A quantile-quantile plot of percentage peak expiratory flow rate (PEFR) change at 2 to 3 weeks following acute asthma care is shown. The line represents the expected normal distribution; therefore, the PEFR change is distributed approximately normal.
Coyle YM, Aragaki CC, Hynan LS, Gruchalla RS, Khan DA. Effectiveness of Acute Asthma Care Among Inner-city Adults. Arch Intern Med. 2003;163(13):1591-1596. doi:10.1001/archinte.163.13.1591
Copyright 2003 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2003
Acute asthma often requires expensive emergency department visits and hospitalizations, especially among economically disadvantaged inner-city adults. However, few studies have examined approaches for improving acute asthma care in this population.
We conducted a cohort study involving patients who were discharged from a public hospital emergency department following acute asthma care between March 31, 1997, and August 5, 1999, to identify processes of care effective for improving peak expiratory flow rate at a 2- to 3-week follow-up. Adult patients who met the predetermined criteria for asthma, who underwent a baseline peak expiratory flow rate reading, and who did not have concurrent acute sinusitis or pneumonia were eligible (N = 448). Of the 365 patients enrolled in the study, 309 (84.7%) completed it. We used a multiple linear regression analysis adjusted for patient risk to assess the association between acute asthma care processes derived from the National Asthma Education Prevention Program guidelines (inhaled β2-agonists, inhaled corticosteroids, systemic corticosteroids, asthma care follow-up, and patient asthma education) and percentage peak expiratory flow rate change at follow-up.
Systemic corticosteroids had a significant effect for increasing percentage peak expiratory flow rate change at the 2- to 3-week follow-up for all asthma exacerbation severity levels (β = 26.1; 95% confidence interval, 1.8-50.5; P = .04) and severity levels specified by the National Asthma Education Prevention Program guidelines (β = 31.6; 95% confidence interval, 8.1-55.1; P = .01).
Outpatient systemic corticosteroids were effective for improving lung function 2 to 3 weeks after acute asthma care, and their use should reduce asthma-related morbidity, especially among economically disadvantaged inner-city adults.
ACUTE ASTHMA often requires expensive emergency department (ED) visits and hospitalizations, especially in economically disadvantaged inner-city adults.1- 4 Many persons with asthma maintain a poor quality of life and are prone to relapse for weeks following acute asthma treatment that is often variable.5- 7 However, although there is substantial evidence that excessive asthma-related morbidity occurs in adults8 and correlates with poverty and urban residence,9 few studies have examined approaches for improving acute asthma care in this population.
Lung function testing is regarded as the best measure for assessing asthma-related morbidity.10 To our knowledge, there are no published studies that fully characterize the processes of care effective for improving lung function after the first 24 hours of treatment for an acute asthma exacerbation.
We conducted a cohort study involving patients who were discharged from the Parkland Health & Hospital System, Dallas (Parkland), ED following usual care for acute asthma between March 31, 1997, and August 5, 1999, to identify processes of care effective for improving lung function at a 2- to 3-week follow-up. Parkland operates a public hospital serving predominantly racial and ethnic minority adults who are indigent and reside in the inner city. Processes of asthma care that were begun following treatment for acute asthma in the ED were selected for testing based on the National Asthma Education Prevention Program (NAEPP) guidelines (inhaled β2-agonists, inhaled corticosteroids, systemic corticosteroids, asthma care follow-up, and patient asthma medication and inhaler use education).11 The NAEPP guidelines provide the most comprehensive recommendations on the likely processes that are essential to acute asthma care based on scientific evidence and expert opinion.11
In addition, because asthma is a medical condition in which the patient's environmental context must be taken into account to effectively evaluate the outcomes of care, the study examined known patient risk factors (age, sex, ethnicity, income, asthma severity, smoking status, upper respiratory tract infection in the past month, asthma knowledge, asthma medication adherence, and indoor allergen and ozone exposure)3- 5,8,12- 18 to promote the linkage between the process of care for acute asthma and lung function.19,20
The prospective cohort for the study involved adults who were discharged from the Parkland ED after usual care for an acute asthma exacerbation between March 31, 1997, and August 5, 1999, to identify the processes of care effective for improving lung function, measured as peak expiratory flow rate (PEFR), at a 2- to 3-week follow-up. The 2- to 3-week follow-up took place at The University of Texas Southwestern Medical Center at Dallas General Clinical Research Center outpatient clinic. Study patients participated in the study once during the study period, related to the first ED visit for acute asthma care at which they were able to meet the study's inclusion criteria. The study cohort averaged 2.2 ED visits per year for acute asthma care during the study period.
