Amoxicillin of minimal benefit for purulent rhinorrhea - Patient-Oriented Evidence that Matters

Journal of Family Practice, August, 2002 by Kenneth G. Schellhase, David L. Fay

De Sutter AI, De Meyere MJ, Christiaens TC, et al. Does amoxicillin improve outcomes in patients with purulent rhinorrhea? A pragmatic randomized double-blind controlled trial in family practice. J Fam Pract 2002; 51:317-23

* BACKGROUND Upper respiratory tract infections (URTIs) involving purulent rhinorrhea frequently lead to inappropriate antibiotic use, in part because of the difficulty in identifying by history and physical examination patients likely to benefit from antibiotics. Previous research on antibiotic use for URTIs has relied on imaging studies not routinely obtained in clinical practice for diagnosis. In contrast, this study evaluated the effect of amoxicillin in patients identified as having URTI based solely on symptoms and clinical examination.

* POPULATION STUDIED The investigators studied 374 patients 12 years and older with URTI and purulent rhinorrhea presenting to 69 different family physicians in Flanders, Belgium. They excluded patients who had symptoms lasting more than 30 days, antibiotic therapy within the past week, evidence of complicated sinusitis, or immune system incompetence. The mean age of the patients was 37 years (SD = 14) for the intervention group and 39 years (SD = 15) for placebo. Slightly more than half of the patients were female. No other demographic information was given.

* STUDY DESIGN AND VALIDITY In this randomized, double-blind, placebo-controlled trial, patients received a 10-day course of either amoxicillin 500 mg orally 3 times daily or placebo. The study was large enough to show a 15% difference in therapy success rates between groups, but not a difference of 10% or less. All subjects had an initial clinical examination (history and physical) and completed a modified "SNOT-20" (sinonasal outcome test) symptom score questionnaire. Over the 10-day treatment course, patients kept a diary of their symptoms, also noting possible adverse drug effects and absence from work or school. Patients could use over-the-counter nasal decongestant drops or pain relievers at their discretion.

On day 10, patients returned for a repeat clinical examination, completion of a survey, and a pill count to assess compliance. Treatment success was defined at day 10 as a score of either 0 (absent) or 1 (very mild) for all of the symptoms defined by the patient to be most important. If the patient had not improved adequately, the physician was free to prescribe an open course of antibiotics without revealing initial treatment allocation.

Patients were randomized by a random-number generator, and randomization allocation was appropriately concealed. Placebo medication was identical to intervention medication in appearance and presentation. Compliance was high, at roughly 90% for both groups. At least partial follow-up data were obtained for approximately 90% of subjects in both groups. Withdrawals were less than 10% in each study group, and an intention-to-treat analysis was used.

This study had 2 key limitations. First, the authors did not clarify if the SNOT-20 survey, or their modification of the survey, was validated. Thus how accurately and reliably the survey will reflect patients' symptoms is unknown. Second, differential use of nasal decongestants between intervention and placebo groups could have an important impact on symptom scores; although the authors stared such use was recorded, no data were presented.

* OUTCOMES MEASURED Primary outcomes were treatment success rate on day 10, and duration of general illness, pain, and purulent rhinorrhea as recorded in the patient diary. Secondary outcomes included incidence of side effects and duration of sick leave.

* RESULTS Whether based on the repeat survey, diary entries, or clinical signs, treatment success rates were not significantly different between groups. The survey showed success for 35% in the amoxicillin group compared with 29% of the placebo-treated patients, which was nor a statistically significant difference. No significant difference was noted between groups for duration of general illness or pain, although duration of purulent rhinorrhea was shorter in the amoxicillin group (75% had complete resolution within 9 days) than the placebo group (75% resolved in 14 days; P = .007). Diarrhea occurred in every 10th person who received amoxicillin rather than placebo (29% compared with 19%, number needed to harm = 10). No significant difference was noted between groups for duration of sick leave.

RECOMMENDATIONS FOR CLINICAL PRACTICE

Overall, treatment of purulent rhinorrhea with amoxicillin does not result in faster resolution of symptoms than treatment with placebo. Although duration of purulent rhinorrhea was reduced with amoxicillin, this effect came at the expense of an increased risk of diarrhea. In the absence of signs of complicated sinusitis, patients who present with URTI and purulent rhinorrhea can be treated safely with decongestants and analgesics, and instructed to return if symptoms worsen or persist beyond 10 to 14 days.