Outcomes Associated With Spirometry for Pediatric Asthma in a Managed Care Organization
http://www.100md.com
《中华首席医学网》
ABSTRACT
BACKGROUND. The National Heart, Lung and Blood Institute asthma guidelines recommend that children with asthma receive spirometry testing "at least every 1 to 2 years to assess the maintenance of airway function."
OBJECTIVE. The purpose of this work was to describe: (1) how often children with asthma receive spirometry testing, (2) what factors are associated with receipt of spirometry testing, and (3) the impact of spirometry testing on subsequent emergency department visits for asthma.
METHODS. We analyzed all pediatric asthma claims data from a university-based managed care organization for a 3-year period (January 2001 to December 2003). We included all of the continuously enrolled patients with active asthma between 7 and 21 years of age. Our outcomes of interest were the presence of 1 claim for spirometry testing (Common Procedural Terminology 94010–6, 94060, 94070, or 94150) and the time to emergency department visit. We used multivariate logistic regression to determine factors associated with receipt of spirometry and survival analyses techniques to assess the association between receipt of spirometry with the likelihood of an emergency department asthma visit in the next year, controlling for patient age, gender, severity of illness, and type of insurance.
RESULTS. There were 2688 eligible children of whom 1509 (56%) were male, 324 (12%) had Medicaid insurance, and 624 (24%) had persistent asthma in the initial year. Of the 2688 children, only 612 (23%) had 1 claim for spirometry testing during the study period. In all of the multivariate logistic analysis models, increased severity of illness was consistently associated with increased likelihood of receiving spirometry testing. Compared with patients without Medicaid insurance, children with Medicaid insurance were consistently less likely to receive spirometry testing. After adjusting for age, gender, severity, and insurance type, receipt of spirometry did not affect the likelihood of future emergency department asthma use.
CONCLUSIONS. Children with Medicaid insurance are less likely to receive spirometry testing. Reasons may be because of access to care, inadequate provider referral for testing, or patient preferences. Objective lung function tests, such as spirometry, are a potentially important component of monitoring chronic disease status. However, it is not clear whether spirometry testing by itself, completed every 1 to 2 years, helps prevent the likelihood of emergency department asthma visits. Compared with guideline recommendations, spirometry is underused; however, additional work is needed to understand how to best integrate such testing to improve asthma outcomes.
Key Words: asthma ? spirometry ? pediatrics ? Medicaid ? managed care
Abbreviations: NHLBI—National Heart, Lung, and Blood Institute ? MCO—managed care organization ? ED—emergency department ? ICD-9-CM—International Classification of Disease-Ninth Revision, Clinical Modification ? HEDIS—Health Plan Employer Data Information Set ? HMO—health maintenance organization ? POS—point-of-service ? CI—confidence interval ? OR—odds ratio
To optimize asthma management and ensure that therapeutic goals are being realized, periodic assessment and monitoring of asthma symptoms and lung function is necessary. The National Heart, Lung, and Blood Institute (NHLBI) Guidelines for the Management of Asthma recommends that children with asthma receive spirometry testing "at least every 1 to 2 years to assess the maintenance of airway function."1
Despite the importance of spirometry, studies suggest that children with asthma infrequently receive such testing. Although specialists, compared with primary care physicians, are more likely to refer patients for such testing,2 the majority of children with asthma do not receive subspecialty care for asthma.3 A national survey of primary care providers reported that only 21% of providers use spirometry routinely.4 Type of medical insurance may affect access to spirometry. Receipt of Medicaid insurance is associated with differences in access to subspecialty care for chronically ill children, including asthma,3,5 which may directly or indirectly affect access to spirometry.
The purpose of this study is to describe which factors are associated with the receipt of spirometry testing for children with asthma in a managed care organization (MCO), and if receipt of this service affects subsequent urgent asthma health care use. We hypothesized that patient and payor characteristics may affect receipt of spirometry testing for children with asthma in an MCO. In addition, we hypothesized that, after controlling for severity of illness, receipt of spirometry may be a marker for quality care, enhancing asthma management and, thus, being associated with decreased emergency department (ED) use.
METHODS
We investigated the claims records from a university-based MCO for a 3-year period between January 1, 2001, and December 31, 2003. The Institutional Review Board of the University of Michigan Health System approved the study protocol. All unique patient identifiers and physician identifiers were removed from the claims database by the MCO before analyses by the investigators.
Subjects
All patients had a claim with an International Classification of Disease-Ninth Revision, Clinical Modification (ICD-9-CM) code for asthma of 493.00 to 493.99 submitted during the study period. Because extended periods of disenrollment from the MCO would affect the likelihood of receipt of spirometry testing, consistent with definitions used by the National Council on Quality Assurance, we excluded patients if there was any disenrollment from the MCO for more than a total of 45 days during the study period. Patients between 7 and 21 years of age were included, whereas children <7 years of age were excluded, because it is technically difficult for young children to complete routine spirometry testing.1,6
Dependent Variable
The outcome of interest was any claim for spirometry testing during each 12-month period. Receipt of spirometry was defined as the presence of 1 claim for testing. These claims were identified using the Common Procedural Terminology (CPT) codes 94010–6, 94060, 94070, and 94150 during the study period. The Healthcare Common Procedure Coding System is more commonly used to report supply and durable medical equipment use; however, it can be used to report health care services. Similarly, the ICD-9-CM is commonly used to identify diagnoses but also includes codes that document procedures. As a result, we also examined the data set for ICD-CM-9 codes (89.38 and 89.37) and Healthcare Common Procedure Coding System codes (A4611 to A4627 and E0424 to E0480), which are associated with spirometry, but did not identify any of these codes in the data set.