The University of Texas Southwestern Medical Center at Dallas institutional review board approved the study. Trained research nurses used a written protocol to identify eligible study patients, perform clinical testing, administer the study's data collection instruments, and assign values to selected measures based on the data collected. Eligible study patients were identified by using the Parkland administrative database and reviewing the paper medical record. The study's data collection involved administering patient questionnaires in English and Spanish, performing clinical testing (allergy skin and PEFR testing) at the General Clinical Research Center outpatient clinic, and reviewing hospital and City of Dallas Ambient Air Quality Surveillance records.
Patient risk related to PEFR change was derived from patient questionnaire data, clinical testing results, and review of the City of Dallas Ambient Air Quality Surveillance records. The appropriate use of the processes of acute asthma care related to the ED visit of interest was based on hospital record (paper medical record and pharmacy database) review supplemented with patient questionnaire data.
By using the Parkland administrative database, we selected all adult patients (>18 years) who were discharged from the Parkland ED with the primary diagnosis of asthma (International Classification of Diseases, Ninth Revision, Clinical Modification codes 493.0, 493.1, 493.2, and 493.9) for a medical record review. Study inclusion criteria were as follows: (1) a physician diagnosis of an acute asthma exacerbation, (2) a clinical diagnosis of asthma, and (3) having undergone a baseline PEFR recording as part of usual care in the Parkland ED. The criteria for a clinical diagnosis of asthma were as follows: (1) having a physician diagnosis of asthma and (2) having an increase in the forced expiratory volume in 1 second or PEFR of 12% or greater after the administration of a β2-agonist in the last year documented in the paper medical record.11,21 This documentation included serial PEFRs obtained during acute asthma care. Exclusion criteria were as follows: (1) a clinical diagnosis of acute sinusitis or (2) pneumonia diagnosed by a chest x-ray film concurrent with the asthma exacerbation being assessed. These 2 comorbid states can delay asthma exacerbation recovery and require specific treatment not being tested in this study. English- and Spanish-speaking patients were enrolled in the study by a combination of telephone and mail contact. Patients were offered a $50 financial incentive for completing the study.
The predictor variables for the study included the potential patient risk factors (age, sex, ethnicity, income, asthma severity, ozone exposure, indoor allergy and exposure, smoking [active and passive], upper respiratory tract infection in the past month, lower asthma knowledge, and asthma medication nonadherence) that could be associated with the outcome variable PEFR and the treatment variables or the acute asthma processes of care.
Six processes of care were selected based on the NAEPP guidelines for the treatment of an acute asthma exacerbation.11 We classified the severity of the asthma exacerbation as mild, moderate, or severe based on the percentage PEFR predicted (>80% indicates mild; 50%-80%, moderate; and <50%, severe).11 Following a written protocol, one of the study's research nurses, blinded to patient outcome, used hospital record data (includes paper medical record and pharmacy database data) and patient questionnaire data to assess the appropriateness of the 6 processes of care. The processes of care were as follows: (1) inhaled β2-agonists at ED discharge, (2) inhaled corticosteroids at ED discharge for patients with mild asthma exacerbations who are taking inhaled corticosteroids, (3) systemic corticosteroids at ED discharge for moderate and severe asthma exacerbations or mild asthma exacerbations treated with systemic corticosteroids within the past month, (4) an asthma care follow-up visit within 5 days of ED discharge, (5) patient education on inhaler device use at ED discharge, and (6) patient education on asthma medication use (review medication list and indications for use) at ED discharge. We used a patient questionnaire to assess the appropriateness of patient asthma education (processes of care 5 and 6 previously described), because the paper medical record did not specify the type of asthma education given. Content for this questionnaire was derived from the NAEPP guidelines for patient education before discharge from the ED after acute asthma care.11 The patient questions used to obtain data for processes of care 5 and 6 were as follows: (1) Did you receive instructions on how to use your asthma medication inhaler when you were treated for an asthma attack (an episode of progressively worsening asthma symptoms, such as a troublesome cough, shortness of breath, chest tightness, or wheezing) in the Parkland ED on (date)? and (2) Did you receive instructions on how to follow a written asthma medication plan when you were treated for an asthma attack in the Parkland ED on (date)?
The percentage predicted PEFR was obtained on presentation for acute asthma care in the ED (zero time) and at the 2- to 3-week follow-up visit as the best of 3 determinations.21 A morning PEFR was obtained at the 2- to 3-week follow-up, which can be considered as an estimate of baseline lung function when a forced expiratory volume in 1 second is not available.10 Study subjects were advised to avoid the use of their short-acting β2-agonist for at least 5 to 6 hours and their long-acting β2-agonist for at least 11 to 12 hours before the 2- to 3-week follow-up PEFR testing. The change in the PEFR was defined as the percentage change between the 2- and 3-week follow-up and the zero-time PEFR.
