions for General Populations or Adults With Asthma

Background. As for the interventions designed specifically for children, self-monitoring or self-management interventions for general populations with asthma (typically for adults but occasionally mixed populations) generally include written instructions on how to alter medications based on information about disease status as measured either by symptoms or peak flow. It has been demonstrated that patients with asthma, regardless of age, often do not use their inhalers properly—even the reputedly easy-to-use dry powder inhalers have correct usage rates that vary widely (5 to 78%).¹²⁵ Because poor inhaler technique can result in decreased delivery of medication to the lungs, increased asthma symptoms, higher costs, and increased side effects,¹²⁵ many patient education interventions are designed to improve inhaler technique and medication adherence, and to decrease asthma symptoms and prevent asthma related co-morbidities.

Study Design Characteristics. We identified 78 studies of self-monitoring, self-management, or patient education interventions for adults—59 of these included a self-monitoring or self-management intervention and 19 focused more on patient education (Table 18). Twenty-four studies compared two or more intervention groups without a control that did not also receive a QI intervention (presented in Table 21). Most of the patient education interventions for adults included at least some minimal component of providing adults with the information that they need to make decisions about their own asthma care. Thus, the self-monitoring, self-management, or patient education interventions for adults were overlapping educational interventions. Therefore, we present them together, grouping them to the extent possible according to common characteristics.

Table 18a

Summary of nurse-led self-monitoring, self-management, or patient education interventions directed at adults with asthma.

The median duration of the follow up period was 11 months (S.D., 9.2 months). The median sample size of these interventions was 100 subjects (Interquartile range: 51, 187). Sixty-three were RCTs, two were quasi-RCTs, and 13 were controlled before-after trials. In univariate analysis, the longest studies were more likely to report a statistically significant improvement in emergency department visits (p=0.02) but not the other outcomes.

Intervention Characteristics. These educational interventions were highly heterogeneous with respect to the content of the education, the provider of the education, the setting in which the education was provided, and the intensity of the education. Some interventions included multimedia tools, others used one-on-one formal teaching sessions, and several include instructional booklets.

Of the self-monitoring, self-management, or patient education studies, 24 (31%) specifically described an underlying conceptual framework or theoretical background as the basis for the intervention—a considerably lower proportion than for pediatric interventions of this type (62%). The controlled before-after studies were somewhat more likely to report a conceptual or theoretical foundation (38%) than RCTs (27%) (p=0.01).

Compared to the pediatric educational interventions, the educational interventions for general populations or adults were less like to be based in primary care clinics (28% versus 17%) and more likely to be based in outpatient specialty clinics (e.g., allergy or pulmonary clinics) (12% versus 30%) (Figure 5 and Figure 13). Most interventions for general populations or adults were taught by nurses/nurse practitioners or physicians (Figure 13).

Figure

Figure 13. Intervention characteristics: setting and educator type.

Thirty-three (42%) of the self-monitoring, self-management, or patient education interventions provided two to five educational sessions. Twenty-seven interventions principally provided individualized instruction, three principally used interactive group teaching sessions, and 27 used combinations of teaching modalities (Figure 14). Five interventions included a computer/web-based component—the use of these technologies was not associated with statistically significant improvements in outcomes or processes of care for general populations or adults with asthma.

Figure

Figure 14. Number of educational sessions provided by educational strategy.

Outcomes Reported. The 78 self-monitoring, self-management, or patient education studies' most frequently reported outcomes were asthma symptoms, urgent care or emergency department visits and hospitalizations due to asthma, and amount of asthma medications used (Table 17). The 42 studies that reported on asthma symptoms were highly heterogeneous—some presented self reported symptoms from patient diaries whereas others evaluated physicians' ratings of asthma symptoms. Similarly, the studies reporting on medication use sometime reported frequency of inhaled β-agonist use per patient whereas others reported the number of prescriptions for oral corticosteroids for the intervention group over the study interval.

Table 17

Outcomes reported by the adult self-monitoring, self-management, or patient education studies.

Interventions Led Principally by Nurses and Pharmacists. 21 interventions relied primarily on nurses to deliver the self-monitoring or self-management program (Table 18a). (Note: many other studies used a combination of physicians, nurses, and other health providers to work in concert to provide various elements of the educational materials—the interventions described in this section were led exclusively or primarily by nurses.) Six of these compared two or more intervention groups without a control that did not also receive a QI intervention (presented in Table 21).¹²⁶–¹³² Several of the nurse-led interventions lacked key characteristics that would seem predictive of being able to produce and detect statistically significant improvements for patients: only five reported basing their educational intervention on a theoretical or conceptual framework¹³¹–¹³⁶ and seven had sample sizes of 65 subjects or less.

