A retrospective study of the use of fluticasone propionate/salmeterol combination as initial asthma controller therapy in a commercially insured population

EXHIBIT 102
Clinical Therapeutics/Volume 30, Number 10, 2008
A Retrospective Study of the Use of Fluticasone Propionate/
Salmeterol Combination as Initial Asthma Controller
Therapy in a Commercially Insured Population
Howard Friedman, PhD1; Teresa Wilcox, PhD2; Gregory Reardon, PhD3; Simone Crespi, MPH4; and Barbara P. Yawn, MD, MSc5 1Analytic Solutions, Inc., New York, New York; 2United BioSource Corporation, Bethesda, Maryland; 3Informagenics, LLC, Worthington, Ohio; 4Schering-Plough Corporation, Kenilworth, New Jersey; and 5Ol-msted Medical Center, Minneapolis, Minnesota ABSTRACT
had an ED visit with an OCS prescription, and 1.5% Background: Asthma management guidelines state
that a low-dose inhaled corticosteroid (ICS) is the Conclusion: More than two thirds of the patients
preferred treatment for mild persistent asthma and who initiated FPS treatment had neither received an ICS that coadministration of a long-acting β -agonist prescription before their first FPS pharmacy claim nor (LABA) should be reserved for patients whose asthma had evidence of asthma severity that would appear to is uncontrolled by single-entity ICS. However, it ap- warrant the use of an ICS/LABA combination. (Clin Ther. pears that many patients in the United States with 2008;30:1908–1917) 2008 Excerpta Medica Inc.
mild persistent asthma are initially treated with com- Key words: asthma guidelines, inhaled cortico-
binations of fluticasone propionate/salmeterol (FPS).
steroids, fluticasone propionate/salmeterol, initial thera- Objective: The aim of this study was to examine
whether use of FPS was consistent with asthma man-agement guidelines.
Methods: A commercial insurance database was
INTRODUCTION
analyzed retrospectively to identify patients aged 12 to Low-dose inhaled corticosteroid (ICS) therapy is rec- 62 years who had ≥1 pharmacy claim for FPS between ommended by the National Asthma Education and October 1, 2004, and September 30, 2006. An index Prevention Program (NAEPP) guidelines as the pre- date corresponding to the date of the first FPS phar- ferred initial treatment for patients with mild persis- macy claim was assigned to each patient. Medi- tent asthma.1 The NAEPP guidelines also state that a cal and pharmacy claims data were analyzed for the long-acting β -agonist (LABA), such as salmeterol or 365-day period before the index date (preindex peri- formoterol, may be used with an ICS for moderate to od). The severity of patients’ asthma was inferred from severe persistent asthma if control is inadequate, as their history of claims. Patients were identified as hav- indicated by manifestations such as frequent symp- ing more severe asthma if, during the preindex peri- toms or nocturnal awakenings or overreliance on a od, they either received >365 doses of short-acting β -agonists (SABAs), an oral corticosteroid (OCS), or an The recommendations for LABA use promulgated emergency department (ED) asthma visit with an OCS by asthma treatment guidelines should be considered in prescription, or were hospitalized for their asthma.
the context of advisory and boxed warnings issued by Results: Among 87,459 patients with new FPS
the US Food and Drug Administration (FDA). In No- claims, 60.8% were female, and the mean age was vember 2005, the FDA issued a public health advisory 37.3 years. Of these patients, 60,453 (69.1%) had no preindex ICS pharmacy claim or claims that would Accepted for publication August 14, 2008. indicate moderate or severe asthma. In the preindex period, only 6.3% had received an ICS, 7.4% had >365 SABA doses, 22.7% had used an OCS, 1.1% 2008 Excerpta Medica Inc. All rights reserved.