Asthma severity was assessed as mild intermittent, mild persistent, moderate persistent, and severe persistent using clinical features of severity. Clinical features of severity, derived from the NAEPP guidelines,11 were obtained by patient questionnaire at the 2- to 3-week follow-up. This questionnaire asked patients the following about their asthma during the past 3 months. (1) How often have you had asthma symptoms (troublesome cough, shortness of breath, chest tightness, and wheezing)? Possible answers were less than once a week, 1 or more times a week but less than once a day, for some period every day, or all of the time (synonymous with continuous asthma symptoms). (2) What best describes your asthma attacks (episodes of progressively worsening asthma symptoms)? Possible answers were as follows: they last a few hours to a few days, they sometimes affect activity and sleep, they always affect activity and sleep, or they are present a lot of the time. (3) Have your physical activities been limited by asthma symptoms? Possible answers were yes and no. (4) Have you had nighttime asthma symptoms? Possible answers were yes and no. Those answering yes were asked the following: How often have you had nighttime asthma symptoms? Possible answers were less than or equal to 2 times a month, more than 2 times a month, more than once a week, or too many times to remember (synonymous with frequent nighttime asthma symptoms).
By using a written protocol, one of the study's research nurses, blinded to patient PEFR outcome, performed the asthma severity classification using a process outlined by the National Institutes of Health in a 1996 report.22 The follow-up to this report (1997 NAEPP guidelines)11 had minor differences in the frequency for some of the clinical features required for placement in the severity categories compared with the 1996 National Institutes of Health report,22 which was not available until after the study was initiated. The presence of a clinical feature of severity was sufficient to place study patients in a severity category. Study patients were assigned to the highest severity category in which a clinical severity feature occurred. While we considered medication use during the same period that we assessed asthma severity, it did not influence the asthma severity class assigned based on the clinical features of severity.
Asthma knowledge was assessed by using an instrument developed by the National Heart, Lung, and Blood Institute.14 This questionnaire was self-administered and consisted of 12 true and false statements that assessed general knowledge of asthma. Based on data from the present study, the reliability (Cronbach α, .45) of this questionnaire was moderate in English and Spanish for the study population (n = 309), with a mean score of 81% (SD, ±13%).
Asthma medication adherence was assessed with an adapted version of a 4-item self-administered medication adherence questionnaire that was valid and reliable.23 Based on data from the present study, the reliability (Cronbach α, .73) of our adapted version was moderately high in English and Spanish for the study population (n = 309), with a mean score of 60% (SD, ±36%).
Study subjects were tested with indoor allergens (standardized mites [Dermatophagoides pteronyssinus and Dermatophagoides farinae], cockroach mix [American and German], mold mix [Alternaria, Hormodendrum, Phoma, Helminthosporium, Aspergillus mix, Penicillium mix, Fusarium, Rhizopus, Mucor, and Pullularia], cat pelt, and dog hair and dander [mixed breeds]) using the skin prick technique.24 Sensitivity to indoor allergens and exposure was determined to be present if study patients had a positive indoor allergen skin test result and reported exposure to the indoor allergen at some time before the 2- to 3-week follow-up using the NAEPP's assessment questions for environmental factors that worsen asthma.11 Because the study was conducted in a semiarid region, we assumed exposure to house dust mites.11
Ozone exposure was assessed by obtaining daily ozone levels from 2 air-monitoring sites located nearest the study subject's residence. Ozone exposure was calculated as the average of the daily ozone levels at both of the air-monitoring sites between the time of the acute asthma care and the 2- to 3-week follow-up.
We effect-coded categories for each of the potential patient risk factors and examined their association with lower percentage PEFR change using a multiple linear regression analysis. Subsequently, we used a multiple linear regression analysis to determine whether the appropriate use of the processes of acute asthma care, adjusted for the patient risk factors identified in the preliminary analysis previously mentioned, predicted higher percentage PEFR change at the 2- to 3-week follow-up.
