Article writing homework help-qu- ct 9
Article writing homework help-qu- ct 9
International Journal for Quality in Health Care 2005; Volume 17, Number 1: pp. 15–22 10.1093/intqhc/mzi015
International Journal for Quality in Health Care vol. 17 no. 1 © International Society for Quality in Health Care and Oxford University Press 2005; all rights reserved 15
Errors in the medication process: frequency, type, and potential
MARIANNE LISBY1, LARS PETER NIELSEN2 AND JAN MAINZ3
1Aarhus University Hospital, Department of Quality improvement and Patient Safety, Aarhus, Aarhus, 2Aarhus University Hospital, Department of Clinical Pharmacology, Aarhus, 3National Indicator Project, County of Aarhus, Aarhus, Denmark
Objective. To investigate the frequency, type, and consequences of medication errors in more stages of the medication process, including discharge summaries.
Design. A cross-sectional study using three methods to detect errors in the medication process: direct observations, unan- nounced control visits, and chart reviews. With the exception of errors in discharge summaries all potential medication error consequences were evaluated by physicians and pharmacists.
Setting. A randomly selected medical and surgical department at Aarhus University Hospital, Denmark.
Study participants. Eligible in-hospital patients aged 18 or over (n = 64), physicians prescribing drugs and nurses dispensing and administering drugs.
Main outcome measures. Frequency, type, and potential clinical consequences of all detected errors compared with the total number of opportunities for error.
Results. We detected a total of 1065 errors in 2467 opportunities for errors (43%). In worst case scenario 20–30% of all evalu- ated medication errors were assessed as potential adverse drug events. In each stage the frequency of medication errors were— ordering: 167/433 (39%), transcription: 310/558 (56%), dispensing: 22/538 (4%), administration: 166/412 (41%), and finally discharge summaries: 401/526 (76%). The most common types of error throughout the medication process were: lack of drug form, unordered drug, omission of drug/dose, and lack of identity control.
Conclusion. There is a need for quality improvement, as almost 50% of all errors in doses and prescriptions in the medication process were caused by missing actions. We assume that the number of errors could be reduced by simple changes of existing procedures or by implementing automated technologies in the medication process.
Keywords: Medication errors, potential adverse drug events, chart review, direct observation, unannounced control visit
Medication errors are a well-known problem in hospitals. Studies have shown that medication errors and adverse drug reactions (ADRs) are one of the main causes for adverse events in hospitals leading to disability and death in up to 6.5% of hospital admissions [1–5].
Not all medication errors are harmful. Previous studies of the association have shown that <1% of all medication errors actually resulted in an adverse drug event, while up to 6.7% were assessed as potential adverse drug events, when exam- ined primarily by chart reviews [3,6,7]. However, a study using direct observation to detect medication errors in the dispens- ing and administration stage assessed 20% of the identified errors as potential adverse drug events . Thus, application of appropriate methods for identifying medication errors and
assessing potential adverse drug events are important in the detection of valid and useful information [9,10].
The literature on medication errors lacks universally accepted definitions of medication errors as well as different methods and criteria, leaving us with an incomplete know- ledge of the actual rate of medication errors [1–5,7,8,11–14]. Likewise, criteria for assessing the potential clinical conse- quences of medication errors vary in different studies [3,8,15]. At present, no studies have investigated medication errors in more stages of the process in the same population including discharge summaries. In addition, no studies have investi- gated medication errors in more stages of the medication process by selecting the most appropriate and valid methods at each stage.
Address reprint requests to Marianne Lisby. E-mail: email@example.com
M. Lisby et al.
The purpose of our study was to perform a systematic, valid, and detailed investigation of the frequency, type, and poten- tial clinical consequences of medication errors in more stages of the medication process, including discharge summaries.
Definition of medication errors
Prescription of drugs can be divided into an intellectual part—decision making, i.e. knowledge of diagnosis, inter- actions, and contraindications, and a technical part including communication of essential information, i.e. drug name, dose, form of administration . Our study focused on medication errors in the technical part.