Notable among the interventions in this category for its rigorous design, sample size, duration, intensity of educational program provided, duration, and reporting is the study by Wilson et al.¹³⁶ This RCT with four arms enrolled 235 patients aged 18–50 with moderate to severe asthma from five Northern California Kaiser Medical Centers. Patients enrolled in the small group education program (N=83) met weekly with a nurse-educator for four 90-minute education sessions for instruction in asthma and asthma management. The intervention was designed based on cognitive learning theory and utilized both verbal and printed instruction materials. Patients enrolled in the individual education program (N=81) met weekly with the nurse-educator for three to five 45-minute meetings where the educator chose among 18 instructional modules (covering the same content included in the small-group program) to develop a program tailored to the needs of the individual. Intervention patients kept symptom-medication diaries. The standard control group (N=71) was given no formal education. An information control group (N=75) was given a workbook containing the same basic information given to the intervention groups. At 12 months, patients in both the group and individualized educational programs were significantly more likely than controls to improve in multiple outcomes (p<0.05 for all comparisons) including asthma symptoms (55% and 50%, respectively compared to 25% in control group), symptom free days (49% and 51%, respectively compared to 26% in the control group), and bedroom control practices (62% and 53%, respectively compared to 32% in the control group). (See Table 18a for more details on this study.)

Table 18b presents the six education interventions led principally by pharmacists. All of these studies reported finding statistically significant improvements in processes or outcomes of care for general populations or adults with asthma. As a group, these interventions had relatively more frequent interactions with patients than other educational interventions—for example, the brief (3 month) study by Barbanel and colleagues provided patients with an initial individualized session followed by weekly calls from a pharmacist to review medications for the next three months.¹³⁷ Similarly, the large study (N=413) by Herborg and colleagues required monthly sessions for patients and their pharmacists for a year. Because of the small number of studies of this category, we did not have sufficient power to assess whether the frequency of patient interactions was consistently associated with the positive outcomes reported among the educational interventions led by pharmacists.

Table 18b

Summary of adult self-monitoring, self-management, or patient education interventions provided by pharmacists.

Other General Population or Adult Self-Monitoring, Self-Management, or Patient Education Interventions. The 33 studies in Table 18c evaluated a broad range of interventions. Among these, the factor most consistently associated with improvements in outcomes for patients was the use of combinations of educational techniques and materials including individualized interactions, group interactions, role-playing, and printed materials among others. Twenty-three of the 27 (85%) self-monitoring, self-management, or patient education interventions for adults that utilized combinations of educational modalities found statistically significant improvements in processes and outcomes of care for patients. In contrast, the interventions that utilized principally a single educational modality were less likely to report statistically significant outcomes: 70% (19/27) of the interventions that exclusively used individualized education, 66% (2/3) of interventions that exclusively used lectures, 33% (2/6) of interventions that exclusively used printed materials, and 33% (1/3) of interventions that exclusively used interactive groups found improvements for patients.

Table 18c

Summary of other self-monitoring, self-management, or patient education interventions directed at adults with asthma.

We did not find that the frequency of patient interactions with the provider of the educational intervention was a key factor in producing improvements for patients. Some interventions provided no direct interaction between healthcare providers but still found statistically significant improvements in patient outcomes (e.g., the study by Legorreta (one of the largest in this systematic review) involved 999 patients who received self-monitoring or self-management and educational materials including written materials, peak flow meters, and a video directly from their HMO, and were found to have improvement in daily use of steroid inhalers and PF meters¹³⁸). In contrast, other interventions provided intense interactions with patients but failed to produce or detect improvements for patients (e.g., the 10 patient study by Grover and colleagues in which patients received 15 individualized educational sessions¹³⁹).

