- Email: [email protected]
Effectiveness of combination therapies in asthma: An observational study
Pulmonary Pharmacology & Therapeutics 22 (2009) 194–198
Contents lists available at ScienceDirect
Pulmonary Pharmacology & Therapeutics journal homepage: www.elsevier.com/locate/ypupt
Effectiveness of combination therapies in asthma: An observational study Samy Suissa a, b, c, Sophie Dell’Aniello a, Pierre Ernst a, c, * a
McGill Pharmacoepidemiology Research Unit, McGill University Health Centre, Montreal, Canada Department of Epidemiology and Biostatistics, McGill University, Montreal, Canada c Department of Medicine, McGill University, Montreal, Canada b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 12 June 2008 Received in revised form 14 November 2008 Accepted 23 November 2008
Background: Asthma guidelines suggest that a long-acting b-agonist be added to the treatment regimen of patients not adequately controlled on a low to moderate dose of inhaled corticosteroids. We compared the effectiveness in a real world setting of two such combinations available in a single inhaler. Methods: We identified patients initiating therapy with either budesonide/formoterol or fluticasone/ salmeterol after May 2001 in a clinical database. We compared asthma medication and health care utilisation over the subsequent year. Results: There were 6918 first-time users of budesonide/formoterol and 16,157 of fluticasone/salmeterol. Overall, there were no differences between the two treatment groups in hospitalisations for asthma (23 and 25 per 1000 per year) or visits to the physician during follow-up. Duration of treatment success defined as the time to occurrence of an exacerbation or the need for different or additional antiinflammatory therapy was also similar in both groups at approximately 8.6 months. The mean duration of persistent treatment was 13% longer with budesonide/formoterol (95% CI 11–16%), during which budesonide/formoterol subjects received 11% less prescriptions for their combination therapy (95% CI 9–13%). Fluticasone/salmeterol users were less likely to require referral to a specialist. Conclusion: We found similar effectiveness for the budesonide/formoterol and fluticasone/salmeterol combinations. Such equivalence appears to be obtained at lower relative doses of inhaled corticosteroids with the budesonide/formoterol combination. Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: Asthma Drug therapy Inhaled corticosteroids Bronchodilators
1. Introduction International and national guidelines recognise inhaled corticosteroids as first line maintenance treatment for persistent asthma [1–3]. If control of asthma is not obtained with low or moderate doses of inhaled corticosteroids, randomised clinical trials have now shown that it is preferable to add a long-acting b-2 agonist bronchodilator such as formoterol or salmeterol rather than increase the dose of inhaled corticosteroids further [4]. Combination products are now available that combine each of these longacting b-2 agonist bronchodilators with an inhaled corticosteroid. Use of such a combination product appears preferable to adding a second inhaler containing the long-acting b-2 agonist bronchodilator since the combination product assures that the patient with asthma is not using the long-acting b-2 agonist bronchodilator alone. Such maintenance therapy with a long-acting b-2 agonist
* Corresponding author. Pulmonary Division, SMBD Jewish General Hospital, 3755 Cote Ste Catherine G-203, Montreal, Que´bec H3T 1E2, Canada. Tel.: þ1 514 340 8222. E-mail address: [email protected] (P. Ernst). 1094-5539/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.pupt.2008.11.008
bronchodilator without a concomitant inhaled corticosteroid has been associated with a deterioration in asthma control [5–7]. We used a clinical general practice research database to carry out an observational study to gather information on how each combination therapy was used in the real world and whether we could provide information on the relative effectiveness of the two combinations.
2. Methods 2.1. Data source We used the United Kingdom General Practice Research Database (GPRD). This database contains over 35 million person-years of data from patient records collected continuously since 1987. Currently, the information is collected on around 3 million patients, approximately 5% of the UK population, by contributing general practices from all around the UK. The information available includes demographics, medical diagnoses, all prescriptions, secondary care referrals, hospitalisations, as well as miscellaneous patient care information.
S. Suissa et al. / Pulmonary Pharmacology & Therapeutics 22 (2009) 194–198
195
The study outcomes measured during the 1-year follow-up include the frequency of the following variables: refills of the two combination therapies; switches between the two combination therapies; hospitalisations for asthma; office visits to the family doctor and specialist referrals for asthma; prescriptions for oral corticosteroids; prescriptions for short-acting b-agonists; prescriptions for long-acting b-agonists; prescriptions for other (oral and nebulised) b-agonists; prescriptions for inhaled and nasal corticosteroids; prescriptions for xanthines and leukotriene receptor antagonists; and prescriptions for antibiotics. Hospitalisation for asthma was defined as a recorded hospitalisation in the physician’s file, along with an asthma diagnosis on the same day or within 7 days, the latter to account for the delay in recording hospitalisation information by the family doctor. An overall outcome called treatment success was created. Treatment success was defined as the period of time from initiation of treatment with combination therapy (cohort entry) until the occurrence of the first of the following: a hospitalisation for asthma, a prescription for an oral corticosteroid, a switch between the two combination therapies, or a prescription for an additional inhaled anti-inflammatory asthma therapy. Complete treatment success was defined by the time to a first prescription for an inhaled shortacting b-agonist within the period of treatment success as defined above.
hospitalisations, as well as other measures of medication and health care use, all in the year prior to cohort entry. These propensity scores were computed based on the probability of receiving budesonide/formoterol versus fluticasone/salmeterol given all potential determinants of initial use and all two way interactions between these determinants. Since the propensity scores in the budesonide/formoterol and fluticasone/salmeterol groups overlapped, we included all subjects in the analysis. The propensity scores were used as a continuous score as well as deciles of the score as adjustment factors in the subsequent analyses that compared the budesonide/formoterol and fluticasone/salmeterol users. A first analysis comparing the efficacy of budesonide/formoterol and fluticasone/salmeterol was based on an intent-to-treat approach, where the outcomes were assessed during the 1-year period after cohort entry, irrespective of whether the patients continued their combination therapy treatment. Since some subjects were not followed for an entire year due to truncating the end of the observation period, these follow-up times were censored at that point. Thus, the frequency of each outcome during follow-up was computed as the number of prescriptions or events divided by the person-time, standardised to a year. Poisson regression methods with overdispersion were used to compare the two groups, after adjusting for the effect of baseline characteristics as assessed by the propensity scores (measured in the year prior to cohort entry) on the outcome rates. A second analysis compared budesonide/formoterol and fluticasone/salmeterol during persistent combination therapy use. To assess treatment persistence with budesonide/formoterol and fluticasone/salmeterol, consecutive prescriptions for these agents were assessed until a prescription was not given after the end of the previous one. A grace period of 30 days was allowed between the end of a prescription and the start of the next to define persistence. The duration of each prescription was estimated from the number of canisters dispensed and the dosing instruction. For varying dosing instruction (2–4 puffs), we used the mean number of puffs per day. In situations where the duration of a prescription was missing or vague, and thus could not be derived from the information available, we imputed the mean duration for the combination therapy type. The duration of treatment persistence, namely the time from treatment initiation (cohort entry) to the end of consecutive use, was compared between the two combination therapies using methods for survival data, namely Kaplan–Meier and Cox’s proportional hazards. Rates of outcomes during persistent therapy were estimated during persistent person-time and compared using Poisson regression methods. The duration of successful treatment was estimated and compared using methods for survival data, namely Cox’s proportional hazards model. We performed several sensitivity analyses. In view of the differing drug formulations, we carried out an analysis restricted to the comparison of cohort members defined by a first budesonide/formoterol or fluticasone/salmeterol in the form of canisters of 120 puffs. Second, we restricted the cohorts to subjects 50 years of age or younger at cohort entry to reduce the likelihood of including COPD patients. Finally, we also analyzed these data in the restricted cohort of users of canisters of 120 puffs and subjects 50 years of age or younger at cohort entry.