A medication error was defined as an error in the medication process: ordering, transcription, dispensing, and administration, and discharge summaries . Errors included wrong as well as missing actions.
Adverse drug events were defined as injuries resulting from medical interventions related to a drug—including both medi- cation errors and ADRs. ADRs were excluded in our study [3,6]. Potential adverse drug events were defined as medication errors with potential for an adverse drug event [3,6].
Design and study population
The study was designed as a cross-sectional study of medi- cation errors and potential harm, and was examined by the use of three methods—direct observation, unannounced control visit, and chart review. The study population con- sisted of: (i) hospital inpatients, aged 18 or over; (ii) nurses dispensing and administering medications; (iii) physicians pre- scribing drugs or secretaries transcribing drugs into the medi- cal record. Prescription errors in the medical record and discharge summaries could be caused by both physicians and secretaries. In our study no distinction was made between these two groups.
The study included regular as well as pro re nata (prn) medi- cations, except from prn medications in discharge summaries. The following drugs were included: tablets, suppositories, mixtures, and injections (intravenous, intramuscular, and sub- cutaneous). Patients administering their own drugs were excluded as well as injections in the unannounced control visit.
The study was conducted at randomly selected medical and a surgical ward at Aarhus University Hospital, Denmark, from January to April 2003.
Physicians were responsible for prescribing drugs and secretaries for transcribing them into medical records and dis- charge summaries. Nurses transcribed drug prescriptions from the medical record into a medication chart and were responsible for dispensing and administering medication. In each ward drugs were administered daily at 8.00 a.m., 12.00 a.m., 5.00 p.m., and 10.00 p.m. by two randomly selected nurses.
Methods, process, and criteria
The criteria for classifying medication errors were based upon American Society of Health System Pharmacists (ASHP) def- initions and Danish legal recommendations for unambiguous prescriptions and correct verification of patient identity . A slight modification of ASHP’s criteria was necessary, in order to avoid overlap of error types and frequencies when separating the dispensing and administration. Definitions of included variables and criteria for errors are shown in Table 1. For each stage in the medication process a structured register form was developed.
The basic data in our study were the number of actual errors divided by the total number of opportunities for errors. An opportunity for error was defined as any drug prescribed, any unordered or omitted drug, and ‘any dose given and any dose omitted’ [11,18]. A drug or a dose could result in more than one error type.
Data collection consisted of five consecutive days (8 h) direct observation in each ward: four days in the daytime and one during evening shift. In the study period, two nurses were observed for ~4 h in each duty. The observed nurses were selected by convenience sampling. All dispensed and adminis- tered drugs were registered and subsequently compared with eligible prescriptions in the nurse medication chart. Any dis- crepancy between the dispensed drugs and the nurse chart was registered as an error according to the criteria in Table 1. The observation was carried out by one of us (M.L.), who did not have any previous knowledge of the patient’s drug use .
Unannounced control visit
An unannounced control visit was made 7 weeks after the observational study. The dispensed drugs were confiscated after dispensing and before administering, and new drugs were immediately dispensed. Two pharmaceutical students identified the confiscated drugs on behalf of recognition. The students’ findings were compared with the drug prescriptions in the medication chart. Any discrepancy between the pre- scriptions in the medication chart and the identified drugs was registered as an error (Table 1). In a pilot test the stu- dents’ recognition of drugs was 92% and 95%, respectively.
The medical records and nurse charts of all patients in the observational study were screened for medication errors. The screening should verify that all eligible prescriptions in the medication chart were identical to the prescriptions in the medical records, and examine whether the prescriptions in the medical record were unambiguous (Table 1). If patients were involved in more than one sample during the observational study, only new and changed prescriptions were screened.
Likewise, discharge summaries were screened for transcrip- tion errors. Any discrepancy in eligible prescriptions and in
Errors in the medication process
communicating essential information between prescriptions in the medical record and discharge summaries was registered as an error (Table 1). All chart reviews were conducted by one of the authors (M.L.)