Volume 30 Number 10
H. Friedman et al.
about the increased risk of deaths and life-threatening Minnesota) was studied to capture data on patients experiences in asthma patients using LABAs.2 This warn- who started FPS therapy between October 1, 2004, ing reflected the findings of the Salmeterol Multicenter and September 30, 2006 (index date). This database Asthma Research Trial (SMART), which was conducted includes patients in exclusive-provider organizations, between 1996 and 2003.3 The SMART study investi- health maintenance organizations, indemnity plans, gated the safety of adding salmeterol or placebo to usual point-of-service plans, and preferred-provider organi- asthma care, which included ICS use (47% of the study zations (specific features of these plans are not noted population) in patients aged ≥12 years who had not in the database). It contains data about enrollment, previously used LABAs. The data revealed a small, but medical and pharmacy claims, and laboratory tests significant, increase in respiratory- and asthma-related for 24 million beneficiaries covered over a period of deaths when compared with placebo (respiratory-related 4.5 years and 14 million beneficiaries continuously deaths: 24 vs 11, relative risk, 2.16 [95% CI, 1.06– enrolled for 12 months. The vast majority of benefi- 4.41]; asthma-related deaths: 13 vs 3, relative risk, 4.37 ciaries are covered by commercial insurance (93%), [95% CI, 1.25–15.34]). In March 2006, the FDA man- with Medicaid (5%) and Medicare (2%) covering the dated boxed warnings for LABAs.4 The current labeling remainder. Approximately 75% of the beneficiaries are for ICS/LABA combination products, including flutica- located in the south or north central United States. sone propionate/salmeterol (FPS), contains a boxed warn- An index date corresponding to the date of the first ing stating that LABAs may increase the risk of asthma- FPS pharmacy claim during the time period between related death and that ICS/LABA combinations should October 1, 2004, and September 30, 2006, was as- be prescribed only for patients who are not adequately signed to each identified patient. Medical and controlled on other asthma-controller medications or pharmacy claims data were analyzed for the 365-day whose disease severity clearly warrants initiation of treat- period before the index date (preindex period).
ment with 2 maintenance therapies.5 Despite the NAEPP Institutional review board approval was not re- guidelines and the boxed warning, it has been reported quired for this retrospective study. The privacy of each that approximately one third of patients with mild per- patient was protected by the assignment of a number sistent asthma are treated with FPS.6 A recent review of to serve as a unique identifier before any data were administrative claims data from the Arkansas Medicaid and State Child Health Insurance Program (SCHIP) found that only 11% of asthma patients who had re- Study Population
ceived prescriptions for FPS had a previous claim for an To be eligible for inclusion in the study (Figure 1),
ICS prescription.7 In addition, fewer than 25% of the patients were required to have at least 1 pharmacy patients who received FPS prescriptions had used phy- claim for FPS during the patient identification period, sician services, rescue medications, or systemic corti- be aged 12 to 62 years on the index date, be eligible costeroids at levels suggestive of moderate or severe for pharmacy and medical benefits, and be continu- persistent asthma during the 6 to 12 months before ously enrolled in the health plan for at least 12 months before the index date. The age range for inclusion was Using data from a large commercial insurance data- selected because all ICSs are approved for use in pa- base, the present study investigated the proportion of tients aged ≥12 years, and because many patients aged asthma patients who were newly prescribed FPS with- >62 years are covered by Medicare rather than com- out evidence of an initial trial of single-entity ICS or mercial insurance. In addition, several exclusion crite- whose resource utilization data did not indicate a dis- ria were applied: any medical service claim coded ease severity that warranted the use of an ICS/LABA during the preindex period (or on the index date) with fixed combination, to determine whether the use of FPS a primary or secondary diagnosis of chronic obstruc- was consistent with asthma management guidelines.
tive pulmonary disease (International Classification of Diseases, Ninth Revision, Clinical Modification MATERIALS AND METHODS
[ICD-9-CM] codes 491.xx, 492.xx, 493.2x, 494.xx, Study Design
496.xx, 770.2), pulmonary hypertension (416.xx), In this retrospective study, a US commercial insur- pulmonary embolism (415.xx), or other pulmonary ance database (Ingenix LabRx, Ingenix, Eden Prairie, circulatory disorder (417.xx); or a filled prescription October 2008
Clinical Therapeutics
lated from the gram weight of albuterol dispensed divided by 1/100 of the gram weight of the canister) or >365 inhalation unit doses (solution or capsule, calculated from the total volume of nebules or cap- sules dispensed) of albuterol or levalbuterol within 365 days before the index date, indicating potential overreliance on a SABA1; ≥1 pharmacy claims for oral corticosteroids (OCSs) within 365 days before the index date; an urgent-care (UC) or emergency- department (ED) visit for asthma (primary or second- ary ICD-9-CM code 493.xx) followed by a prescrip-tion for an OCS (within 7 days of the visit) within 365 days before the index date; or a hospital admis- sion for asthma within 365 days before the index date. Previous claims for leukotriene modifiers or theophylline were not used to categorize patients for more severe disease; these drugs are not preferred treatments, and there are no guidelines recommended for patients whose disease is not well controlled with these medications.1 Patients classified as no ICS use/milder disease had an index FPS prescription but did not meet any of the previously noted criteria.