Similarly, no particular intervention setting was consistently associated with improvements in outcomes. In contrast, three of the self-monitoring, self-management, or patient education studies were home-based and none found improvements in outcomes for patients.¹⁴⁰–¹⁴²

Three of the included studies^{133, 143, 144} based their self-management intervention on the Wheezers Anonymous program. Developed by Winder and colleagues,^{143, 144} Wheezers Anonymous was based on two effective self-monitoring or self-management programs for children (Living with Asthma¹⁴⁵ and the Family Asthma Program¹⁴⁶) and was designed to provide comprehensive asthma education and self-monitoring or self-management skills with minimal leader training for use in clinical settings (Table 18c). It uses a combination of teaching modalities including lectures, videotaped segments, and discussion sections.¹⁴³ The relatively small study by Kotses and colleagues, found improvements in peak flow measures among recipients of Wheezers Anonymous compared to control subjects.¹⁴⁴ However, the other studies did not report consistent improvement in clinical, functional, or health services utilization measures among recipients of Wheezers Anonymous compared to control subjects.^{133, 143}

Conclusions. Self-monitoring, self-management, or patient education interventions can be effective for improving the processes and outcomes of care for adults with asthma. In particular, interventions that utilize combinations of educational modalities and those that rely on pharmacists to lead the educational effort have been associated with statistically significant improvements for patients.

Synthesis of Evidence From Self-Monitoring, Self-Management, or Patient Education Interventions for General Populations or Adults With Asthma

In this section we present the results of our evaluation of the association of study design characteristics (e.g., whether the authors specified a theoretical/conceptual framework, year of study, sample size, country), intervention characteristics (e.g., the number of educational sessions provided to students, the setting in which the intervention took place), and improvements in the outcomes of interest for the self-monitoring, self-management, or patient education interventions. We present our analyses according to each of the primary outcome types.

Clinical Outcomes. Sufficient numbers of studies presented results for percent predicted FEV₁ and peak flow that these data could be synthesized quantitatively.

Percent predicted FEV ₁ . Among the self-monitoring, self-management, or patient education interventions for general populations or adults with asthma, 17 reported change in percent predicted FEV₁ from which we were able to calculate individual study and a summary standardized mean difference (between intervention and control subjects at the end of the study) (Figure 17-see below). These studies were statistically homogeneous and produced a weighted mean difference of 2.92 percent change in FEV₁ (95% CI: 0.92, 4.92; p=0.004). This corresponds to a summary mean difference of 0.13 (95% CI of 0.03, 0.23) favoring the intervention groups (p=0.01). The funnel plot from these 17 studies (Figure 15) does not suggest substantial publication bias. Of all the quantitative analyses in this report, the data included in this calculation are the most robust given the relatively large number of studies reporting the same outcome in the same way. However, the statistically significant effect reflects a clinical improvement of only borderline significance.

Figure

Figure 17. Self-monitoring, self-management, or patient education studies reporting percent of predicted FEV₁. Note: The effect sizes presented here are the random effects standardized mean difference (SMD).

Figure

Figure 15. Funnel plot: FEV₁ percent predicted.

We sought study design and intervention characteristics associated with the greatest improvements in percent predicted FEV₁. For this analysis, we used the standardized mean difference as the dependent variable in a weighted least squares regression. We found that the more recent the year of publication, the greater the likelihood of finding improvements in percent predicted FEV₁ (p=0.004) (Table 19). We denoted the studies with interventions performed since 2000 in Figure 17 with an asterisk. We cannot determine the critical characteristics that distinguish these more recent studies and that might be associated with the improved spirometric measures.

Table 19

Association of study/intervention characteristics and improvements in FEV₁ percent predicted.

In none of our analyses was the study design, the use of a theoretical/conceptual framework, the duration of the intervention, the frequency of interactions with the patient, the type of provider of the education (e.g., nurse, physician), setting of the educational intervention (e.g., home, school), the type of educational materials used associated with improved outcomes (data not shown).

Peak flow measurements. Among the self-monitoring, self-management, or patient education interventions for general populations or adults, 16 studies reported mean peak flow at the end of the study period. Typically, they reported mean peak flow for the intervention and control groups and did not specify time of day. If they did present mean morning and mean evening measurements, we used mean morning values in our summary analysis because these were the most frequently reported. The weighted mean difference in peak flow from these 16 (statistically heterogeneous) was 27.95 L/min (95% CI: 10.75, 45.15; p=0.01). The standardized mean difference of 0.26 (95% CI of 0.10, 0.42) favored the intervention groups (p=0.001) (Figure 18-see below). As with the FEV₁ results, this represents a change of only modest clinical relevance. The funnel plot (Figure 16) for this analysis does not suggest substantial publication bias.

