The survey of look alike / sound alike ( LASA ) drugs available in hospitals in Thailand

A cross-sectional survey was designed to study look-alike, sound-alike (LASA) drugs in hospitals in Thailand. The questionnaires were developed and mailed to 1,380 hospitals throughout Thailand. The return rate was 11.16% or 154 hospitals, consisting of 5 tertiary hospitals (3.25%), 3 university hospitals (1.95%), 16 secondary hospitals (10.39%), 96 primary hospitals (62.34%), 26 private hospitals (16.88%) and 8 others (5.20%). A total of 5,327 pairs of drugs were identified as LASA drugs, including 3,695 tablets/capsules (Ranitidine-Roxithromycin pair in the highest frequency), 944 injections (DiazepamFurosemide pair in the highest frequency), 307 liquid dosage forms (Alum milk-Milk of magnesia pair in the highest frequency), 367 external drugs (0.02% Triamcinolone cream and 0.1% Triamcinolone cream pair in the highest frequency) and 14 pairs of chemotherapeutic agents. This LASA report could be integrated into a suitable program used in hospitals in order to identify and prevent medication errors in the future.


INTRODUCTION
The confusion of similar drug names is one of the most common causes of medication errors worldwide that threatens patients' safety (WHO, 2007a;Lambert et al., 2001Lambert et al., , 1999;;Basco et al., 2010;Phatak et al., 2005).With tens of thousands of drugs currently available, both of brandname and generic in the market, the potential for medication error due to confusing drug names is significant (WHO, 2007a;The Joint Commission, 2001).Causes of look-alike, sound-alike (LASA) medication errors were identified; and include illegible handwriting, unfamiliarity with drug names, similarity in the spelling and/or pronunciation of drug names, newly available products, similar packaging or labeling, similar clinical use, similar strength, dosage forms, frequency of administration, incorrect selection of a similar name from a computerized product list, and the failure of *Corresponding author.E-mail: chanatee_1@yahoo.com.Tel: 6645353630, 66846068745.manufacturers and regulatory authorities to recognize the potential for error and to conduct rigorous risk assessments, both for nonproprietary and brand names, prior to approving new product names (WHO, 2007a;Lambert et al., 2001;McCoy, 2005;Hoffman and Proulx, 2003).More than 33,000 trademarked and 8,000 nonproprietary medication names were reported in the United States alone in 2004, and an estimated 24,000 therapeutic health products were reported in the Canadian market (WHO, 2007a).On June 2011, the Institute for Safe Medication Practices (ISMP) (2011a, b) reported a listing of confusing drug names involved in medication errors that were reported through the ISMP National Medication Errors Reporting Program (ISMP MERP) (ISMP, 2011a).The United States Pharmacopeia (USP) also publishes a list of look-alike and sound-alike drug names periodically.Name pairs that have been included in LASA medication error reports are listed alphabetically (Cohen et al., 2007;Carothers, 1995).The aim of our list is to help healthcare providers and the public to determine which medications require special safeguards to reduce the risk of errors (ISMP, 2011).Sauberan et al. (2010) reported cases of medication dispensing errors with look-alike, sound alike medication in neonatal care units, where multidisciplinary collaboration within the system helped the pharmacy identify, resolve and prevent errors related to medication storage, labeling, delivery, knowledge, and administration documentation.Santell et al. (2009) found that look-alike, sound-alike drug names is one of the problems associated with computer entry and actions at the university of Pittsburgh Medical, causing of order-entry errors and adverse drug events.Number of errors from brand or generic name look-alike are 887 (1.6%) from Medmarx®, 22 (3.1%) from University of Pittsburgh Medical Center (UPMC).Actions included adding warning notes and mixed-case or uppercase lettering in the pharmacy computer system and informing pharmacy staff through e-mail notices of the potential for drug name confusion errors when drug products are added to the formulary or when drug mix-up errors occur.Additionally, the USP operates MEDMARX, an internet-accessible, anonymous error-reporting program specifically designed for hospital and health care system (Hicks et al., 2004;Nosek et al., 2010).A 2008 MEDMARX report by the USP identified 1,470 drugs involved in LASA errors, including categorizing medication errors (Hicks et al., 2008).A retrospective study revealed that one of the causes of medication errors in the inpatient department of Mahasarakham Hospital, Thailand, was LASA drugs, such as the confusing similar-sound of dobutamine injection and dopamine injection (Pattanajak, 2006).
In 2005, the World Health Organization (WHO) lanched the World Alliance for Patient Safety and identified six action areas which look-alike, sound-alike medication names is one of the inaugural patient safety solutions (WHO, 2007b).Similarly, in Thailand, the Thai Patient Safety Goals 2008 program was developed to improve medication safety with focus on LASA drugs.The program is based on WHO Collaborating Centre for Patient Safety Solutions (2007a, b) and is still used as a guideline in Thailand (The Healthcare Accreditation Institute, 2011).However, there have been few studies specifically addressing look-alike, sound-alike medication in Thailand, therefore this study surveys the pair list of LASA drugs in hospitals in Thailand.