Potential clinical consequences
Apart from errors registered in discharge summaries, all errors identified in the present study were assessed according to pre- defined criteria for potential clinical consequences (Appendix 1). A four-scale unambiguous classification system was developed including the following categories—potentially fatal, potentially serious, potentially significant, and potentially non-significant. Definitions of potential fatal and potential serious errors were in accordance with international definitions of potential adverse drug events [3,6].
Two expert physicians in each ward and a group of three experienced pharmacists assessed the potential clinical consequences of identified medication errors. The physicians independently assessed errors in their own ward, while the pharmacists assessed errors in both wards using the audit principle.
The data were analysed by using SPSS 11.0 (SPSS Inc., Chicago, IL, USA) and NCSS 2000 (McGraw-Hill Co.). Frequencies were described as percentages and the descriptive data as mean or median, if appropriate. Comparison of frequencies between the medical and surgical wards was made using the χ2 test or Fisher’s exact test. Parametric data were compared using the Student’s t-test while non-parametric data were com- pared using the Mann–Whitney non-parametric test. Friedman’s
two-way ANOVA was used in order to compare physicians’ and pharmacists’ assessment of the clinical consequences.
Statistical significance was defined at a level of 0.05 and data were described with a confidence interval of 95%.
The study was approved by the Danish Data Protection Agency. Nurses in both wards were informed verbally and in writing about the purpose of the observational study, but not about the unannounced control visit and chart reviews. Physi- cians were not informed about the study. The investigator was ethically obliged to interfere immediately if a medication error was observed and the interference would proceed before the nurse administered the medication. All medication errors prevented by the investigator would be registered as a medi- cation error.
Opportunities for errors were independently summed up for each stage in the medication process. In total 2467 opportunities for errors were registered of which 1067 (43%) errors were detected. Doses and prescriptions were equally distributed between the two wards with 1209 in the medical ward and 1258 in the surgical ward, in 27 and 37 patients, respectively. The estimated median error rate per patient in the first sample was 17 (11–24) errors per patient in the medical ward and 13 (7–22) errors per patient in the surgical ward. There was no statistical difference between the error rate per patient in the medical and the surgical wards (P = 0.13).
Table 1 Criteria for medication errors
Stage Definition Error types ………………………………………………………………………………………………………………………………………………………………………………………………
Ordering Unambiguous prescription Omission of: drug name; drug formulation; route; dose; dosing regime; date; signature; treatment time for antibiotics
Transcription An identical copy of prescription in medical record
Discrepancy in: drug name; drug formulation; route; dose; dosing regime; omission of drug; unordered drug
Dispensing Dispensed medication is concordant with prescribed drug in nurse medication chart
Unordered drug (wrong drug); unordered dose; omission of dose; wrong dose; wrong drug formulation
Administering The right medication to the right patient in the right way and at the right time
Wrong: administration technique (inj.); route; time (± 60 min); delivery (dose not delivered directly to the patient); unordered drug; unordered dose; omission of dose; lack of identity control
Discharge summaries Eligible prescriptions in medical record are identical to prescriptions in discharge summaries
Discrepancy in: drug name; drug formulation; route; dose; regime; omission of drug; unordered drug
M. Lisby et al.
All eligible patients where medication was dispensed by a nurse were included in the study. In the surgical ward five discharge summaries were excluded because the patients did not have any regular drug prescriptions and one medical record was not accessible in the medical ward.
The included patients consisted of 14 men (52%) and 13 women (48%) in the medical ward and 16 men (43%) and 21 women (57%) in the surgical ward. The mean age in the medical ward was 55 years (95% CI: 48–62) and in the surgical ward, 62 years (95% CI: 56–68). In the first study sample the median days of admission was 4 days (range 2–7) in the medical ward and 4 days (range 2–9) in the surgical ward. The number of opportuni- ties for errors per patient in the medical ward was 44 (range 22– 59) and in the surgical ward 29 (range 15–46). There were no statistically significant differences between the wards.