Figure

Figure 18. Self-monitoring, self-management, or patient education studies reporting mean peak flow. Note: The effect sizes presented here are the random effects standardized mean difference (SMD).

Figure

Figure 16. Funnel plot: mean peak flow.

We sought study design and intervention characteristics associated with the greatest improvements in mean peak flow. For this analysis, we used the standardized mean difference as the dependent variable in a weighted least squares regression. We did not find a combination of the intervention or design characteristics that explained a statistically significant proportion of the variation in mean peak flow (R²=0.29) (Table 20).

Table 20

Association of study/intervention characteristics and improvements in mean peak flow.

Functional Status Outcomes. Nine studies reported mean days lost from work or school. For those studies that did not specifically report the timeframe over which the school absenteeism was recorded (e.g., days absent per month versus per school year); we assumed that the reported days of missed school were for the entire study duration. When interpreting these data, we note that not all authors specifically identified these data as mean days lost from school and some likely represent other parameters such as total days lost from school for the whole cohort. We calculated the standardized mean difference between intervention and control groups in terms of the mean number of days lost from work or school (Figure 19). The nine studies were statistically heterogeneous (I²=73%) and they did not find any statistically significant difference in days lost from work or school between intervention and control subjects (weighted mean difference of -0.19 days absent per month; 95% CI: -0.40, 0.02; p=0.08; standardized mean difference of -0.21; 95% CI: -0.44, 0.02; p=0.08).

Figure

Figure 19. Self-monitoring, self-management, or patient education studies reporting mean days lost from work or school. Note: The effect sizes presented here are the random effects standardized mean difference (SMD).

We performed logistic regression, using any statistically significant functional status outcome as the dependent variable and study/intervention characteristics as the independent variables. In none of our analyses was the study design, year of intervention, the use of a theoretical/conceptual framework, the duration of the intervention, the frequency of interactions with the patient, the type of provider of the education (e.g., nurse, physician), setting of the educational intervention (e.g., home, school), or the type of educational materials used associated with improved outcomes in any of our analyses (data not shown).

Health Services Utilization Outcomes. Twenty-one studies reported urgent care/emergency department visit data in such a way that they could be combined quantitatively (Figure 20). These studies were highly heterogeneous (I²=98.9%). They did not find a statistically significant difference in urgent care/emergency department visits between intervention and control subjects (weighted mean difference -0.23 visits per month; 95% CI: -0.64, 0.18; p=0.26; standardized mean difference -0.48; 95% CI: -1.11, 0.14; p=0.13).

Figure

Figure 20. Self-monitoring, self-management, or patient education studies reporting urgent care/emergency department visits. Note: The effect sizes presented here are the random effects standardized mean difference (more...)

Similarly, 24 studies reported hospitalization rates (Figure 21); however, these were also highly heterogeneous (I²=99.4%) and they did not find a statistically significant difference in hospitalizations between intervention and control subjects (weighted mean difference -0.34 hospital days per month; 95% CI: -0.99, 0.31; p=0.3; standardized mean difference -0.58; 95% CI of -1.53, 0.37; p=0.23).

Figure

Figure 21. Self-monitoring, self-management, or patient education studies reporting hospitalizations. Note: The effect sizes presented here are the random effects standardized mean difference (SMD).

We performed logistic regression, using any statistically significant health services utilization outcome as the dependent variable and study/intervention characteristics as the independent variables. We did not find study design, intervention, or patient characteristics that were associated with improved outcomes in any of our analyses (data not shown).

Guideline Adherence Outcomes. The 40 studies reporting on guideline adherence measures were heterogeneous in terms of the specific outcomes evaluated. We performed logistic regression, using any statistically significant guideline adherence outcome as the dependent variable and study/intervention characteristics as the independent variables. We did not find study design, intervention, or patient characteristics that were associated with improved outcomes in any of our analyses (data not shown).

Conclusions. Self-monitoring, self-management, or patient education interventions for general populations or adults with asthma can result in improvements in FEV₁ and peak flow. However, these improvements are of borderline clinical importance. Our analyses suggest that more recent studies may result in greater improvements in these spirometric measures. Whether year of the intervention reflects other key covariates that could affect spirometric outcomes, is likely.