Independent Variables
Independent variables included patient age, gender, severity of illness, and insurance type. Results were analyzed using 2 and multivariate logistic regression. Independent variables included patient factors, such as age in years and gender.
We determined severity of illness during the first full year of availability of claims data using modified criteria from the Health Plan Employer Data Information Set (HEDIS). Patients had persistent asthma (versus intermittent asthma) if they had claims consistent with any ED visit for asthma, any hospitalization for asthma at any time during a calendar year, or 4 physician-office or outpatient clinic visits for asthma with 2 asthma medication-dispensing events during a calendar year.7
We controlled for type of insurance, because insurance arrangements may affect access to subspecialist care and, hence, the receipt of spirometry.3 Although primary care providers may use spirometry in their own office, anecdotally the majority of primary care providers may rely and/or review the results of spirometry testing performed during subspecialist visits. The data were obtained from an MCO that requires patients in the traditional health maintenance organization (HMO) plan to have a referral from their primary care physician to see a specialist before the MCO will authorize payment. However, the primary care physician does not have to obtain referral preauthorization from the MCO.
The MCO has a point-of-service (POS) option, which allows a patient to see a subspecialist without the approval of a primary care physician. As a result, the subjects could potentially have 1 of 4 types of insurance: traditional HMO, POS, a program provided for students, and Medicaid. Because Medicaid has been described previously as a variable associated with differential receipt of health care services,3,5 we dichotomized the insurance variable as Medicaid versus non-Medicaid insurance.
Data Analyses
The Pearson 2 or Student t test (2-sided) were used to examine univariate associations between the dependent and independent variables. We examined associations between patient characteristics and spirometry use. Because the receipt of spirometry in a given year is a dichotomous outcome, we used multivariate logistic regression to determine the adjusted odds of receipt of spirometry for each variable (SAS 8.0, SAS Institute, Inc, Cary, NC). Separate analyses were conducted for each of the years (2001–2003) in our data set.
Survival Analysis
We examined the relationship of a spirometry claim with any subsequent ED asthma visit using separate Kaplan-Meier analyses.8 We hypothesized that receipt of spirometry may help improve assessment of asthma symptoms, improve management, and decrease the likelihood of a subsequent ED asthma visit. We did not assess the association between spirometry and asthma hospitalization because of the relative infrequency of occurrence in the data set (Table 1).
Because the NHLBI guidelines suggest that spirometry be performed at least once every 2 years, we used the first 2 years of the data set (2001–2002) to identify those patients who received spirometry or did not receive spirometry. Because asthma symptoms and ED events are affected by seasonality, to standardize the time periods, we used the complete last year of data (January 1, 2003, to December 31, 2003) to determine whether an ED asthma visit occurred. We defined an ED asthma visit as any set of claims with any of the Common Procedural Terminology codes (99281–8) for ED care and an ICD-9-CM code for asthma (493.xx).
The log-rank test was used to assess whether ED asthma visits varied significantly according to the receipt of spirometry in a 1-year time frame. To determine whether the relationship between receipt of spirometry and the likelihood of a subsequent ED asthma visit was confounded by other independent variables, we performed a Cox proportional hazards analysis. We controlled for the following variables: age, gender, severity of illness, and Medicaid insurance. Hazard ratios are presented as relative hazards with 95% confidence intervals (CIs).
RESULTS
We examined 181616 asthma claim records received for 9945 pediatric patients having 1 claim with any ICD-9-CM code for asthma at any point during their enrollment in the MCO. Of the 9945 patients, there were 7758 children with no enrollment breaks >45 days. Of the 7758 children, 4037 were >7 years of age. Of these 4037 children, only 2688 had active asthma, identified by 1 claim within the study period. All of the analysis includes this final set of 2688 patients.
Patient Characteristics
The 2688 patients had a mean age of 12.1 years (±3.5 years), and 1509 (56%) were male. There were 2167 patients (81%) enrolled in the traditional HMO plan; 324 (12%) had Medicaid insurance; 192 (7%) had a POS option. Only 5 (<1%) were enrolled in the student insurance plan. Using criteria based on the HEDIS definition of persistent asthma, we found that 657 (24%) of children had persistent asthma in 2001. The number of patients with persistent asthma increased to 820 (31%) in 2002 and 890 (33%) in 2003.
The 2688 patients received care from >356 different primary care practice sites in 2001, 378 sites in 2002, and 360 in 2003. The median number of patients per site and intraquartile range of patients for each of the years is listed in Table 1.
