The impact of an electronic prescribing and administration system on the safety and quality of medication administration

Objective To assess the effect of an electronic prescribing and administration system on the safety and quality of medication administration in a UK hospital. Setting Surgical ward in a teaching hospital. Method Data were collected before and after introducing a closed‐loop system comprising electronic prescribing, automated dispensing, barcode patient identification and electronic medication administration records (ServeRx, MDG Medical). We observed medication administration during drug rounds and assessed medication administration error (MAE) rates for ward‐stock and non‐ward‐stock drugs, accuracy of medication administration documentation, timeliness of administration, administration of medication from unlocked areas and supervision of patients taking oral medication by nursing staff. Key findings Pre‐ and post‐intervention MAE rates were 6.4 and 2.3% respectively for ward‐stock drugs (95% confidence interval for the difference (CI) ?5.8 to ?2.4%), and 14.6 and 13.7% for non‐ward‐stock drugs (CI ?6.5 to 4.7%). Excluding omissions due to unavailability, pre‐ and post‐intervention MAE rates were 6.2 and 2.2% respectively for ward‐stock drugs (CI ?5.7 to ?2.3%), and 9.2 and 3.5% for non‐ward‐stock drugs (CI ?9.3 to ?2.1%). Pre‐intervention, 2086 doses (96.3%) were documented correctly and 1557 (95.9%) post‐intervention (CI ?1.6 to 0.8%). There were five clinically significant documentation discrepancies pre‐intervention (0.2%), and 33 (2.0%) afterwards (CI 1.1 to 2.5%). Timeliness of administration improved post‐intervention (P < 0.001; Chi‐square test), as did administration of medication from unlocked areas (CI 4.7 to 7.3%) and supervision of patients taking oral medication (CI 17 to 23%). Conclusion Reductions in MAEs, excluding omissions due to unavailability, occurred for both ward‐stock and non‐ward‐stock drugs. The system also improved timeliness and security of drug administration. However, there was an increase in potentially significant documentation discrepancies.     

A multicentre study of dispensing errors in British hospitals

Nineteen hospital pharmacies in Britain participated in a study of dispensing errors. All dispensing errors (errors detected outside the department) and a sample of dispensing incidents (errors detected as part of the checking process) were recorded using a standard data collection form over a period of six months. One hundred and seventy-eight errors and 1,500 incidents were analysed. The overall error rate was 18.1/100,000 dispensed items (range 2.9–49.1/100,000). The mean error rate was 11.5/100,000 items in hospitals where all dispensing is checked and 35/100,000 items in hospitals where only items dispensed by non-pharmacists are checked. There were also differences in type between errors and incidents ?31.7 per cent of incidents involved incorrect directions on the label compared with 17.4 per cent of errors; 34.2 per cent of incidents involved the wrong drug or drug strength being dispensed compared with 45.5 per cent of errors. One hundred and forty-five different drugs were involved in the 178 errors, with three drugs being involved in six errors each (flucloxacillin, glyceryl trinitrate and frusemide). The results of this study can contribute to the preparation of risk management strategies for dispensing.     

Incorrect drug selection at the point of dispensing: a study of potential predisposing factors

Objective To determine potential predisposing factors to medication errors involving confusion between drug names, strengths and dosage forms. Methods The study analysed medication errors reported over the period January 2005 to December 2008 from the two main dispensaries of a 1200‐bed NHS Foundation Hospital Trust in London. Dispensing incidents considered for analysis included all incidents involving drug name, strength and dosage label and content errors. Statistical analyses were performed using Statistica. Dispensing frequencies of the prescribed and wrongly dispensed drugs were compared by means of Wilcoxon signed‐rank test, and the extent of correlation between dispensing frequency and error frequency was assessed using Spearman's rank correlation coefficient. Key findings The Trust recorded a total of 911 dispensing errors between 2005 and 2008. The most significant category, which accounted for 211 (23.2%) of the reported errors, involved errors in drug selection. Drug‐selection errors were not random events because the plot of error frequency against the average yearly dispensing frequency for the 1000 most issued drugs showed little evidence of association (r = 0.19, P(α) = 0.03). There was, however, an increased likelihood of drug‐selection errors occurring when the prescribed drug was dispensed with relatively low frequency and shows a significant orthographic similarity to another drug which has a higher dispensing frequency. Conclusion The majority of drug‐selection errors would seem to be caused by insufficient attention paid to the specified drug strength. Dispensing frequency is an important factor influencing the likelihood of a drug‐selection errors occurring, but it is also shown here that a large proportion of the drug‐selection errors involved specifications exhibiting high orthographic similarity.     

