Su-Jau T. Yang, PhD

To describe the relationships between persistent asthma defined by administrative versus survey data and their stability over time.

 Longitudinal survey and retrospective administrative database.

 Administrative data were used to identify patients meeting the Healthcare Effectiveness Data and Information Set (HEDIS) criteria for persistent asthma in year 1 (2006). At the end of year 2 and on 3 occasions during year 3, patients were mailed a survey to define persistent asthma based on symptoms and medication use in the prior month and exacerbations in the prior 12 months. Administrative data were also used to define medical utilization for asthma in year 3.

 Of 13,833 eligible patients, 2895 (20.9%) returned the survey; 2751 of these respondents reported physician-diagnosed asthma, of whom 2517 (91.5%) had survey-defined persistent asthma. Patients having survey-defined persistent asthma (68.0%) were more likely to requalify as having HEDIS-defined persistent asthma in year 2 than patients not having survey-defined persistent asthma (22.2%). However, 81.6% of survey respondents who did not requalify as having HEDIS-defined persistent asthma in year 2 had survey-defined persistent asthma. Patients with survey-defined persistent asthma in year 2 had significantly more medical utilization for asthma in year 3 than patients without survey-defined persistent asthma. Approximately 82% of the 799 patients completing all 4 surveys had persistent asthma on all surveys.

 HEDIS-defined persistent asthma is generally consistent with survey-defined persistent asthma. Persistent asthma usually remains persistent over a 3-year period, indicating that it is a stable characteristic of asthma for most patients. The low survey response rate suggests that further population-based studies will be necessary to confirm the validity and generalizability of our study findings regarding persistent asthma.

	Healthcare Effectiveness Data and Information Set (HEDIS)-defined persistent asthma seems to reflect clinical persistent asthma.

		The 2-year HEDIS definition is more specific for persistent asthma than the 1-year definition.

		Persistent asthma is associated with increased medical utilization compared with nonpersistent asthma.

		Persistent asthma seems to be persistent over a 3-year period and is an appropriate target for intervention.

	Persistent asthma is clinically defined based on the frequency and severity of asthma symptoms and on pulmonary function test results when available.

 It is important to  differentiate persistent from intermittent asthma because persistent asthma requires daily controller therapy, while intermittent asthma can be treated with rescue therapy as needed.

	The National Committee for Quality Assurance

 has developed an administrative data—based definition of persistent asthma as part of the Healthcare Effectiveness Data and Information Set (HEDIS) quality measurement program. This definition relies on asthma-coded medical visits and asthma-related pharmacy utilization and is meant to define a population for whom asthma controller therapy is indicated. However, studies

 have shown that only a modest relationship exists between this administrative data—based definition of persistent asthma and a definition based on patient report of symptoms. Moreover, findings from other studies

 suggest that 24% to 45% of patients who qualify as having persistent asthma in one year based on the HEDIS definition do not qualify again the next year. This raises the question of how persistent is persistent asthma. In other words, is persistent asthma a sufficiently stable characteristic of asthma to warrant the recommendation for long-term controller therapy? We are unaware of studies that have specifically addressed this question.

	Because data suggest that patients who qualify for administrative data—based persistent asthma for 2 consecutive years have increased markers of asthma morbidity and are more likely to receive controller therapy,

 the HEDIS definition of persistent asthma for the purposes of assessing appropriate controller use has been modified to require qualification in each of 2 successive years.2 However, no studies to date have investigated the relationship of this 2-year administrative data—based definition of persistent asthma to persistent asthma as defined by patient report of symptom frequency, severity, and treatment. The objectives of this study were (1) to investigate the relationship of the 2-year HEDIS definition of persistent asthma to persistent asthma as defined by patient report, (2) to compare medical utilization for patients with versus without persistent asthma, and (3) to determine the persistence of persistent asthma over time.