For each filled prescription of an albuterol canister, the Ingenix database reports the total metric quantity dispensed (in grams of albuterol) and a National Drug Code (NDC) value for each prescription fill dispensed. Figure 1. Flowchart of patient selection from a Using the NDC value, the Multum drug database (Denver, Colorado) was cross-referenced for the total weight of albuterol contained in each package unit that was dispensed. The ratio of grams dispensed per fill (from Ingenix) to grams per canister dispensed who had ≥1 pharmacy claims for flutica- (from Multum) was used to estimate number of can- isters dispensed per fill. However, because this was a dose combination between October 1, 2004, and September 30, 2006.
retrospective study, the reliability of these values could not be evaluated.
Analysis
for >1 dose strength of FPS on the index date. Only 1 dose In addition to recording patients’ age and sex at the strength was allowed, in order to analyze FPS use index date and classifying them as either ICS use/more based on dosage. This criterion excluded only 1.1% of severe disease or no ICS use/milder disease, several other variables of interest were identified: dose of FPS Patients were classified into either of 2 study groups at the index date; medical and prescription history for based on their claims experience in the preindex pe- the 365 days before the index date, including use of riod: ICS use/more severe disease or no ICS use/milder SABAs, LABAs, ICSs, and OCSs; and asthma-related disease. Those classified as ICS use/more severe dis- ease were required to satisfy at least 1 of the following Continuous variables were characterized by mean criteria: at least 1 pharmacy claim for an ICS with- and standard deviation, and categorical variables in 365 days before the index date; >365 doses of al- were characterized by frequency and percentage buterol from an inhaler (2 inhalations per dose, calcu- within each category. All analyses were performed Volume 30 Number 10
H. Friedman et al.
using SAS version 8.2 software (SAS Institute Inc., Cary, North Carolina).
Table I. Demographic characteristics and flutica- Study Population
The population of eligible patients (N = 87,459) had a mean age of 37.3 years, and 60.8% were female (Table I). Patients were from many regions of the
United States: 24% from the Southeast, 23% from the Great Lakes region, 23% from the South Central re- gion, 13% from the Northeast, 11% from the Moun- tain states, 5% from the Western states, and <1% from unknown regions of the United States. On the index date, 93.8% of the FPS prescriptions were for the 100/50-μg (43.5%) or 250/50-μg (50.3%) doses.
Outcomes
Patients’ preindex disease severity indicators are summarized in Table II. Figure 2 shows the distribu-
tion of patients according to preindex ICS use or se- verity classification. Of the total eligible population, 69.1% of patients were classified as no ICS use/milder disease (ie, had no evidence during the preindex period of ICS use or disease severity warranting the use of 2 controller medications). Only 6.3% of patients in the total population had a claim for an ICS during the preindex period. Furthermore, during the preindex In the subpopulation of patients with evidence of period, 7.4% of patients had evidence of overreliance previous ICS use or more severe disease in the pre- on SABA treatment, and 22.7% were prescribed an index period, 20.4% were prescribed an ICS, 73.5% OCS. Rates of unscheduled asthma-related hospitaliza- were prescribed an OCS, and 23.9% were estimated tions, and ED or UC visits (plus an OCS prescription to be overreliant on SABA use. In addition, 8.5% had within 7 days), were very low: 1.5% and 1.1%, respec- asthma-related hospitalizations, ED visits, or UC vis- tively. As shown in Figure 3, patients using FPS at the
its. Each estimate was calculated from patient counts higher doses during the preindex period were more shown in Table II by dividing the number of patients