Self-Monitoring, Self-Management, or Patient Education Trials Comparing Two Intervention Groups Without a Control Group

Background. Several of the included studies compared two or more intervention groups without including a control group that did not also receive a QI intervention. For example, if a study provided a 10-session asthma patient education program to the experimental group and a two-session asthma education program to the “control group,” we considered this to be a trial comparing two intervention groups (sometime called a study with an “active control group”). In this section, we present the results of these studies.

Results. We found 35 trials that compared two or more intervention groups without a control group that did not also receive a QI intervention (Table 21). They all included self-monitoring, self-management, or patient education interventions. Typically, these comparative studies were designed such that one group received a less intense version of what the second group received. Perhaps not surprisingly, these studies generally found improvements in both groups over baseline. Thirteen of these 35 (37%) also found differences between the two groups.

Pediatric interventions. Eleven studies focused exclusively on children and all of these were either self-monitoring, self-management, or patient education interventions and two also included organizational change components.^{57, 64} Five studies found statistically significant differences between interventions groups. ^{55, 56, 62, 64, 197} Notable among these was the study of a patient education intervention called A.C.T. (Asthma Care Training) for Kids, an intensive educational program for children with severe asthma whose content is based on programs with demonstrated effectiveness, that was compared to a less intense lecture and discussion program.⁶² The article by Lewis et al. describes the conceptual rationale for the A.C.T. program and its contents in more detail than most of the included articles in this review.⁶² Briefly, the program emphasizes that children can control their disease rather than being controlled by it. Using the analogy of driving safely, medications and other asthma prevention and treatment techniques are color-coded: green for daily medications used to “keep going and prevent symptoms,” yellow for “caution” to be used when mild symptoms develop, and red to “stop” the disease when severe symptoms occur.⁶² Participants are taught about underlying asthma pathophysiology and symptomatology, reducing asthma triggers, relaxation skills and breathing techniques, medication use, and decision making skills. They found that the A.C.T. recipients had statistically significantly fewer emergency department visits (2.3 versus 3.7, p<0.001) compared with traditional patient education recipients and that, although there were no differences in terms of number of hospitalizations, the A.C.T. recipients tended to have shorter hospital stays (0.67 versus 1.54 days/child/yr, p<0.001).⁶² The costs of the program were estimated at $125 per A.C.T. recipient and $37.50 per traditional patient education recipient. Given the lower emergency department and hospital costs in the A.C.T. group, the authors estimated a net incremental savings of $180 per child per year for the sponsoring institution.⁶²

We did not find that increasing either the frequency or the duration of contact between patients and health educators was consistently associated with improvement in outcomes. Specifically, five of nine studies compared two (or more) interventions of differing interventional frequency or duration and found no statistically significant improvements in any outcomes. There were no other distinguishing characteristics of the interventions that were associated with improved outcomes.

General population or adult interventions. We found 24 studies that compared two or more QI interventions without a control group that did not also receive a QI intervention for general/adult populations. Interestingly, only two of these were U.S.-based interventions.^{198, 199} Eight of sixteen studies reported significantly greater improvements in the intensive treatment arm.^{127, 128, 200}–²⁰⁷ For example, the study by McLean et al.²⁰⁵ compared a lower intensity pharmacist-provided asthma education/management program to a more intense pharmacist-provided asthma education program that emphasized asthma action plans, the use of peak flow meters, and other self-monitoring or self-management techniques. This study found that the intensive education group experienced significantly greater improvement than the less intense intervention: 11% improvement in peak flows (p=0.0002), 50% overall decrease in asthma symptoms (p<0.01 for most symptoms evaluated), improvement in QOL (<0.05), and decrease in medical visits (but not emergency department visits or hospital visits) between groups.²⁰⁵ The various interventions presented in these studies were heterogeneous with respect to material presented, type of educator, and frequency of interaction with the patients—thus, no particular intervention characteristics were associated with the most successful programs.

Conclusions. Among the trials comparing multiple QI interventions without a control group that did not also receive a QI intervention, we found that, in general, all intervention groups tended to improve from baseline, but only 37% of studies of this type reported statistically significant differences between groups in the outcomes evaluated. The heterogeneity of studies of this type limits conclusions about commonalities of interventions associated with clinically meaningful improvements.

