Errors in the administration of intravenous medication in Brazilian hospitals

Aim. To verify the frequency of errors in the preparation and administration of intravenous medication in three Brazilian hospitals in the State of Bahia. Background. The administration of intravenous medications constitutes a central activity in Brazilian nursing. Errors in performing this activity may result in irreparable damage to patients and may compromise the quality of care. Design. Cross‐sectional study, conducted in three hospitals in the State of Bahia, Brazil. Methods. Direct observation of the nursing staff (nurse technicians, auxiliary nurses and nurse attendants), preparing and administering intravenous medication. Results. When preparing medication, wrong patient error did not occur in any of the three hospitals, whereas omission dose was the most frequent error in all study sites. When administering medication, the most frequent errors in the three hospitals were wrong dose and omission dose. Conclusions. The rates of error found are considered low compared with similar studies. The most frequent types of errors were wrong dose and omission dose. The hospitals studied showed different results with the smallest rates of errors occurring in hospital 1 that presented the best working conditions. Relevance to clinical practice. Studies such as this one have the potential to improve the quality of care.     

Prevention of potential errors in resuscitation medications orders by means of a computerised physician order entry in paediatric critical care

INTRODUCTION: Computerised physician order entry with clinical decision support system (CPOE+CDSS) is an important tool in attempting to reduce medication errors. The objective of this study was to evaluate the impact of a CPOE+CDSS on (1) the frequency of errors in ordering resuscitation (CPR) medications and (2) the time for printing out the order form, in a paediatric critical care department (PCCD). METHODS: Setting: An 18-bed PCCD in a tertiary-care children's hospital. Design: Prospective cohort study. Measures: Compilation and comparison of number of errors and time to fill in forms before and after implementation of CPOE+CDSS. Time to fill in conventional, simulated and CPOE forms was measured and compared. RESULTS: There were three reported incidents of errors among 13,124 CPR medications orders during the year preceding implementation of CPOE+CDSS. These represent errors that escaped the triple check by three independent staff members. There were no errors after CPOE+CDSS was implemented (100% error reduction for 46,970 orders). Time to completion of drug forms dropped from 14 min 42 s to 2 min 14s (p < 0.001). CONCLUSIONS: CPOE+CDSS completely eliminated errors in filling in the forms and significantly reduced time to completing the form.     

Response to Medication Dosing Alerts for Pediatric Inpatients Using a Computerized Provider Order Entry System

Objective

Medication dosing errors are of particular concern in hospitalized children. Avoidance of such errors is essential to quality improvement and patient safety. Computerized provider order entry (CPOE) systems with clinical decision support (CDS) have the potential to reduce medication errors. The objective of this study was to evaluate provider response to the dosing alerts in a CPOE system with CDS for pediatric inpatients and to identify differences in provider response based on clinician specialty.

Patients and methods

We conducted a retrospective analysis of all medication dosing alerts over a 1-year period (January 1 through December 31, 2008) for all pediatric inpatients at Hospital for Special Surgery. Alerts were analyzed with respect to medication dosing, prescriber, and action taken by the prescriber after the alert was triggered (i.e., accepted suggested change, ignored recommendation/overrode, or cancelled the order).

Results

During the study period, 18,163 medication orders were placed and 1,024 dosing alerts were fired. Overdosing of medications accounted for 91% of the alerts and underdosing 9%. The pediatric-trained providers ignored more alerts and cancelled fewer orders than the non-pediatric-trained providers (p<0.001). Both groups changed the order similarly based on CDS recommendations.

Conclusions

Differences in response to CDS were found between pediatric-trained and non-pediatric-trained providers caring for pediatric patients; however, both groups changed orders based on CDS similarly. CPOE with built-in CDS may be of particular value when providers with different specialties and types of training are caring for pediatric patients.     

Diffusion of pharmacist interventions within the framework of clinical pharmacy activity in the clinical ward

BACKGROUND: Computerised physician order entry (CPOE) and the integration of a pharmacist in clinical wards have been shown to prevent medication errors. OBJECTIVES: The objectives were to describe interventions performed by a clinical pharmacist integrated into clinical wards with CPOE, to assess the acceptance of interventions by prescribers, and to describe factors associated with acceptance. METHODS: A 3-week prospective study was conducted in five wards of a 2000-bed French teaching hospital. RESULTS: During pharmacist review of medication orders and participation on physician rounds, six resident pharmacists provided interventions either conveyed orally to prescribers, using the computer system, or combining both methods. There were 221 pharmacist interventions concerning drug-drug interactions (27%), drug monitoring (17%) and computer-related problems (16%). Pharmacist interventions consisted of change of drug choice or dose adjustment (49%), drug monitoring (17%) and administration modality optimisation (14%). Interventions were provided solely via computer systems in 67% of cases. The rate of intervention acceptance was 47.1%. In multivariate analysis, acceptance was significantly associated with oral transmission (odds ratio [OR] = 6.46; 95% confidence interval [95% CI] [1.65-25.24]; p < 0.01), change of drug choice or dose adjustment recommendations (OR = 3.81; 95% CI [1.63-8.86]; p < 0.01) and administration modality optimisation recommendations (OR = 9.51; 95% CI [3.02-29.93]; p < 0.01). CONCLUSION: Communication method and nature of recommendation are factors associated with pharmacist intervention acceptance. CPOE is necessary to develop clinical pharmacy practice. However, only the integration of the pharmacist on the ward can guarantee a high level of acceptance of pharmacist interventions by prescribers.     