Spirometry Claims
Of the 2688 patients, 612 (22.8%) had a claim for spirometry testing during any of the 3 years (2001–2003) in our study period. In the first year (2001), 267 (9.9%) had spirometry, 469 (17.5%) had it in the second year (2002), and 493 (18.3%) had it in the third year (2003). In unadjusted analysis, receipt of spirometry was associated with greater asthma disease severity, older patient age, and non-Medicaid insurance status (Table 2).
Multivariate Analyses
After adjustment for all other variables, increased severity of asthma was significantly associated with receipt of spirometry testing (odds ratio [OR]: 9.63; 95% CI: 7.27–12.8). This finding was consistent in all 3 of the analyses (Table 3), although the strength of the association declined over time.
Children with Medicaid insurance were less likely to receive spirometry testing than those with non-Medicaid insurance (OR: 0.47; 95% CI: 0.29–0.75). This association was consistent in all 3 of the analyses as well. Patient age and gender were not associated with receipt of spirometry.
Survival Analyses
The receipt of spirometry was not associated with any difference in the likelihood of a subsequent ED asthma visit (P = .79) using the log rank test (Fig 1). In multivariate analysis by Cox proportional hazards regression, after adjusting for severity of illness, patient gender, age, and Medicaid insurance, the receipt of spirometry was not associated with an increase or decrease in ED asthma visits (Table 4). However, receipt of Medicaid insurance was associated with an increased risk of a subsequent ED asthma visit (relative hazard: 2.37; 95% CI: 1.66–3.39).
DISCUSSION
National guidelines recommend routine spirometry for children with asthma. We found that the likelihood for receiving spirometry is associated with several factors. In our sample, children with more severe asthma are more likely to receive spirometry, whereas children with Medicaid insurance are less likely to receive spirometry compared with those with other insurance. However, we did not find that receipt of spirometry affected the likelihood of future ED asthma visits.
Increased severity of asthma may be associated with receipt of spirometry, because management may be more complicated and require objective measures of lung function. The NHLBI asthma guidelines recommend that patients with severe persistent asthma and those patients requiring high-dose corticosteroids be seen by a subspecialist, which may also increase the likelihood of spirometry.2,4
The finding that Medicaid status is associated with a decreased likelihood of receiving spirometry may reflect health care system, physician, or patient factors associated with the referral process. In terms of health care system-related factors, parents of children with asthma have identified barriers to care that include accessibility of the health care provider, which may be compounded by financial constraints for transportation.9
Provider behavior may also account for the differences in receipt of spirometry. Although the number of subspecialists participating in the Medicaid and non-Medicaid plans in this MCO was similar, subspecialists may have been more reluctant to accept Medicaid referrals, which may affect the rate of spirometry.10 However, the services provided by the MCO in our study are based around a single, large tertiary care and specialty center. Anecdotally, specialists and scheduling staff are generally not aware of patient's type of insurance when accepting referrals. Patient-related factors may include patient adherence with referral for spirometry. Several studies have noted that referral completion for specialized asthma care was more likely for patients with private insurance compared with Medicaid.11,12
Theoretically, with more frequent symptom monitoring, spirometry may enhance asthma management and decrease the likelihood of subsequent ED asthma use. For example, previous studies have shown an association between forced expiratory volume and the risk of subsequent asthma attacks.13,14 Spirometry may be able to detect changes in forced expiratory volume and, with proper management, decrease the likelihood of an asthma exacerbation. In our analyses, we found no such difference. However, we defined patients as having received spirometry if any spirometry claim was present for a 2-year period. Although the guidelines suggest spirometry once every 1 to 2 years, more frequent testing and monitoring may be necessary for some patients. Alternatively, physicians may be monitoring symptoms using subjective measures, such as parent or patient report of the frequency of daytime and nighttime symptoms, which may obviate the need for regular spirometry use.
These analyses are based on claims data with patients nonrandomly selected to receive spirometry. An ideal design would be a prospective, randomized study to estimate the effects of spirometry testing. In our data set, physicians selected which patients would be referred for spirometry. It may be possible that the group of patients that received spirometry may have had their management modified (eg, controller medicines initiated) based on the results of the spirometry, which led to a decreased likelihood of ED visits. Thus, without spirometry, the survival curves of those patients potentially could have been worse.
Limitations
There are several other limitations to this study. We assumed that spirometry and enhanced asthma management are associated with decreased ED use for asthma. However, there are many factors associated with ED use, such as other aspects of physician management, parental attitudes, and access to primary care, that we were not able to control for in our analyses.15–17
There are limitations in our ability to control for severity of illness. For example, we included all of the children with a diagnosis of asthma, which may include those with a concurrent chronic disease, such as cystic fibrosis. If children with multiple chronic conditions are more likely to receive spirometry, as well as have a greater likelihood of ED asthma visits, the "protective" effect of spirometry may be masked. In addition, although the HEDIS criteria to determine severity of illness is very sensitive, is has a relatively low specificity.18
Although referral to spirometry may "cluster" by physician practice, we were unable to identify subjects in relation to a specific provider or site of primary care practice. For example, in 2002, 372 patients (14.6%) had claims provided by multiple primary care sites. As a result, we were unable to adjust for correlation among patients seen by the same provider or by the same primary care site.
