For a one-year period, we collected data from a managed care database for all prescriptions filled for a three- to five-day course of canadian azithromycin therapy or a five- to 14-day course of clarithromycin therapy. The following groups of people were excluded from the study:

  • patients younger than 18 years of age
  • patients who were receiving concurrent antibiotic therapy
  • patients taking immunosuppressive medications, such as cyclosporine, tacrolimus, or chronic corticosteroid therapy (i.e., a prednisone equivalent of 20 mg/day or more)
  • patients taking medications for human immunodeficiency virus (HIV) infection
  • patients receiving chronic antibiotic therapy, including those taking azithromycin 1,200 mg/week or a single dose of azithromycin
  • patients receiving treatment for Helicobacter pylori

Even though we gathered data from December 1999 to January 2001, we compiled azithromycin and clarithromycin prescriptions from January 2000 through December 2000 for this retrospective study. The months of December 1999 and January 2001 were used to gather additional data.

Table 1    Cost-Minimization Analysis

Azithromycin Clarithromycin generic
Drug cost



Cost of other antibiotics



Cost of doctor visits



Cost of laboratory tests



Total cost/number
of prescriptions



For each subject, we collected the following information (Figure 1):

  • member identification number
  • personal information (date of birth/age/sex)
  • the treatment drug and pre-study and post-study antibiotics, with date of prescription, quantity, days of supply, and cost
  • concurrent medications with date of prescriptions

For statistical analyses, we used Student’s t-test to measure failure rates of the two study drugs.

Figure 1 A data collection sheet was used to gather

Figure 1 A data collection sheet was used to gather information about the patients who received therapy. DS = days of supply;Qty = quantity;Rx = prescription.

With the cost-minimization analysis for subjects who met the primary objective, we calculated the total cost of antibiotics for each patient, then summed total antibiotic costs in the azithromycin and clarithromycin study arms. We incorporated the cost of doctor visits and laboratory tests if therapy with two or more antibiotics was unsuccessful. We estimated the costs of $50 per doctor visit and $60 per laboratory test.


From January through December 2000, we assessed 633 prescriptions: 465 for azithromycin and 168 for clarithromycin. Of the 465 azithromycin prescriptions, 427 (92%) met the primary objective and 38 (8%) met the secondary objective. Of 168 prescriptions, 150 (89.3%) met the primary objective and 18 prescriptions (10.7%) met the secondary objective. Figure 2 illustrates the stratification of azithromycin and clarithromycin prescriptions.

Figure 2 Stratification of azithromycin and clarithromycin prescriptions

Figure 2   Stratification of azithromycin and clarithromycin prescriptions. AB = antibiotic; AZI = azithromycin; CLAR = clarithromycin; Rx = prescription.

The failure rates of azithromycin and clarithromycin were similar for both primary and secondary objectives. For the primary objective, the failure rate for azithromycin was 14.8%; for clarithromycin, it was 16% (P = .7137). For the secondary objective, the failure rates and clarithromycin were 34.2°% and 22.2°%, respectively (P = .3621).

We conducted a cost-minimization analysis for the prescriptions that met the primary criteria. After summing the cost of the study drugs, other antibiotics, doctor visits, and laboratory tests, we divided this value by the number of prescriptions. Table 1 (see page 659) clearly illustrates that clarith-romycin was twice as expensive as azithromycin.