2.4. Data analysis
3. Results
Logistic regression was used to develop propensity scores based on factors in the prior year that were associated with the initial use of budesonide/formoterol versus fluticasone/salmeterol. The factors examined include markers of asthma severity, namely the number of prescriptions of b-agonists, theophylline, oral or nebulised b-agonists, oral corticosteroids, the number of asthma
We identified the source population of all 635,861 subjects in the GPRD with a diagnosis of asthma between January 1987 and December 2004. From this source population, we identified all 39,533 subjects without a COPD diagnosis who received one of the two combination therapies in question. After excluding subjects who had less than 1 year of available information before their first
2.2. Study design To assess the determinants of the use of these combination therapies, we first identified all patients in the database with a diagnosis of asthma between January 1, 1987 and December 31, 2005. For all subjects, we obtained all prescription drug information, hospitalisation data, and the date of death, if applicable. In particular, we identified the occurrence of the first prescription for a single inhaler combination therapy, either budesonide/formoterol or fluticasone/salmeterol. The date of the first of these combination therapy prescriptions was named the index date and identified the start of follow-up for the two cohorts thus formed, initial users of either budesonide/formoterol or fluticasone/salmeterol. We excluded subjects who also had a diagnosis of COPD and those with less than 1 year of available information before their index date. We also excluded subjects initially prescribed fluticasone/salmeterol before budesonide/formoterol came on the market (from March 1999 to May 2001) so as to compare these two therapies during the time period when both options were available. Markers of disease severity, including the use of inhaled corticosteroids and longacting b-agonists, prior to the date of the first use of a combination therapy were examined during the prior 1-year period. To assess the determinants, patterns and impact of these combination therapies, we used the two cohorts (initial users of either budesonide/formoterol or fluticasone/salmeterol) identified previously and followed the subjects for 1 year or until the end of the study period, December 2005, whichever was the earliest. All prescriptions for asthma medications given during the follow-up, including oral corticosteroids and short-acting beta-agonists, were identified. 2.3. Outcomes
196
S. Suissa et al. / Pulmonary Pharmacology & Therapeutics 22 (2009) 194–198
Table 1 Baseline demographic characteristics and health care utilisation during the year prior to cohort entry (initiation of combination therapy).
Age (years) Gender (% male) Years registered in practice
Budesonide/ formoterol n ¼ 6918
Fluticasone/ salmeterol n ¼ 16157
%
%
Mean or rate
Mean or rate
43.8 40.7
42.5 42.4
14.9
Health care utilisation in the year prior to index date Asthma hospitalisation 2.5 0.033 Other hospitalisation 10.7 0.17 Asthma referral 1.0 0.011 Number of GP visits 9.79 Antibiotic prescriptions 56.2 1.36
14.4 3.0 10.6 1.7 58.3
0.040 0.17 0.020 10.31 1.48
combination therapy prescription and those for whom the first combination therapy was before May 2001, we identified our study cohort members of 6918 first-time budesonide/formoterol users and 16,157 first-time fluticasone/salmeterol users. In these two cohorts, the budesonide/formoterol prescriptions were in the form of dry powder inhalers of 120 doses (95%) and 60 doses (5%) that were prescribed between 1 and 8 inhalations per day. These prescriptions, when clearly identified, included primarily the Symbicort Turbuhaler 160/4.5 mg (69%), the Turbuhaler 80/4.5 mg (26%) and the Turbuhaler 320/9 mg (5%). The fluticasone/salmeterol prescriptions were in dry powder or metered-dose inhalers of 120 inhalations (68%) and 60 inhalations (32%) with 1 to 8 prescribed inhalations per day. These prescriptions were for a dry powder inhaler with 50 mg of salmeterol and100, 250 and 500 mg of fluticasone (32%), or a pressurised inhaler with 25 mg of salmeterol and 50, 125 and 250 mg of fluticasone (68%). The daily dose was missing in 42% of budesonide/ formoterol prescriptions as compared with 14% for fluticasone/ salmeterol. Table 1 displays the characteristics of budesonide/formoterol and fluticasone/salmeterol users during the baseline year prior to cohort entry (the first combination therapy prescription), while Table 2 displays the patterns of use of asthma medications during this baseline year. These tables show that the mean age at cohort entry is approximately 43 years, while 42% of the subjects are men. Approximately 3% of patients had been hospitalised for asthma during the baseline year and subjects had visited their family doctor (GP) on average 10 times. Use of most medications, except possibly nasal corticosteroids was more common in the fluticasone/
salmeterol group. These differences in baseline factors are adjusted for in the analysis with the propensity scores when examining differences in outcomes during the 1-year follow-up period. The biggest difference between users of each combination was the type of inhaled corticosteroid which had been used previously; fluticasone in those initiating the combination of fluticasone/salmeterol and budesonide in those initiating therapy with budesonide/ formoterol. Table 3 describes the cohorts in terms of follow-up time and duration of persistent use. It shows that over 85% of the patients were followed up for the entire 12-month period (mean 11.6 months). Twenty percent of budesonide/formoterol patients only received a single combination therapy prescription during the year, as compared to 15% for the fluticasone/salmeterol users. The mean duration of persistent use is similar at around 5 months allowing for a 30-day grace period. Duration of treatment success is similar in the two groups while complete treatment success lasted longer in the budesonide/formoterol group. Table 4 displays the rates of asthma drug prescriptions and other health care resource use during the year following the first prescription of budesonide/formoterol and fluticasone/salmeterol. Budesonide/formoterol subjects received 13% fewer prescriptions for their combination therapy than the fluticasone/salmeterol subjects during the year (adjusted rate ratio 0.87; 95% CI 0.84– 0.89). Moreover, budesonide/formoterol subjects received 6% fewer total prescriptions for all asthma medications than the fluticasone/ salmeterol subjects during the year (adjusted rate ratio 0.94; 95% CI 0.92–0.97). The budesonide/formoterol group did receive more subsequent prescriptions for long-acting b-agonists and inhaled corticosteroids, however. The rates of asthma hospitalisation and GP visits were similar for the two groups. The rate of referrals for asthma was higher by 22% (95% CI 12–34%) while the number of antibiotic prescriptions was reduced by 6% (95% CI 2–10%) among users of budesonide/formoterol. Table 5 displays the rates of these outcomes with persistent budesonide/formoterol and fluticasone/salmeterol use during the year following the first prescription. Budesonide/formoterol subjects received 11% fewer prescriptions for their combination inhaler than the fluticasone/salmeterol subjects during persistent use (adjusted rate ratio 0.89; 95% CI 0.87–0.91). Moreover, budesonide/formoterol subjects received 7% fewer total prescriptions for asthma medications than the fluticasone/salmeterol subjects during persistent use (adjusted rate ratio 0.93; 95% CI 0.90–0.95). During persistent budesonide/formoterol use, the rate of asthma hospitalisation was the same as for fluticasone/salmeterol patients
