Abstract
Asthma management guidelines recommend the use of preventive medications in sufficient amounts to control asthma symptoms. The validity of a medication use index as a proxy for asthma severity has not been established. We recruited 1,279 Ontario adults with asthma or parents of children with asthma from a community-based surveillance program in 1995–96. Participants completed a telephone questionnaire at baseline, 3 and 6 mo. The questionnaire gathered information about asthma medication use, health care utilization, and symptoms. Asthma was classified as mild in 28%, moderate in 49%, or severe in 23% of patients based on the amount and types of medication used. There were significant differences among groups in health resource use such that adults with higher medication use visited primary care physicians and specialists more frequently, had pulmonary functions tests more frequently, and were admitted to hospital more frequently. The findings among children were similar. There were weak positive correlations between medication use and symptom frequency in adults and children. We conclude that a medication use index may be useful in population-based research where clinical asthma severity data are lacking. Such an index is distinct from but is related to disease control.
Keywords: asthma; severity of illness; medications; disease management; health services
The past two decades have witnessed a dramatic shift in the management of asthma and, consequently, our definitions of disease severity. When treatment was bronchodilator-driven, many patients with asthma suffered from persistent airflow obstruction at times of disease “stability,” and forced expiratory volume in one second (FEV1) was used to quantify disease severity. Asthma is now managed preventively; anti- inflammatory medications and adjunctive agents are used on a scheduled basis with the goal of preventing all or almost all symptoms of asthma and producing normal or nearly normal spirometry. In this setting, the amount of medication required to achieve these goals could be regarded as an indicator of disease severity. By contrast, the frequency of symptoms or exacerbations is an indicator of disease control.
Contemporary national and international guidelines for asthma management promulgate a relatively uniform set of recommendations for increasing amounts and types of preventive medication for increasing degrees of disease severity (1-5). Ascertaining asthma severity, however, is not straightforward. Existing indices of chronic asthma severity have relied on multiple dimensions, including symptoms, pulmonary function, and medication use, alone or in various combinations (1, 2, 6– 11). These scales are not easily adapted to population-based research when access to detailed clinical information is not available. Population-based observational research is challenged by the need to control for disease severity, a common and important confounder in assessments of patient outcomes. Assuming that prescriptions for asthma medications reflect the physician's multidimensional assessment of a patient's underlying severity, medication use could serve as an approximate index of asthma severity in population research. This hypothesis has not been tested. In this study, we gathered information about asthma medication use in a large population of adults and children with asthma and assessed the performance of a medication use index by comparing it with urgent and nonurgent health care use data as independent proxies for asthma severity. We also explored the relationship between medication use and symptom frequency as an index of disease control.
The study received ethics approval and participants provided written consent. The methods used to enroll patients have been published elsewhere (12, 13). Patients filling inhaler prescriptions in community pharmacies in southern Ontario reported medication use, health services use, and symptoms through telephone interviews over a 6-mo period. Health services use was measured as part of a larger cost of illness study (14-16). Among the 1,588 registrants who completed the study by March 31, 1996, 1,279 patients had probable asthma, defined as having a prescription for an inhaler and experiencing shortness of breath (SOB), wheeze, or recurrent cough in the past. To exclude probable cases of chronic obstructive pulmonary disease, patients older than 55 yr of age with a smoking history of 20 pack-years or more (e.g., 2 packs a day for 10 yr) and patients using oxygen were excluded. There were 940 subjects 15 yr of age and older and 339 subjects younger than 15 yr of age.
Participants underwent structured telephone interviews at 1, 3, and 6 mo postregistration. For children younger than 15 yr of age, information was collected from parents. In exceptional cases, older children were interviewed directly. The types and dosages of current asthma medications were reported at the baseline (1-mo) interview. Subsequent interviews recorded changes to the regimens. Evidence suggests that chronic medications are recalled more reliably (17, 18). The frequencies of respiratory-related visits to family doctors and respiratory specialists, pulmonary function tests, respiratory emergency department (ED) visits, respiratory admissions, and spacer and peak flow meter use occurring between interviews were reported at each interview. Frequencies of SOB, wheeze, and cough were reported for the 4-wk period before each interview. In a small subsample of adult study patients (n = 65), FEV1 results were extracted from medical charts.