	This study was approved by the Kaiser Permanente SouthernCalifornia Institutional Review Board. Using computerized Kaiser Permanente Southern California inpatient, outpatient, pharmacy, and membership records, we identified patients aged 18 to 56 years with persistent asthma in 2006 (study year 1) based on their meeting 1 or more of the following HEDIS criteria: (1) 4 or more asthma medication dispensings, (2) 1 or more emergency department visits or hospitalizations with a principal diagnosis of asthma, or (3) 4 or more asthma outpatient visits with 2 or more asthma medication dispensings. Patients with continuous membership in 2007 who also met 1 or more of these HEDIS criteria in 2007 (study year 2) were considered to requalify for 2-year HEDIS-defined persistent asthma (

). Only patients with continuous membership in Kaiser Permanente Southern California from 2006 to 2008 were included in the study.

	The following additional information was captured from the Kaiser Permanente Southern California records: (1) sex, (2) age, (3) smoking status, (4) Medicare or Medi-Cal insurance, (5) hospitalization with asthma coded (

International Classification of Diseases, Ninth Revision, Clinical Modification

] code 493.xx) as the primary diagnosis or as the

	second diagnosis with another respiratory primary diagnosis, (6) emergency department visit with asthma coded (

 code 493.xx) as the primary diagnosis or as the second diagnosis with another respiratory primary diagnosis, (7) outpatient visit with asthma coded (

 code 493.xx) as a diagnosis, (8) any oral corticosteroid dispensing, (9) dispensings of more than 6 short-acting b-agonist canisters in a 12-month period, (10) any asthma controller dispensing, and (11) address. We successfully linked more than 95% of the addresses of our enrollees to census block groups and assigned block group—level income and education estimates in 2006 to patients by using the proprietary  demographic estimates supplied by Nielsen Claritas (http://www.claritas.com). Census-based income and education variables have been widely used as proxies for the socioeconomic status of individuals in many health-related investigations.

 Using this method, we determined (1) the proportion of the population older than 25 years in the patient’s census block having more than a high school education and (2) the median household income in the patient’s census block. Medical utilization for asthma was assessed during 2006 (study year 1) and during 2008 (study year 3) (Table 1).

	In November 2007, all eligible patients were mailed a survey to define persistent asthma based on symptoms and medication use in the prior month and exacerbations in the prior 12 months. The survey used the Expert Panel Report 3 definition of persistent asthma, which includes impairment and risk domains.

 The impairment domain includes the frequency and intensity of symptoms and functional limitations that the patient is currently experiencing. The risk domain includes the likelihood of asthma exacerbations based on the number of exacerbations in the prior 12 months. Four yes or no questions asked whether the patient in the past 30 days had (1) “experienced asthma symptoms at least 3 times per week,” (2) “used your rescue inhaler for symptoms at least 3 times per week,” (3) “been awakened by asthma symptoms at least 3 times,” and (4) had asthma that “interfered with your normal activity.” One question asked “In the past 12 months, how many times have you been given oral corticosteroids (such as prednisone, methylprednisolone, medrol) for a flare up of your asthma?” The final question regarding persistent asthma asked “Over the past 30 days, have you taken any of the following medications every day for your asthma?” and was followed by a list of controller medications with a yes or no option for each. A comparison of the HEDIS and survey definitions of persistent asthma is given in 

	Patients who completed the baseline surveys were sent follow-up surveys in April, July, and October 2008 (Table 1). Only survey respondents who reported physician-diagnosed asthma on the baseline survey were included in the study.

	The main analyses were directed at (1) identifying the relationships between HEDIS-defined persistent asthma (hereafter HEDIS persistent asthma) and survey-defined persistent asthma (hereafter survey persistent asthma) among patients completing the survey, (2) comparing medical utilization for asthma among patients with versus without persistent asthma, and (3) determining the persistence of persistent asthma over time. Hypothesis testing for comparisons of characteristics among survey responders versus  nonresponders was by means of 

Survey Persistent Asthma Versus HEDIS Persistent Asthma. 

The presence or absence of survey persistent asthma on the baseline survey administered at the end of 2007 was assessed relative to requalification or nonrequalification for HEDIS persistent asthma in 2007 using 

 Persistent Asthma Versus Medical Utilization.