Table 2 Baseline use of asthma medications during the year prior to cohort entry (initiation of combination therapy). Budesonide/formoterol n ¼ 6918
Fluticasone/salmeterol n ¼ 16157
%
Mean or rate
Mean in users
%
Mean or rate
Mean in users
83.33 23.23 6.94 2.34 2.46 4.05 76.70 47.72 25.31 6.91 0.40 15.09 27.94
4.28 1.07 0.29 0.14 0.10 0.20 3.35 1.88 1.09 0.36 0.010 0.44 0.58
5.14 4.59 4.12 6.19 4.02 4.91 4.36 3.93 4.33 5.26 2.71 2.92 2.09
85.32 36.06 7.39 3.06 3.55 4.85 80.13 57.65 5.90 20.89 0.14 15.58 31.50
4.66 1.61 0.29 0.20 0.16 0.23 3.61 2.30 0.24 1.07 0.003 0.47 0.73
5.46 4.47 3.99 6.56 4.50 4.73 4.50 3.98 4.03 5.13 1.96 3.04 2.33
b2-Agonists Short-acting Long-acting Other b-agonistsa Xanthines Ipratropium bromide Leukotriene receptor antagonists Inhaled corticosteroids Beclomethasone Budesonide Fluticasone Mometasone Nasal corticosteroids Oral corticosteroids a
Oral or nebulised b-agonist or in combination with ipratropium.
S. Suissa et al. / Pulmonary Pharmacology & Therapeutics 22 (2009) 194–198 Table 3 Durations of complete follow-up, persistent use and treatment success for budesonide/formoterol and fluticasone/salmeterol users in the year after the cohort entry (initiation of combination therapy).
Number of subjects Person-years of follow-up, total Follow-up time, mean (months) 12 months (%) 6 to <12 months (%) <6 months (%) Number of prescriptions of the cohort-defining combination therapy, mean Duration of persistent use, mean (months)a Duration of treatment success, mean (months) Duration of complete treatment success, mean (months) Switch to the other combination therapy (%) a
Budesonide/ formoterol
Fluticasone/ salmeterol
6918 6606.4 11.6 87.0 9.1 3.9 4.9
16157 15423.3 11.6 86.7 9.4 3.9 5.6
5.5 8.7 4.7
5.6 8.6 4.2
5.2
2.9
Includes a 30 day grace period after the last prescription.
(adjusted rate ratio 1.02; 95% CI 0.93–1.11), and the number of antibiotic prescriptions was reduced by 8% (95% CI 4–13%). The duration of treatment success was similar in the two groups, 261.5 days (136.2) for users of budesonide/formoterol and 256.7 (136.6) for users of fluticasone/salmeterol. Complete treatment success was on average longer by 4.9 days in the budesonide/formoterol group and this was statistically significant (p ¼ 0.013; 95%CI 1.0–8.7). A switch from budesonide/formoterol to fluticasone/salmeterol was more likely to occur than the other way around (adjusted hazard ratio 1.78, 95% CI 1.56–2.06). The mean daily dose of budesonide during follow-up was 281.8 mg (SD 217.3) in the budesonide/formoterol cohort, while the mean daily dose of fluticasone in the fluticasone/salmeterol cohort was 264.7 mg (SD 245.4). Results were not substantially affected when the analyses were restricted to subjects younger than 50 years old and/or subjects prescribed inhalers with 120 doses only. 4. Discussion We compared outcomes in patients with asthma, treated in general practices in the UK, who initiated treatment with a combination of a long-acting b-agonist bronchodilator with an inhaled corticosteroid, either budesonide/formoterol or fluticasone/salmeterol. At the start of single inhaler combination therapy
197
(combination therapy), patients prescribed fluticasone/salmeterol were more severe as attested to by greater use of asthma medications in the prior year. These baseline differences in asthma severity were adjusted for using propensity scores and outcomes during the subsequent year were examined. Overall, there were no important differences in hospitalisations for asthma or visits to the physician for asthma, or for all causes, between the two treatment groups during follow-up. Duration of treatment success defined as the time to occurrence of an exacerbation or the need for different or additional anti-inflammatory therapy was also similar in both groups at approximately 8.6 months. Including the need for a prescription for a short-acting b-agonist in the definition of treatment success, the duration was greater for the budesonide/formoterol group (4.7 versus 4.2 months). This was also associated with fewer prescriptions for a short-acting b-agonist and for the combination therapy itself in the budesonide/formoterol group. Users of budesonide/formoterol had fewer total prescriptions for asthma medications during follow-up whether considering initial or persistent users of combination therapy, while the number of prescriptions for oral corticosteroids was similar. This apparently similar efficacy was obtained at a lower equivalent dose of inhaled corticosteroids in the budesonide/formoterol group if one considers fluticasone to be approximately twice as potent on a microgram per microgram basis [8,9]. On the other hand, more patients treated with budesonide/formoterol were referred to specialists for asthma and this was more marked for persistent users possibly suggesting an insufficient dose of inhaled corticosteroids for more severe patients. In support of this users of budesonide/formoterol were more likely to be prescribed an additional corticosteroid or long-acting b-agonist. Our study is observational in nature and subject to certain biases. Which subjects received a particular combination did not occur randomly and was likely related to patient and physician characteristics. To adjust for physician preferences we compared patients treated in the same practice. As for patient characteristics, those initially prescribed fluticasone/salmeterol appeared overall to be more severe and this difference was adjusted for using information on medication use and health care utilisation in the year prior to starting combination therapy. We cannot be sure that differences observed during combination therapy were not due to unmeasured differences between patients and their physician. We consider this unlikely, however, since the biggest difference between the groups initiating combination therapy, was the type of corticosteroid the patient had been prescribed previously and there was very little switching from one combination therapy to another, suggesting
Table 4 Crude and adjusted rate ratios of asthma medication and health care use from intention-to-treat analysis in the year after the start of combination therapy for budesonide/ formoterol relative to fluticasone/salmeterol. Rate per year of: Combination therapy of the same type
Budesonide/formoterol (person-years of follow-up ¼ 6606.36)
Fluticasone/salmeterol (person-years of follow-up ¼ 15423.33)
Crude rate ratio
Adjusted rate ratioa
4.05
4.86
0.83
0.87 (0.84–0.89)
Asthma medication use Short-acting b agonist Long-acting b agonist Inhaled corticosteroid Oral corticosteroid Total number of asthma medications
3.38 0.25 0.68 0.49 10.45
3.85 0.26 0.64 0.58 12.04
0.88 0.96 1.07 0.85 0.87
0.97 (0.94–1.00) 1.26 (1.17–1.35) 1.17 (1.10–1.24) 1.02 (0.96–1.08) 0.94 (0.92–0.97)
Health care utilisation Asthma hospitalisations Asthma referrals Asthma visits GP visits Number of antibiotic prescriptions
0.023 0.011 1.44 10.92 1.24
0.025 0.011 1.50 11.36 1.40
0.91 1.03 0.96 0.96 0.89
1.01 1.22 1.01 0.99 0.94
a
Adjusted for deciles of propensity score and propensity score.