Frequency of medication errors varied between the differ- ent stages in the medication process. In Table 2 the frequency of medication errors are shown, at different stages in the medication process and the two wards, respectively. Differ- ences between the number of prescriptions in the ordering and transcription stage were due primarily to unordered drugs in the medication chart. Yet, differences between the number of dispensed and administered doses in the observational study were due to lack of administering doses to patients attending medical examinations.
In discharge summaries more than half (215/401 or 54%) of all medication errors were related to a new drug prescribed during admission or changes in previous drug prescriptions.
Likewise, the type of medication error in the different stages of the medication process is shown in Table 3. It is noticeable that error types like ‘lack of identity control’, ‘wrong time’ and ‘wrong delivery’ in the administration stage could be mutually dependent as lack of one identity control would automatically affect all doses the patient was about to receive with an error. Further analysis of these error types showed that ‘lack of identity control’ affected 21 of 43 (49%) deliveries in the medical ward and 21 of 56 (38%) deliveries in the surgical ward. ‘Wrong time’ affected six of 43 (14%) deliv- eries in the medical department and 1 of 56 (2%) deliveries in the surgical ward and finally, ‘wrong delivery’ affected two of 43 (5%) deliveries in the medical ward. Wrong administration
technique only included injections. Of the 24 injections observed in our study, five errors were detected.
Identified medication errors in the stages from prescription to administration were assessed retrospectively by two physi- cians in each ward and a panel of pharmacists. In a worst case scenario 20–30% of all errors identified at each stage were assessed as potential adverse drug events as shown in Table 4. The most common ATC groups involved in potential adverse drug events were drugs for: infectious diseases (J), heart and circulation (C), and central nervous system (N). Apart from the transcription stage, errors were most frequently seen in regular medications.
Our findings of 43% errors in the medication process indicate a need for improvement in more stages of the medication process. None of the errors identified affected the patients’ health but one-fifth was assessed as being potentially serious or fatal in a worst case scenario. The high percentage of iden- tified errors must be viewed in the light of the detailed and systematic examination of errors and types of error at each stage of the medication process.
Ordering and transcription stage
Previous studies have suggested a need for a unified medi- cation system to eliminate errors at the ordering and tran- scription stage [19–21]. This medication chart, paper or electronic, should clearly state the components needed to fulfil requirements for unambiguous prescription—especially drug form and route, as these were the most frequent types of error in our study. The high frequency of discrepancies in drug form between medical records and medication charts were caused by nurses’ interpretation of drug prescriptions, and lack of drug formulation in the medical record. Often, these interpretations are correct and improve the quality of the drug prescription, but these actions are beyond nurse authority and could, ultimately, result in fatal consequences for the patients.
Table 2 Frequency of medication errors in all stages of the medication process
ntotal = the total opportunities of errors in each stage (prescriptions and doses) and n = the total amount of detected errors in each stage of the medication process. 1Dispensing stage observational study. 2Dispensing stage—unannounced control visit. 3Discharge summaries.
Ordering (ntotal = 433), n (%)
Transcription (ntotal = 558), n (%)
Disp. obs.1 (ntotal = 419), n (%)
Disp. contr. (ntotal = 119),
Administration (ntotal = 412), n (%)
Discharge sum.3 (ntotal = 526), n (%)
Medical 85 (41) 122 (51) 7 (3) – 103 (52) 188 (64) Surgical 82 (37) 188 (59) 10 (4) 5 (10) 63 (29) 213 (91) Overall 167 (39) 310 (56) 17 (4) 5 (4) 166 (41) 401 (76)
Errors in the medication process
Table 4 Potential clinical consequences of errors detected in the ordering, transcribing, dispensing, and administration stages
Only worst case scenarios are shown in each stage of the medication process. 1One missing assessment. 2Dispensing stage observational study. 3Dispensing stage—unannounced control visit.