In addition, the data are from only 1 MCO and may not be generalizable to other MCOs. However, the frequency of patients receiving spirometry (10–18%) is similar to the percentage of physicians that report using spirometry regularly (21%) in 3 MCOs.4
Implications
Despite these limitations, there are important study implications. In our sample, spirometry testing seems to be underused compared with the national recommendations. Less than 1 of 3 patients with asthma received any spirometry testing during our 3-year study period. Furthermore, the type of insurance is associated with the likelihood or receipt of spirometry testing.
Several studies have compared Medicaid and non-Medicaid patients and noted differences in patient outcomes for asthma and other diseases. This study demonstrates that this disparity in receipt of spirometry testing is greater for children with Medicaid within the same MCO. Medicaid status may be a marker for other factors at the provider, patient, or health care system level that lead to poor outcomes. Inappropriate access to spirometry testing may be one reason for these differences in outcomes. However, it is not clear from this analysis whether spirometry testing is associated with the likelihood of subsequent ED asthma visits. Nevertheless, there may be other benefits of spirometry that were not assessed in our analysis. For example, spirometry testing may help identify those patients with unrecognized persistent symptoms and may help lead to the appropriate prescription of daily asthma controller medications. Improved daily management may improve control of symptoms, daily functioning, and quality of life, which are difficult to assess using administrative claims data.
Nevertheless, there is a dearth of evidence that supports the regular use of spirometry. This lack of evidence and the results of this study question the blanket recommendation that all patients should receive regular spirometry testing. In our study, there seems to be no difference for those patients who received spirometry in the last 2 years and those patients who did not receive spirometry in terms of ED asthma visits. It is possible that a subset of patients may not need such regular testing. An additional rigorous, prospective study is needed to either support the current recommendation or suggest alternate strategies for the best use of spirometry testing.
FOOTNOTES
The authors have indicated they have no financial relationships relevant to this article to disclose.
This work was presented in part at the annual meeting of the Society for Pediatrics Research; May 14, 2005; Washington, DC.
REFERENCES
National Institutes of Health (National Heart Lung and Blood Institute). Expert Panel Report 2: Guidelines for the Diagnosis and Management of Asthma. NIH Publication No. 97–4051. National Institutes of Health: Washington, DC; 1997:26–28
O'Dowd LC, Fife D, Tenhave T, Panettieri RA. Attitudes of physicians toward objective measures of airway function in asthma. Am J Med. 2003;114 :391 –396
Cabana MD, Bruckman D, Rushton JL, Bratton SL, Green L. Receipt of asthma subspecialty care for children in a managed care organization. Ambul Pediatr. 2002;2 :456 –461
Finkelstein JA, Lozano P, Shulruff R, et al. Self-reported physician practices for children with asthma: are national guidelines followed? Pediatrics. 2000;106 :886 –896
Szilagyi PG. Managed care for children: effect on access to care and utilization of health services. Future Child. 1998;8 :39 –59
Pfaff JK, Morgan WJ. Pulmonary function in infants and children. Pediatr Clin North Am. 1994;41 :401 –423
National Committee for Quality Assurance. HEDIS 2003, Volume 2. Washington, DC: National Committee for Quality Assurance; 2003:25–28
Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53 :457 –481
Mansour ME, Lanphear BP, DeWitt TG. Barriers to asthma care in urban children: parent perspectives. Pediatrics. 2000;106 :512 –519
Access of Medicaid recipients to outpatient care. N Engl J Med. 1994;330 :1426 –1430
Forrest CB, Glade GB, Baker AE, et al. Coordination of specialty referrals and physician satisfaction with referral care. Arch Pediatr Adolesc Med. 2000;154 :499 –506
Yoon EY, Davis MM, Van Cleave J, Maheshwari S, Cabana MD. Factors associated with non-attendance at pediatric subspecialty asthma clinics. J Asthma. 2005;42 :555 –559
Fuhlbrigge AL, Kitch BT, Paltiel AD, et al. FEV(1) is associated with risk of asthma attacks in a pediatric population. J Allergy Clin Immunol. 2001;107 :61 –67
Kitch BT, Paltiel AD, Kuntz KM, et al. A single measure of FEV1 is associated with risk of asthma attacks in long-term follow-up. Chest. 2004;126 :1875 –1882
Adams RJ, Smith BJ, Ruffin RE. Factors associated with hospital admissions and repeat emergency department visits for adults with asthma. Thorax. 2000;55 :566 –573
Wakefield M, Staugas R, Ruffin R, Campbell D, Beilby J, McCaul K. Risk factors for repeat attendance at hospital emergency departments among adults and children with asthma. Aust N Z J Med. 1997;27 :277 –284
Cabana MD, Bruckman D, Bratton SL, Kemper AR, Clark NM. Association between outpatient follow-up and pediatric emergency department asthma visits. J Asthma. 2003;40 :741 –749
Cabana MD, Slish KK, Nan B, Clark NM. Limits of the HEDIS criteria in determining asthma severity for children. Pediatrics. 2004;114 :1049 –1055
a Department of Pediatrics, University of California San Francisco, San Francisco, California
b Department of Pediatrics, University of Michigan Health Care System, Ann Arbor, Michigan
c Department of Biostatistics, University of Michigan School of Public Health, Ann Arbor, Michigan
d Department of Pediatrics, University of Utah, Salt Lake City, Utah(Michael Cabana, MD, MPHa,)
BACKGROUND. The National Heart, Lung and Blood Institute asthma guidelines recommend that children with asthma receive spirometry testing "at least every 1 to 2 years to assess the maintenance of airway function."