One thousand sixty-six adult patients were discharged from the Parkland ED with the primary diagnosis of asthma based on review of International Classification of Diseases, Ninth Revision, Clinical Modification codes in the Parkland administrative database. On review of the paper medical record, it was found that 470 of these patients met the study's inclusion criteria. Twenty-two of these patients were excluded from the study because acute sinusitis or pneumonia was diagnosed by a chest x-ray film, concurrent with the asthma exacerbation being assessed. Thus, 448 patients were eligible for the study. Three hundred sixty-five (81.5%) of the eligible study patients were enrolled in the study by a combination of telephone and mail contact. Three hundred nine (84.7%) of the study patients returned for the 2- to 3-week follow-up. Patients eligible but not enrolled in the study (n = 83) and those who did not return for the 2- to 3-week study follow-up (n = 56) were similar to the study patients in terms of sex. However, the eligible study patient population had more white persons and fewer African American and Hispanic persons than the study population (χ23 = 9.42, P = .02), and they were significantly older (39 years) than the study population (37 years) (t1340 = 2.18, P = .03).
The range of the percentage PEFR change for this study population was from −165 to 590 (median, 80; semi-interquartile range, 77.5). The middle 80% of the data lay between −60 (10th percentile) and 240 (90th percentile). This distribution was slightly positively skewed (0.46), with an outlier at 590. The quantile-quantile normal plot for the percentage change in PEFR at the 2- to 3-week follow-up is provided in Figure 1.
Table 1 presents the association of selected potential patient risk factors (age, sex, ethnicity, income, asthma severity, ozone exposure, indoor allergy and exposure, smoking status, upper respiratory tract infection in the past month, lower asthma knowledge, and medication nonadherence) with percentage PEFR change and the associated P values. The only patient risk factors that were associated with an increase in percentage PEFR change were sex, ethnicity, and asthma severity (Table 1).
The study population's asthma severity for the 4 different levels of severity was as follows: mild intermittent, 4.8%; mild persistent, 33.6%; moderate persistent, 32.7%; and severe persistent, 28.9%.
Adjusting for patient sex, ethnicity, and asthma severity, only the appropriate use of systemic corticosteroids at ED discharge had a significant effect for increasing percentage PEFR at the 2- to 3-week follow-up, as noted in Table 2. However, the appropriate use of all the processes of acute asthma care was positively associated with an increase in the percentage PEFR. In addition, because ethnicity was an important patient risk factor related to the other category only (n = 9), we reran the analyses, removing the 9 study patients from the other ethnicity category, which did not change the results.
When we stratified for asthma severity (mild intermittent and mild persistent, and moderate and severe persistent) in the multiple linear regression analysis, none of the processes of care are significant for increasing percentage PEFR at the 2- to 3-week follow-up (Table 2).
Only 4 study subjects were taking oral corticosteroids when the PEFR was obtained at the 2- to 3-week follow-up. Removing these 4 study patients from the regression analysis did not change the results previously presented.
Because of small numbers for the appropriate and inappropriate use of inhaled corticosteroids, with restriction to the mild exacerbation category as required by the NAEPP guidelines,11 we reran the analyses for inhaled corticosteroids at ED discharge without restriction to specific asthma exacerbation severity category. These results indicated that inhaled corticosteroids at ED discharge (used appropriately in 6 patients and inappropriately in 303 patients) again were not significant for increasing PEFR change (β = −6.2; 95% confidence interval, −30.0 to 17.6; P = .61). However, when we reran the analysis for a course of oral corticosteroids at ED discharge (used appropriately in 173 patients and inappropriately in 136 patients) without restriction to asthma exacerbation severity, as required by the NAEPP guidelines,11 it again was significant for increasing PEFR change (β = 26.1; 95% confidence interval, 1.8-50.5; P = .04).
The goal of the study was to identify processes of acute asthma care, derived from the NAEPP guidelines, that were effective for improving lung function in inner-city adults 2 to 3 weeks after discharge from the ED. Overall, the study population was economically disadvantaged and primarily African American and Hispanic (Table 1). Thus, the study's results offer guidance for improving acute asthma care in this population, an area of clinical practice in which little guidance is available.
The study supports the use of systemic corticosteroids for the treatment of adult outpatients released from an acute-care setting for all levels of asthma exacerbation severity (Table 2). This is consistent with past research,25 which shows that the use of systemic corticosteroids 24 hours after the treatment of an asthma exacerbation prevented relapse in children and adults and improved lung function in children. What is disturbing is that although the NAEPP guidelines exist, adherence to them has been less than adequate for prescribing a short course of systemic corticosteroids in patients following their treatment for acute asthma, and adherence varies among some ethnic minorities.26,27 In a survey by Emerman and colleagues26 in 1996, 75% of 800 randomly selected emergency physicians reported that they prescribed outpatient systemic corticosteroids in concordance with the NAEPP guidelines. McDermott and colleagues27 explored physician adherence in treating Latino and African American patients with asthma who were attending a public hospital ED. Their findings were that Latino patients received prednisone, when indicated, less often compared with African American patients.27 The results of our study and of these previous studies indicate a need to develop physician education interventions that promote the use of outpatient systemic corticosteroids in patients following their treatment for acute asthma.
