Journal of the American Pharmacists Association
TOOLS FOR ADVANCING PHARMACY PRACTICERetrospective Detection of Potential Medication Errors Involving Drugs with Similar Names
Section snippets
Objectives
The objectives of this study were to estimate the frequencies of potential errors involving drugs with similar names using a retrospective claims database and to validate frequencies of potential errors as a measure of drug name similarity by examining the association between the frequencies of potential errors and edit distance and normalized edit distance. The following hypotheses were formulated for the latter objective:
H1: A negative association exists between edit distance and frequencies
Methods
Idaho Medicaid claims data from 1993 to 2000 were analyzed for the purpose of this study. A computer program for measuring edit distance was developed by the programmer working with the Drug Utilization Review Board at the Idaho State University College of Pharmacy. Accuracy of the computer program was verified by comparing edit distance values for selected drug pairs with values obtained for those pairs from an external source.15 Normalized edit distance was computed by dividing the edit
Results
Based on screening of the Idaho Medicaid claims database using our potential error criteria, 106 of 300 drug pairs (35.33%) were involved in at least one episode of adjacent claims for both drugs. A total of 1,138 potential errors were observed; 17.54% were potential initial errors while the remaining 82.46% were potential refill errors (Table 2). The mean number (± SD) of total potential errors observed for any drug pair was 3.79 ± per drug pair. The mean number (± SD) of potential errors for a
Discussion
Using the method developed in this study, a claims database can be used to capture potential initial or refill errors that occur as a result of similarity in drug names. Algorithms used to detect potential errors were designed to identify claims sequences with high likelihood of potential errors. Besides focusing on adjacent claims sequences during an index period, the possibility of capturing look-alike/sound-alike medication errors was increased by using a confirmation period subsequent to
Limitations
Our method for identifying look-alike/sound-alike medication errors is based on important assumptions, including isolated incidence of look-alike/sound-alike errors, and rectification of such errors during the next refill. It did not capture situations in which drug errors lasted for the duration greater than the maximum index period of 60 days. Identification of this type of error was beyond the scope of error patterns defined in this study. The results of this study cannot be generalized for
Conclusion
Error criteria developed in this study were successfully applied to a retrospective claims database to detect potential initial and refill errors that involved drugs with similar names. Negative correlations observed between measures of orthographic drug name similarity and potential error frequencies confirmed the validity of our measure for use in look-alike/sound-alike medication error research.
Using study error criteria, pharmacovigilance researchers, health systems, or managed care
CE Credit
To obtain 1.5 contact hours of continuing education credit (0.15 CEUs) for completing “Retrospective Detection of Potential Medication Errors Involving Drugs with Similar Names,” complete the assessment exercise and CE registration form and return them to APhA. A statement of credit will be awarded to respondents achieving a grade of 70% or better. APhA continuing education policy provides you with two opportunities to successfully complete this continuing education examination. Please note
References (21)
- et al.
Effect of orthographic and phonological similarity on false recognition of drug names
Soc Sci Med
(2001) - Allan EL, Barker KN. Fundamentals of medication error research. Am J Hosp Pharm....
- et al.
Medication errors: experience of the United States Pharmacopoeia (USP) MEDMARX reporting system
J Clin Pharmacol
(2003) - et al.
Selected medication error data from USP's MEDMARX program for 2002
Am J Health Syst Pharm
(2004) Preventing medication errors
US Pharm
(Aug 1995)United States Pharmacopoeia. Use caution—avoid confusion
USP Qual Rev. USP
(Mar 2001)- et al.
Look-alike and sound-alike drug names: the problem and the solution
Hosp Pharm
(1992) Drug names that look and sound alike
Hosp Pharm
(1997)Predicting look-alike/sound-alike medication errors
Am J Hosp Pharm
(1997)- et al.
Similarity as a risk factor in drug-name confusion errors: the look-alike (orthographic) and sound-alike (phonetic) model
Med Care
(1999)
Cited by (0)
Continuing education credits: See learning objectives and test questions at the end of this article, which is number 202-000-05-197-H01 in APhA’s educational programs. CE answer sheets are located at the end of this article. To take the CE test for this article online, go to www.pharmacist.com/education.cfm, and follow the links to the APhA CE center.
Disclosure: The authors declare no conflicts of interest or financial interests in any product or services mentioned in this article, including grants, employment, gifts, stock holdings, or honoraria.
Acknowledgments: To Craig Kelley, BS, for computer programming and technical support; and Joseph Thomas III, PhD, for input during preparation of this manuscript.
Previously presented at the American Pharmaceutical Association Annual Meeting, March 15–19, 2002, Philadelphia, Pa.