likely to have ICS use/more severe disease than those with each criteria by the total number of ICS use/more using a lower dose, although fewer than half of all pa- severe disease patients (27,006); however, proportions tients using even the highest strength of FPS had do not total 100% because the groups are not mutu- ally exclusive (eg, a patient could have received The assessment of preindex prescription histories both an OCS and an ICS prescription in the preindex in the total population revealed that 48.8% of eligible patients had no evidence of previous asthma medi- As shown in Figure 4, ICS prescribing was reason-
cations during the 365 days. Furthermore, 17.5% of ably constant throughout most of the study period. eligible patients had received a SABA as the only However, Figure 4 appears to show trends toward a
asthma medication before the initiation of FPS small decrease in the proportion of patients with pre- (Table III). For the Medicaid subpopulation with
index ICS prescriptions after the FDA issued its advi- FPS claims in our study, only 9.9% had pre- sory about LABAs in November 2005 and a small index ICS claims, and 10.5% had preindex asthma- increase in preindex ICS prescriptions after the FDA related hospitalizations or ED/UC claims coded for issued its mandate for boxed warnings for LABAs in October 2008
Clinical Therapeutics
Table II. Assessment of preindex disease severity indicators in the total population examined from a retro- spective analysis of a commercial insurance database. Data were analyzed for patients aged 12 to 62 years who had ≥1 pharmacy claims for fluticasone propionate/salmeterol (FPS) fixed-dose com-bination between October 1, 2004, and September 30, 2006.
UC/ED visits with OCS within 7 days of index date ICS = inhaled corticosteroid; Rx = prescription; OCS = oral corticosteroid; SABA = short-acting β -agonist; UC/ED = urgent care or emergency department.
* Patients may have had >1 indicator of more severe disease.
† Defined as ≥365 doses (according to the preindex period) of albuterol inhaler (2 inhalations per dose), or ≥365 inhalation unit doses of albuterol or levalbuterol within 365 days preindex. DISCUSSION
warning campaigns were instituted for a number of Within a large population of asthma patients enrolled drugs. The authors concluded that although prescrip- in managed care plans, 93.7% were newly prescribed tions for these drugs were sometimes reduced, such an FPS regimen without filling a single-entity ICS pre- campaigns were not always effective. For example, a scription within the previous 365 days. Furthermore, time-series analysis of prescriptions for propoxyphene 69.1% of all patients who were prescribed FPS in this was conducted in the mid-1970s after both the FDA study population did not have asthma resource utili- and drug manufacturer initiated a campaign to inform zation indicating sufficient disease severity to warrant physicians of the high risk of habituation and overdose. The refill rates of propoxyphene remained near a constant For the patients prescribed FPS in the Medicaid 43% in the 8 quarters before and the 3 quarters after an subpopulation, only a small percentage (9.9%) had a FDA warning on no-refills.9 A recent retrospective previous ICS claim. These findings for the Medicaid study found that a greater proportion of Medicaid subpopulation are consistent with those of a study patients (48%) than commercially insured patients that examined Medicaid and SCHIP claims in Arkan- (11%) had preindex ICS use and discontinued using sas,7 in which only 11% of new FPS users had a previ- LABAs after the FDA advisory (28% and 17%, re- ous ICS claim and <25% had claims for physician spectively).10 These results are in contrast to those of services or medications at levels suggestive of moder- the present study, which found a relatively flat trend for ICS prescriptions before and after the FDA advi- One area of interest in the current study was the sory and boxed warnings for LABAs. Differences be- temporal relationship between FDA announcements tween the present study results and those of the previ- regarding FPS use and prescribing trends for this agent. ous study might stem from differences in the study Soumerai et al8 reviewed studies that evaluated the designs. For instance, the previous study required pa- prescribing patterns of physicians after government tients to have prescriptions for a LABA (including Volume 30 Number 10