 Hypothesis testing was conducted for the relationship between the presence of survey persistent asthma on the baseline survey at the end of 2007 (year 2) or HEDIS requalification in 2007 and medical utilization for asthma in 2008 (year 3). This was performed by means of 

 Concordance between survey persistent asthma at baseline and on follow-up surveys was evaluated. Among patients who completed all 4 surveys, the proportions of patients with persistent asthma on 0, 1, 2, 3, or 4 surveys were determined.

	Nominal 2-tailed statistical significance for all analyses was set at 

 <.05. All analyses were conducted using SAS version 9.13 for Windows (SAS Institute Inc, Cary, North Carolina).

	A total of 13,833 patients aged 18 to 56 years met the HEDIS criteria for persistent asthma in 2006 and were continuously enrolled in Kaiser Permanente Southern California from 2006 to 2008. Of these, 2895 (20.9%) returned the November 2007 baseline survey; 144 denied a history of physician-diagnosed asthma, leaving a sample of 2751 survey respondents.

	Compared with nonrespondents, respondents were more likely to be female, were older, lived in neighborhoods (census blocks) with higher education and household income, and were less likely to be smokers (

). Respondents experienced fewer exacerbations requiring emergency department visits or oral corticosteroid dispensings and were less likelyto receive more than 6 canisters of short-acting ß-agonists in 2006. In contrast, respondents were more likely to have asthma outpatient visits, to receive asthma controllers,  and to requalify for HEDIS persistent asthma in 2007.

	Of 2751 respondents to the baseline survey, 799 (29.0%) completed all 4 surveys. A total of 455 (16.5%) completed 3 surveys, 422 (15.3%) completed 2 surveys, and 1075 (39.1%) completed only the baseline survey.

Survey Persistent Asthma Versus HEDIS Persistent Asthma

	A total of 2517 among 2751 respondents (91.5%) had survey persistent asthma on the baseline survey. These 2517 patients met the following persistent asthma criteria: (1) symptoms with or without other criteria (74.7%), (2) medications without qualifying symptoms (24.5%), or (3) exacerbations without symptom or medication criteria (0.8%).

	A total of 1764 among 2751 respondents (64.1%) requalified for HEDIS persistent asthma in 2007. Of those who requalified, 97.1% reported survey persistent asthma 

). Patients with survey persistent asthma were significantly more likely to requalify for HEDIS persistent asthma in 2007 than patients without survey persistent asthma (68.0% vs 22.2%, 

 <.001). However, 805 of 987 patients (81.6%) who did not requalify as having HEDIS persistent asthma in 2007 (but who had qualified in 2006) reported survey persistent asthma in 2007.

 Persistent Asthma Versus Medical Utilization

	Compared with patients without survey persistent asthma, patients with survey persistent asthma in year 2 (2007) had significantly more year 3 (2008) asthma hospitalizations or emergency department visits, oral corticosteroid dispensings, and shortacting ß-agonist canisters (

). Similarly, among patients with survey persistent asthma, those who requalified for HEDIS persistent asthma in year 2 had significantly more year 3 medical utilization for asthma (asthma emergency department visits or hospitalizations, oral corticosteroid dispensings, and short-acting ß-agonist canisters) than patients who did not requalify.

	The relationships between survey persistent asthma at baseline and at the first follow-up survey 5 months later are summarized in 

. For this analysis, persistent asthma was based only on impairment and not on risk (exacerbations) because the interval between surveys (5 months) was too short to identify persistent asthma based on exacerbations over the interval (12 months) recommended in the Expert Panel Report 3 guidelines to assess risk.

 Ninety-four percent of 1199 patients who reported persistent asthma at baseline

	also reported persistent asthma on the first follow-up survey.Of 91 patients who did not report persistent asthma on the baseline survey, more than half (51.6%) reported persistent

	asthma on the first follow-up survey (Table 6).

Articles

HEDIS-defined persistent asthma is generally consistent with survey-defined persistent asthma over a 3-year period.

Persistent Asthma Defined Using HEDIS Versus Survey Criteria

 To compare the relationship of surrogate markers of asthma control to subsequent asthma exacerbations.