Ordering1 n = 167 n (%)
Transcription1 n = 310 n (%)
Disp. obs.1,2 n = 17 n (%)
Disp. contr.3 n = 5 n (%)
Administration1 n = 166 n (%)
Fatal 3 (2) 6 (2) 0 0 2 (1) Serious 30 (18) 65 (21) 4 (25) 1 (20) 33 (20) Significant 67 (40) 127 (41) 9 (56) 2 (40) 53 (32) Non-significant 66 (40) 111 (36) 3 (19) 2 (40) 77 (46)
Table 3 Detected error types in the medication process
*N = total number of prescriptions or doses affected with an error. n = the total sum of error detected in the identified prescriptions/doses with errors (N). For example, 167 (329) means that a total of 329 errors were detected in 167 prescriptions. Unfilled cells indicate error types which were not included in the actual stage of the medication process. 1Transcription stage. 2Dispensing stage observational study. 3Dispensing stage—unannounced control visit. 4Administration stage. 5Discharge summaries. 6Wrong administration techniques. 7Lack of identity control.
Ordering *N = 167 (n = 329)
Transcription1 *N = 310 (n = 370)
Disp. obs2 *N = 17 (n = 17)
Disp. cont3 *N = 5 (n = 5)
Administration4 *N = 166 (n = 190)
Dis. sum.5 *N = 401 (n = 447)
n/N n/N n/N n/N n/N n/N ……………………………………………………………………………………………………………………………………………………………………………………………….
Drug name 4/167 8/310 3/401 Drug formulation 125/167 90/310 0 0 64/401 Omission of route 114/167 21/310 2/401 Omission of dose 30/167 7/17 5/5 1/166 Dosing regime 18/167 34/310 27/401 Treatment time 33/167 Date 1/167 Signature 4/167 Discrepancy in dose 50/310 42/401 Unordered drug 123/310 5/17 0 0 27/401 Omission of drug 44/310 0 0 282/401 Unordered dose 0 0 1/166 Wrong dose 5/17 0 Wrong adm.6 technique 8/166 Wrong route 0 Lack of id.7 control 150/166 Wrong time 18/166 Wrong delivery 12/166
M. Lisby et al.
Dispensing and administration stage
Compared with previous studies of medication errors identified through observational studies, error rates in our study varied considerably even when adjusted to Barker’s original criteria . Possible explanations of the low frequency of dispensing errors identified in our study, could be differences in study population or observed drug forms as well as differences in the drug distribution system. Furthermore, it can be assumed that the inclusion of specialized pharmaceutical variables such as excess of intravenous drug duration, incorrect dilution of drug, wrong storage of drug, and use of expired drug have contributed to a higher error rate in previous studies [11,17]. Our aim was to explore whether dispensing and administra- tion of medication were concordant with prescribed medication in the medication charts in respect of drug name, dose, drug form, time, drug route, administration technique, and giving drugs to the right patient, including identity control before administering medication. Pharmaceutical variables such as wrong storage and use of expired drugs were not included as these were controlled by pharmacists twice a week, in each ward. Compared with similar studies observing other than unit dose systems, error rates from <1 and up to 25% have been found depending on the amount and definitions of included variables [8,11,22–25]. In our study the most com- mon error types at the dispensing stage were unordered or omitted doses while lack of identity control and wrong time was most common when administering medication. These findings are generally in line with previous studies, indicating that unordered dose and wrong time were among the most frequent error types [8,22–25]. Lack of identity control is not an ASHP criterion and has to our knowledge only been inves- tigated in a few previous studies emphasizing the importance of proper identification of patients before administering med- ication [26,27]. Unlike the study in the UK, we did not observe any comparison between the patients’ identity number at the wristband and the identity number written at the medication chart, not even when patients were unconscious or demented . Further we found that 150 of 412 (36%) doses were administered without any previous verbal verification of the patients’ identity. In worst case scenario approximately 20% of these doses were assessed as a potential adverse event. Controlling the patient’s identity before administered medica- tion was not a standard routine in the included wards, and furthermore, the recommendations from the Danish National Board of Health are ambiguous. There is thus a real risk gap in the medication process that needs to be bridged by improved procedures or new technologies such as bar code medication administration although this could introduce new paths for medication errors and adverse drug events .