OBJECTIVE. The purpose of this work was to describe: (1) how often children with asthma receive spirometry testing, (2) what factors are associated with receipt of spirometry testing, and (3) the impact of spirometry testing on subsequent emergency department visits for asthma.
METHODS. We analyzed all pediatric asthma claims data from a university-based managed care organization for a 3-year period (January 2001 to December 2003). We included all of the continuously enrolled patients with active asthma between 7 and 21 years of age. Our outcomes of interest were the presence of 1 claim for spirometry testing (Common Procedural Terminology 94010–6, 94060, 94070, or 94150) and the time to emergency department visit. We used multivariate logistic regression to determine factors associated with receipt of spirometry and survival analyses techniques to assess the association between receipt of spirometry with the likelihood of an emergency department asthma visit in the next year, controlling for patient age, gender, severity of illness, and type of insurance.
RESULTS. There were 2688 eligible children of whom 1509 (56%) were male, 324 (12%) had Medicaid insurance, and 624 (24%) had persistent asthma in the initial year. Of the 2688 children, only 612 (23%) had 1 claim for spirometry testing during the study period. In all of the multivariate logistic analysis models, increased severity of illness was consistently associated with increased likelihood of receiving spirometry testing. Compared with patients without Medicaid insurance, children with Medicaid insurance were consistently less likely to receive spirometry testing. After adjusting for age, gender, severity, and insurance type, receipt of spirometry did not affect the likelihood of future emergency department asthma use.
CONCLUSIONS. Children with Medicaid insurance are less likely to receive spirometry testing. Reasons may be because of access to care, inadequate provider referral for testing, or patient preferences. Objective lung function tests, such as spirometry, are a potentially important component of monitoring chronic disease status. However, it is not clear whether spirometry testing by itself, completed every 1 to 2 years, helps prevent the likelihood of emergency department asthma visits. Compared with guideline recommendations, spirometry is underused; however, additional work is needed to understand how to best integrate such testing to improve asthma outcomes.
Key Words: asthma ? spirometry ? pediatrics ? Medicaid ? managed care
Abbreviations: NHLBI—National Heart, Lung, and Blood Institute ? MCO—managed care organization ? ED—emergency department ? ICD-9-CM—International Classification of Disease-Ninth Revision, Clinical Modification ? HEDIS—Health Plan Employer Data Information Set ? HMO—health maintenance organization ? POS—point-of-service ? CI—confidence interval ? OR—odds ratio
To optimize asthma management and ensure that therapeutic goals are being realized, periodic assessment and monitoring of asthma symptoms and lung function is necessary. The National Heart, Lung, and Blood Institute (NHLBI) Guidelines for the Management of Asthma recommends that children with asthma receive spirometry testing "at least every 1 to 2 years to assess the maintenance of airway function."1
Despite the importance of spirometry, studies suggest that children with asthma infrequently receive such testing. Although specialists, compared with primary care physicians, are more likely to refer patients for such testing,2 the majority of children with asthma do not receive subspecialty care for asthma.3 A national survey of primary care providers reported that only 21% of providers use spirometry routinely.4 Type of medical insurance may affect access to spirometry. Receipt of Medicaid insurance is associated with differences in access to subspecialty care for chronically ill children, including asthma,3,5 which may directly or indirectly affect access to spirometry.
The purpose of this study is to describe which factors are associated with the receipt of spirometry testing for children with asthma in a managed care organization (MCO), and if receipt of this service affects subsequent urgent asthma health care use. We hypothesized that patient and payor characteristics may affect receipt of spirometry testing for children with asthma in an MCO. In addition, we hypothesized that, after controlling for severity of illness, receipt of spirometry may be a marker for quality care, enhancing asthma management and, thus, being associated with decreased emergency department (ED) use.
METHODS
We investigated the claims records from a university-based MCO for a 3-year period between January 1, 2001, and December 31, 2003. The Institutional Review Board of the University of Michigan Health System approved the study protocol. All unique patient identifiers and physician identifiers were removed from the claims database by the MCO before analyses by the investigators.
Subjects
All patients had a claim with an International Classification of Disease-Ninth Revision, Clinical Modification (ICD-9-CM) code for asthma of 493.00 to 493.99 submitted during the study period. Because extended periods of disenrollment from the MCO would affect the likelihood of receipt of spirometry testing, consistent with definitions used by the National Council on Quality Assurance, we excluded patients if there was any disenrollment from the MCO for more than a total of 45 days during the study period. Patients between 7 and 21 years of age were included, whereas children <7 years of age were excluded, because it is technically difficult for young children to complete routine spirometry testing.1,6
Dependent Variable
The outcome of interest was any claim for spirometry testing during each 12-month period. Receipt of spirometry was defined as the presence of 1 claim for testing. These claims were identified using the Common Procedural Terminology (CPT) codes 94010–6, 94060, 94070, and 94150 during the study period. The Healthcare Common Procedure Coding System is more commonly used to report supply and durable medical equipment use; however, it can be used to report health care services. Similarly, the ICD-9-CM is commonly used to identify diagnoses but also includes codes that document procedures. As a result, we also examined the data set for ICD-CM-9 codes (89.38 and 89.37) and Healthcare Common Procedure Coding System codes (A4611 to A4627 and E0424 to E0480), which are associated with spirometry, but did not identify any of these codes in the data set.