 Administrative data were used to identify patients who met the Healthcare Effectiveness Data and Information Set (HEDIS) criteria for persistent asthma in 2006 and 2007. The following potential surrogate markers of asthma control assessed in 2007 were compared for their ability to predict asthma exacerbations in 2008 (defined as oral corticosteroid dispensing or an asthma hospitalization or emergency department visit): dispensing of any controller, unweighted medication ratio (the ratio of controller to total medication), weighted medication ratio, and the number of short-acting Î²-agonist (SABA) canisters dispensed. Weighted medication ratios were weighted for controller potency and for doses per container.

Meeting the HEDIS criteria for persistent asthma were 8634 patients (60.5% female) aged 18 to 56 years (mean age, 42.7 years), of whom 6.5% experienced emergency hospital care and 27.3% received oral corticosteroids in 2008. The largest effect size for predicting reduced emergency hospital care was for the number of SABA canisters dispensed (odds ratio [OR], 0.49; 95% confidence interval [CI], 0.40-0.60), followed by the unweighted medication ratio (OR, 0.54; 95% CI, 0.40-0.72), and then the weighted medication ratio (OR, 0.57; 95% CI, 0.45-0.73). Dispensing of any controller was associated with a nonsignificant increased risk of emergency hospital care (OR, 1.41; 95% CI, 0.95-2.09).

The number of SABA canisters dispensed is most strongly related to improved asthma outcomes, followed by the unweighted medication ratio; dispensing of any controller is least related. Health plans can use the number of SABA canisters dispensed and the unweighted medication ratio for asthma population management or for provider quality-of-care assessment to reduce asthma exacerbations, which exact a high economic and humanistic cost.

	The following relationships of asthma control to subsequent asthma exacerbations were observed:

		The number of short-acting β-agonist canisters dispensed is most useful to identify at-risk patients for targeted intervention in population management.

		The ratio of controller to total medication can be used as an asthma quality-of-care marker.

		Dispensing of any controller does not seem to be a good marker for population management or for quality of care.

Asthma is a common chronic medical condition that is associated with substantial individual morbidity and cost to society. 

The traditional approach to improving disease outcomes has been by one-on-one physician—patient interactions. Computerization of medical utilization and claims data (“administrative data”) has allowed 2 more recent approaches for asthma control. The first is the use of administrative data to define asthma quality-of-care or performance measures. These measures are used by managed care organizations and others to improve the performance of their providers, with the expectation that improved performance on the measures will lead to improved asthma outcomes and lower costs.

 The most widely used asthma quality-of-care measure is the one developed by the National Committee for Quality Assurance

 to evaluate the performance of health plans, the Healthcare Effectiveness Data and Information Set (HEDIS) measure (“HEDIS measure”). This measure identifies patients with “persistent asthma” based on administrative data criteria and then determines the proportion of these patients who receive at least 1 controller medication dispensing the following year. Although used for approximately a decade, few studies have attempted to validate the HEDIS measure as being associated with improved subsequent outcomes. A study

 in children demonstrated such a relationship, but 3 studies

 in children and adults failed to show that improved performance on this measure is associated with improved asthma outcomes. After a study by Mosen et al

 demonstrated a relationship between the number of consecutive years of HEDIS qualification and markers of persistent asthma, the National Committee for Quality Assurance

modified the definition of HEDIS persistent asthma to require that the administrative data criteria for persistent asthma be met for both the measurement year (the year the measure is being assessed) and the premeasurement year (“2-year HEDIS population”). To date, the HEDIS measure has not been systematically evaluated in this new 2-year HEDIS population.

The second use of computerized administrative data has been for population management. Patients with inadequately controlled asthma are identified based on administrative data criteria, and targeted intervention by mail, telephone, or in-person visits are instituted. In addition to the HEDIS measure of any controller use, 2 other measures have been suggested for use as quality-of-care measures or as measures to identify patients with uncontrolled asthma in need of intervention. The first is the medication ratio, which is the ratio of units of controllers dispensed to the total units of controllers and units of short-acting β-agonists (SABAs) dispensed. A higher medication ratio has been associated with reduced subsequent exacerbations and with better patient- reported outcomes.