METHODS
This was a cross-sectional survey research.The questionnaires were sent along with an accompanying letter by mail to 1,380 hospitals through out Thailand during December 2009 to January 2010.The questionnaire was designed to gather data in the area of personnel affiliation, the LASA drug list generic/brand name and the form of drug, including tablets/capsules, injectable drugs, liquid preparations, external use preparations and chemotherapy agents, as well as, a medication error categorization index (category A-I) for each LASA pair, according to the National Coordinating Council for Medication Error, Reporting and Prevention (NCC MERP 's, 1998-2012) index for categorizing medication errors.Category A represents circumstances or events that have the capacity to cause error, category B indicates that a medication error occurred but it did not reach the patient, categories C and D designate medication errors that reached the patient but did not result in patient harm.The varying level of patient harm is reflected in categories E, F, G and H. Category E refers to errors that may have contributed to or resulted in temporary harm to the patient and required intervention.Category F contains errors that may have contributed to or resulted in temporary harm to the patient and required initial or prolonged hospitalization.Category G labels all errors that may have contributed to or resulted in permanent patient harm.Category H errors require intervention necessary to sustain life and any medication error that resulted in, or may have contributed to a patient's death is classified as category I (NCC MERP, 1998MERP, to 2012;;Hartwigh et al., 1991;Dubey et al., 2006).All of the error reports were submitted using voluntary error reporting programs in each hospital, by health care professionals (pharmacists, nurses and physicians).