Discharge summaries had the highest percentage of errors constituting almost half of all errors detected in the present study. Previous studies investigating medication in discharge summaries and in medical records have shown discrepancy in 16–36% of prescribed drugs [12–14]. Still, more than twice as
many errors were identified in our study, presumably as a consequence of the stringent and detailed criteria of the present design. Whether these criteria were too idealistic, in comparison with clinical practice could be discussed. Yet, the systematic information collected in the present study points out weaknesses in existing practice. For example, more than two-thirds of the identified errors were caused by lack of transcribing eligible prescriptions into discharge summa- ries, due to lack of discontinuing expired drug prescriptions. These findings, among others, stress a need for general and unambiguous guidelines for drug prescriptions in discharge summaries.
This study has some limitations. The validity of the modified observation length was checked by an unannounced control visit in each ward, indicating no influential consequences. The study was conducted in one medical and one surgical ward in a single university hospital and therefore results are not to be extrapolated to other hospitals or cultures. However, many of the results were in accordance with former studies thus indi- cating a general line with our findings.
Errors were observed in almost every second handling of medication in the present study. Quality improvements are required in most of the stages of the medication process. Several of the identified errors and error types, at least in theory, could be avoided by automated solutions like compu- terized order entry, electronic discharge summaries, and bar code medication administration [19,20,29,30]. Based upon the low frequency of errors detected at the dispensing stage in combination with the overall uncertainty of the effect of a unit dose system, it remains uncertain whether unit dose systems would be an effective solution in the dispensing stage [11,22,23,25]. Other interventions must also be considered, i.e. evidence-based clinical guidelines for safe medication practice, unified medication charts, unambiguous recommen- dations for controlling patient identity as well as unam- biguous recommendations for drug prescriptions in discharge summaries.
1. Thomas EJ, Studdert DM, Burstin HR. Incidence and types of adverse events and negligent care in Utah and Colorado. Med Care 2000; 38: 261–271.
2. Brennan TA, Leape LL, Laird N, Herbert L, Localio AR, Lawthers AG. Incidence of adverse events and negligence in hospitalized patients. Results of the Harvard Medical Practice Study I. N Engl J Med 1991; 324: 370–375.
3. Bates DW, Cullen DJ, Laird N et al. Incidence of adverse drug events and potential adverse drug events. Implications for pre- vention. ADE Prevention Study Group. J Am Med Assoc 1995; 274: 29–34.
Errors in the medication process
4. Leape LL, Brennan TA, Laird N, Lawthers AG, Localio AR, Barnes BA. The nature of adverse events in hospitalized patients. Results of the Harvard Medical Practice Study II. N Engl J Med 1991; 324: 377–384.
5. Wilson RM, Runciman WB, Gibberd RW, Harrison BT, Newby L, Hamilton JD. The Quality in Australian Health Care Study. Med J Aust 1995; 163: 458–471.
6. Bates DW, Boyle DL, Vander Vliet MB, Schneider J, Leape L. Relationship between medication errors and adverse drug events. J Gen Intern Med 1995; 10: 199–205.
7. Kaushal R, Bates DW, Landrigan C, McKenna KJ, Clapp MD, Federico F et al. Medication errors and adverse drug events in paediatric inpatients. J Am Med Assoc 2001; 285: 2114–2120.
8. Tissot E, Cornette C, Demoly P, Jaquet M, Barale F, Capalleier G. Medication errors at the administration stage in an intensive care unit. Intensive Care Med 1999; 25: 353–359.
9. Flynn EA, Barker KN, Pepper GA, Bates DW, Mikeal RL. Comparison of methods for detecting medication errors in 36 hospitals and skilled-nursing facilities. Am J Health Syst Pharm 2002; 59: 436–446.
10. Gandhi TK, Seger DL, Bates DW. Identifying drug safety issues: from research to practice. Int J Qual Health Care 2000; 12: 69–76.
