Objective: Recommended preventive care following an asthma admission includes prescribing controller medications and encouraging outpatient follow-up. We sought to determine (1) the proportion of patients who receive controller medications or attend follow-up after asthma admission and (2) what factors predict these outcomes.
Methods: South Carolina Medicaid data from 2007-2009 were analyzed. Patients who were included were 2 to 18 years old, and had at least one admission for asthma. Variables examined were: age, gender, race, and rural location. Outcome variables were controller medication prescription and follow-up appointment. Any claim for an inhaled corticosteroid (ICS) or ICS/long-acting beta-agonist in the 2 months after admission was considered appropriate. Any outpatient visit for asthma in the 2 months after admission was considered appropriate. Bivariate analyses used chi-square tests. Logistic regression models identified factors that predict controller medications and follow-up.
Results: Five hundred five patients were included, of whom 60% were male, 79% minority race/ethnicity, and 58% urban. Rates of receiving controller medications and attending follow-up appointments were low, and an even lower proportion received both. Overall, 52% received a controller medication, 49% attended follow-up, and 32% had both. Multivariable analyses demonstrated that patients not of minority race or ethnicity were more likely to receive controller medications (odds ratio, 1.7; 95% confidence interval, 1.1-2.6).
Conclusions: Patients with asthma admitted for acute exacerbations in South Carolina have low rates of controller medication initiation and follow-up attendance. Minority race/ethnicity patients are less likely to receive controller medications. To decrease rates of future exacerbations, inpatient providers must improve the rates of preventive care delivery in the acute care setting with a focus on racial/ethnic minority populations.
Asthma is the most prevalent chronic condition affecting children.1,2 The prevalence of asthma continues to increase.1,3 In addition, morbidity and mortality associated with asthma continues to increase, including the number of hospitalizations, missed school days, and missed parental work days.1,4 Asthma is one of the most common reasons for hospital admission among children.1,5 Appropriate use of controller medications can reduce asthma exacerbations, acute care visits for asthma, and hospitalizations.5,7
Acknowledging the high rates of acute care visits for asthma, the Institute of Medicine has identified asthma as an ambulatory care–sensitive condition.8
Despite this, current evidence suggests that pediatric asthma patients underutilize preventive care in the ambulatory setting, including controller medications.9–12 Asthma patients who are admitted to the hospital for acute exacerbations are at higher risk for future exacerbations13,14; therefore, it is paramount that these patients receive appropriate preventive care. Consequently, physicians who care for children with asthma in the hospital setting have a unique and critically important opportunity to provide preventive care in the acute care setting.
The National Heart, Lung, and Blood Institute (NHLBI) guidelines for the treatment of asthma published in 2007 recommend the use of long-term controller medications for all those with persistent asthma.15 This is consistent with the 1997 NHLBI asthma guidelines which state that “persistent asthma is most effectively controlled with daily anti-inflammatory therapy”.16 The guidelines also recommend that the initiation of controller medication therapy be considered after any asthma admission.15 Even stronger, the Global Initiative for Asthma (GINA) 2008 guidelines recommend initiation or continuation of controller medications for all patients before discharge from the acute care setting.17 In addition to initiating or continuing controller medications at hospital discharge, recommended preventive care before discharge from an asthma hospitalization includes arranging close outpatient follow-up.15
Despite these recommendations, controller medications, particularly inhaled corticosteroids (ICS) continue to be underutilized among pediatric asthma patients across all levels of health care utilization.4,9,18–20 Surprisingly few studies have looked specifically at ICS prescribing practices at hospital discharge. Lantner et al21 studied a population of adult and pediatric asthma patients admitted in the Chicago area and found that only 41% of pediatric patients were discharged with a prescription for an ICS, and patients with Medicaid were less likely to be discharged with a prescription for an ICS. We have shown previously that among pediatric asthma patients admitted to a single institution, 73% were prescribed an ICS at hospital discharge.9
Controller medication use has been shown to be lower among African American children with asthma.22 African American children are hospitalized for asthma at a rate that is 3.5 times greater than that of white children, and their mortality rate from asthma is four times that of white children.23 In addition, minority race/ethnicity patients are more likely to be readmitted within 180 days of an asthma hospitalization,24 making the appropriate use of controller medications and establishing longitudinal outpatient care critical for this patient population. To our knowledge, no prior study has investigated racial disparities in controller medication use specifically at hospital discharge.