This ratio has been used in a weighted form (weighted for doses per container and drug potency) 

 although these 2 ways of using the ratio measure have not been directly compared to date. The second is the number of SABA canisters dispensed per year (“SABA measure”), which has been shown to be a long-term measure of asthma control.

To our knowledge, this SABA measure has not been compared with the HEDIS measure or the medication ratio measure as a predictor of improved asthma outcomes.

The objective of this study was to compare the ability of the HEDIS measure, the medication ratio measure (weighted and unweighted), and a SABA measure to predict asthma exacerbations (defined in 2 different ways) in the 2-year HEDIS population. Asthma exacerbations are particularly appropriate outcomes to use in validating surrogate markers of asthma control because of their large contribution to the cost of asthma care

 and their substantial effect on quality of life. Identifying valid asthma control measures based on administrative data that are associated with reduced asthma exacerbations should allow health plans to use these measures to provide outreach to patients in need of intervention and to offer “inreach” to providers whose patients may be at increased risk for poor asthma outcomes.

This study was approved by the Kaiser Permanente Southern California (KPSC) Institutional Review Board. Using computerized KPSC inpatient, outpatient, pharmacy, and membership records, we identified patients aged 18 to 56 years with continuous enrollment (

1 gap of <45 days’ coverage) in 2006 and with persistent asthma in 2006 based on meeting 1 or more of the following HEDIS criteria: 4 or more asthma medication dispensings, 1 or more emergency department visits or hospitalizations with a principal diagnosis of asthma, or 4 or more asthma outpatient visits with 2 or more asthma medication dispensings. Patients with continuous enrollment in 2007 who requalified in 2007 as having HEDISdefined persistent asthma (ie, met 

1 of these criteria in 2007) who also remained continuously enrolled KPSC members in 2008 are the subjects of this study.

The outcome of interest was asthma exacerbations in 2008. These were identified in the following 2 ways: (1) emergency hospital care for asthma, defined as having at least 1 hospitalization or emergency department visit for which asthma (

] code 493.xx) is the primary diagnosis or is a secondary diagnosis with another respiratory diagnosis coded as the primary diagnosis, or (2) a filled prescription of an oral corticosteroid prescribed within 7 days after an outpatient encounter for asthma (

Nebulized medications were excluded as potential markers of asthma control because dispensed quantities could not be reliably translated into medication units. The following candidate markers of quality asthma care or uncontrolled asthma in 2007 were evaluated as predictors of asthma exacerbations in 2008:

HEDIS Measure (Dispensing at Any Time and in Any Amount of Any Asthma Controller, as Defined by HEDIS,4 Which Is the Current HEDIS Asthma Quality-of-Care Measure). 

These controllers include inhaled corticosteroids, inhaled corticosteroid/long-acting β-agonist combinations,leukotriene modifiers, theophylline, and cromolyn sodium. They do not include SABAs, long-acting β-agonists alone, or oral corticosteroids.

Unweighted Medication Ratio (Unweighted Ratio of Controller to Total Asthma Medication).

 The ratio was calculated for each patient as the number of controller units dispensed during 2007 divided by the sum of the controller units and SABA inhaler canisters dispensed during 2007. For inhalers, 1 controller unit was 1 canister. For oral medications, 1 controller unit was 1 prescription of an amount lasting 30 days or less. Medication ratios range from 0 (no controllers) to 1.0 (no SABAs).

Weighted Medication Ratio (Weighted Ratio of Controller to Total Asthma Medication). 

This ratio was calculated as for the unweighted medication ratio except that controllers were weighted for medication potency and for the number of doses per container to account for potentially important differences in these characteristics between products. The specific weighting algorithm for individual products is available from the authors by request. Examples of weights for common controllers among our patients are 1.0 (montelukast sodium), 1.7 (beclomethasone dipropionate hydrofluoroalkane 80), 2.0 (budesonide 180), and 4.0 (fluticasone propionate 220 and mometasone furoate 220).

SABA Measure (Number of SABA Canisters Dispensed in 2007). 

