An assessment of linezolid utilization in selected Canadian provinces

protein synthesis with activity against Gram-positive bacteria, including multidrug-resistant, methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant Enterococcus faecium (VREF) and Enterococcus faecalis (1-3). Linezolid was approved for use in Canada on April 4, 2001, for treatment of adults with VREF, group A and B streptococcus, methicillin-susceptible and -resistant S aureus, and penicillin-susceptible Streptococcus pneumoniae. Linezolid is available in both intravenous (IV) and 100% bioavailable oral forms. Currently, microbial resistance to linezolid is not significant in Canada (4). Rates of inappropriate antibiotic use are high in Canada (5-10). This is particularly concerning because it has been demonstrated that inappropriate use increases costs related to treatments (5-7). The present study’s primary objective was to characterize linezolid utilization in selected provinces in Canada and determine its rate of appropriate use as defined in the Infectious Diseases Pharmacy Specialty Network (ID-PSN) recommendations. A secondary objective was to examine associated resource utilization.

chosen for feasibility reasons -it was likely that linezolid use would be observed in these provinces due to the fact that they have higher MRSA incidence rates than the other Canadian provinces (11).
No sample size calculations were conducted.Because there had been relatively limited use of the product over the 12-month period before the study, it was estimated that 10 treatment courses per site would be the inclusion criteria and provide a representative sample size.This is consistent with published antibiotic drug utilization review (DUR) studies conducted in Canada (5)(6)(7)9,10,12).A pharmacist at each site was designated as the study investigator (see Appendix).The study protocol was approved by the governing ethics review board of each site.Study procedures were standardized among sites by the use of a single instruction guide at each site, a uniform training session for each pharmacist before initiation of data collection and by the provision of telephone support throughout the study to address questions.

Patient population
A pharmacy database at each site was used to identify hospitalized patients who received IV and/or oral linezolid treatment for an active infection.The course of linezolid may have been prescribed during hospitalization or at discharge from hospital.Each chart identified through the database search was screened for study eligibility.The inclusion and exclusion criteria were minimal to mimic real-life linezolid use.Patients were eligible for the study if they were at least 18 years of age, hospitalized at a study site and had received at least one linezolid treatment course (IV and/or oral) between October 1, 2001, andDecember 31, 2002.To discriminate between the number of linezolid treatment courses received by a patient, a single course was defined as continuous treatment with an interruption of no greater than 72 h.Patients were excluded from the study if they were enrolled in a clinical trial during hospitalization or treatment.If a site could not identify 10 eligible patients, a different site with higher linezolid use was asked to recruit more patients, with up to 20 patients per site.

Procedure
Data relevant to the determination of the level of appropriate linezolid use were abstracted from each chart.These data included the following: site and type of infection; culture and sensitivity results; antimicrobials prescribed before, simultaneous to or succeeding linezolid treatment for the same infection (start and stop dates, dose and route of administration); details of linezolid use (start and stop dates, dose and route of administration); and antibiotic contraindications (allergy history and significant rifampin drug interactions).Antibiotic contraindications also included a history of vancomycin intolerance.This intolerance was defined as any of the following: nephrotoxicity due to vancomycin (defined as two consecutive tests showing an absolute rise in serum creatinine levels [sCr] by 44 µmol/L or a relative rise in sCr by 50% from baseline, where baseline was the first sCr measured from the day of initiating vancomycin therapy); signs and symptoms of ototoxicity attributed to vancomycin; and vancomycin hypersensitivity reactions.The definition of vancomycin intolerance was based on the clinical judgment of the study investigators.Data on basic demographics and resource utilization (defined as duration of hospitalization) were also collected.

Data analysis
Data from all sites were pooled for all analyses except for the appropriate utilization analysis, in which case total and site-specific analyses were conducted.For provincial appropriateness ratings, data from Alberta and British Columbia were combined to represent the western provinces.
The proportion of treatment courses were classified as appropriate, partially appropriate or inappropriate.Treatment course classification was based on a decision algorithm shown in Figure 1.The decision algorithm was based on predefined treatment recommendations suggested by the ID-PSN (Table 1), which were reviewed and refined by the clinical expert and study advisors (see Appendix).Appropriateness of linezolid use was evaluated based on infection site, pathogen, sensitivities, contraindications, dose, dosage interval and duration of treatment.A treatment course was deemed completely appropriate if all criteria were met.A course was deemed partially appropriate if the dose and/or duration of treatment criteria were not met.A treatment course was considered inappropriate if it violated the criteria for appropriate indication, pathogen, sensitivities or contraindications.
If a decision on the level of appropriateness could not be made because a treatment course did not fit the decision algorithm, a blinded adjudication process was used.In summary, two study advisors (see Appendix) were independently presented with all clinical data collected for each treatment course that did not fall within one of the appropriateness categories; the advisors then classified them according to the classification system.For treatment courses where disagreement remained, the data were forwarded to the clinical expert (see Appendix) for a final decision.The total duration of hospitalization was calculated as the number of days between hospital admission and discharge.Infection-related hospitalization was estimated and defined as the number of days from the first day of anti-infective treatment received in hospital until the date of successful infection treatment in hospital, the date of discharge or until the patient died in hospital, whichever occurred first.