(0.92–1.10) (1.12–1.34) (0.97–1.04) (0.96–1.01) (0.90–0.98)
198
S. Suissa et al. / Pulmonary Pharmacology & Therapeutics 22 (2009) 194–198
Table 5 Crude and adjusted rate ratios of asthma medication and health care use during persistent use of budesonide/formoterol or fluticasone/salmeterol in the year after the start of combination therapy. Rate per year of: Combination therapy of the same type
Budesonide/formoterol (person-years of follow-up ¼ 3127.92)
Fluticasone/salmeterol (person-years of follow-up ¼ 7453.54)
Crude rate ratio
Adjusteda rate ratio
5.61
6.52
0.86
0.89 (0.87–0.91)
Asthma medication use Short-acting b-agonist Long-acting b-agonist Inhaled corticosteroid Oral corticosteroid Total number of asthma medications
3.74 0.16 0.42 0.56 12.19
4.21 0.17 0.43 0.65 14.13
0.89 0.96 0.98 0.86 0.86
0.98 1.20 1.05 1.05 0.93
(0.94–1.01) (1.11–1.30) (0.98–1.12) (0.99–1.12) (0.90–0.95)
Health care utilisation Asthma hospitalisations Asthma referrals Asthma visits GP visits Number of antibiotic prescriptions
0.027 0.018 1.84 11.70 1.29
0.031 0.013 1.83 12.19 1.51
0.88 1.37 1.00 0.96 0.85
1.02 1.56 1.02 0.99 0.92
(0.93–1.11) (1.43–1.71) (0.99–1.06) (0.96–1.01) (0.87–0.96)
a
Adjusted for deciles of propensity score and propensity score.
that physician’s prescribing habits were the major determinant of which combination therapy was prescribed. We found that patients prescribed budesonide/formoterol used, on average, a lower dose of inhaled corticosteroid when considering equipotent doses of budesonide and fluticasone. One must note, however, that the actual number of inhalations per day was missing for a significant proportion of users of budesonide/formoterol and had to be estimated. The greater frequency of missing dosing information for budesonide/formoterol may have resulted from publication of studies of this medication which have used variable as opposed to fixed doses usually recommended for fluticasone/salmeterol [10,11]. The information for the present study was obtained from the population-based General Practice Research Database (GPRD), which has been described in detail [12]. More than 6 million people in the UK are enrolled with over 400 general practitioners who use office computers and have agreed to provide data for research purposes. The physicians generate prescriptions directly with their study computer; this information is automatically transcribed into the computer record. The recorded information on drug exposure and diagnoses has been validated and is of high quality [13]. This database has been the source for over 300 publications in the peerreviewed medical literature. In summary our results suggest similar efficacy for the combinations of an inhaled corticosteroid and a long-acting b-agonist, fluticasone/salmeterol and budesonide/formoterol. Such similar efficacy appears to be obtained at lower relative doses of inhaled corticosteroid in the budesonide/formoterol group. This research was funded by a grant from AstraZeneca and grants from the Canadian Institutes of Health Research (CIHR). SS is the recipient of a Distinguished Investigator award from the CIHR. The database was funded by grants from CIHR, the Fonds de la recherche en sante´ du Que´bec (FRSQ) and the Canadian Foundation for Innovation (CFI).
References [1] Expert Panel Report 3 (EPR-3): Guidelines for the diagnosis and management of asthma – summary report 2007. J Allergy Clin Immunol 2007;120(5 Suppl):S94–138. [2] Global Initiative for Asthma. Pocket guide for asthma management and prevention. 2006. [3] Lemiere C, Bai T, Balter M, Bayliff C, Becker A, Boulet LP, et al. Adult asthma consensus guidelines update 2003. Can Respir J 2004;11(Suppl A): 9A–18A. [4] Greenstone IR, Ni Chroinin MN, Masse V, Danish A, Magdalinos H, Zhang X, et al. Combination of inhaled long-acting beta2-agonists and inhaled steroids versus higher dose of inhaled steroids in children and adults with persistent asthma. Cochrane Database Syst Rev 2005;(4). CD005533. [5] Lemanske Jr RF, Sorkness CA, Mauger EA, Lazarus SC, Boushey HA, Fahy JV, et al. Inhaled corticosteroid reduction and elimination in patients with persistent asthma receiving salmeterol: a randomized controlled trial. JAMA 2001;285(20):2594–603. [6] Lazarus SC, Boushey HA, Fahy JV, Chinchilli VM, Lemanske Jr RF, Sorkness CA, et al. Long-acting beta2-agonist monotherapy vs continued therapy with inhaled corticosteroids in patients with persistent asthma: a randomized controlled trial. JAMA 2001;285(20):2583–93. [7] Mann M, Chowdhury B, Sullivan E, Nicklas R, Anthracite R, Meyer RJ. Serious asthma exacerbations in asthmatics treated with high-dose formoterol. Chest 2003;124(1):70–4. [8] Boulet LP, Becker A, Berube D, Beveridge R, Ernst P. Canadian asthma consensus report, 1999. Canadian Asthma Consensus Group. Can Med Assoc J 1999;161(11 Suppl):S1–61. [9] National Asthma Education Program NHLBI. Practical guide for the diagnosis and management of asthma. US Department of Health and Human Services 1997:1–52. [10] Aalbers R, Backer V, Kava TT, Omenaas ER, Sandstrom T, Jorup C, et al. Adjustable maintenance dosing with budesonide/formoterol compared with fixed-dose salmeterol/fluticasone in moderate to severe asthma. Curr Med Res Opin 2004;20(2):225–40. [11] Bateman ED, Boushey HA, Bousquet J, Busse WW, Clark TJ, Pauwels RA, et al. Can guideline-defined asthma control be achieved?: The Gaining Optimal Asthma ControL Study. Am J Respir Crit Care Med 2004;170(8):836–44. [12] Garcia Rodriguez LA, Perez GS. Use of the UK General Practice Research Database for pharmacoepidemiology. Br J Clin Pharmacol 1998;45(5): 419–25. [13] Jick SS, Kaye JA, Vasilakis-Scaramozza C, Garcia Rodriguez LA, Ruigomez A, Meier CR, et al. Validity of the general practice research database. Pharmacotherapy 2003;23(5):686–9.