Organizational Change Interventions for Children With Asthma

Background. The studies of organizational change interventions designed to improve the processes and outcomes of care for children with asthma fall broadly into two categories: those that augment the care provided in general pediatrics clinics and those that provide an increased level of care in schools. Whereas the interventions described in the section on patient education (Table 12b) were asthma education programs in schools that often just taught children about the disease and its management, the school-based interventions described in this section all added increased levels of asthma care provided in the school setting. Asthma in children affects schools in several ways: Ongoing disease management typically happens while children are in school and life-threatening acute asthma symptoms can occur at school. Asthma is a leading cause of illness-related school absenteeism and this absenteeism often necessitates “make up” school work for these children and their teachers. Thus, schools are a natural site for asthma-based education and management programs.

Results. We found 13 studies of organizational change strategies designed specifically for children with asthma (Table 24).^{108, 115, 117, 119, 124, 239}–²⁴¹ Two of these studies compared two or more intervention groups without a control that did not also receive a QI intervention and are presented with studies of that design (Table 21). As a group of interventions, these were relatively heterogeneous and not as likely as other types of QI interventions to report improvements in outcomes for patients: three studies reported improvements in clinical outcomes, three studies reported improvements in health services utilization, and three studies reported improvements in functional status. They were more likely to report improvements in terms of the number of asthma patients receiving inhaled controller medications.

Table 24

Organizational change interventions for children with asthma.

Asthma specialty clinics/care. Nine studies compared usual asthma care to asthma care augmented by providers with special training (i.e., nurses, ^{108, 117, 118, 225}–^{227, 235} pharmacists,^{115, 241} psychologists and physiotherapists).¹²⁴

The Pediatric Asthma Care PORT (Patient Outcomes Research Team), a RCT designed to assess the effectiveness of three strategies (physician-peer leaders (PLE), peer leaders in combination with asthma nurse visits (PACI), and usual care) to implement guidelines for childhood asthma found that, at the end of the two year follow up period, compared with the usual-care arm (N=199), patients randomized to the PLE arm (N=226) had annual increase of 6.5 symptom-free days (SFD), whereas those randomized to the PACI arm (N=213) had an annual increase of 13.3 SFD over usual care patients.^{225, 226} The average number of physician visits during the two year trial was higher in the PACI arm as was regular use of inhaled controller medications (not found for PLE patients).^{226, 227} Total treatment and intervention costs per year per patient were $1,292 for PACI, $504 for PLE, and $385 for usual care. The incremental cost-effectiveness compared to usual care was$18/SFD for PLE and $69/SFD for PACI. ²²⁵

The RCT by Kelly and colleagues of low income children in Virginia who were randomized to receive comprehensive education and management by a physician allergist with monthly phone follow up by an outreach nurse found that children in the intervention group were less likely to visit an emergency department or require hospitalization than control children (95% versus 23%; p<0.001).¹¹⁹ The rate of smoking was high in both the intervention and control households (47% and 50%, respectively). Subgroup analysis suggested a trend toward greater effectiveness of the intervention to decrease hospitalizations among children residing in smoke-free households.¹¹⁹

School-based directly-observed therapy programs. Two of the included studies evaluated the use of directly observed therapy for children with asthma. Long-term asthma control therapies work best when taken on a consistent basis, but compliance with this treatment is a known problem—schools provide an opportunity for directly-observed therapy among pediatric patients. Anderson and colleagues compared outcomes among asthma patients attending the Kunsberg school in Denver (designed specifically for children with chronic health conditions) to outcomes among a matched group of children with asthma attending regular schools.²³⁹ The Kunsberg school had a nurse administrator, two nurses, a social worker, and teachers familiar with methods for managing children with chronic medical conditions who provided a range of patient and parent asthma educational interventions, case management including communication with primary care providers, and daily directly observed medical therapy. At the Kunsberg school, nurses reported that 89% of children with asthma received their inhaled corticosteroids on a daily monitored basis.²³⁹ Compared with controls, children enrolled in the Kunsberg school had fewer hospitalizations, ED visits, and follow up visits for their asthma.²³⁹ A survey of school nurse and parents indicated that improved medication compliance resulted from directly-observed therapy and an overall increased structure of care resulted in better outcomes.²³⁹ Given the resources available to children in this type of school, it may be difficult to generalize their results to other school settings.

Another study, a RCT by Halterman and colleagues specifically evaluated the role of directly-observed therapy.²⁴⁰ Their study population of children aged 3 to 7 years with asthma from urban and primarily minority and low-income demographic groups received their daily dose of inhaled corticosteroids from the school nurse. The control group continued to receive their inhalers at home. Overall, children in the intervention group had a reported 84% compliance rate with inhaled corticosteroids versus 63% among the control group. Children in the intervention group missed significantly fewer days of school and had more symptom-free days than did children in the control group. In a post-hoc analysis, it was found that all statistically significant improvements that were observed in the study were found in those patients not exposed to second-hand smoke. (Overall, 44% of children with asthma enrolled in the study lived in a home with at least one smoker.)