H. Friedman et al.
One limitation of the present study was a lack of information on direct clinical measures of asthma se- verity. However, the proxy definition of mild asthma in our study was based on prescriptions for SABAs (ie, <365 unit doses) and OCSs (ie, none) in the 365-day preindex period, and on the frequency of exacerba-tions, ED visits, or hospitalizations (ie, none) in the preindex period. Similar proxy measures of asthma severity were used in an observational study based on claims analyses.11 The proxy measures of asthma se- verity used in the present study appear to be consis-tent with asthma management guidelines that dis- tinguish between levels of asthma severity (mild, moderate, and severe) in previously untreated patients and levels of asthma control (well controlled, partly controlled, uncontrolled) that patients achieve with their current therapy.12 Patients in the present study who were overreliant on SABAs or had asthma-related ED visits or hospitalizations were, at best, partly con- trolled based on guideline definitions and were classi- use or severity classification in the total fied as having more severe asthma that may have warranted treatment with FPS as initial controller therapy. In contrast, patients defined as having mild asthma in the present study had low numbers of pre- aged 12 to 62 years who had ≥1 pharma- index SABA claims and no preindex ED visits or hos- pitalizations, and the vast majority had no preindex ICS prescription. Based on the guidelines, these pa- tients should have received prescriptions for an ICS before FPS therapy was initiated unless they presented Methods section for the definition of more severe disease.) to their physicians with signs or symptoms of poor asthma control.
Different criteria for classification of asthma severity have been used elsewhere. In The Epidemiology and FPS) in the preindex period and for an anti- Natural History of Asthma: Outcomes and Treatment asthmatic drug in the postindex period. This require- Regimens (TENOR) study,13 3% of patients were clas- ment would select patients with more severe asthma sified by their physicians as having mild persis- than those in the present study, which required only tent asthma that they considered difficult to treat. The TENOR study found that 6% of mild asthma patients Although the cost of a drug (ie, copayments or had ED visits within the previous 3 months.This fig- cost-sharing provisions) may have affected physician ure is >5 times higher than the proportion of patients and patient behavioral choices, such as which drug to in the present study (1.1%) with preindex ED visits. prescribe or how often to use the medication, the da- An evaluation of asthma severity classification of tabase does not include this information and the cur- TENOR patients based on criteria of the 2002 NAEPP rent study does not address these issues. It is also not guidelines found that 29% of patients had mild per- known how accurately pharmacy claims data reflect sistent asthma, whereas only 4% of TENOR patients patients’ adherence to prescribed treatment. In addition, had mild persistent asthma based on the criteria of the the prescribing information of physicians by specialty Global Initiative for Asthma guidelines.14 was not captured in the database and is beyond the scope Findings in the present study do not preclude the of this study. Further research on this topic is needed.
possibility that some patients may have had no prein- October 2008
Clinical Therapeutics
No ICS use/milder diseaseICS use/more severe disease Figure 3. Index fluticasone propionate/salmeterol (FPS) fixed-dose combination doses in the total study population from a retrospective analysis of a commercial insurance database. Data were analyzed for patients aged 12 to 62 years who had ≥1 pharmacy claims for FPS between October 1, 2004, and September 30, 2006. (See the Materials and Methods section for the definition of more severe dis-ease.) ICS = inhaled corticosteroid.
Table III. One-year preindex prescription history in the total population examined in a retrospective analysis of a commercial insurance database. Data were analyzed for patients aged 12 to 62 years who had ≥1 pharmacy claims for fluticasone propionate/salmeterol (FPS) fixed-dose combination between October 1, 2004, and September 30, 2006.
SABA = short-acting β -agonist; LABA = long-acting β -agonist; ICS = inhaled corticosteroid; OCS = oral corticosteroid. Volume 30 Number 10
H. Friedman et al.
Figure 4. Inhaled corticosteroid prescription trends relative to US Food and Drug Administration (FDA) ad- visory and long-acting β -agonist (LABA) boxed warnings from a retrospective analysis of a com- mercial insurance database. Data were analyzed for patients aged 12 to 62 years who had ≥1 phar-macy claims for fluticasone propionate/salmeterol fixed-dose combination between October 1, 2004, and September 30, 2006. Point A represents FDA advisory on LABA safety; point B represents FDA-mandated boxed warnings for LABAs.