Contents lists available at ScienceDirect
Pulmonary Pharmacology & Therapeutics journal homepage: www.elsevier.com/locate/ypupt
Effectiveness of combination therapies in asthma: An observational study Samy Suissa a, b, c, Sophie Dell’Aniello a, Pierre Ernst a, c, * a
McGill Pharmacoepidemiology Research Unit, McGill University Health Centre, Montreal, Canada Department of Epidemiology and Biostatistics, McGill University, Montreal, Canada c Department of Medicine, McGill University, Montreal, Canada b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 12 June 2008 Received in revised form 14 November 2008 Accepted 23 November 2008
Background: Asthma guidelines suggest that a long-acting b-agonist be added to the treatment regimen of patients not adequately controlled on a low to moderate dose of inhaled corticosteroids. We compared the effectiveness in a real world setting of two such combinations available in a single inhaler. Methods: We identified patients initiating therapy with either budesonide/formoterol or fluticasone/ salmeterol after May 2001 in a clinical database. We compared asthma medication and health care utilisation over the subsequent year. Results: There were 6918 first-time users of budesonide/formoterol and 16,157 of fluticasone/salmeterol. Overall, there were no differences between the two treatment groups in hospitalisations for asthma (23 and 25 per 1000 per year) or visits to the physician during follow-up. Duration of treatment success defined as the time to occurrence of an exacerbation or the need for different or additional antiinflammatory therapy was also similar in both groups at approximately 8.6 months. The mean duration of persistent treatment was 13% longer with budesonide/formoterol (95% CI 11–16%), during which budesonide/formoterol subjects received 11% less prescriptions for their combination therapy (95% CI 9–13%). Fluticasone/salmeterol users were less likely to require referral to a specialist. Conclusion: We found similar effectiveness for the budesonide/formoterol and fluticasone/salmeterol combinations. Such equivalence appears to be obtained at lower relative doses of inhaled corticosteroids with the budesonide/formoterol combination. Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: Asthma Drug therapy Inhaled corticosteroids Bronchodilators
1. Introduction International and national guidelines recognise inhaled corticosteroids as first line maintenance treatment for persistent asthma [1–3]. If control of asthma is not obtained with low or moderate doses of inhaled corticosteroids, randomised clinical trials have now shown that it is preferable to add a long-acting b-2 agonist bronchodilator such as formoterol or salmeterol rather than increase the dose of inhaled corticosteroids further [4]. Combination products are now available that combine each of these longacting b-2 agonist bronchodilators with an inhaled corticosteroid. Use of such a combination product appears preferable to adding a second inhaler containing the long-acting b-2 agonist bronchodilator since the combination product assures that the patient with asthma is not using the long-acting b-2 agonist bronchodilator alone. Such maintenance therapy with a long-acting b-2 agonist
* Corresponding author. Pulmonary Division, SMBD Jewish General Hospital, 3755 Cote Ste Catherine G-203, Montreal, Que´bec H3T 1E2, Canada. Tel.: þ1 514 340 8222. E-mail address: [email protected] (P. Ernst). 1094-5539/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.pupt.2008.11.008
bronchodilator without a concomitant inhaled corticosteroid has been associated with a deterioration in asthma control [5–7]. We used a clinical general practice research database to carry out an observational study to gather information on how each combination therapy was used in the real world and whether we could provide information on the relative effectiveness of the two combinations.
2. Methods 2.1. Data source We used the United Kingdom General Practice Research Database (GPRD). This database contains over 35 million person-years of data from patient records collected continuously since 1987. Currently, the information is collected on around 3 million patients, approximately 5% of the UK population, by contributing general practices from all around the UK. The information available includes demographics, medical diagnoses, all prescriptions, secondary care referrals, hospitalisations, as well as miscellaneous patient care information.
S. Suissa et al. / Pulmonary Pharmacology & Therapeutics 22 (2009) 194–198
195
The study outcomes measured during the 1-year follow-up include the frequency of the following variables: refills of the two combination therapies; switches between the two combination therapies; hospitalisations for asthma; office visits to the family doctor and specialist referrals for asthma; prescriptions for oral corticosteroids; prescriptions for short-acting b-agonists; prescriptions for long-acting b-agonists; prescriptions for other (oral and nebulised) b-agonists; prescriptions for inhaled and nasal corticosteroids; prescriptions for xanthines and leukotriene receptor antagonists; and prescriptions for antibiotics. Hospitalisation for asthma was defined as a recorded hospitalisation in the physician’s file, along with an asthma diagnosis on the same day or within 7 days, the latter to account for the delay in recording hospitalisation information by the family doctor. An overall outcome called treatment success was created. Treatment success was defined as the period of time from initiation of treatment with combination therapy (cohort entry) until the occurrence of the first of the following: a hospitalisation for asthma, a prescription for an oral corticosteroid, a switch between the two combination therapies, or a prescription for an additional inhaled anti-inflammatory asthma therapy. Complete treatment success was defined by the time to a first prescription for an inhaled shortacting b-agonist within the period of treatment success as defined above.
hospitalisations, as well as other measures of medication and health care use, all in the year prior to cohort entry. These propensity scores were computed based on the probability of receiving budesonide/formoterol versus fluticasone/salmeterol given all potential determinants of initial use and all two way interactions between these determinants. Since the propensity scores in the budesonide/formoterol and fluticasone/salmeterol groups overlapped, we included all subjects in the analysis. The propensity scores were used as a continuous score as well as deciles of the score as adjustment factors in the subsequent analyses that compared the budesonide/formoterol and fluticasone/salmeterol users. A first analysis comparing the efficacy of budesonide/formoterol and fluticasone/salmeterol was based on an intent-to-treat approach, where the outcomes were assessed during the 1-year period after cohort entry, irrespective of whether the patients continued their combination therapy treatment. Since some subjects were not followed for an entire year due to truncating the end of the observation period, these follow-up times were censored at that point. Thus, the frequency of each outcome during follow-up was computed as the number of prescriptions or events divided by the person-time, standardised to a year. Poisson regression methods with overdispersion were used to compare the two groups, after adjusting for the effect of baseline characteristics as assessed by the propensity scores (measured in the year prior to cohort entry) on the outcome rates. A second analysis compared budesonide/formoterol and fluticasone/salmeterol during persistent combination therapy use. To assess treatment persistence with budesonide/formoterol and fluticasone/salmeterol, consecutive prescriptions for these agents were assessed until a prescription was not given after the end of the previous one. A grace period of 30 days was allowed between the end of a prescription and the start of the next to define persistence. The duration of each prescription was estimated from the number of canisters dispensed and the dosing instruction. For varying dosing instruction (2–4 puffs), we used the mean number of puffs per day. In situations where the duration of a prescription was missing or vague, and thus could not be derived from the information available, we imputed the mean duration for the combination therapy type. The duration of treatment persistence, namely the time from treatment initiation (cohort entry) to the end of consecutive use, was compared between the two combination therapies using methods for survival data, namely Kaplan–Meier and Cox’s proportional hazards. Rates of outcomes during persistent therapy were estimated during persistent person-time and compared using Poisson regression methods. The duration of successful treatment was estimated and compared using methods for survival data, namely Cox’s proportional hazards model. We performed several sensitivity analyses. In view of the differing drug formulations, we carried out an analysis restricted to the comparison of cohort members defined by a first budesonide/formoterol or fluticasone/salmeterol in the form of canisters of 120 puffs. Second, we restricted the cohorts to subjects 50 years of age or younger at cohort entry to reduce the likelihood of including COPD patients. Finally, we also analyzed these data in the restricted cohort of users of canisters of 120 puffs and subjects 50 years of age or younger at cohort entry.