CYP2D6‐Inhibiting Drugs and the Increased Risk of Fall‐Related Injuries Due to Newly Initiated Opioid Treatment – A Swedish,Register‐Based Case‐Crossover Study

It has been shown that newly initiated opioid therapy increases the risk of fall‐related injuries. Yet, it remains to be determined whether drug–drug interactions can affect this negative effect, for instance with drugs inhibiting cytochrome P4502D6 (CYP2D6) that metabolizes codeine and also has a partial effect on tramadol and oxycodone. Our aim was to investigate how CYP2D6‐inhibiting drugs contribute to explaining the risk of fall‐related injuries for newly initiated opioid treatments with codeine, tramadol or oxycodone. Data from a Swedish national case‐cross over study were revisited. This study identified a total of 167,257 fall‐related injuries leading to hospitalization that occurred between 1 May 2006 and 31 December 2009 and linked information about dispensed drugs to them. Use of newly dispensed opioids in the 28 days before fall‐related injury with and without CYP2D6‐inhibiting drugs was compared with an earlier control period. For codeine, there was a two‐times increased risk with concomitant CYP2D6‐inhibiting drug use (OR, 1.76; 95% CI 1.40–2.20) and a three‐times risk increase without (OR, 3.17; 95% CI 2.88–3.50). For tramadol, the risks were doubled when CYP2D6‐inhibiting drugs were used (OR, 2.19; 95% CI 1.84–2.60) and tripled without their use (OR, 3.04; 95% CI 2.82–3.27). The risks were about the same for oxycodone, morphine, fentanyl and buprenorphine irrespective of CYP2D6‐inhibiting drug use. In newly initiated opioid therapies, drug–drug interactions from concomitant use of CYP2D6‐inhibiting drugs are associated with a lower risk of fall‐related injury for codeine and tramadol that undergo metabolism via CYP2D6, but not for other opioids.     

The Frequency and Potential Causes of Dispensing Errors in a Hospital Pharmacy

Objectives: To determine the frequency and types of dispensing errors identified both at the final check stage and outside of a UK hospital pharmacy, to explore the reasons why they occurred, and to make recommendations for their prevention.Method: A definition of a dispensing error and a classification system were developed. To study the frequency and types of errors, pharmacy staff recorded details of all errors identified at the final check stage during a two-week period; all errors identified outside of the department and reported during a one-year period were also recorded. During a separate six-week period, pharmacy staff making dispensing errors identified at the final check stage were interviewed to explore the causes; the findings were analysed using a model of human error.Main outcome measures: Percentage of dispensed items for which one or more dispensing errors were identified at the final check stage; percentage for which an error was reported outside of the pharmacy department; the active failures, error producing conditions and latent conditions that result in dispensing errors occurring.Results: One or more dispensing errors were identified at the final check stage in 2.1% of 4849 dispensed items, and outside of the pharmacy department in 0.02% of 194,584 items. The majority of those identified at the final check stage involved slips in picking products, or mistakes in making assumptions about the products concerned. Factors contributing to the errors included labelling and storage of containers in the dispensary, interruptions and distractions, a culture where errors are seen as being inevitable, and reliance on others to identify and rectify errors.Conclusion: Dispensing errors occur in about 2% of all dispensed items. About 1 in 100 of these is missed by the final check. The impact on dispensing errors of developments such as automated dispensing systems should be evaluated.     

The Effect of Systematic Medication Review in Elderly Patients Admitted to an Acute Ward of Internal Medicine