RESULTS
After a month of distribution, 154 completed questionnaires were returned (11.16% response rate).Data from questionnaires were evaluated.All respondents were affiliated with hospitals of various types, as shown in Table 1.Response was received, with most frequently response from primary hospitals (62.34%).Ninety percent (90.91%) of hospitals found LASA medication errors a relevant problem.
The study revealed two distinct types of drug similarity, based on the source of confusion.Similar spelling or pronunciation (LASA) caused visual and auditory confusion, while similar packaging/labeling (LA1) or similar tablets/capsules (LA2) caused visual-only confusion.
The most frequently occurring similarity was LASA stemming from similar spelling or pronunciation, as shown in Table 2. From the surveyed medication error categories, category B error occurred but did not reach the patient and was most prevalently reported (Table 3).No category I (fatal) errors were reported.Harmful medication errors (category F, G) are shown in Table 4. From all the reported drugs, 3,695 pairs were tablets/   capsules (Ranitidine and Roxithromycin pair in the highest frequency) (Table 5), 944 were injections (Diazepam injection and Furosemide injection pair in the highest frequency) (Table 6), 307 were liquid dosage forms (Alum milk and Milk of magnesia pair in the highest frequency) (Table 7), 367 were external drugs (0.02% Triamcinolone cream and 0.1% Triamcinolone cream pair in the highest frequency) (Table 8) and 14 pairs were chemotherapeutic agents (Table 9).There were 323 pairs of high-alert medications, all in injection preparation form.
However, the list of potential high-alert medications can vary by individuals and organizations.This study distinquished 6 distinct variations (Warinchumrab hospital, 2009;ISMP, 2011b) shown in Tables 10 to 14; including narrow therapeutic index medications (110, 34.06%), highly concentrated electrolytes (36, 11.15%), emergency      , 2005).This study found the source of confusion to include not only drug names but also look-alike product labeling and packaging (LA1), and look-alike tablets/capsules (LA2), confirming earlier reports (McCoy, 2005;Levine and Cohen, 2007;Cohen, 2007b).Category E, F and G (harmful) errors were reported in tablet/capsule preparations; however no category H and I (fatal) errors have been reported in our study.This could be the result of the obstacle of system, especially in light of 2006 MEDMARX® report whose 176,407 examined records 14 cases, where medication errors may have caused or contributed to patient deaths.The report also noted that the percentage of harmful errors has remained above 1% for more than seven years (Hicks et al., 2008).This research identified 5,327 LASA pairs, most of which came from tablet/capsule preparations (3,695 pairs, 69.36%).This could likely be explained by the prevalence of these forms of drugs on the market in general; with 15,404 and 4,287 preparations, tablets and capsules make up 43 and 12% of drugs available on the market, respectively (Thai Drug Control Division, 1999).Of all these, Ranitidine-Roxithromycin pair was reported the most frequently.Reported LASA-related errors fell into categories A, B, C, D, E, F and G.The highest harmful medication error was G, found only in tablet/capsule preparations.The study made a distinction in drug similarity, based on the source of confusion: similar spelling/pronunciation (LASA), similar packaging/labeling (LA1) and similar tablets/capsules (LA2).Two pairs of high-alert tablet drugs with a narrow therapeutic index are Warfarin (3 mg)-Warfarin (5 mg) pair and Warfarin (2 mg)-Warfarin (3 mg) pair; fortunately their LASAassociated error cateogories were only A and B.
From the 944 reported injection LASA pairs, Diazepam-Furosemide pair occurred in the highest frequency (34.55%).Diazepam is a high-alert drug with a narrow therapeutic range; however, the associated error categories were only A and B. From the 307 liquid dosage forms drug pairs, Alum milk-Milk of Magnesia pair was reported frequently.Among the 367 external drug pairs, the one occurred most frequently was 0.02% Triamcinolone cream-0.1% Triamcinolone cream.This study also identified confusing medication names and packaging in oncology medications.The names of several chemotherapy agents can look or sound alike (Carboplatin and Cisplatin, Vinblastine and Vincristine) or can be confused with unrelated medications (for example, Methotrexate and Multivitamin, found in this study), as noted also by Schulmeister (2006).Most of the high-alert drugs were found in injection preparations, a trend also shown in ASHP and other literature reports (ASHP, 2008;Hicks and Becker, 2006;Taxis and Barber, 2003a;Taxis and Barber, 2003b;Parshuram et al., 2008).James et al. (2009) investigated dispensing errors in the UK, US, Australia, Spain and Brazil.The most common dispensing errors identified by community and hospital pharmacies were dispensing the wrong drug, strength, form or quantity, or labeling medication with incorrect directions.They revealed that factors subjectively reported as contributing to dispensing errors were look-alike, sound-alike drugs, low staffing and computer software.High workload, interruptions, distractions and inadequate lighting were objectively shown to increase the occurrence of dispensing errors (James et al., 2009).Other factors associated with dispensing errors may be communication failures, problems related to package labels, work overload, the physical structure of the working environment, distraction and interruption, the use of incorrect and outdated information sources and the lack of patient knowledge and education about the drugs they use (Anacleto et al., 2005).The ISMP's landmark article by Leape et al, 1995. on systems analysis of adverse drug events defined broad categories, where the underlying problems that result in medication errors indentified