11. Allan EL, Barker KN. Fundamentals of medication error research. Am J Hosp Pharm 1990; 47: 555–571.
12. Morrill GB, Barreuther C. Screening discharge prescriptions. Am J Hosp Pharm 1988; 45: 1904–1905.
13. Wilkin TJ, Christopher LJ, Crooks J. Transcription errors in the drug information supplied from hospital to general practitioner. Health Bull (Edinb) 1978; 36: 13–16.
14. Wilson S, Warwick R, Chapman M, Miller R. General practi- tioner—hospital communications: A review of discharge sum- maries. J Qual Clin Pract 2001; 21: 104–108.
15. Folli HL, Poole RL, Benitz WE, Russo JC. Medication error prevention by clinical pharmacists in two children’s hospitals. Pediatrics 1987; 79: 718–722.
16. Dean BS, Barber ND, Schachter M. What is a prescribing error? Qual Health Care 2000; 9: 232–237.
17. ASHP. ASHP Standard definition of a medication error. Am J Hosp Pharm 1982; 39: 321.
18. Barker KN. Data collection techniques: observation. Am J Hosp Pharm 1980; 37: 1235–1243.
19. Bates DW, Leape LL, Cullen DJ et al. Effect of computerized physician order entry and a team intervention on prevention of serious medication errors. J Am Med Assoc 1998; 280: 1311–1316.
20. Bates DW, Teich JM, Lee J et al. The impact of computerized physician order entry on medication error prevention. J Am Med Inform Assoc 1999; 6: 313–321.
21. Vere DW. Errors of complex prescribing. Lancet 1965; 13: 370–373.
22. Dean BS, Allan EL, Barber ND, Barker KN. Comparison of medication errors in an American and a British hospital. Am J Health Syst Pharm 1995; 52: 2543–2549.
23. Taxis K, Dean BS, Barber ND. Hospital drug distribution system in the UK and Germany—a study of medication errors. Pharm World Sci 1999; 21: 25–31.
24. Barker KN, McConnell WE. The problems of medication. Am J Hosp Pharm 1962; 19: 361–369.
25. Borel JM, Rascati KL. Effect of an automated nursing unit- based drug-dispensing device on medication errors. Am J Health Syst Pharm 1995; 52: 1875–1879.
26. Byrne AK. Errors in giving medication. Am J Nurs 1953; 53: 829–831.
27. Wirtz V, Taxis K, Barber ND. An observational study of intra- venous medication errors in the United Kingdom and in Germany. Pharm World Sci 2003; 25: 104–111.
28. Patterson ES, Cook RI, Render M. Improving patient safety by identifying side effects from introducing bar coding in medication administration. J Am Med Inform Assoc 2002; 9: 540–553.
29. Thornton PD, Simon S, Mathew TH. Towards safer drug pre- scribing, dispensing and administration in hospitals. J Qual Clin Pract 1999; 19: 41–45.
30. Bates DW. Using information technology to reduce rates of medication errors in hospitals. Br Med J 2000; 320: 788–791.
Accepted for publication 27 September 2004
M. Lisby et al.
Appendix 1 Definition of potential clinical consequences
Category Definition Definition of keywords ………………………………………………………………………………………………………………………………………………………………………………………………..
Potentially fatal Medication errors judged to imply a potential clinical risk for causing the death of the patient
Fatal refers to medication errors that could lead to the death of the patient
Potentially serious Medication errors judged to imply a potential clinical risk of injuring the patient
Injury includes medication errors that would require active treatment to restore the health of the patient. A potentially serious error would lead to either permanent or temporary disability
Potentially significant Medication errors judged to imply a potential clinical risk of being inconvenient for the patient— without causing any harm or injury
Inconvenient refers to unpleasant consequences of wrong dose/drug or omission of dose/drug that could lead to pain, dizziness. It also refers to any monitoring of the patient such as extra blood test, measurement of blood pressure
Potentially non-significant Medication errors judged to be without any potential clinical risk for the patient
Without clinical risk refers to medication errors that did not lead to any injury or inconvenience for the patient