Patients were classified as mild, moderate, or severe, based on reported daily medication use over the study period. Medication use has been recommended for ascertaining severity in the absence of clinical information (19, 20). In this paradigm, the daily amount and type of medication are regarded as a proxy for severity whereas the frequency of symptoms is regarded as an indicator of disease control. Classification was based on clinical expert opinion and Canadian guidelines available during the study (21). The criteria used to classify asthma medication use appear in Table 1.
| Intensity | Definition | |
|---|---|---|
| Mild | Bronchodilator monotherapy; inhaled bronchodilator use did not exceed 4 puffs/day over the study period | |
| Moderate | Bronchodilator monotherapy exceeded 4 puffs/day OR | |
| Inhaled anti-inflammatory (budesonide, beclomethasone, flunisolide, triamcinolone, fluticasone, cromolyn, or nedocromil) monotherapy did not exceed the equivalent of beclomethasone 800 μg/day | ||
| OR | ||
| Treatment with two types of medication, neither of which was an oral corticosteroid. If one or both medications were inhaled anti-inflammatory agents, the combined dose did not exceed the equivalent of beclomethasone 800 μg/day | ||
| Severe | Treatment with one or two types of medication, where one was an inhaled anti-inflammatory drug with a dosage greater than the equivalent of beclomethasone 800 μg/day | |
| OR | ||
| Treatment with three or more types of medication where one was an inhaled anti-inflammatory | ||
| OR | ||
| Treatment with oral corticosteroids |
Differences in the proportions of patients reporting a health service across the medication use categories were assessed with a chi-square test for visits to family doctors and specialists, pulmonary function tests, ED visits, admissions, spacer use, and peak flow meter use (22). Symptoms were analyzed as the proportion of patients reporting shortness of breath, wheeze, or cough and as the average daily frequency of each symptom for the month before the interview. A composite daily symptom score was calculated by summing the total reported number of all three symptoms per day. Differences in the proportions of patients reporting symptoms by medication use level were assessed with a chi-square test. Medication use level was correlated with health care visits and with symptom frequency using Spearman correlation coefficients (22).
Table 2 reports the study sample characteristics. A majority of patients were women, and 74% were 15 yr of age or older. As classified according to the medication index, mild and moderate patients accounted for 77% of the sample and 79% reported using inhaled anti-inflammatory medication. Eighty-nine percent of the sample had a drug plan to pay for medications, and 18% were current smokers.
| Sex | ||
| Male | 530 (41%) | |
| Female | 749 (59%) | |
| Age | ||
| 4 yr or less | 88 (7%) | |
| 5 to 14 | 251 (20%) | |
| 15 to 34 | 342 (27%) | |
| 35 to 54 | 361 (28%) | |
| 55 and older | 237 (19%) | |
| Drug plan | ||
| Yes | 1,132 (89%) | |
| No | 138 (11%) | |
| Unknown | 9 (1%) | |
| Medication use level | ||
| Mild | 352 (28%) | |
| Moderate | 630 (49%) | |
| Severe | 297 (23%) | |
| Medication class | ||
| BD alone | 255 (20%) | |
| IAI alone | 119 (9%) | |
| BD + IAI | 849 (66%) | |
| BD + IAI + OC | 55 (4%) | |
| Current smoker | ||
| Yes | 226 (18%) | |
| No | 1,053 (83%) |
Figures 1A and 1B depict the proportion of patients reporting various categories of asthma-related health resource use classified according to the medication index for adults and children, respectively. For all categories of health resource, the proportions of adult patients reporting use of that resource increased with higher levels of medication use. Of the health service use variables, family doctor visits were most common, with 32 to 51% of adults reporting at least one asthma-related family doctor visit during the 6-mo study period. Hospital admissions for asthma were rare; none were reported by patients classified as mild, 2% of adults classified as moderate reported an admission, and 5% of adults classified as severe according to medication use reported an admission. There was a statistically significant difference among classification levels (p < 0.0001) for family doctor visits, specialist visits, pulmonary function tests, and hospital admissions. Statistically significant differences among levels (p < 0.0001) were also observed for spacer use, which varied from 10 to 40% in adults and for peak flow meter use, which varied from 2 to 16%.
Fig. 1. Health resource use by medication use level. The proportions of patients (%) reporting each type of health resource use is displayed for adults and children. Panel A: Adults. *Differences between levels were significant, chi-square p < 0.0001. Panel B: Children. *Differences between levels were significant, chi-square p < 0.05. **Differences between levels were significant, chi-square p < 0.01. ▪ mild; ▩ moderate; □ severe.