dex ICS claims or asthma history but presented on the step-up therapy is not often applied as recommended index date with asthma of a severity that warranted in asthma management guidelines or evaluated in con- initial treatment with 2 controller medications. Some trolled clinical trials. For example, a review of medical patients may have had poor asthma control because records in Olmsted County, Minnesota, found that of treatment failure, whereas others may have had prescriptions for asthma medications were episodic, poor asthma control because of nonadherence to the with 52% of medication changes occurring during prescribed treatment. Medication nonadherence is a office visits after an asthma flare-up, and 14% occur- major obstacle to asthma control, and it can be a ring in the ED. In addition, 64% of the medication problem with ICS therapy, specifically. Previous stud- changes were step-ups, whereas 9% were step-downs ies have shown that, in practice, the underuse of ICSs in current medication dosage and 4% were step- and overuse of β -agonists, whether short- or long- downs in the class of prescribed medication.21 The use acting, are common.15,16 In spite of this, the clinical of FPS for patients who have not stepped up through benefits of stepping up from low-dose single-entity single-entity ICS may not be necessary unless the pa- ICS to combined ICS/LABA therapy in patients not tients’ asthma severity clearly warrants initiation of adequately controlled with single-entity ICS are well treatment with 2 controller therapies.22–24 established and reflected in the NAEPP guidelines.1 Many patients with asthma are prescribed ICS/ Clinical studies have shown that the addition of a LABA combination therapy before it is necessary, LABA to an ICS in patients whose asthma is not ade- when viewed in the context of NAEPP guidelines and quately controlled with ICS monotherapy improves the FDA boxed warning on LABA-containing combi- asthma control by a greater extent than is achieved nation products. Although the reasons for potential by increasing the ICS dose17–19 or adding an antileu- overuse of ICS/LABA combination therapy have not kotriene.20 In real-world clinical practice, however, yet been elucidated, the practice of reserving the use of October 2008
Clinical Therapeutics
4. US Food and Drug Administration. Advair Diskus, Advair these combination agents for patients whose disease HFA, Brovana, Foradil, Perforomist, Serevent Diskus, and is not adequately controlled with single-entity ICS Symbicort information (long acting beta agonists). http:// per asthma management guidelines is not followed www.fda.gov/cder/drug/infopage/LABA/default.htm. 5. Advair Diskus (fluticasone propionate and salmeterol CONCLUSION
inhalation powder) [prescribing information]. Research More than two thirds of the patients who initiated Triangle Park, NC: GlaxoSmithKline; 2008.
FPS treatment had neither received an ICS prescrip- 6. Nathan RA. Utilizing mometasone furoate dry powder tion before the first FPS pharmacy claim nor had evi- inhaler to get back to basics with asthma guidelines. dence of asthma severity that would appear to war- Presented at: Aspen Allergy Conference; July 26–29, 2006; rant the use of an ICS/LABA combination.
7. Helm ME, Vargas PA, Jones SM. Advair prescribing pat- ACKNOWLEDGMENTS
terns in a state Medicaid and child health insurance program. J Allergy Clin Immunol. 2007;119(Suppl):S170.
This study was sponsored by Schering-Plough Corpo- 8. Soumerai SB, McLaughlin TJ, Avorn J. Improving drug ration. Editorial assistance was provided by Ken Kauff- prescribing in primary care: A critical analysis of the man, BSc, and Erin P. Scott, PhD; this assistance was experimental literature. Milbank Q. 1989;67:268–317.
funded by Schering-Plough. Drs. Friedman and Rear- 9. Soumerai SB, Avorn J, Gortmaker S, Hawley S. Effect of don provide consulting services to Schering-Plough. government and commercial warnings on reducing pre- Ms. Crespi is an employee of Schering-Plough. Dr. Wil- scription misuse: The case of propoxyphene. Am J Public cox has received grant/research support from Adams Respiratory Therapeutics, Altana AG (Nycomed), Astra- 10. Baker D, Rasu R, Burke S, Toe D. Evaluation of asthma- Zeneca, Boehringer Ingelheim, Forest Laboratories, related utilization in a managed care population following Inc., GlaxoSmithKline, Merck & Company, Novartis, a FDA public health advisory on long-acting β2-agonists. Ortho-McNeil, Pfizer Inc., Sepracor Inc., and Schering- Presented at: Academy of Managed Care Pharmacy 19th Plough, and is a stock shareholder of GlaxoSmithKline. Annual Meeting; April 13, 2007; San Diego, Calif.