2.4. Data analysis
3. Results
Logistic regression was used to develop propensity scores based on factors in the prior year that were associated with the initial use of budesonide/formoterol versus fluticasone/salmeterol. The factors examined include markers of asthma severity, namely the number of prescriptions of b-agonists, theophylline, oral or nebulised b-agonists, oral corticosteroids, the number of asthma
We identified the source population of all 635,861 subjects in the GPRD with a diagnosis of asthma between January 1987 and December 2004. From this source population, we identified all 39,533 subjects without a COPD diagnosis who received one of the two combination therapies in question. After excluding subjects who had less than 1 year of available information before their first
2.2. Study design To assess the determinants of the use of these combination therapies, we first identified all patients in the database with a diagnosis of asthma between January 1, 1987 and December 31, 2005. For all subjects, we obtained all prescription drug information, hospitalisation data, and the date of death, if applicable. In particular, we identified the occurrence of the first prescription for a single inhaler combination therapy, either budesonide/formoterol or fluticasone/salmeterol. The date of the first of these combination therapy prescriptions was named the index date and identified the start of follow-up for the two cohorts thus formed, initial users of either budesonide/formoterol or fluticasone/salmeterol. We excluded subjects who also had a diagnosis of COPD and those with less than 1 year of available information before their index date. We also excluded subjects initially prescribed fluticasone/salmeterol before budesonide/formoterol came on the market (from March 1999 to May 2001) so as to compare these two therapies during the time period when both options were available. Markers of disease severity, including the use of inhaled corticosteroids and longacting b-agonists, prior to the date of the first use of a combination therapy were examined during the prior 1-year period. To assess the determinants, patterns and impact of these combination therapies, we used the two cohorts (initial users of either budesonide/formoterol or fluticasone/salmeterol) identified previously and followed the subjects for 1 year or until the end of the study period, December 2005, whichever was the earliest. All prescriptions for asthma medications given during the follow-up, including oral corticosteroids and short-acting beta-agonists, were identified. 2.3. Outcomes
196
S. Suissa et al. / Pulmonary Pharmacology & Therapeutics 22 (2009) 194–198
Table 1 Baseline demographic characteristics and health care utilisation during the year prior to cohort entry (initiation of combination therapy).
Age (years) Gender (% male) Years registered in practice
Budesonide/ formoterol n ¼ 6918
Fluticasone/ salmeterol n ¼ 16157
%
%
Mean or rate
Mean or rate
43.8 40.7
42.5 42.4
14.9
Health care utilisation in the year prior to index date Asthma hospitalisation 2.5 0.033 Other hospitalisation 10.7 0.17 Asthma referral 1.0 0.011 Number of GP visits 9.79 Antibiotic prescriptions 56.2 1.36
14.4 3.0 10.6 1.7 58.3
0.040 0.17 0.020 10.31 1.48
combination therapy prescription and those for whom the first combination therapy was before May 2001, we identified our study cohort members of 6918 first-time budesonide/formoterol users and 16,157 first-time fluticasone/salmeterol users. In these two cohorts, the budesonide/formoterol prescriptions were in the form of dry powder inhalers of 120 doses (95%) and 60 doses (5%) that were prescribed between 1 and 8 inhalations per day. These prescriptions, when clearly identified, included primarily the Symbicort Turbuhaler 160/4.5 mg (69%), the Turbuhaler 80/4.5 mg (26%) and the Turbuhaler 320/9 mg (5%). The fluticasone/salmeterol prescriptions were in dry powder or metered-dose inhalers of 120 inhalations (68%) and 60 inhalations (32%) with 1 to 8 prescribed inhalations per day. These prescriptions were for a dry powder inhaler with 50 mg of salmeterol and100, 250 and 500 mg of fluticasone (32%), or a pressurised inhaler with 25 mg of salmeterol and 50, 125 and 250 mg of fluticasone (68%). The daily dose was missing in 42% of budesonide/ formoterol prescriptions as compared with 14% for fluticasone/ salmeterol. Table 1 displays the characteristics of budesonide/formoterol and fluticasone/salmeterol users during the baseline year prior to cohort entry (the first combination therapy prescription), while Table 2 displays the patterns of use of asthma medications during this baseline year. These tables show that the mean age at cohort entry is approximately 43 years, while 42% of the subjects are men. Approximately 3% of patients had been hospitalised for asthma during the baseline year and subjects had visited their family doctor (GP) on average 10 times. Use of most medications, except possibly nasal corticosteroids was more common in the fluticasone/
salmeterol group. These differences in baseline factors are adjusted for in the analysis with the propensity scores when examining differences in outcomes during the 1-year follow-up period. The biggest difference between users of each combination was the type of inhaled corticosteroid which had been used previously; fluticasone in those initiating the combination of fluticasone/salmeterol and budesonide in those initiating therapy with budesonide/ formoterol. Table 3 describes the cohorts in terms of follow-up time and duration of persistent use. It shows that over 85% of the patients were followed up for the entire 12-month period (mean 11.6 months). Twenty percent of budesonide/formoterol patients only received a single combination therapy prescription during the year, as compared to 15% for the fluticasone/salmeterol users. The mean duration of persistent use is similar at around 5 months allowing for a 30-day grace period. Duration of treatment success is similar in the two groups while complete treatment success lasted longer in the budesonide/formoterol group. Table 4 displays the rates of asthma drug prescriptions and other health care resource use during the year following the first prescription of budesonide/formoterol and fluticasone/salmeterol. Budesonide/formoterol subjects received 13% fewer prescriptions for their combination therapy than the fluticasone/salmeterol subjects during the year (adjusted rate ratio 0.87; 95% CI 0.84– 0.89). Moreover, budesonide/formoterol subjects received 6% fewer total prescriptions for all asthma medications than the fluticasone/ salmeterol subjects during the year (adjusted rate ratio 0.94; 95% CI 0.92–0.97). The budesonide/formoterol group did receive more subsequent prescriptions for long-acting b-agonists and inhaled corticosteroids, however. The rates of asthma hospitalisation and GP visits were similar for the two groups. The rate of referrals for asthma was higher by 22% (95% CI 12–34%) while the number of antibiotic prescriptions was reduced by 6% (95% CI 2–10%) among users of budesonide/formoterol. Table 5 displays the rates of these outcomes with persistent budesonide/formoterol and fluticasone/salmeterol use during the year following the first prescription. Budesonide/formoterol subjects received 11% fewer prescriptions for their combination inhaler than the fluticasone/salmeterol subjects during persistent use (adjusted rate ratio 0.89; 95% CI 0.87–0.91). Moreover, budesonide/formoterol subjects received 7% fewer total prescriptions for asthma medications than the fluticasone/salmeterol subjects during persistent use (adjusted rate ratio 0.93; 95% CI 0.90–0.95). During persistent budesonide/formoterol use, the rate of asthma hospitalisation was the same as for fluticasone/salmeterol patients
Table 2 Baseline use of asthma medications during the year prior to cohort entry (initiation of combination therapy). Budesonide/formoterol n ¼ 6918
Fluticasone/salmeterol n ¼ 16157
%
Mean or rate
Mean in users
%
Mean or rate
Mean in users
83.33 23.23 6.94 2.34 2.46 4.05 76.70 47.72 25.31 6.91 0.40 15.09 27.94
4.28 1.07 0.29 0.14 0.10 0.20 3.35 1.88 1.09 0.36 0.010 0.44 0.58
5.14 4.59 4.12 6.19 4.02 4.91 4.36 3.93 4.33 5.26 2.71 2.92 2.09
85.32 36.06 7.39 3.06 3.55 4.85 80.13 57.65 5.90 20.89 0.14 15.58 31.50
4.66 1.61 0.29 0.20 0.16 0.23 3.61 2.30 0.24 1.07 0.003 0.47 0.73
5.46 4.47 3.99 6.56 4.50 4.73 4.50 3.98 4.03 5.13 1.96 3.04 2.33
b2-Agonists Short-acting Long-acting Other b-agonistsa Xanthines Ipratropium bromide Leukotriene receptor antagonists Inhaled corticosteroids Beclomethasone Budesonide Fluticasone Mometasone Nasal corticosteroids Oral corticosteroids a
Oral or nebulised b-agonist or in combination with ipratropium.