The objective of this study was to determine, using recent statewide Medicaid data, the proportion of pediatric asthma patients who receive controller medications or attend outpatient follow-up visits after an asthma admission. The first year of data used in this analysis coincides with the release of the most recent NHLBI guidelines, so we also sought to determine if the delivery of preventive care has improved since the guideline release. We focused specifically on hospitalization for asthma as a sentinel event that should trigger controller medication initiation. We hypothesized that the rates of controller medication use and outpatient follow-up remain suboptimal. Our secondary objective was to determine what demographic factors predict the delivery of preventive care among hospitalized asthma patients. We hypothesized that younger patients and minority race/ethnicity patients would have lower rates of controller medication initiation and outpatient follow-up.
We conducted a retrospective analysis using 2007-2009 South Carolina Medicaid data. Patients aged 2 to 18 years with at least 1 calendar year of continuous Medicaid enrollment were included in the dataset. We isolated all encounters with a primary diagnosis of asthma using International Classification of Diseases, 9th Revision (ICD-9) codes 493.00-493.02, 493.10-493.493.12, 493.20-493.22, 493.8, and 493.9. From this group with asthma, we eliminated all patients with the following comorbid conditions (patients were eliminated if the ICD-9 code for any of these conditions were present in the dataset): exercise-induced asthma; cystic fibrosis; bronchiolitis; respiratory syncytial virus; moderate, severe, or profound mental retardation; congenital heart disease; sickle cell disease; chronic respiratory disease arising in the perinatal period; tracheostomy; gastrostomy, or any other artificial opening of the gastrointestinal tract. These comorbidities have been used as exclusionary criteria in previous pediatric asthma studies.25
Inpatient visits were identified using Current Procedural Terminology codes 99221-99223, 99217-99223, and 99234-99236. We limited our analysis to patients without a claim for an ICS or ICS/long-acting beta agonist (LABA) in the 2 months preceding the month of the asthma hospitalization. We believed this was important because it showed us how well physicians recognize hospitalizations as sentinel events for diagnosing persistent asthma and initiating or reinstating controller medication therapy in patients without recent controller medication use. In addition, for those who had multiple admissions in the dataset we analyzed their most recent visit.
The pharmacy claims file was refined to include only ICS and ICS/LABA agonist because the NHLBI guidelines consider these to be preferred for the treatment of persistent asthma.15 In addition, ICS have been shown to be less expensive and more effective than montelukast for pediatric patients with mild to moderate persistent asthma.26 We identified specific drugs using National Drug Codes as well as all possible permutations of each drug name (both generic and trade names). The pharmacy data were then merged with the patient-level asthma visit file using unique, encrypted identifiers for each subject in the dataset.
Because of the constraints of the Health Insurance Portability and Accountability Act in the South Carolina Medicaid system, we did not have access to exact dates of visits and pharmacy claims. Instead we were provided with month/year for each claim as well as a sequencing number so that we could order claims within each month.
Our primary outcome variable was a claim for any ICS or ICS/LABA in the hospitalization month or the month thereafter. Our secondary outcome variable was any outpatient visit with a primary diagnosis of asthma in either the month of or the month after the hospitalization. Only outpatient visits that occurred after the hospitalization were considered appropriate follow-up. This included visits with a primary care physician (PCP) or subspecialist. In addition we analyzed the two outcome variables in combination.
