
INTRODUCTION
Hypokalemia, one of the most prevalent laboratory abnormalities in clinical practice, is defined as a potassium value of less than 3.6 mmol/L. Hypokalemia occurs in over 20% of hospitalized patients. It can be caused by the loss of total body potassium, usually from the gastrointestinal tract as a result of vomiting, diarrhea, or a surgical fistula, or from the kidney as a result of renal disease, diuretic administration, or increased aldosterone production. Hypokalemia can also develop as a result of a shift from extracellular fluid into cells, as seen in patients with alkalosis or with diabetes after the administration of insulin for hyperglycemia. Clinically, patients with mild hypokalemia may have no symptoms, but those with more profound hypokalemia can develop generalized weakness, tetany, and cardiac arrhythmias. In addition, hypokalemia may increase the toxicity of digoxin, whereas prolonged hypokalemia can cause secondary renal tubular damage, which may be permanent.
The lack of a commercially available vancomycin solution for oral administration poses problems for adults and children who are unable to swallow capsules. Until the time of this study, vancomycin oral liquid was prepared on an “as needed” basis by the pharmacy staff at Fraser Health. These solutions were given short expiration dates, and refrigeration was specified for storage. To the authors’ knowledge, there are no published stability studies for vancomycin oral solutions prepared in equal volumes of Ora-Sweet and distilled water.
Regression analysis of the peak area ratios of vancomycin to internal standard versus vancomycin standard concentrations demonstrated linearity over the range of the concentrations, with coefficient of determination (r2) greater than 0.998 (n = 4). The intraday and interday CVs were less than 10% and therefore within acceptable limits: 2.7% and 5.5%, respectively, at 0.31 mg/mL; 7.22% and 7.37%, respectively, at 0.62 mg/mL; 2.17% and 8.76%, respectively, at 1.25 mg/mL; and 0.70% and 5.24%, respectively, at 2.5 mg/mL. The intraday and interday accuracies were at least 90% and therefore also within acceptable limits: 96.19% and 99.77%, respectively, at 0.31 mg/mL; 92.22% and 92.22%, respectively, at 0.62 mg/mL; 92.13% and 92.20%, respectively, at 1.25 mg/mL; and 94.57% and 96.17%, respectively, at 2.5 mg/mL. The lower limit of quantitation was 0.25 mg/mL, and the lower limit of detection was 0.0025 mg/mL. The retention times were 1.57 min for vancomycin and 2.59 min for metronidazole (Figures 1A and 1B).
Preparation of Study Samples
The vancomycin study samples were thawed, mixed (by vortex for 10 s), and diluted to 2.5 mg/mL in HPLC-grade water (Fisher Scientific, Richmond, British Columbia; lot 092106). A 0.5-mL aliquot of each sample was then diluted with 0.5 mL of internal standard (metronidazole 1 mg/mL). The final theoretical concentrations were 1.25 mg/mL for vancomycin and 0.5 mg/mL for the internal standard. Each sample was passed through a 0.45-^m microfilter before withdrawal of a 10-^L sample for injection onto the column.

Preparation of Vancomycin
Stock solutions of vancomycin 25 mg/mL were prepared by reconstituting commercially available vancomycin for injection (10 g/vial) (Pharmaceutical Partners of Canada, Richmond Hill, Ontario; lot 204369, expiry December 2009) with sterile water for injection (Baxter Corporation, Missis- sauga, Ontario; lot W9F15AO, expiry January 2011) and then diluting with a mixture of equal volumes of Ora-Sweet vehicle (Paddock Laboratories Inc, Minneapolis, Minnesota; lot 9085961, expiry February 2012) and distilled water (Ripple FX Water Inc, New Westminster, British Columbia; lot 9147102, expiry February 27, 2010). Portions of this preparation were transferred to fourteen 15-mL opaque blue polyethylene unit- dose cups with aluminum seal (Medical Packaging Inc, Ringoes, New Jersey; lot FD15MB0705) and six 50-mL amber polyvinyl chloride (recycle code 3) prescription bottles (Richards Packaging Inc, RIGO Products Division, Gloucester, Ontario). The volume dispensed was 5 mL into each unit-dose cup and 25 mL into each bottle. Half of the unit-dose cups and half of the plastic bottles were kept at room temperature (25°C) and the remainder of the containers were stored in the refrigerator (4°C).
INTRODUCTION
Orally administered vancomycin is indicated for the treatment of Clostridium difficile infections and is often used as the first-line agent in severe cases. With the rising incidence of C. difficile infection in the health care setting, usage of oral vancomycin has also increased. Typically, oral vancomycin is administered in capsule form. However, vancomycin in a liquid form is required for patients who have nasogastric and other feeding tubes and also for many patients in the elderly and pediatric populations.
As the body of drug information continues to grow in size and complexity, drug information databases have become essential tools for pharmacists in their role as medication experts. The results of this study confirmed the perception that pharmacists rely heavily on such databases, with the majority of study participants reporting that they consulted drug information databases at least daily. This finding further emphasizes the need for active engagement of end-users in objective comparisons between databases. A previous analysis ranked the databases used in this study as the top 3 online drug information databases in terms of scope, completeness, and ease of use. However, that earlier study yielded a different overall rank ordering: Clinical Pharmacology, Micromedex, and Lexi-Comp. Although the rankings in that study were not significantly different among the databases, the difference in rank order was most likely due to the study’s substantially different (and, in our view, less applicable) methodology and to changes to the databases that have occurred over the ensuing 3 years. We found that Lexi-Comp Online was the highest-quality, best-performing, and most preferred online drug information database among those evaluated. Clinical Pharmacology’s lack of Canadian content was identified as the most important disadvantage of this database, which highlights the relevance of completing a study like this from a Canadian perspective. It has been suggested that users’ evaluations of a database may be favourably affected by familiarity with the database. In the current study, participants reporting prior access to a particular database tended to rank that database higher than participants reporting no prior access to the database. However, this difference was not statistically significant, and, despite 88% of study participants reporting prior access to Micromedex, this database was still clearly the least preferred of the 3 databases.