[More] [Minimize]The patterns of resource use observed in children were similar to that observed in adults for most categories. Children demonstrated higher rates of specialist visits, ED visits, and spacer use. As with adults, the proportion of patients reporting the health resource use increased with increasing medication use in all categories. There was a statistically significant difference between medication use levels (p < 0.05) for family doctor visits and specialist visits. Statistically significant differences among levels (p < 0.01) were also observed for spacer use, which varied from 32 to 54% in children and for peak flow meter use, which varied from 4 to 21%.
The correlations between medication use level and the number of visits for various categories of health service use for adults and children are presented in Table 3. In adults, correlations were weak but statistically significant for all categories of health services except ED visits. In children, weak but significant correlations were observed for all health services except pulmonary function tests and admissions.
| Outcome | Adults (n = 940) | Children (n = 339) | ||||||
|---|---|---|---|---|---|---|---|---|
| Correlation | p | Correlation | p | |||||
| Health service | ||||||||
| Family doctor/pediatrician visit | 0.16 | < 0.0001 | 0.15 | 0.0069 | ||||
| Specialist visit | 0.22 | < 0.0001 | 0.16 | 0.0032 | ||||
| Pulmonary function test | 0.20 | < 0.0001 | 0.04 | 0.4269 | ||||
| Emergency room visit | 0.04 | 0.2407 | 0.12 | 0.0281 | ||||
| Admission | 0.13 | < 0.0001 | 0.04 | 0.4376 | ||||
| Symptoms | ||||||||
| Shortness of breath | 0.17 | < 0.0001 | 0.17 | 0.0022 | ||||
| Wheeze | 0.10 | 0.0014 | 0.14 | 0.0091 | ||||
| Cough | 0.14 | < 0.0001 | 0.10 | 0.0758 | ||||
| Composite score—Baseline | 0.14 | < 0.0001 | 0.10 | 0.0670 | ||||
| Composite score—Month 3 | 0.12 | 0.0004 | 0.13 | 0.0175 | ||||
| Composite score—Month 6 | 0.13 | < 0.0001 | 0.15 | 0.0055 | ||||
| Average composite score | 0.16 | < 0.0001 | 0.14 | 0.0124 | ||||
As seen in Figures 2A and 2B, there was a trend toward increasing proportions of patients reporting symptoms with increasing medication use in adults and children, respectively. The results depict the reporting by any patients of SOB, wheeze, and cough. In adults, approximately 80 to 90% of patients reported symptoms, depending on the level of medication use. The proportions of patients reporting symptoms were stable over time. In children, fewer symptoms were reported, with the proportions varying from approximately 70 to 85%. There was little change in reporting over time, although the difference between medication use levels was statistically significant at the 6-mo interview (p < 0.05). Although the trend toward increased proportions of patients reporting symptoms with increased medication use was present, it appeared less pronounced and less significant than observed with health resource use.
Fig. 2. Symptom frequency by medication use level. The proportions of patients (%) reporting shortness of breath, wheeze, or cough is displayed for adults and children. Panel A: Adults. Panel B: Children. *Differences between levels were significant, chi-square p < 0.05. ▪ mild; ▩ moderate; □ severe.
[More] [Minimize]The correlations between medication use level and the frequency of daily symptoms are presented in Table 3. Correlations were low but statistically significant at all study time points in adults. A smaller sample size in children resulted in reduced statistical power. Consequently, the correlation between baseline symptom frequency and medication use did not reach statistical significance.
In the subsample of adult patients for whom FEV1 data were available, the expected trend of reduced FEV1 percentage of predicted with increased medication use was observed. The Spearman correlation coefficient was −0.12 and was not statistically significant (p = 0.31); however, these FEV1 tests were performed an average of 1.5 yr before the study.