RESULTS
The study enrolled 95 patients who received 99 linezolid courses for treatment of 103 infections.Most patients were from sites in Ontario and Quebec (38.9% and 36.8%,respectively).Patient demographics are summarized in Table 2. Bone and joint surgeries comprised 41.2% of procedures performed in patients with surgeries in the previous six months.Debridement and/or irrigation of various sites comprised 35.9% of all procedures performed in patients with surgery during hospitalization.For the nine patients who expired during hospitalization, the causes of death were bacteremia (three patients), respiratory conditions (respiratory failure, bronchoaspiration, invasive aspergillosis and idiopathic pulmonary fibrosis, one patient each) and cancer relapse (one patient).One death was due to an unknown cause.
The pathogens identified by type of infection are shown in Table 3.The major use of linezolid was for treatment of skin and soft tissue infections (SSTI) (32.0%), followed by bacteremia (15.5%).The most prevalent pathogen was MRSA, identified in 44.7% of infections.
In 55.8% of patients, anti-infective treatment began before admission and continued following discharge from hospital.An allergy history to one or more antibiotic was reported in 52.6% of patients.Allergies to penicillin, vancomycin and the cephalosporins were the most common.A history of vancomycin intolerance was reported in 24% of patients and 18.9% of patients were unable to receive rifampin due to potential drug interactions.
The mean duration of total anti-infective treatment was 34.8 days (range one to 180 days).The mean durations of in-hospital and out-of-hospital anti-infective treatment were 24.2 days and 19.0 days, respectively.The duration of linezolid treatment and hospitalization are presented in Table 4.The majority of patients received some form of anti-infective drug before initiating linezolid treatment.In these patients, vancomycin and ciprofloxacin were the most common agents.Also, a small proportion received at least one other antiinfective following linezolid treatment.The most commonly initiated anti-infectives were vancomycin, rifampin and ciprofloxacin.Linezolid was used concomitantly (more than a two-day overlap) with other anti-infective drugs in 27.4% of patients.
Linezolid treatment courses lasted a mean of 14.4 days, with a minimum treatment period of one day (eg, due to treatment interruption for surgery) and a maximum of 57 days.Linezolid was primarily prescribed in the oral form.In 52.6% of patients, treatment of the infection with linezolid was successful.Linezolid was discontinued in 44.2% of patients, and the most common reason for discontinuation was the switch to another anti-infective (9.5%).
Patients spent a mean of 40.6 days in hospital and antiinfective treatment was administered 58.5% of this time.Of the 31.6% of patients who still required anti-infective treatment following discharge from hospital, the majority were discharged on linezolid in the oral form.Readmittance to the same hospital due to an event deemed to be associated with linezolid-treated infection occurred for 6.3% of patients and resulted in a mean of 9.0 additional days in hospital.
For the different linezolid formulations, the rate of appropriate use is presented in Table 5.The total linezolid treatment course was deemed appropriate in 53% of the 99 courses evaluated, partially appropriate in 1% of courses and inappropriate in 46% of courses.In 93.5% of courses deemed inappropriate, recommended first-line therapies were not attempted before linezolid.When the form of linezolid used was considered in addition to the sequence in which it was given, the highest rate of appropriate utilization occurred when linezolid was given following other anti-infectives in the IV to oral combination and during the switch to the oral form directly from other anti-infectives.
The rate of appropriate use for each infection type is shown in Figure 2. The highest rate of appropriate use included that of linezolid treatment for osteomyelitis, prosthetic joint infections, other infection types (due to Enterococcus species and other pathogens) and for treatment of patients with multiple infections.The rate of appropriate linezolid use also varied provincially and among sites.Quebec had the highest rate of appropriate linezolid use, followed by the western provinces and finally Ontario (59%, 57% and 44%, respectively).The rate of appropriate linezolid use was as low as 25% at one site and as high as 75% at another site.
Adjudication to determine the level of appropriateness was necessary in approximately 85% of the treatment courses.The first round of adjudication resolved 57% of these cases and the remainder were resolved by the next and final round of adjudication.Cases were adjudicated for the following reasons: infection and treatment characteristics; nonconfirmation of a pathogen; treatment of Enterococcus species infections when ampicillin susceptibilities were unknown or where no ampicillin contraindications existed; and for treatment of SSTI due

