Introduction

Top Abstract Introduction Subjects and methods Results Discussion References

International guidelines for the treatment of asthma regard inhaled budesonide and beclomethasone as equipotent.¹ This view is based on the results of randomised clinical trials comparing inhaled beclomethasone and budesonide which have mostly failed to detect differences in potency.^2-5 However, these studies have been limited by the small numbers of patients studied, the mild to moderate nature of their asthma, differences in inhaler devices used, differing end points and duration of treatment, crossover designs with carryover effects, previous use of inhaled corticosteroids, concomitant drug treatment, and the difficulties resulting from the flat dose-response relation for inhaled corticosteroids in the therapeutic range for provocation and lung function end points. ^{2 6 7} Other problems include the uncertain size of any differences when calculating sample sizes at the design stage of a clinical trial and the variability of asthma within and between patients. These limitations illustrate the difficulties in designing a definitive randomised clinical trial to further explore the relative potency of different inhaled corticosteroids in asthma. However, the topic remains of substantial interest not only for clinical reasons such as helping practitioners with anticipated inhaled corticosteroid dose equivalencies, advancing the debate about appropriate doses of inhaled corticosteroids,¹ and possibly explaining the anecdotal variability in usual prescribed doses in different countries, but also because of the costs entailed if these drugs are not equipotent.

The existence of primary care computerised databases containing standardised prescription and clinical information provides an opportunity to study relative potency of drugs in a pragmatic way using epidemiological methods. We used such an approach in this study accessing the database of the Royal New Zealand College of General Practitioners Research Unit, which includes patient identifier codes (all data are anonymised), consultation dates and free text, prescribing dates, and prescription details. ^{8 9}

	Subjects and methods

Top Abstract Introduction Subjects and methods Results Discussion References

Twenty eight general practices from the database were specified in the study protocol for this study as their computerised documentation of consultations and prescribing were known to be complete. In these 28 practices computers are the sole source of complete patient records. The study protocol was approved by the Southern Regional Health Authority (Otago) Ethics Committee in August 1996. The main objective was to use the inhaled corticosteroid prescribing data to compare the potency of budesonide and beclomethasone in general practice. All patients prescribed inhaled budesonide or beclomethasone formulations available in New Zealand from 1 July 1994 to 30 June 1995 were identified in the database.

A dataset was prepared including patient identifier code, date of birth, sex, consultation data, and prescribing date and information, including other drug treatment. The protocol specified age bands of two years up to the age of 14 and then bands of five years for analyses of inhaled corticosteroids by dose. Tabulations and histograms showing the proportions receiving one or more prescriptions above 800 µg per day (age to 14 years) or above 1500 µg per day (for age 15 years or over) in each age group were also defined in the protocol.

Prescribed daily doses of corticosteroids were calculated from the number of micrograms per inhalation for each drug and the number of inhalations indicated per day. Prescribed daily doses were based on the minimum dose indicated in the prescribing notes when a flexible regimen was prescribed. For example, if 2-4 puffs 2-4 times daily was prescribed the daily dose was calculated using 2 puffs twice daily. Qualifications on a few prescriptions that stipulated changing the dose when peak flow rate dropped below a given value were not taken into account. Corticosteroids that are inhaled through a metered dose inhaler and are available in New Zealand are Becotide (beclomethasone, Glaxo Wellcome; 50 µg, 100 µg, 250 µg), Respocort (beclomethasone, 3M; 100 µg, 250 µg), Atomide (beclomethasone, Douglas; 50 µg, 100 µg, 250 µg), and Pulmicort (budesonide, Astra; 200 µg). The breath activated devices are Becodisk (beclomethasone, Glaxo Wellcome; 100 µg, 200 µg, 400 µg), Pulmicort Turbuhaler (budesonide, Astra; 100 µg, 200 µg, 400 µg), Respocort (beclomethasone, 3M; 50 µg, 100 µg, 250 µg).

A scatter plot of the daily dose prescribed by age for patients receiving beclomethasone or budesonide alone showed higher doses and greater variation in the average daily dose with increasing age. These characteristics suggested that a log transformation of the data might be appropriate for subsequent analyses of the average daily dose of inhaled corticosteroids. Linear regression was therefore used to analyse the log data incorporating extra sum of squares techniques to determine whether the addition of a set of explanatory variables significantly improved the regression model. A search for evidence of bias towards usage of budesonide in patients with known or potential markers of more severe asthma was undertaken. This covered the number of courses of oral corticosteroids and antibiotics, the amount of oral corticosteroids prescribed, and admissions, outpatient consultations, emergency room visits, and overall consultation rates. Analyses for secondary care were based on linkage to hospital data whenever possible because data on hospital contacts in the database of the Royal New Zealand College of General Practitioners Research Unit are incomplete. We also analysed the doses of inhaled corticosteroids for patients changing from budesonide to beclomethasone and from beclomethasone to budesonide treatment.

