Clinical Drug Experience Knowledgebase

METHODS

COUGH STOP was an investigator-initiated study. We conducted a prospective, randomized, double blind, placebo controlled trial to test the hypothesis that ipratropium would improve cough symptoms in otherwise healthy subjects with acute bronchitis. Subjects were enrolled between October 9th 2002 and November 1 2003, and were included if they had a cough with or without sputum production for less than 30 days duration; were age 18 through 65; and were willing to follow up by phone for a brief interview at 2, 4, and 8 days after enrollment. Subjects were excluded if they had a history of COPD; asthma; or other lung disease; had localized lung findings on exam to suggest pneumonia or asthma; chest X-ray (if done) with evidence of pneumonia; purulent nasal discharge or other evidence of bacterial sinus infection; evidence of streptococcal pharyngitis; temperature greater than 101.5 in the preceding 72 hours; treatment of a respiratory tract infection in the last 30 days; pregnancy; breast feeding; actively trying to become pregnant; history of heart failure; history of renal failure or insufficiency with a creatinine greater than 2.0 mg/dl; history of psychiatric illness other than minor depression; currently incarcerated; or were unwilling to sign the consent form.

The study was reviewed and approved by the IRB for our facility, and approved by the FDA by way of an IND application. Subjects were either referred to one of the study investigators by a practicing physician or nurse practitioner at our facility, or were enrolled by a member of our team directly. Subjects were given a copy of the consent form, asked to review it and ask questions. Those who signed the consent form were randomly assigned to treatment with ipratropium or placebo metered dose inhaler (MDI). The assignment key was maintained by the pharmacy and by one of the investigators who had no interaction with the subjects or the data collection process. Subjects received training on how to use an MDI with a spacer device from a member of the study team. Subjects were instructed to take two puffs from the study MDI four times a day for the next 8 days. After the first 8 days they were instructed to use the study MDI on an as needed basis if they thought it was benefiting them.

During a structured enrollment interview subjects were asked the following questions: How long have you been coughing (number of days)? What other symptoms do you have? Are you producing sputum? Do you smoke? Are you coughing at night? If "Yes," is it interrupting your sleep? Have you stopped working/daily activities? Are you having fevers? If "Yes," how high are they? What over-the-counter medications are you using?

During the follow up telephone interviews, on days 2, 4, and 8 after enrollment, subjects were asked a similar but shorter set of questions: Are you still coughing? Is your cough better, worse, the same? Are you producing sputum? Is your sputum production better, worse, the same? Are you coughing at night? If "Yes," is it interrupting your sleep? Have you returned to work/daily activities? Has your sense of well being returned? Are you having fevers? If "Yes," how high are they? Are you having any side effects with the medication? If "Yes," what are they? Are you having any problems using your inhaler? If "Yes," please describe. What over-the-counter medications are you using? Subjects were not given any medical advice during the follow up interviews and were advised to address any medical questions with their primary care physician.

Medical records were reviewed after two months of enrollment for, frequency of return clinic visits, medication prescriptions, especially antibiotics, and complications related to study enrollment. Subjects who had either a subsequent antibiotic prescription or clinic visit for respiratory complaints in the 60 days following study enrollment had an audit of their medical records. Subjects who were prescribed antibiotics in this period had their records reviewed for the following: which antibiotic was prescribed; how long after enrollment it was prescribed; and why it was prescribed. This data was collected in a separate database and reviewed in a blinded fashion by the entire research team for inclusion in the final analysis. All decisions were unanimous, and based on the following factors: the diagnosis at the time of the prescription was consistent with that of an upper respiratory tract infection (URI); the antibiotic prescribed was consistent with treatment for URI; in cases where the diagnosis at the time of prescription could not be determined, the subject was included in the secondary analysis if the antibiotic prescribed was consistent with treatment of a URI. Subjects were excluded from the analysis if their diagnosis at the time of the prescription was not consistent with URI or if the antibiotic prescribed was not consistent with treatment of a URI.

Similarly, subjects who had clinic visits in the subsequent sixty-days had their records audited. Data was again collected in a separate database and reviewed in a blinded fashion by the entire team. Again, all decisions were unanimous, and criteria for inclusion were that the diagnosis at the time of the visit was consistent with that of a URI.

Statistical Analysis Power analysis indicated that a sample size of approximately 100 subjects in the treatment and control arms would be sufficient to detect a 33% reduction of cough frequency in the experimental arm receiving ipratropium, assuming a =.05 and b =.20, using a two- tailed z-statistic. Calculation of standard effect size for this power analysis was based upon extrapolation from Hueston's and Melbye's data13 15 on albuterol and fenoterol use in acute bronchitis, assuming a similar effect of ipratropium. Effect size for this calculation was based upon published values of relative risk in the references. Power analysis was not performed for other outcome variables.

Data was analyzed on an intention-to-treat basis, according to randomized assignment. The ipratropium and placebo arms were compared for: self-reported cough severity [same, better, worse]; cough duration [present, not present] for days 2, 4, 8; , well-being [well, not-well] for days 2, 4, 8; work absence [attended work, did not work] for days 2, 4, 8. Proportions and rates for non-continuous grouped variables were compared using chi-square test. Normally distributed continuous variables, such as cough duration, were compared between treatment arms using a two-tailed t-test.