Demographic variables used in the analyses were age, race, gender, and county of residence. “Age” indicated the patient’s age in years on entry into the cohort. We divided patients into three age categories: 2 to 6 years old, 7 to 12 years old, and 13 years and older. We believed this represented three distinct groups of patients: preschool/kindergarten, school-aged, and adolescents. Race and ethnicity are self-identified at the time of application for Medicaid. In this dataset, race/ethnicity was defined as white, black, Hispanic, or other. In our final analysis, we dichotomized race to nonminority (white) versus minority race or ethnicity (black, Hispanic, or other). County of residence was used to classify patients as rural or urban. We used the Metropolitan Statistical Area definition of urban for the analysis, and counties not qualified as urban were labeled rural.27 We included a severity marker based on short-acting beta agonist (SABA) pharmacy claims, a method that has been used in previously published studies.28 Patients’ conditions were labeled as severe if they had six or more pharmacy claims for SABAs in the calendar year of the hospitalization.
All statistical analyses were performed using SAS statistical software, version 9.2 (SAS Inc). All variables were categorized as previously described. We used chi-square tests to compare characteristics of patients who did and did not receive each of the outcome variables (controller medication, follow-up appointment, and controller medication + follow-up appointment). We did not adjust for multiple comparisons. We calculated the adjusted relationship among all variables for each of the outcome variables using logistic regression. We built logistic regression models for each of the outcome variables including all demographic variables and the severity marker.
The Institutional Review Board of the Medical University of South Carolina reviewed and approved this study.
Overall, 34,389 patients with asthma were identified in the dataset. Many patients were noted to have an extremely high number of visits in the study period. Because of concerns that these patients did not represent typical asthma patients, we excluded those in the 99th percentile for total number of visits, thereby excluding 348 patients with more than 67 visits for asthma in 3 years. We then identified 666 patients with at least one admission for asthma. After excluding those with an ICS or ICS/LABA claim in the 2 months before admission, 505 patients were included in the final analysis cohort. The mean age of these patients was 7.6 years (median 7). Table 1 shows complete demographic data.
Recommended Preventive Care
Fifty-one percent of patients received an ICS or an ICS/LABA in either the month of or the month after the hospitalization. In addition, 49% of patients had a follow-up appointment in either the month of or the month after the hospitalization. Interestingly, there was only a partial overlap between those who received appropriate prescriptions and those who attended follow-up. When analyzed together, we found that 32% of patients hospitalized for an asthma exacerbation received both an ICS or an ICS/LABA and attended a follow-up appointment. To evaluate if there has been any change in the rates of controller medication filling since the release of the 2007 NHLBI guidelines, we calculated rates by year. In bivariate analysis using chi-square, we found no significant relationship between controller medication filling and year of admission. Rates by year (2007, 2008, 2009) were 54%, 57%, and 45%, respectively (P=.08).
Bivariate Analysis (Table 2)
For controller medications, the analyses demonstrated that minority race/ethnicity patients were less likely to have a pharmacy claim for an ICS or ICS/LABA in the 2 months following an asthma hospitalization than majority race/ethnicity patients. No other statistically significant differences were found.
Multivariable Analysis (Table 3)
After controlling for all covariates, minority race/ethnicity patients were less likely than majority race/ethnicity patients to have a pharmacy claim for an ICS or ICS/LABA in the 2 months following an asthma hospitalization. Nonminority race/ethnicity patients had 70% greater odds of receiving a controller medication after hospitalization for asthma. No other statistically significant predictors were found.
The current analysis demonstrates that only 52% of children with asthma admitted for acute exacerbation receive preferred controller medications during the month of hospital admission or in the month thereafter. In addition we show that fewer than 50% of patients follow-up with an outpatient provider in the same period. Minority race/ethnicity patients are less likely to receive controller medications after asthma admission. Our finding of overall low rates of controller medication dispensing and outpatient follow-up suggest a need to improve the delivery of preventive care in the acute care setting among patients hospitalized for acute asthma exacerbations, with a focus on the minority race/ethnicity population.