S. Suissa et al. / Pulmonary Pharmacology & Therapeutics 22 (2009) 194–198 Table 3 Durations of complete follow-up, persistent use and treatment success for budesonide/formoterol and fluticasone/salmeterol users in the year after the cohort entry (initiation of combination therapy).
Number of subjects Person-years of follow-up, total Follow-up time, mean (months) 12 months (%) 6 to <12 months (%) <6 months (%) Number of prescriptions of the cohort-defining combination therapy, mean Duration of persistent use, mean (months)a Duration of treatment success, mean (months) Duration of complete treatment success, mean (months) Switch to the other combination therapy (%) a
Budesonide/ formoterol
Fluticasone/ salmeterol
6918 6606.4 11.6 87.0 9.1 3.9 4.9
16157 15423.3 11.6 86.7 9.4 3.9 5.6
5.5 8.7 4.7
5.6 8.6 4.2
5.2
2.9
Includes a 30 day grace period after the last prescription.
(adjusted rate ratio 1.02; 95% CI 0.93–1.11), and the number of antibiotic prescriptions was reduced by 8% (95% CI 4–13%). The duration of treatment success was similar in the two groups, 261.5 days (136.2) for users of budesonide/formoterol and 256.7 (136.6) for users of fluticasone/salmeterol. Complete treatment success was on average longer by 4.9 days in the budesonide/formoterol group and this was statistically significant (p ¼ 0.013; 95%CI 1.0–8.7). A switch from budesonide/formoterol to fluticasone/salmeterol was more likely to occur than the other way around (adjusted hazard ratio 1.78, 95% CI 1.56–2.06). The mean daily dose of budesonide during follow-up was 281.8 mg (SD 217.3) in the budesonide/formoterol cohort, while the mean daily dose of fluticasone in the fluticasone/salmeterol cohort was 264.7 mg (SD 245.4). Results were not substantially affected when the analyses were restricted to subjects younger than 50 years old and/or subjects prescribed inhalers with 120 doses only. 4. Discussion We compared outcomes in patients with asthma, treated in general practices in the UK, who initiated treatment with a combination of a long-acting b-agonist bronchodilator with an inhaled corticosteroid, either budesonide/formoterol or fluticasone/salmeterol. At the start of single inhaler combination therapy
197
(combination therapy), patients prescribed fluticasone/salmeterol were more severe as attested to by greater use of asthma medications in the prior year. These baseline differences in asthma severity were adjusted for using propensity scores and outcomes during the subsequent year were examined. Overall, there were no important differences in hospitalisations for asthma or visits to the physician for asthma, or for all causes, between the two treatment groups during follow-up. Duration of treatment success defined as the time to occurrence of an exacerbation or the need for different or additional anti-inflammatory therapy was also similar in both groups at approximately 8.6 months. Including the need for a prescription for a short-acting b-agonist in the definition of treatment success, the duration was greater for the budesonide/formoterol group (4.7 versus 4.2 months). This was also associated with fewer prescriptions for a short-acting b-agonist and for the combination therapy itself in the budesonide/formoterol group. Users of budesonide/formoterol had fewer total prescriptions for asthma medications during follow-up whether considering initial or persistent users of combination therapy, while the number of prescriptions for oral corticosteroids was similar. This apparently similar efficacy was obtained at a lower equivalent dose of inhaled corticosteroids in the budesonide/formoterol group if one considers fluticasone to be approximately twice as potent on a microgram per microgram basis [8,9]. On the other hand, more patients treated with budesonide/formoterol were referred to specialists for asthma and this was more marked for persistent users possibly suggesting an insufficient dose of inhaled corticosteroids for more severe patients. In support of this users of budesonide/formoterol were more likely to be prescribed an additional corticosteroid or long-acting b-agonist. Our study is observational in nature and subject to certain biases. Which subjects received a particular combination did not occur randomly and was likely related to patient and physician characteristics. To adjust for physician preferences we compared patients treated in the same practice. As for patient characteristics, those initially prescribed fluticasone/salmeterol appeared overall to be more severe and this difference was adjusted for using information on medication use and health care utilisation in the year prior to starting combination therapy. We cannot be sure that differences observed during combination therapy were not due to unmeasured differences between patients and their physician. We consider this unlikely, however, since the biggest difference between the groups initiating combination therapy, was the type of corticosteroid the patient had been prescribed previously and there was very little switching from one combination therapy to another, suggesting
Table 4 Crude and adjusted rate ratios of asthma medication and health care use from intention-to-treat analysis in the year after the start of combination therapy for budesonide/ formoterol relative to fluticasone/salmeterol. Rate per year of: Combination therapy of the same type
Budesonide/formoterol (person-years of follow-up ¼ 6606.36)
Fluticasone/salmeterol (person-years of follow-up ¼ 15423.33)
Crude rate ratio
Adjusted rate ratioa
4.05
4.86
0.83
0.87 (0.84–0.89)
Asthma medication use Short-acting b agonist Long-acting b agonist Inhaled corticosteroid Oral corticosteroid Total number of asthma medications
3.38 0.25 0.68 0.49 10.45
3.85 0.26 0.64 0.58 12.04
0.88 0.96 1.07 0.85 0.87
0.97 (0.94–1.00) 1.26 (1.17–1.35) 1.17 (1.10–1.24) 1.02 (0.96–1.08) 0.94 (0.92–0.97)
Health care utilisation Asthma hospitalisations Asthma referrals Asthma visits GP visits Number of antibiotic prescriptions
0.023 0.011 1.44 10.92 1.24
0.025 0.011 1.50 11.36 1.40
0.91 1.03 0.96 0.96 0.89
1.01 1.22 1.01 0.99 0.94
a
Adjusted for deciles of propensity score and propensity score.