Our data suggest that medication use may be related to the complex construct of disease severity. In our study population, there was a relationship between levels of medication use and health care use such that patients with higher levels of use (and presumably more severe disease) saw physicians more frequently and were more likely to be hospitalized for asthma (adults) or to visit the emergency room for care (children). If indeed medication use is an indicator of disease severity, our data suggest that disease severity and disease control are not independent variables in a population sense. Patients with more dose-intensive medication use who may have had more severe disease were more likely to suffer symptoms of asthma than patients with less medication use and presumably milder disease. However, the correlation between medication use level and symptoms appeared to be weaker than the correlation between medication use level and health services use. These findings are similar to evidence that physician-assessed severity and patients' own perceptions of disease severity are more highly correlated with a National Asthma Education Program (NAEP) medication-based index of severity than with asthma symptoms (20, 23). Altogether, our data suggest an approximate but imperfect correspondence between medication prescribed for asthma management and actual disease severity. Additional research is required to fully explore the validity and reliability of a medication-based index by comparisons with multiple markers of disease severity, including pulmonary function, bronchial hyperresponsiveness, diurnal variation, symptom characteristics, and health-related quality of life.
We believe that this medication use index could prove useful as an estimate of a dimension of asthma severity in population health care research and health care use planning. Asthma is the most common chronic respiratory disease affecting 5 to 10% of the general population in most developed nations (24– 30). Although asthma is rarely fatal, it accounts for considerable chronic morbidity and substantial direct and indirect costs to society (14, 31, 32). We have previously shown that our medication use index correlates with productivity time losses and societal costs of asthma (14-16).
If asthma management goals were always achieved, there would be no correlation between disease severity and symptom control and an asthma medication index would be a clearer measure of underlying severity. In reality, many patients are inadequately controlled (33). In these situations, the index could misclassify as moderate, patients with severe underlying severity who are undertreated. It is not surprising that our data revealed weak correlations between disease severity, as measured by medication use, and the frequency of symptoms. In practice, there will always be some discordance between the need for medication and the amount of medication consumed. Such discordance will most likely occur when patients are newly diagnosed, when disease severity is worsening, and when medications are withdrawn to determine the least amount of medication required for long-term control. Discordance may also stem from patient noncompliance and insufficient awareness by physicians of their patients' medication needs (34, 35). Indeed, the weak but positive correlations between our medication-based index and symptom frequency suggest that overtreatment did not occur in the population studied.
Our data also reveal suboptimal asthma management in the population studied. Among adults with more intensive medication use and presumably more severe disease, only one in four saw a specialist during the 6-mo monitoring period and only one in seven had pulmonary function measured. Among children with more intensive medication use, only one in three was seen by a specialist and only one in five made use of a peak flow meter. If one were designing an intervention to improve asthma management and control, a medication-based index could be used to target the interventions toward those with most severe disease, and presumably, greatest need.
Some limitations to our study should be noted. First, the classification boundaries in our index may not always parallel treatment recommendations for step-up therapy in current guidelines. Our goal was to divide our population into categories for the purpose of facilitating the analysis of epidemiologic data, and our results suggest that our chosen cut-points have some validity. As the treatment perspective changes or as one applies the index to a different population with different health or prescribing practices, the index might require modification. Second, enrollment was limited to patients living in southern Ontario and our findings might not be representative of the health status in other regions of the country or in countries where health care delivery is markedly different. Nonetheless, we believe that the general conclusion that a medication usage index is a useful indicator of a dimension of asthma severity in population research is broadly applicable to Western nations where contemporary asthma consensus guidelines have been adopted. Third, our study population sample was not randomly chosen. The sampling plan included pharmacies based in low-, medium-, and high-density population regions, and pharmacists enrolled consecutive patients as much as possible. We believe that this stratified community pharmacy-based approach is superior to limiting enrollment to physicians' practices or a single health plan, but acknowledge that it skews enrollment to patients with more severe disease. Fourth, we relied on patient reports to determine a diagnosis of asthma and cannot rule out the possibility that some study subjects suffered from other respiratory or even nonrespiratory diseases. Finally, medication and health resource use data in adults were derived from patient self-reports with recall interval periods that varied from 2 to 3 mo. Self-reports of asthma medication use have been shown to be reliable over the short term (36). We previously compared self-reports of health care use in our adult study population with data contained in the provincial claims database. Self-reports were reliable for admissions, specialist visits, and primary care visits (37). It is possible that the parental proxy report of medication use led to misclassification. However, because children were enrolled at the time that their parents purchased their asthma medications, it is expected that these parents would be familiar with their child's medication regimen. The accuracy of parental proxy report of asthma health services use and symptoms is unknown. That the trends in children were the same as those observed in adults suggests that parents may be reliable proxy reporters. Weaker statistical findings in the pediatric subgroup may be due to the reduced power resulting from the smaller sample size.