	Results

Top Abstract Introduction Subjects and methods Results Discussion References

The computerised records of 128 585 patients and their 626 744 prescriptions were covered in this study. During the study period 6582 patients received 18 168 prescriptions for inhaled corticosteroids; 5930 (90.1%) of them who received 16 725 prescriptions (92.1%) had computer records that included dosing instructions. These 5930 patients were the focus of the subsequent analysis. In all, 4295 (72.4%) were exclusively prescribed beclomethasone and 1532 (25.8%) budesonide. Of the other 103 (1.7%) patients, 61 changed from beclomethasone to budesonide treatment and 32 from budesonide to beclomethasone treatment, with 10 changing more than once. The proportion changing in one direction was not significantly different from the proportion changing in the other (difference 0.65% (95% confidence interval -0.2% to 1.4%); P=0.08).

Prescribed daily doses
Table 1 shows the average daily dose in the prespecified age categories in patients who were exclusively prescribed budesonide or beclomethasone. Overall, the daily prescribed dose was higher for patients taking budesonide (mean 979 µg) than for those taking beclomethasone (mean 635 µg), a difference of 344 µg (95% confidence interval 313 to 376 µg). Further investigations using multiple regression analysis confirmed the significance of this difference (P<0.001) and its consistency across age bands, with some increase in doses with age for both corticosteroids (figure). The higher doses of budesonide prescribed were evident whether mean or geometric mean data were compared.

View larger version (28K): [in this window] [in a new window]   Relation between age and average daily prescribed dose of beclomethasone and budesonide. Regression lines and geometric mean doses with 1SE (dotted lines) are shown

Asthma severity analyses
A total of 1063 patients received 2221 prescriptions for prednisone (table 2). We performed specific analyses focusing on patients aged 35 years or under to limit confounding from patients with chronic bronchitis or emphysema.

	Discussion

Top Abstract Introduction Subjects and methods Results Discussion References

This epidemiological study has taken advantage of the opportunity provided by the database of the Royal New Zealand College of General Practitioners Research Unit to examine the relative potency of inhaled budesonide and beclomethasone in clinical practice. We found that the average prescribed daily dose of budesonide was about 50% higher than the average prescribed daily dose of beclomethasone. However, several potential sources of bias need to be addressed before budesonide is accepted as having a lower potency.

Potential sources of confounding
The main question is whether the findings are confounded by severity of asthma. We evaluated the recognised^10-12 and other possible markers of asthma severity. A detailed analysis which included the number of prescriptions for oral corticosteroids, number of inhaled corticosteroid prescriptions per patient, and antibiotic use did not show evidence of severity bias. Likewise, the increased rate of secondary care events in patients prescribed lower but not higher doses of budesonide did not support preferential prescribing of high dose budesonide to patients with more severe asthma.

Titration of dose according to response
The interpretation of our findings is based on the established clinical practice in New Zealand of general practitioners adjusting the dose of inhaled corticosteroid according to changes in the severity of their patients' asthma. This practice was formally recommended by the Department of Health in asthma guidelines sent to all medical practitioners in New Zealand in 1991¹³ and reinforced as part of a major initiative by the Asthma Foundation of New Zealand in 1992.¹⁴ It was also strongly promoted by the academic profession and pharmaceutical companies and, as a result, had become common practice for general practitioners before our study began.

Systemic effects and potency in vitro
Our findings are consistent with comparative studies investigating adverse systemic effects, which have fewer limitations than the comparative efficacy studies. Budesonide has been reported to cause less suppression of the pituitary axis¹⁵ and less suppression of serum markers for bone and collagen turnover¹⁶ than beclomethasone. Our findings are also consistent with in vitro data in which receptor affinity of 17-beclomethasone monopropionate (the active metabolite of beclomethasone dipropionate) is 1.4 times greater than that of budesonide.¹⁷

Conclusions
In conclusion, our results suggest that a higher dose of inhaled budesonide should be advised in the various asthma guidelines to give a treatment effect similar to a lower dose of inhaled beclomethasone. Confirmation of our results using similar computerised databases in other countries would not be difficult and is now required. Pending such epidemiological evaluation, international asthma guidelines may need to be modified. In addition, the fact that equipotency of these inhaled corticosteroids was generally accepted on the basis of clinical trials with low statistical power and results showing no difference raises new questions about the way in which clinical potency of inhaled steroids is assessed. We suggest that assessment of the potency of newly developed inhaled steroids in clinical trials will require confirmation in appropriately designed epidemiological studies based in general practice.