Independent Variables
Independent variables included patient age, gender, severity of illness, and insurance type. Results were analyzed using 2 and multivariate logistic regression. Independent variables included patient factors, such as age in years and gender.
We determined severity of illness during the first full year of availability of claims data using modified criteria from the Health Plan Employer Data Information Set (HEDIS). Patients had persistent asthma (versus intermittent asthma) if they had claims consistent with any ED visit for asthma, any hospitalization for asthma at any time during a calendar year, or 4 physician-office or outpatient clinic visits for asthma with 2 asthma medication-dispensing events during a calendar year.7
We controlled for type of insurance, because insurance arrangements may affect access to subspecialist care and, hence, the receipt of spirometry.3 Although primary care providers may use spirometry in their own office, anecdotally the majority of primary care providers may rely and/or review the results of spirometry testing performed during subspecialist visits. The data were obtained from an MCO that requires patients in the traditional health maintenance organization (HMO) plan to have a referral from their primary care physician to see a specialist before the MCO will authorize payment. However, the primary care physician does not have to obtain referral preauthorization from the MCO.
The MCO has a point-of-service (POS) option, which allows a patient to see a subspecialist without the approval of a primary care physician. As a result, the subjects could potentially have 1 of 4 types of insurance: traditional HMO, POS, a program provided for students, and Medicaid. Because Medicaid has been described previously as a variable associated with differential receipt of health care services,3,5 we dichotomized the insurance variable as Medicaid versus non-Medicaid insurance.
Data Analyses
The Pearson 2 or Student t test (2-sided) were used to examine univariate associations between the dependent and independent variables. We examined associations between patient characteristics and spirometry use. Because the receipt of spirometry in a given year is a dichotomous outcome, we used multivariate logistic regression to determine the adjusted odds of receipt of spirometry for each variable (SAS 8.0, SAS Institute, Inc, Cary, NC). Separate analyses were conducted for each of the years (2001–2003) in our data set.
Survival Analysis
We examined the relationship of a spirometry claim with any subsequent ED asthma visit using separate Kaplan-Meier analyses.8 We hypothesized that receipt of spirometry may help improve assessment of asthma symptoms, improve management, and decrease the likelihood of a subsequent ED asthma visit. We did not assess the association between spirometry and asthma hospitalization because of the relative infrequency of occurrence in the data set (Table 1).
Because the NHLBI guidelines suggest that spirometry be performed at least once every 2 years, we used the first 2 years of the data set (2001–2002) to identify those patients who received spirometry or did not receive spirometry. Because asthma symptoms and ED events are affected by seasonality, to standardize the time periods, we used the complete last year of data (January 1, 2003, to December 31, 2003) to determine whether an ED asthma visit occurred. We defined an ED asthma visit as any set of claims with any of the Common Procedural Terminology codes (99281–8) for ED care and an ICD-9-CM code for asthma (493.xx).
The log-rank test was used to assess whether ED asthma visits varied significantly according to the receipt of spirometry in a 1-year time frame. To determine whether the relationship between receipt of spirometry and the likelihood of a subsequent ED asthma visit was confounded by other independent variables, we performed a Cox proportional hazards analysis. We controlled for the following variables: age, gender, severity of illness, and Medicaid insurance. Hazard ratios are presented as relative hazards with 95% confidence intervals (CIs).
RESULTS
We examined 181616 asthma claim records received for 9945 pediatric patients having 1 claim with any ICD-9-CM code for asthma at any point during their enrollment in the MCO. Of the 9945 patients, there were 7758 children with no enrollment breaks >45 days. Of the 7758 children, 4037 were >7 years of age. Of these 4037 children, only 2688 had active asthma, identified by 1 claim within the study period. All of the analysis includes this final set of 2688 patients.
Patient Characteristics
The 2688 patients had a mean age of 12.1 years (±3.5 years), and 1509 (56%) were male. There were 2167 patients (81%) enrolled in the traditional HMO plan; 324 (12%) had Medicaid insurance; 192 (7%) had a POS option. Only 5 (<1%) were enrolled in the student insurance plan. Using criteria based on the HEDIS definition of persistent asthma, we found that 657 (24%) of children had persistent asthma in 2001. The number of patients with persistent asthma increased to 820 (31%) in 2002 and 890 (33%) in 2003.
The 2688 patients received care from >356 different primary care practice sites in 2001, 378 sites in 2002, and 360 in 2003. The median number of patients per site and intraquartile range of patients for each of the years is listed in Table 1.
Spirometry Claims
Of the 2688 patients, 612 (22.8%) had a claim for spirometry testing during any of the 3 years (2001–2003) in our study period. In the first year (2001), 267 (9.9%) had spirometry, 469 (17.5%) had it in the second year (2002), and 493 (18.3%) had it in the third year (2003). In unadjusted analysis, receipt of spirometry was associated with greater asthma disease severity, older patient age, and non-Medicaid insurance status (Table 2).