In the conduct of epidemiologic research in the asthma population, it is essential to measure and adjust for underlying disease severity. We conclude that with further validity testing, the medication use index described here may prove to be a valuable tool for classifying asthma disease severity in an era of increased emphasis on population studies for assessing health outcomes and for conducting clinical evaluation and health services research.
The assistance of the Pharmacy Medication Monitoring Program Advisory Board is gratefully acknowledged. The PMMP Advisory Board includes Ms. Kathryn A. Gaebel, Dr. Charles H. Goldsmith, Dr. Mitchell A. H. Levine, Dr. Donald Willison, Department of Clinical Epidemiology and Biostatistics, McMaster University and the Center for Evaluation of Medicines, St. Joseph's Hospital, Hamilton, Ontario, and Dr. Jeffrey W. Poston, Canadian Pharmacists Association.
| 1. | National Institutes of Health. National Heart, Lung, and Blood Institutes. International Consensus Report on Diagnosis and Treatment of Asthma. U.S. Department of Health and Human Services, Bethesda, MD. NIH Publication No. 92-3091, 1992. |
| 2. | National Institutes of Health. National Heart, Lung, and Blood Institutes. National Asthma Education and Prevention Program Expert Panel Report 2: Guidelines for the Diagnosis and Management of Asthma. U.S. Department of Health and Human Services, Bethesda, MD. NIH Publication No. 97-4051, 1997. |
| 3. | 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–S61. |
| 4. | British Thoracic Society. Guidelines for management of asthma in adults. I - Chronic persistent asthma. Br Med J 1990;301:651–653. |
| 5. | Bernstein IL. Asthma in adults: diagnosis and treatment. In: Middleton E Jr, Reed CE, Ellis EF, editors. Allergy: principles and practice, 2nd ed. St. Louis: Mosby; 1983. |
| 6. | Jones KP, Bain DJG, Middleton M, Mullee MACorrelates of asthma morbidity in primary care. Br Med J3041992361364 |
| 7. | Woolcock AJ, Jenkins CRAssessment of bronchial responsiveness as a guide to prognosis and therapy in asthma. Med Clin North Am741990753765 |
| 8. | Richards JM, Bailey WC, Windsor RA, Martin B, Soong SJSome simple scales for use in asthma research. J Asthma251988363371 |
| 9. | Brooks SM, Bernstein IL, Raghuprusad PK, Maccia CA, Mieczkowski LAssessment of airway hyperresponsiveness in chronic stable asthma. J Allergy Clin Immunol8519901726 |
| 10. | Juniper EF, O'Byrne PM, Guyatt GH, Ferrie PJ, King DRDevelopment and validation of a questionnaire to measure asthma control. Eur Respir J141999902907 |
| 11. | Ellman MS, Viscoli CM, Sears MR, Taylor DR, Beckett WS, Horwitz RIA new index of prognostic severity for chronic asthma. Chest1121997582590 |
| 12. | Borden EK, Lee JGA methodological study of post-marketing drug evaluation using a pharmacy-based approach. J Chron Dis351982803816 |
| 13. | Willison DJ, Gaebel KA, Borden EK, Levine MAH, Poston JW, Goldsmith CH, Wong BExperience in the development of a postmarketing surveillance network: the Pharmacy Medication Monitoring Program. Ann Pharmacother29199512081213 |
| 14. | Ungar WJ, Coyte PC, Chapman KR, MacKeigan L, The Pharmacy Medication Monitoring Program Advisory Board. The patient level cost of asthma in adults in south central Ontario. Can Respir J 1998;5:463–471. |
| 15. | Ungar WJ, Coyte PC, The Pharmacy Medication Monitoring Program Advisory Board. A prospective study of the patient-level cost of asthma care in children. Pediat Pulmonol 2001;32:101–108. |
| 16. | Ungar WJ, Coyte PC, The Pharmacy Medication Monitoring Program Advisory Board. Measuring productivity loss days in asthma patients. Health Economics 2000;9:37–46. |
| 17. | West SLA comparison of data sources for drug exposure ascertainment in pharmacoepidemiologic studies with emphasis on self-reported information. Pharmacoepidemiol Drug Safety61997215218 |