the following proximal causes of medication errors: lack of knowledge of the drug, lack of information about the patient, violations of rules, slips and memory lapses, transcription errors, faulty identity checking, faulty interaction with other services, faulty dose checking, infusion pump and parenteral delivery problems, inadequate patient monitoring, preparation errors and lack of standardization (Cohen, 2007a).Risk reduction strategies include being aware of medications which look or sound like other drugs, installing pop-up alerts in computer systems, prescribing medications both by their generic and trade names, placing eye-catching labels and warning stickers on storage bins, storing medication in nonadjacent areas, and advising patients to be alert for potential mix-ups with look-alike, sound-alike medications (Schulmeister, 2006).Further strategies could invole the use of uppercase letters which successfully alert health professionals to potential for error with several generic name pairs, for example; acetoHEXAMIDE, acetaZOLAMIDE; lists of similar name pairs (Cohen et al., 2007), annual review and revision of a list of look-alike, sound-alike drugs and a proactive implementation of safety strategies to prevent Another important factor to consider is multidisciplinary collaboration within the system; increasing organisations' ability to identify, resolve and prevent errors related to medication storage, labeling, delivery, knowledge, and administration documentation (McCoy, 2005;Cohen, 2007a;Sauberan et al., 2010).The World Health Organization is currently developing global patient safety taxonomy that could facilitate consistent data collection and assist the development of error-reduction strategies (James et al., 2009).Little research is available on the topic, but regulators and the industry have begun discussing what level of similarity leads to confusion and errors (Cohen et al., 2007).The Food and Drug Administration (FDA) also has a role in preventing drug confusion.The FDA Office of Drug Safety, Division of Medication Errors and Technical Support (DMETS), reviews all trademarks prior to marketing and disapproves trademarks proposals that are deemed too similar to existing drug names (Cohen et al., 2007).Patient safety clearly requires improved methods for naming pharmaceutical products.FDA, the pharmaceutical industry, USP, ISMP, and others need to reach consensus and collaborate on a guidance document for product naming (Cohen et al., 2007).Such guidance should propose a common nomenclature with Although there is no single or simple answer to reducing medication errors, a coordinated effort to research problems associated with drug naming would produce timely and measurable results (Cohen et al., 2007).Like other voluntary error reporting systems, this study is limited by the voluntary nature in capturing medication errors, therefore significant errors may go unreported and thus may not be reflected in this report, a limitation that is also applicable to UPMC or MEDMARX data (Santell et al., 2009).While a detailed analysis of the of the root causes of LASA medication errors is outside the scope of this study, discernible patterns indicate that the potential for LASA drug errors is present at every step of the patient care process; prescription, dispensing and administration of drugs.Recommendations for preventing LASA medication errors at the prescription stage include maintaining awareness of LASA information, clearly indicating drug name, strength and dosage form on prescriptions, and abstaining from verbal-only orders.At the dispensing stage, pharmacy and therapeutic committees should regularly provide LASA information to health-care staff and consider the possibility of name confusion when adding a new drug product to the hospital formulary, avoiding LASA in both presentation and packaging.Furthermore, when it comes to computerized systems, the appearance of LASA drug names could be emphasized with highlighting, boldface, unique colors, uppercase lettering or a mix of upper and lowercase lettering at different in parts of the name, such as vinCRIStine, vinBLAStine.Pharmacy departments could store LASA products in different locations, or use alert stickers on shelves or places where LASA drugs are present.An independent double-check would also be useful to ensure LASA-error prevention at the dispensing process.
At the drug administration step, LASA-errors could be prevened by nurses reading prescriptions more carefully and considering them in the context of patient status and diagnoses.It is also important that appropriate systems for reporting errors and potentially hazardous LASA conditions are present and their use is encouraged.Moreover, professional staff could educate patients and their families about the confusing names and indication of their medication, while encouraging patients to ask about their medication if it is unfamiliar, or looks or sounds different.

Conclusion
A cross-sectional survey was developed to study lookalike, sound-alike (LASA) drugs in hospitals in Thailand.This LASA report could be developed and integrated to suitable programs used in hospitals in order to identify and prevent medication errors in the future.

Table 1 .
Affiliation of the hospital.

Table 2 .
Types of LASA (n = 6,550, one pair of LASA may have more than one type of LASA).

Table 3 .
Cross-tabulation of type of medication by error category index (n = 3,161; some pairs of LASA did not report error category index).
specialty hospitals and military hospitals, because they did not return the questionaire.The Joint Commission on Accreditation of Healthcare Organizations now requires that accredited health care organizations develop and maintain programs to minimize these LASA medication risks; a good example of this is the annual report list of

Table 10 .
Top ten of LASA pairs in high-alert narrow therapeutic index medications (N =110).