Multivariate Analyses
After adjustment for all other variables, increased severity of asthma was significantly associated with receipt of spirometry testing (odds ratio [OR]: 9.63; 95% CI: 7.27–12.8). This finding was consistent in all 3 of the analyses (Table 3), although the strength of the association declined over time.
Children with Medicaid insurance were less likely to receive spirometry testing than those with non-Medicaid insurance (OR: 0.47; 95% CI: 0.29–0.75). This association was consistent in all 3 of the analyses as well. Patient age and gender were not associated with receipt of spirometry.
Survival Analyses
The receipt of spirometry was not associated with any difference in the likelihood of a subsequent ED asthma visit (P = .79) using the log rank test (Fig 1). In multivariate analysis by Cox proportional hazards regression, after adjusting for severity of illness, patient gender, age, and Medicaid insurance, the receipt of spirometry was not associated with an increase or decrease in ED asthma visits (Table 4). However, receipt of Medicaid insurance was associated with an increased risk of a subsequent ED asthma visit (relative hazard: 2.37; 95% CI: 1.66–3.39).
DISCUSSION
National guidelines recommend routine spirometry for children with asthma. We found that the likelihood for receiving spirometry is associated with several factors. In our sample, children with more severe asthma are more likely to receive spirometry, whereas children with Medicaid insurance are less likely to receive spirometry compared with those with other insurance. However, we did not find that receipt of spirometry affected the likelihood of future ED asthma visits.
Increased severity of asthma may be associated with receipt of spirometry, because management may be more complicated and require objective measures of lung function. The NHLBI asthma guidelines recommend that patients with severe persistent asthma and those patients requiring high-dose corticosteroids be seen by a subspecialist, which may also increase the likelihood of spirometry.2,4
The finding that Medicaid status is associated with a decreased likelihood of receiving spirometry may reflect health care system, physician, or patient factors associated with the referral process. In terms of health care system-related factors, parents of children with asthma have identified barriers to care that include accessibility of the health care provider, which may be compounded by financial constraints for transportation.9
Provider behavior may also account for the differences in receipt of spirometry. Although the number of subspecialists participating in the Medicaid and non-Medicaid plans in this MCO was similar, subspecialists may have been more reluctant to accept Medicaid referrals, which may affect the rate of spirometry.10 However, the services provided by the MCO in our study are based around a single, large tertiary care and specialty center. Anecdotally, specialists and scheduling staff are generally not aware of patient's type of insurance when accepting referrals. Patient-related factors may include patient adherence with referral for spirometry. Several studies have noted that referral completion for specialized asthma care was more likely for patients with private insurance compared with Medicaid.11,12
Theoretically, with more frequent symptom monitoring, spirometry may enhance asthma management and decrease the likelihood of subsequent ED asthma use. For example, previous studies have shown an association between forced expiratory volume and the risk of subsequent asthma attacks.13,14 Spirometry may be able to detect changes in forced expiratory volume and, with proper management, decrease the likelihood of an asthma exacerbation. In our analyses, we found no such difference. However, we defined patients as having received spirometry if any spirometry claim was present for a 2-year period. Although the guidelines suggest spirometry once every 1 to 2 years, more frequent testing and monitoring may be necessary for some patients. Alternatively, physicians may be monitoring symptoms using subjective measures, such as parent or patient report of the frequency of daytime and nighttime symptoms, which may obviate the need for regular spirometry use.
These analyses are based on claims data with patients nonrandomly selected to receive spirometry. An ideal design would be a prospective, randomized study to estimate the effects of spirometry testing. In our data set, physicians selected which patients would be referred for spirometry. It may be possible that the group of patients that received spirometry may have had their management modified (eg, controller medicines initiated) based on the results of the spirometry, which led to a decreased likelihood of ED visits. Thus, without spirometry, the survival curves of those patients potentially could have been worse.
Limitations
There are several other limitations to this study. We assumed that spirometry and enhanced asthma management are associated with decreased ED use for asthma. However, there are many factors associated with ED use, such as other aspects of physician management, parental attitudes, and access to primary care, that we were not able to control for in our analyses.15–17
There are limitations in our ability to control for severity of illness. For example, we included all of the children with a diagnosis of asthma, which may include those with a concurrent chronic disease, such as cystic fibrosis. If children with multiple chronic conditions are more likely to receive spirometry, as well as have a greater likelihood of ED asthma visits, the "protective" effect of spirometry may be masked. In addition, although the HEDIS criteria to determine severity of illness is very sensitive, is has a relatively low specificity.18
Although referral to spirometry may "cluster" by physician practice, we were unable to identify subjects in relation to a specific provider or site of primary care practice. For example, in 2002, 372 patients (14.6%) had claims provided by multiple primary care sites. As a result, we were unable to adjust for correlation among patients seen by the same provider or by the same primary care site.