11. Colice G, Wu EQ, Birnbaum H, et al. Healthcare and Dr. Yawn has received research grants from Astra- workloss costs associated with patients with persistent Zeneca, GlaxoSmithKline, Boehringer Ingelheim, Pfizer, asthma in a privately insured population. J Occup Environ and Schering-Plough for work in asthma and chronic obstructive pulmonary disease; she has served as an 12. Global Initiative for Asthma (GINA). GINA report: Global advisor to Schering-Plough, AstraZeneca, Merck & strategy for asthma management and prevention. http:// Company, Boehringer Ingelheim, and Pfizer.
www.ginasthma.com/Guidelineitem.asp??l1=2&l2= 1&intId=60. Accessed December 6, 2006.
REFERENCES
13. Dolan CM, Fraher KE, Bleecker ER, et al, for the TENOR 1. National Heart, Lung, and Blood Institute. Expert Panel Study Group. Design and baseline characteristics of the Report 3 (EPR3): Guidelines for the diagnosis and man- epidemiology and natural history of asthma: Outcomes agement of asthma. http://www.nhlbi.nih.gov/guidelines/ and Treatment Regimens (TENOR) study: A large cohort asthma/index.htm. Accessed April 16, 2008.
of patients with severe or difficult-to-treat asthma. 2. US Food and Drug Administration. FDA Public Health Ann Allergy Asthma Immunol. 2004;92:32–39.
Advisory: Serevent Diskus (salmeterol xinafoate inhala- 14. Miller MK, Johnson C, Miller DP, et al, for the tion powder), Advair Diskus (fluticasone propionate and TENOR Study Group. Severity assessment in asthma: salmeterol inhalation powder), Foradil Aerolizer (formot- An evolving concept. J Allergy Clin Immunol. 2005;116: erol fumarate inhalation powder). http://www.fda.gov/ cder/drug/advisory/LABA.htm. Accessed December 4, 15. Rabe KF, Vermeire PA, Soriano JB, Maier WC. Clinical management of asthma in 1999: The Asthma Insights and 3. Nelson HS, Weiss ST, Bleecker ER, et al, for the SMART Reality in Europe (AIRE) study. Eur Respir J. 2000;16:802– Study Group. The Salmeterol Multicenter Asthma Re- search Trial: A comparison of usual pharmacotherapy for 16. Adams RJ, Fuhlbrigge A, Guilbert T, et al. Inadequate use asthma or usual pharmacotherapy plus salmeterol [pub- of asthma medication in the United States: Results of the lished correction appears in Chest. 2006;129:1393]. Chest. asthma in America national population survey. J Allergy Clin Immunol. 2002;110:58–64.
Volume 30 Number 10
H. Friedman et al.
DS, et al, for the Formoterol and Corticosteroids Establishing Therapy (FACET) International Study Group. Effect of inhaled formoterol and budesonide on exacerbations of asthma [published correction appears in N Engl J Med. 1998;338:139]. N Engl J Med. 1997;337:1405–1411.
quet J, et al, for the GOAL Investiga-tors Group. Can guideline-defined asthma control be achieved? The Gaining Optimal Asthma Control study. Am J Respir Crit Care Med. 2004;170:836–844.
berg C. Addition of salmeterol to low-dose fluticasone versus higher-dose fluticasone: An analysis of asthma exacerbations. J Allergy Clin Immunol. 2001;107:783–789.
20. Ringdal N, Eliraz A, Pruzinec R, et al, for the International Study Group. The salmeterol/fluticasone combi-nation is more effective than flutica- sone plus oral montelukast in asth-ma. Respir Med. 2003;97:234–241.
et al. Asthma treatment in a pop- ulation-based cohort: Putting step-up and step-down treatment chang- es in context. Mayo Clin Proc. 2007;82:414–421.
ment of asthma. Am Fam Physician. 2006;74:232.
right time. Chest. 2006;130(Suppl 1): 29S–40S.
Services. Judicious use of Advair. http://www.dhs.state.mn.us/main/groups/business_partners/documents/ pub/dhs_id_053529.pdf. Accessed December 4, 2006.
Address correspondence to: Howard Friedman, PhD, Analytic Solutions,
Inc., 26 Prince Street, Suite 2B, New York, NY 10012. E-mail: howard@
analytic-consulting.com
October 2008

Source: http://www.glaxowhistleblowers.com/pdf/exhibits/102.pdf