DISCUSSION
The present study characterized linezolid utilization in nine tertiary care centres across four provinces in Canada: Quebec, Ontario, Alberta and British Columbia.Overall, the rate of appropriate linezolid use was 53%.The mean duration of hospitalization was 40.6 days.Linezolid was primarily used to treat SSTI and bacteremias, with the most common pathogen being MRSA.The oral form of linezolid was most commonly prescribed following IV treatment with another anti-infective.
The level of appropriateness of linezolid utilization was assessed against recommendations developed by the ID-PSN pharmacists, with details finalized by the study advisors and the clinical expert.The recommendations reflected an evidence-based review of the existing clinical literature at the time the present DUR was conducted (2001), with the intent of publishing the guidelines.Infections considered by the guidelines included complicated and uncomplicated SSTI, pneumonia, infections due to vancomycin-resistant enterococci, osteomyelitis and prosthetic joint infections.Linezolid is not approved in Canada for use in pneumonia (caused by penicillin-resistant S pneumoniae and multiresistant species), prosthetic joint infections or for osteomyelitis.However, it was thought by the study advisors that, in light of the limited treatment alternatives available for these potentially life-threatening infections, the use of linezolid may be appropriate in certain settings when used after failed attempts by all other available agents.
The ID-PSN guidelines used to define appropriate use in the present study specify the prescribing of other anti-infective drugs before linezolid, except in certain situations.The majority of inappropriate linezolid use (93.5%) was due to use of linezolid as first-line therapy.Efforts could be made to clarify and emphasize proper first-line antibiotic therapy to improve linezolid use.The inclusion of an adjudication process and the large number of adjudicated cases (85%) in the present study reflects recognition by the clinical expert and study advisors that clinical practice does not always concretely fit into guidelines.As experience with linezolid increases, refinements to the guidelines will help to more completely define appropriate linezolid use.
The rate of inappropriate linezolid use in the present study is similar to those found in Canadian DUR studies of other antibiotics.The rate of inappropriate vancomycin use has been reported as 42% (7), 58.2% (6) and 65% (5).For other antibiotics, the rate of inappropriate use has been reported 42% for cefoxitin (6), 62% for metronidazole (10) and 80% for ciprofloxacin (8).Implementing drug utilization programs may reduce inappropriate antibiotic use.Several of these programs have been successfully implemented in Canada (13)(14)(15).The use of drug utilization programs may have accounted for some of the variations in inappropriate linezolid use at the different sites across Canada.
High rates of appropriate linezolid use were determined when linezolid was prescribed following treatment with another anti-infective drug.This high rate of appropriate use may be partially attributed to the stepdown to oral linezolid from IV treatment at the time of discharge to eliminate the use of IV treatment in outpatients.While this study was not designed to test whether this stepdown facilitated discharge from hospital, several Canadian studies have demonstrated that stepdown treatments reduce the duration of hospitalization (5,12,16,17).Indeed, Li et al (18) have shown that the duration of hospitalization is reduced during stepdown linezolid treatment compared with IV vancomycin treatment.
There are several limitations to the present study.Although the ID-PSN recommendations for appropriate use of linezolid were based on the literature available, they have not been

Figure 1 )
Figure 1) Appropriate use decision algorithm template Walker et al Can J Infect Dis Med Microbiol Vol 17 No 3 May/June 2006 180 Utilization of linezolid in selected Canadian provinces Can J Infect Dis Med Microbiol Vol 17 No 3 May/June 2006 181

Figure 2 )
Figure 2) Linezolid utilization classification by infection type

TABLE 1
Utilization of linezolid in selected Canadian provinces Can J Infect Dis Med Microbiol Vol 17 No 3 May/June 2006 179 Summary of Infectious Disease Pharmacy Specialty Network recommendations for appropriate use of linezolid* *Please see disclaimer at the end of the article; † Could be longer, if clinically necessary.IV Intravenous form; PO Oral form; PRSP Penicillin-resistant Streptococcus pneumoniae; PSSP Penicillin-susceptible S pneumoniae

TABLE 3
Percentage of infections with major pathogen identified

TABLE 4
Anti-infective therapies and length of hospitalization (n=95)

TABLE 5
Linezolid utilization classification by form of linezolid prescribed