Hospitalizations for asthma should be recognized as red flags for uncontrolled disease. Taking into account the low rates of controller medication initiation and outpatient follow-up in response to hospitalizations in this study, it seems that hospitalizations are frequently accepted by both providers and patients as part of the normal disease course of asthma, not as signals of management failure or persistent disease. This may be due, in part, to the fact that although the GINA guidelines and Healthcare Effectiveness Data and Information Set (HEDIS) criteria clearly state that patients with asthma who are admitted to the hospital qualify as persistent asthmatics and should begin controller medications, the NHLBI guidelines do not make clear recommendations regarding the initiation of controller therapy at the time of hospital discharge. Further supporting the use of ICS for all patients after asthma admission, a recent study by Martinez et al29 has shown the efficacy of ICS controller therapy even for patients with mild intermittent asthma. It is possible that a hesitancy to diagnose persistent asthma in response to a hospitalization to preserve diagnostic specificity has led to a loss of sensitivity, resulting in far too many asthma patients being underclassified and consequently, undertreated.
Asthma patients admitted to the hospital represent a captive audience and provide inpatient providers with an important opportunity to contribute to the management of their chronic disease through the initiation of long-term controller medications. In addition, inpatient providers must take a proactive approach to facilitating outpatient follow-up appointments using techniques such as scheduling appointments for the patient before discharge, direct communication with the PCP regarding hospitalization and discharge medications, and stressing the importance of attending the follow-up appointment with the patient and family.
Our study demonstrates that minority race/ethnicity patients (including African Americans and Hispanics) were less likely to receive and/or fill a prescription for a controller medication after an asthma hospitalization among Medicaid enrollees in South Carolina. In 2002, Lieu et al22 reported similar findings. Their analysis of parentally reported data showed that among asthma patients enrolled in Medicaid in California, Washington, and Massachusetts, African American and Latino children were 30% less likely to use preventive asthma medications compared with white children.22 Our study complements this finding by identifying greater racial disparities among hospital discharge practices, perhaps when asthma patients are at their most vulnerable, therefore identifying a specific place where we can aim to improve care. Our evidence of racial disparities in controller medication use suggests the need to address this issue in a culturally competent manner.
Because our dataset is limited to pharmacy claims data, we were unable to determine if the disparity in controller medication dispensing occurred because of differences in rates of prescribing or in rates of prescription filling. Differences could arise in rates of provider prescribing because of poorer access to outpatient care among minority populations. We believe this disparity is less likely to be the result of differences in access to care. Given that all patients in this study were covered under Medicaid, we found no rural/urban disparity in controller medication dispensing, and no difference in outpatient follow-up rates between minorities and nonminorities.
On the contrary, the disparity in dispensing could be the result of differences in rates of prescription filling. Perhaps prescriptions were written at similar rates but minority race/ethnicity patients were less likely to fill their prescriptions. This theory is supported by a prior study that identified both prescriptions written for and prescriptions filled by a cohort of pediatric asthma patients.30 They found that between Medicaid patients and non-Medicaid patients in a single Health Maintenance Organization, the rates of controller medication prescribing were equal, but rates of controller medication dispensing among the Medicaid population were statistically significantly lower.30 In this prior study, 70% of Medicaid patients were nonwhite, while only 35% of non-Medicaid patients were nonwhite.30 This finding lends credence to our theory that prescriptions were written at similar rates for both minority and nonminority race/ethnicity patients but that minority patients were less likely to fill their prescriptions.
This begs the question, why would minority patients be less likely to fill prescriptions than nonminority patients? It cannot be simply because of poor compliance, because we found equal rates of outpatient follow-up between the two groups. Prior studies have suggested that African Americans have concerns about the safety of long-term controller medications as well as doubts about their usefulness.10,31 This suggests that the possible lower rate of prescription filling among minorities could be related to differences in health beliefs. This highlights the importance of delivering culturally competent care to overcome these barriers and ideally improve the rate of controller medication use among minority asthma patients.