The total number of inhalational SABA canisters dispensed in 2007 was used for this marker.

Optimal cutoffs for the medication ratio and SABA measures to predict emergency hospital care and asthma exacerbations as defined by oral corticosteroid dispensing were determined. For the medication ratios, cutoffs between 0.3 and 0.9 (at 0.1 intervals) were examined; for the SABA measure, levels of 4 to 12 canisters per 12 months were examined. These levels were chosen to include likely optimal cutoffs based on results of prior studies.

 For each level of the predictor, odds ratios (ORs) and 95% confidence intervals (CIs) were calculated by means of simple logistic regression analysis. The optimal cutoff was determined based on (1) a significant association with a reduced risk of the outcome (upper 95% CI <1) and (2) the lowest OR associated with a significant reduction.

Optimal cutoffs for the unweighted medication ratio, weighted medication ratio, and SABA measure were then compared with the results for the HEDIS measure (dispensing of any controller) and with each other based on the magnitude of the significant effect sizes (calculated as 1 minus the OR for significant relationships). Predictor status of all patients was defined and was used in the analyses of each measure. All analyses were performed using commercially available statistical software (SAS version 9.13 for Windows; SAS Institute, Cary, NC).

Of 2,365,947 KPSC members 18 years and older in 2006, a total of 14,856 (age range, 18-56 years) met the HEDIS criteria for having persistent asthma in 2006 and were conntinuously enrolled in 2007 and 2008. Of these patients, 8634 (58.1%) requalified as having persistent asthma in 2007 (mean [SD] age, 42.7 [10.9] years; 60.5% female). Of these 8634 patients, 558 (6.5%) experienced an exacerbation as defined by emergency hospital care for asthma in 2008, and 2361 (27.3%) experienced at least 1 asthma exacerbation as defined by oral corticosteroid dispensing linked to an asthma encounter.

 Odds ratios for emergency hospital care decreased with increasing unweighted and weighted medication ratio cutoffs (

). Odds ratios for emergency hospital care for any given ratio cutoff were lower (effect size larger) for the unweighted ratio compared with the weighted ratio. The optimal cutoff for the unweighted and weighted ratios was 0.9 or higher, with an OR of 0.54 (95% CI, 0.40- 0.72) for the unweighted ratio and an OR of 0.57 (95% CI, 0.45-0.73) for the weighted ratio.

Oral Corticosteroid Dispensing—Defined Asthma Exacerbations.

 Odds ratios for oral corticosteroid dispensing— defined asthma exacerbations also decreased with increasing unweighted and weighted medication ratio cutoffs and for any given ratio cutoff were lower (effect size larger) for the unweighted ratio compared with the weighted ratio (

). However, unweighted ratio cutoffs of 0.3 or higher and 0.4 or higher were associated with increased risk of asthma exacerbations, and only an unweighted ratio cutoff of 0.9 or higher was associated with significantly reduced risk of oral corticosteroid dispensing (OR, 0.78; 95% CI, 0.68-0.90). In contrast, weighted ratio cutoffs between 0.3 and 0.8 were associated with increased risk of asthma exacerbations as defined by oral corticosteroid dispensing, and a weighted ratio cutoff of 0.9 or higher neither significantly increased nor decreased the risk.

All levels of SABA dispensing cutoffs were associated with reduced risk of emergency hospital care and oral corticosteroid dispensing—defined asthma exacerbations, although ORs were lower for emergency hospital care than for asthma exacerbations as defined by oral corticosteroid dispensing (

). The optimal cutoff for emergency hospital care was less than 11, with an OR of 0.49 (95% CI, 0.40-0.60) and for oral corticosteroid dispensing was less than 12, with an OR of 0.64 (95% CI, 0.56-0.73).

Fewer short-acting agonist canisters dispensed (but not dispensing of any controller) and higher ratio of controller to total medication are associated with reduced asthma exacerbations.

Su-Jau T. Yang, PhD

Articles

Persistent Asthma Defined Using HEDIS Versus Survey Criteria

Relationship of Asthma Control to Asthma Exacerbations Using Surrogate Markers Within a Managed Care Database