Conclusions. From the few studies of this type, we conclude that augmenting usual asthma care with additional specialty asthma clinics staffed by asthma-trained nurses and pharmacists can be effective. Directly-observed therapy may increase the rate of inhaled corticosteroid use and improve clinical outcomes among school children with asthma, particularly those who are not exposed to second hand smoke.

Organizational Change Interventions for Adults With Asthma

Background. Most of the QI strategies that involve organizational change are designed to create new opportunities for patients to receive comprehensive education and monitoring of symptoms—often by creating asthma specialty clinics in the general practice setting or by creating asthma-specific clinics in other settings (e.g., pharmacies, specialists' offices). For example, recognizing that most patients receive their asthma care in primary care settings, the development of asthma clinics within the general practices of the United Kingdom has been encouraged (a 1993 national survey found that 77% of general practices in the U.K. ran asthma clinics).²⁴³

Results. We found 14 studies of organizational change strategies for general populations of adults with asthma (Table 25). Three of these studies compared two or more intervention groups without a control that did not also receive a QI intervention and are presented with studies of that design (Table 21).

In contrast to the organizational change interventions for children with asthma, the studies of organization change interventions for general populations or adults with asthma reported considerably more improvements in clinical outcomes. We cannot clearly identify the distinguishing characteristics between the organizational change interventions for pediatric versus adult populations; however, the organizational change studies in this section were more likely to augment providers' roles (e.g., adding a teaching role to pharmacists already providing routine pharmacy care for patients) or augment the types of providers encountered while receiving “usual care” (e.g., by adding multidisciplinary teams to routine clinical practice). This is in contrast to the addition of specialty care clinics where patients receive care that is distinctly separate from their routine health care encounters (these specialty clinics were more common for children).

For example, none of the three articles that described nurse-run asthma clinics within general care practices found statistically significant improvements in processes or outcomes of care for asthma patients.^{155, 158, 244} However, several of the other organizational change interventions did result in improvements in the processes and outcomes of care for patients with asthma. For example, the ambitious intervention described by Cordina and colleagues provided training for pharmacists in Malta to assume a greater role in patient education, monitoring of asthma status (including monthly review of patients' medication use, peak flow records, and symptoms), and to recommend treatment changes to patients' physicians (as needed).¹⁵⁹ This represented “a major change in conventional practice for Maltese pharmacists” and resulted in a lower rate of self-reported hospitalizations (no patients in the intervention group versus eight patients in the control group) and nighttime wheezing (20% versus 36%).

Two interventions associated with improvements in clinical outcomes for patients utilized computer tools to facilitate asthma management.^{180, 190} One such intervention, described by Johnson and colleagues,¹⁸⁰ was a comprehensive asthma disease management program in which subjects received asthma education and providers and case managers received computer generated communications regarding whether the patient had an action plan, received influenza vaccinations, had rescue inhalers, and reported on patients' use of daily controller medications. This study reported that 12 months after the intervention, the intervention group (196 patients) had fewer emergency department visits (118 versus 305, p<0.0001) and hospitalizations (39 versus 114, p<0.0001) than the control group (196 subjects) albeit no differences in use of asthma medications or preventative vaccinations.¹⁸⁰

Three of the studies of organizational change interventions reported smoking rates for participants—in general, these rates were relatively high and recalcitrant to change. For example, in the 1994 study of the implementation of the Scottish health board's asthma education and treatment protocol in general practices, 20–25% of asthma patients studied reported smoking at baseline and there was no statistically significant reduction in smoking as a result of the program.²⁴³ Similarly, the intervention reported by Garrett and colleagues of patients (principally from lower socioeconomic groups) receiving care in a South Auckland asthma community education center found even higher baseline smoking rates (33%–34%), which did not change after nine months enrollment in the education center.^{152, 153}

Conclusions. Interventions that augment the “routine care” being provided to patients both in their physicians' offices and in pharmacies were associated with improved outcomes for patients. For example, technologies that enhance communication of patient information to providers can result in statistically significant improvements in outcomes of care for patients.