Other studies28,29 have shown that patient asthma education during and immediately following acute asthma treatment, and short-term follow-up asthma care following acute asthma treatment, improves the clinical outcome of asthma; however, this improvement is only in terms of health care use.30 Although our study did find these interventions positively associated with improved lung function, these associations were not statistically significant for improving lung function 24 hours after acute asthma care. Therefore, our study indicates a need for future research that seeks to determine whether these interventions are effective for improving lung function after acute asthma care, especially among populations at high risk for asthma-related morbidity.
It is possible that the study's other 5 processes of care (Table 2), although positively associated with increasing lung function 2 to 3 weeks following acute asthma care, were not statistically significant for this effect, because measures for the study's predictor and outcome variables threatened its construct validity.31 The predictor variable measures most likely involved were indoor allergy exposure, the processes of acute asthma care related to patient education measures, and the outcome measure PEFR.
The study's indoor allergen exposure measure was based on patient report of home characteristics for pet, cockroach, and mold allergen and the assumption that those who were allergic to dust mites had dust mite exposure, because the study was conducted in a semiarid region.11 The NAEPP recommends that dust mite allergen exposure can be assumed for those who are allergic to dust mites and live in semiarid regions, especially when the measurement of indoor allergen levels is cost prohibitive or impractical.11 Therefore, the most likely threat to the validity of the indoor allergy exposure measure was the approach used to assess dust mite exposure, because Chew and colleagues32 reported, after our study was completed, that patient-reported home characteristics significantly predicted levels of indoor allergens (cat, dust mite, and cockroach) in the home, especially at the high and intermediate levels.
The measures for the processes of acute asthma care related to patient education lacked information on the extent and quality of the education given, which was beyond the scope of the study to include. Therefore, future research that can control for the quality of this education will be required to definitively determine its value in the treatment of acute asthma for primarily ethnic minority and economically disadvantaged populations.
The PEFR was selected as the outcome for this observational study because the PEFR is part of usual acute asthma care in our hospital's ED. As a result, measurements of PEFR in the ED could have threatened the construct validity of the study for 2 reasons. First, a morning PEFR reading10 was not performed at the baseline assessment in the ED as it was at the 2- to 3-week follow-up. As previously mentioned, the PEFR may be considered as an estimate of baseline lung function when a forced expiratory volume in 1 second is not available.10 Second, the baseline assessment for the PEFR may have been a source of technical variation for assessing PEFR change, because several observers, although trained to perform PEFR testing, did so in a busy ED.21 Interrater reliability testing to estimate the extent of this bias was not performed in this study.
Another potential limitation of the study is its possible lack of generalizability, because we excluded patients who had acute sinusitis or pneumonia concurrent with their acute asthma exacerbation being assessed (n = 22). In addition, our study was conducted in a predominantly economically disadvantaged ethnic minority population (Table 1). Therefore, although the purpose of the study was to identify approaches for improving acute asthma care in a primarily disadvantaged ethnic minority population, it is unknown if the study's results would have been similar if it had been conducted in patient populations with different ethnic and socioeconomic backgrounds.
The study indicates that outpatient systemic corticosteroids, regardless of asthma exacerbation severity, are an effective intervention for improving lung function in adults greater than 24 hours after acute asthma care. Therefore, the consistent use of systemic corticosteroids following acute asthma care would be expected to substantially reduce asthma-related morbidity and the related health care costs. Furthermore, we expect this intervention to have its greatest benefit among populations that frequently require ED visits and hospitalizations to manage their asthma, such as economically disadvantaged inner-city adults.
Corresponding author and reprints: Yvonne Marie Coyle, MD, Division of General Internal Medicine, The University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd, Dallas, TX 75390-9103 (e-mail: email@example.com).
Accepted for publication September 19, 2002.
This study was supported by grant U18 HS094601-03 from the Agency for Healthcare Research and Quality, Rockville, Md (from September 30, 1996, to September 29, 1999).
We thank The University of Texas Southwestern Medical Center at Dallas General Clinical Research Center for providing allergy skin testing supplies and research clinic space to perform study procedures and data collection; and Patsy Hargrave, LVN, and Elizabeth McDonald, RN, the study's research nurses, for patient recruitment, testing, and data collection.
The contents of this study are solely the responsibility of the authors and do not necessarily represent the official views of the Agency for Healthcare Research and Quality.