| 18. | Harlow SD, Linet MSAgreement between questionnaire data and medical records, the evidence for accuracy of recall. Am J Epidemiol1291989233248 |
| 19. | Cockroft DWAsthma control and disease severity. J Allergy Clin Immunol98199610161018 |
| 20. | Osborne ML, Vollmer WM, Pedula KL, Wilkins J, Buist AS, O'Hollaren MLack of correlation of symptoms with specialist-assessed long-term asthma severity. Chest11519998591 |
| 21. | Hargreave FE, Dolovich J, Newhouse MTeds. The assessment and treatment of asthma: a conference report. J Allergy Clin Immunol85199010981111 |
| 22. | Hirsch RP, Riegelman RK. Statistical first aid: interpretation of health research data. Boston: Blackwell Scientific Publications; 1992. |
| 23. | Nguyen BP, Wilson SR, German DFPatients' perceptions compared with objective ratings of asthma severity. Ann Allergy Asthma Immunol771996209215 |
| 24. | Weiss KB, Gergen PJ, Wagener DKBreathing better or wheezing worse? The changing epidemiology of asthma morbidity and mortality. Ann Rev Publ Health141993491513 |
| 25. | Mannino DM, Homa DM, Pertowski CA, Ashizawa A, Nixon LL, Johnson CA, Ball LB, Jack E, Kang DS. Surveillance for Asthma— United States, 1960–1995. Morbid Mortal Weekly Rep, Surveillance Summaries 1998;47(SS-1):1–28. http://www.cdc.gov/epo/mmwr/preview/mmwrhtml/00052262.htm |
| 26. | Anderson HR, Butland BK, Strachan DPTrends in prevalence and severity of childhood asthma. Br Med J308199416001604 |
| 27. | Yunginger JW, Reed CE, O'Connell J, Melton LJ, O'Fallon WMA community-based study of the epidemiology of asthma: incidence rates 1964–1983. Am Rev Respir Dis461992888894 |
| 28. | Lava J, Moore R, Li F, El-Saadany S. Childhood Asthma in Sentinel Health Units. Report of the Student Lung Health Survey 1995–1996. Sentinel Health Unit Surveillance System - Asthma Component. Laboratory Centre for Disease Control, Respiratory Disease Division, Health Canada, Ottawa, ON. Catalogue no. H49-120/1998E, 1998. |
| 29. | Millar WJ. Childhood asthma. Health Reports 1998;10:9–21. Statistics Canada Catalogue No. 82-003. |
| 30. | Wilkins K, Mao YTrends in rates of admission to hospital and death from asthma among children and young adults in Canada during the 1980s. Can Med Assoc J481993185190 |
| 31. | Krahn M, Berka C, Langlois P, Detsky ADirect and indirect costs of asthma in Canada, 1990. Can Med Assoc J1541996821831 |
| 32. | Weiss KB, Gergen PJ, Hodgson TAAn economic evaluation of asthma in the United States. N Engl J Med3261992862866 |
| 33. | Chapman KR, Ernst P, Grenville A, Dewland P, Zimmerman S. Control of asthma in Canada: failure to achieve guideline targets. Can Respir J 2001;8(Suppl A):35A–40A. |
| 34. | Bauman A, Cooper C, Bridges-Webb C, Tse M, Miles D, Bhasale A, Pollock MAsthma management and morbidity in Australian general practice: the relationship between patient and doctor estimates. Respir Med891995665672 |
| 35. | Mawhinney H, Spector SL, Kinsman RA, Siegel SC, Rachelefsky GS, Katz RM, Rohr ASCompliance in clinical trials of two non-bronchodilator, anti-asthma medications. Ann Allergy661991294299 |
| 36. | Anie KA, Jones PW, Hilton SR, Anderson HRA computer-assisted telephone interview technique for the assessment of asthma morbidity and drug use in adult asthma. J Clin Epidemiol491996653656 |
| 37. | Ungar WJ, Coyte PC, The Pharmacy Medication Monitoring Program Advisory Board. Health services utilization reporting in respiratory patients. J Clin Epidemiol 1998;51:1335–1342. |
The analysis was undertaken with funding from the Hospital for Sick Children Research Institute. The Pharmacy Medication Monitoring Program Bronchial Inhalers Study was supported by grants from Ciba Canada Inc. and Glaxo Wellcome Canada Inc.
This article has an online data supplement, which is accessible from this issue's table of contents online at www.atsjournals.org