This study has several limitations. We analyzed a retrospective cohort of patients using administrative claims data. ICD-9 codes are inherently inaccurate for the identification of subjects; however, we believe that our sample size is large enough to help minimize the effect of potential inaccuracies in coding. We analyzed Medicaid data from one state so our finding may not be generalizable to other geographic locations. However, we believe that our large sample size and the diversity in our state help to curtail this issue. Given the nature of the data, we are unable to determine if the prescriptions for controller medications were written by hospitalists or PCPs. Also, the pharmacy claims data contain only prescriptions that were filled; therefore, we are unable to determine to what extent the low rate of controller medication use is the result of low rates of prescribing, patients not filling prescriptions, or a combination of these two factors. Given previously published data that suggest suboptimal prescribing of controller medications by hospitalists, it is reasonable to assume that a proportion of the underutilization of controller medications in this cohort is the result of underprescribing.9
Pediatric hospital care providers are poised to contribute to the improved delivery of preventive care to hospitalized asthma patients. In accordance with the GINA guidelines, we suggest uniform and more aggressive prescribing of controller medications at hospital discharge and, perhaps in the future, dispensing controller medications before discharge to eliminate barriers related to prescription filling. In addition, a focus on addressing health beliefs pertinent to chronic asthma care in a culturally competent manner will be vital to address the racial/ethnic disparities in preventive care delivery for hospitalized asthma patients.
RESEARCH SUPPORT: This study was supported by a National Research Service Award (HRSA T32HP10255).
FINANCIAL DISCLOSURE: None.
- Akinbami LJ,
- Moorman JE,
- Garbe PL,
- Sondik EJ
- Warman KL,
- Silver EJ,
- Stein RE
- Adams RJ,
- Fuhlbrigge A,
- Finkelstein JA,
- et al
- Lieu TA,
- Quesenberry CP Jr.,
- Capra AM,
- Sorel ME,
- Martin KE,
- Mendoza GR
- 8.↵Agency for Healthcare Research and Quality. AHRQ Quality Indicators-Guide to Prevention Quality Indicators. Rockville, MD: AHRQ; 2004. AHRQ publication No. 02–R0203; revision 4 ed.
- Lintzenich A,
- Teufel RJ,
- Basco WT Jr.
- Mansour ME,
- Lanphear BP,
- DeWitt TG
- Zorc JJ,
- Chew A,
- Allen JL,
- Shaw K
- Zorc JJ,
- Scarfone RJ,
- Li Y,
- et al
- Crane J,
- Pearce N,
- Burgess C,
- Woodman K,
- Robson B,
- Beasley R
- 15.↵National Asthma Education and Prevention Program, National Heart Lung and Blood Institute. Expert Panel Report 3: Guidelines for the Diagnosis and Management of Asthma. Bethesda MD: National Institutes of Health; 2007;08–5846.
- 16.↵National Heart, Lung, and Blood Institute. Expert Panel Report 2: Guidelines for the Diagnosis and Managment of Asthma. 1997. http://www.nhlbi.nih.gov/guidelines/archives/epr-2/asthgdln_archive.pdf. Accessed June 2, 2011.
- 17.↵Global Strategy for Asthma Management and Prevention. Global Initiative for Asthma (GINA); 2009. http://www.ginasthma.org/guidelines-gina-report-global-strategy-for-asthma.html
- Lieu TA,
- Lozano P,
- Finkelstein JA,
- et al
- Akinbami LJ,
- Schoendorf KC
- Wallace JC,
- Denk CE,
- Kruse LK
- 27.↵US Department of Commerce Economics and Statistics Administration, US Census Bureau. South Carolina: core based statistical areas and counties. Available at: http://ftp2.census.gov/geo/maps/metroarea/stcbsa_pg/Nov2004/cbsa2004_SC.pdf. Accessed June 3, 2011
- Leickly FE,
- Wade SL,
- Crain E,
- Kruszon-Moran D,
- Wright EC,
- Evans R 3rd.
- Copyright © 2011 by the American Academy of Pediatrics