In addition, the data are from only 1 MCO and may not be generalizable to other MCOs. However, the frequency of patients receiving spirometry (10–18%) is similar to the percentage of physicians that report using spirometry regularly (21%) in 3 MCOs.4
Implications
Despite these limitations, there are important study implications. In our sample, spirometry testing seems to be underused compared with the national recommendations. Less than 1 of 3 patients with asthma received any spirometry testing during our 3-year study period. Furthermore, the type of insurance is associated with the likelihood or receipt of spirometry testing.
Several studies have compared Medicaid and non-Medicaid patients and noted differences in patient outcomes for asthma and other diseases. This study demonstrates that this disparity in receipt of spirometry testing is greater for children with Medicaid within the same MCO. Medicaid status may be a marker for other factors at the provider, patient, or health care system level that lead to poor outcomes. Inappropriate access to spirometry testing may be one reason for these differences in outcomes. However, it is not clear from this analysis whether spirometry testing is associated with the likelihood of subsequent ED asthma visits. Nevertheless, there may be other benefits of spirometry that were not assessed in our analysis. For example, spirometry testing may help identify those patients with unrecognized persistent symptoms and may help lead to the appropriate prescription of daily asthma controller medications. Improved daily management may improve control of symptoms, daily functioning, and quality of life, which are difficult to assess using administrative claims data.
Nevertheless, there is a dearth of evidence that supports the regular use of spirometry. This lack of evidence and the results of this study question the blanket recommendation that all patients should receive regular spirometry testing. In our study, there seems to be no difference for those patients who received spirometry in the last 2 years and those patients who did not receive spirometry in terms of ED asthma visits. It is possible that a subset of patients may not need such regular testing. An additional rigorous, prospective study is needed to either support the current recommendation or suggest alternate strategies for the best use of spirometry testing.
FOOTNOTES
The authors have indicated they have no financial relationships relevant to this article to disclose.
This work was presented in part at the annual meeting of the Society for Pediatrics Research; May 14, 2005; Washington, DC.
REFERENCES
National Institutes of Health (National Heart Lung and Blood Institute). Expert Panel Report 2: Guidelines for the Diagnosis and Management of Asthma. NIH Publication No. 97–4051. National Institutes of Health: Washington, DC; 1997:26–28
O'Dowd LC, Fife D, Tenhave T, Panettieri RA. Attitudes of physicians toward objective measures of airway function in asthma. Am J Med. 2003;114 :391 –396
Cabana MD, Bruckman D, Rushton JL, Bratton SL, Green L. Receipt of asthma subspecialty care for children in a managed care organization. Ambul Pediatr. 2002;2 :456 –461
Finkelstein JA, Lozano P, Shulruff R, et al. Self-reported physician practices for children with asthma: are national guidelines followed? Pediatrics. 2000;106 :886 –896
Szilagyi PG. Managed care for children: effect on access to care and utilization of health services. Future Child. 1998;8 :39 –59
Pfaff JK, Morgan WJ. Pulmonary function in infants and children. Pediatr Clin North Am. 1994;41 :401 –423
National Committee for Quality Assurance. HEDIS 2003, Volume 2. Washington, DC: National Committee for Quality Assurance; 2003:25–28
Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53 :457 –481
Mansour ME, Lanphear BP, DeWitt TG. Barriers to asthma care in urban children: parent perspectives. Pediatrics. 2000;106 :512 –519
Access of Medicaid recipients to outpatient care. N Engl J Med. 1994;330 :1426 –1430
Forrest CB, Glade GB, Baker AE, et al. Coordination of specialty referrals and physician satisfaction with referral care. Arch Pediatr Adolesc Med. 2000;154 :499 –506
Yoon EY, Davis MM, Van Cleave J, Maheshwari S, Cabana MD. Factors associated with non-attendance at pediatric subspecialty asthma clinics. J Asthma. 2005;42 :555 –559
Fuhlbrigge AL, Kitch BT, Paltiel AD, et al. FEV(1) is associated with risk of asthma attacks in a pediatric population. J Allergy Clin Immunol. 2001;107 :61 –67
Kitch BT, Paltiel AD, Kuntz KM, et al. A single measure of FEV1 is associated with risk of asthma attacks in long-term follow-up. Chest. 2004;126 :1875 –1882
Adams RJ, Smith BJ, Ruffin RE. Factors associated with hospital admissions and repeat emergency department visits for adults with asthma. Thorax. 2000;55 :566 –573
Wakefield M, Staugas R, Ruffin R, Campbell D, Beilby J, McCaul K. Risk factors for repeat attendance at hospital emergency departments among adults and children with asthma. Aust N Z J Med. 1997;27 :277 –284
Cabana MD, Bruckman D, Bratton SL, Kemper AR, Clark NM. Association between outpatient follow-up and pediatric emergency department asthma visits. J Asthma. 2003;40 :741 –749
Cabana MD, Slish KK, Nan B, Clark NM. Limits of the HEDIS criteria in determining asthma severity for children. Pediatrics. 2004;114 :1049 –1055
a Department of Pediatrics, University of California San Francisco, San Francisco, California
b Department of Pediatrics, University of Michigan Health Care System, Ann Arbor, Michigan
c Department of Biostatistics, University of Michigan School of Public Health, Ann Arbor, Michigan
d Department of Pediatrics, University of Utah, Salt Lake City, Utah(Michael Cabana, MD, MPHa,)