(0.92–1.10) (1.12–1.34) (0.97–1.04) (0.96–1.01) (0.90–0.98)
198
S. Suissa et al. / Pulmonary Pharmacology & Therapeutics 22 (2009) 194–198
Table 5 Crude and adjusted rate ratios of asthma medication and health care use during persistent use of budesonide/formoterol or fluticasone/salmeterol in the year after the start of combination therapy. Rate per year of: Combination therapy of the same type
Budesonide/formoterol (person-years of follow-up ¼ 3127.92)
Fluticasone/salmeterol (person-years of follow-up ¼ 7453.54)
Crude rate ratio
Adjusteda rate ratio
5.61
6.52
0.86
0.89 (0.87–0.91)
Asthma medication use Short-acting b-agonist Long-acting b-agonist Inhaled corticosteroid Oral corticosteroid Total number of asthma medications
3.74 0.16 0.42 0.56 12.19
4.21 0.17 0.43 0.65 14.13
0.89 0.96 0.98 0.86 0.86
0.98 1.20 1.05 1.05 0.93
(0.94–1.01) (1.11–1.30) (0.98–1.12) (0.99–1.12) (0.90–0.95)
Health care utilisation Asthma hospitalisations Asthma referrals Asthma visits GP visits Number of antibiotic prescriptions
0.027 0.018 1.84 11.70 1.29
0.031 0.013 1.83 12.19 1.51
0.88 1.37 1.00 0.96 0.85
1.02 1.56 1.02 0.99 0.92
(0.93–1.11) (1.43–1.71) (0.99–1.06) (0.96–1.01) (0.87–0.96)
a
Adjusted for deciles of propensity score and propensity score.
that physician’s prescribing habits were the major determinant of which combination therapy was prescribed. We found that patients prescribed budesonide/formoterol used, on average, a lower dose of inhaled corticosteroid when considering equipotent doses of budesonide and fluticasone. One must note, however, that the actual number of inhalations per day was missing for a significant proportion of users of budesonide/formoterol and had to be estimated. The greater frequency of missing dosing information for budesonide/formoterol may have resulted from publication of studies of this medication which have used variable as opposed to fixed doses usually recommended for fluticasone/salmeterol [10,11]. The information for the present study was obtained from the population-based General Practice Research Database (GPRD), which has been described in detail [12]. More than 6 million people in the UK are enrolled with over 400 general practitioners who use office computers and have agreed to provide data for research purposes. The physicians generate prescriptions directly with their study computer; this information is automatically transcribed into the computer record. The recorded information on drug exposure and diagnoses has been validated and is of high quality [13]. This database has been the source for over 300 publications in the peerreviewed medical literature. In summary our results suggest similar efficacy for the combinations of an inhaled corticosteroid and a long-acting b-agonist, fluticasone/salmeterol and budesonide/formoterol. Such similar efficacy appears to be obtained at lower relative doses of inhaled corticosteroid in the budesonide/formoterol group. This research was funded by a grant from AstraZeneca and grants from the Canadian Institutes of Health Research (CIHR). SS is the recipient of a Distinguished Investigator award from the CIHR. The database was funded by grants from CIHR, the Fonds de la recherche en sante´ du Que´bec (FRSQ) and the Canadian Foundation for Innovation (CFI).
References [1] Expert Panel Report 3 (EPR-3): Guidelines for the diagnosis and management of asthma – summary report 2007. J Allergy Clin Immunol 2007;120(5 Suppl):S94–138. [2] Global Initiative for Asthma. Pocket guide for asthma management and prevention. 2006. [3] Lemiere C, Bai T, Balter M, Bayliff C, Becker A, Boulet LP, et al. Adult asthma consensus guidelines update 2003. Can Respir J 2004;11(Suppl A): 9A–18A. [4] Greenstone IR, Ni Chroinin MN, Masse V, Danish A, Magdalinos H, Zhang X, et al. Combination of inhaled long-acting beta2-agonists and inhaled steroids versus higher dose of inhaled steroids in children and adults with persistent asthma. Cochrane Database Syst Rev 2005;(4). CD005533. [5] Lemanske Jr RF, Sorkness CA, Mauger EA, Lazarus SC, Boushey HA, Fahy JV, et al. Inhaled corticosteroid reduction and elimination in patients with persistent asthma receiving salmeterol: a randomized controlled trial. JAMA 2001;285(20):2594–603. [6] Lazarus SC, Boushey HA, Fahy JV, Chinchilli VM, Lemanske Jr RF, Sorkness CA, et al. Long-acting beta2-agonist monotherapy vs continued therapy with inhaled corticosteroids in patients with persistent asthma: a randomized controlled trial. JAMA 2001;285(20):2583–93. [7] Mann M, Chowdhury B, Sullivan E, Nicklas R, Anthracite R, Meyer RJ. Serious asthma exacerbations in asthmatics treated with high-dose formoterol. Chest 2003;124(1):70–4. [8] Boulet LP, Becker A, Berube D, Beveridge R, Ernst P. Canadian asthma consensus report, 1999. Canadian Asthma Consensus Group. Can Med Assoc J 1999;161(11 Suppl):S1–61. [9] National Asthma Education Program NHLBI. Practical guide for the diagnosis and management of asthma. US Department of Health and Human Services 1997:1–52. [10] Aalbers R, Backer V, Kava TT, Omenaas ER, Sandstrom T, Jorup C, et al. Adjustable maintenance dosing with budesonide/formoterol compared with fixed-dose salmeterol/fluticasone in moderate to severe asthma. Curr Med Res Opin 2004;20(2):225–40. [11] Bateman ED, Boushey HA, Bousquet J, Busse WW, Clark TJ, Pauwels RA, et al. Can guideline-defined asthma control be achieved?: The Gaining Optimal Asthma ControL Study. Am J Respir Crit Care Med 2004;170(8):836–44. [12] Garcia Rodriguez LA, Perez GS. Use of the UK General Practice Research Database for pharmacoepidemiology. Br J Clin Pharmacol 1998;45(5): 419–25. [13] Jick SS, Kaye JA, Vasilakis-Scaramozza C, Garcia Rodriguez LA, Ruigomez A, Meier CR, et al. Validity of the general practice research database. Pharmacotherapy 2003;23(5):686–9.