Quantification and classification of errors associated with hand-repackaging of medications in long-term care facilities in Germany

Objective: The aim of this study was to quantify and classify errors associated with the repackaging of residents' medications in long-term care facilities in Germany.Methods: This was a prospective 8-week study conducted in 3 long-term care facilities. Pill organizers, each of which contained all repackaged solid oral dosage forms of long-term medications for a particular resident for an entire day, were inspected and checked against residents' medication sheets by the investigator-pharmacist. On agreement between the pharmacist and the registered nurse responsible for residents' medications, all errors were rectified before medications were administered. The primary study measure was the overall rate of incorrectly repackaged medications relative to all repackaged medications. Secondary measures were the proportion of all pill organizers with medication errors and the proportion of residents who would have been affected by these errors. Errors were categorized by type as follows: wrong time of administration, wrong dose, wrong medication, omission of a medication, extra dose, incorrect halving of tablets, and damaged medication.Results: One hundred ninety-six residents were included in the study, representing 8798 daily pill organizers and 48,512 inspected medications. Residents received a mean of 5.4 solid oral dosage forms of long-term medications per day. Six hundred forty-five errors were detected, for an error rate of 1.3%; the errors involved 7.3% of daily pill organizers and 53.0% of residents. The largest proportion of errors involved incorrect halving of tablets (49.1%), followed by omission of a medication (22.0%), extra dose (9.8%), wrong time of administration (8.4%), damaged medication (6.4%), wrong dose (4.2%), and wrong medication (0.2%). These results may underestimate true rates of repackaging errors across long-term care facilities in Germany, as the conditions in the 3 facilities in this study were near-optimal in terms of the environment, process, and quality of repackaging.Conclusions: Among 48,512 medications inspected over 8 weeks in 3 German long-term care facilities, the rate of repackaging errors was 1.3%, involving 7.3% of daily pill organizers and the medications of 53.00% of residents. The largest proportion of errors involved incorrect halving of tablets.     

The Effect of Computerized Provider Order Entry on Medical Students’ Ability to Write Orders

Background: Computerized provider order entry (CPOE) is being implemented at increasing numbers of U.S. hospitals, yet the effects of CPOE on medical student education are largely unstudied. Purpose: The objective is to investigate the effects of CPOE on medical students’ ability to write orders for patients. Methods: One hundred forty-three medical students who began their Basic Medicine clerkship between March 2003 and April 2004 were asked to write mock admission orders for a patient with pneumonia after the 1st month of their clerkship. Students had spent the month at 1 of 3 hospitals: 1 using CPOE, 1 paper orders, and 1 that began using CPOE midway through this study. Admission orders were scored for the presence of specific orders and features. Results: One hundred twenty students attempted to write admission orders. Students who trained at hospitals using CPOE and those who trained at hospitals using paper orders included expected basic, lifesaving, and higher level orders at similar rates. No significant differences in order clarity or inclusion of unnecessary orders were found for the 2 groups. No significant differences were found when controlling for school year and 4 modifiable rotation features. Conclusions: When admission order completeness and quality for medical students who trained at hospitals using CPOE were compared to those who trained using handwritten orders, no important differences were found.     

Prevalence of errors in a pediatric hospital medication system: implications for error proofing

The purpose of this study was to determine the prevalence of errors in the medication system of a pediatric teaching hospital. Error was defined broadly to capture all deviations in the process from medication order through administration of the dose. The long-term goal was to provide direction to efforts to error-proof the system. The sample was 3,312 medication orders written during 669 patient-days for which a total of 11,978 doses were passed. Errors were categorized as intercepted errors (intercepted through the normal processes of the medication system) or administration errors (errors that involve the patient with or without adverse sequelae). Errors were also categorized as errors in primary activities (e.g., prescribing or preparing the medication for administration) or supporting activity (e.g., transferring the order to another record). A total of 784 errors were identified; 98% were intercepted and 2% were administration errors. More errors (71%) occurred in supporting activities than in primary activities. Medication systems are complex processes. Errors are imbedded in the medication system and are typically intercepted before patients are involved. Intercepting errors involves additional work that adds to an already cumbersome process. Error proofing will be different for errors in primary activities and for supporting activities.     

Autologous donation error rates in Canada

BACKGROUND: Although certain transfusion risks are eliminated by the use of autologous blood, clerical errors may still occur. In addition, because of differences in donor selection criteria and donor-patient expectations, the consequences of certain errors may be different in autologous and allogeneic donations. STUDY DESIGN AND METHODS: In January 1996, autologous donation error rates in Canada from 1989 to November 1995 were estimated by 1) a detailed questionnaire sent to hospitals supplied by the Canadian Red Cross, Blood Services, Transfusion Center of Quebec at Montreal autologous donation program (n = 31), 2) a review of that institution's quality assurance non- compliance reports, and 3) a detailed questionnaire sent to other Canadian Red Cross centers with autologous donation programs (n = 16) and hospital-based autologous programs in Canada (n = 3). The total number of autologous donations collected was determined from Canadian Red Cross annual reports and information supplied by hospital-based programs. RESULTS: There were 113 errors reported for 16,873 units collected by the Montreal center (1/149 units) based on collection center and hospital data. The most frequent errors were the late receipt of units for surgery (25% of errors) or the receipt of units in the wrong hospital (23%). Other Canadian programs reported 166 errors for approximately 53,500 units collected (1/322 units). However, this figure was based mainly on collection center, and not hospital, data. The most frequent errors were in labeling (48%) and component preparation (25%). One unit of autologous fresh-frozen plasma was transfused to the wrong recipient. Errors were more frequent if components were produced, if units were drawn in hospitals for interhospital transfer, or it units were shipped between Red Cross centers. CONCLUSION: Errors are not infrequent in autologous donation programs. Autologous transfusion should not be considered as being without risk.