Abstract: Elderly patients are vulnerable to medication errors and adverse drug events due to increased morbidity, polypharmacy and inappropriate interactions. The objective of this study was to investigate whether systematic medication review and counselling performed by a clinical pharmacist and clinical pharmacologist would reduce length of in‐hospital stay in elderly patients admitted to an acute ward of internal medicine. A randomized, controlled study of 100 patients aged 70 years or older was conducted in an acute ward of internal medicine in Denmark. Intervention arm: a clinical pharmacist conducted systematic medication reviews after an experienced medical physician had prescribed the patients’ medication. Information was collected from medical charts, interview with the patients and database registrations of drug purchase. Subsequently, medication histories were conferred with a clinical pharmacologist and advisory notes recommending medication changes were completed. Physicians were not obliged to comply with the recommendations. Control arm: medication was reviewed by usual routine in the ward. Primary end‐point was length of in‐hospital stay. In addition, readmissions, mortality, contact to primary healthcare and quality of life were measured at 3‐month follow‐up. In the intervention arm, the mean length of in‐hospital stay was 239.9 hr (95% CI: 190.2–289.6) and in the control arm: 238.6 hr (95% CI: 137.6–339.6), which was neither a statistical significant nor a clinically relevant difference. Moreover, no differences were observed for any of the secondary end‐points. Systematic medication review and medication counselling did not show any effect on in‐hospital length of stay in elderly patients when admitted to an acute ward of internal medicine.     

Exposure to phototoxic NSAIDs and quinolones is associated with an increased risk of melanoma

Purpose

Ultraviolet radiation exposure is the most important exogenous risk factor for cutaneous malignancies. It is possible that phototoxic drugs promote the development of cutaneous melanoma (CM) by intensifying the effect of ultraviolet light on the skin. We investigated the association between the use of common systemic phototoxic drugs and development of CM.

Methods

This study was a case–control study in a Dutch population-based cohort. The drug dispensing data was obtained from PHARMO, a Dutch drug dispensing and hospital admissions registry, and linked to PALGA, the nationwide pathology network of the Netherlands. The cases were patients diagnosed with pathologically confirmed primary CM between 1991 and 2004. Controls were sampled from the PHARMO population. Exposure to systemic phototoxic drugs was measured and included antimicrobial agents, diuretics, antipsychotic drugs, antidiabetic drugs, cardiac drugs, antimalarials and nonsteroidal anti-inflammatory drugs (NSAIDs). A multivariate conditional logistic regression analysis was performed to study the association between exposure to phototoxic drugs and CM.

Results

The study population included 1,318 cases and 6,786 controls. Any phototoxic drug during the study period was dispensed for 46 % of the cases and 43 % of the controls (p?=?0.012). The use of quinolones [odds ratio (OR) 1.33, 95 % confidence interval (CI) 1.01–1.76] and propionic acid derivative NSAIDs (OR 1.33, 95 % CI 1.14–1.54) had a positive association with CM.

Conclusions

Our study shows that the use of phototoxic drugs is associated with an increased risk of developing CM. Even a short-term use of phototoxic quinolones and propionic acid derivative NSAIDs may increase the risk for CM. Patient education to promote sun-protective behaviour is essential to avoid immediate adverse effects and possible long-term effects of phototoxic drugs.     

Parenteral drug administration errors by nursing staff on an acute medical admissions ward during day duty.

BACKGROUND: Parenteral therapy is a route of administration for drugs which are poorly absorbed via the oral route and it can provide a rapid response during an emergency. However, poorly prepared and/or administered parenteral therapy can cause potential harm to patients such as thrombus formation, severe hypersensitivity reactions and infection. Very few studies have investigated the incidence of medication errors associated with parenteral drug administration. OBJECTIVES: To determine the error rate during preparation and administration of parenteral medications by nursing staff and to propose strategies to reduce the error rate during parenteral administration. METHODS: A direct, disguised observation technique was used. The first author (JB) observed and recorded errors that occurred during the preparation and administration of parenteral medications on an admissions ward between 8.00 am and 4.30 pm from Monday to Friday for a 4-week period during December 1998. The staff were told that the observer was timing the administration; therefore they were not aware of the true nature of the study. This study was approved by the hospital audit committee. RESULTS: Drug administration was witnessed for a 4-week period providing 107 opportunities for error. 27 errors were observed which equated to an error rate of 25.2% [95% confidence interval (CI) 17.0 to 33.5%] including wrong time errors. Excluding wrong time errors, the most frequently occurring type of error, reduced the error rate to 10.3% (95% CI 3.8 to 14.9%). DISCUSSION: The error rate was lower than reported in the literature, this may be due to different methodologies, small sample size or effective nursing training and operating procedures. In the observed hospital, only nursing staff who have completed a training package are allowed to administer parenteral medications. CONCLUSION: Based on our small study, and 2 previous small studies, we can conclude that parenteral medication administration errors are common in the UK; however, these studies are too small-scale to detect rare and serious errors.     

Incidence,type and causes of dispensing errors: A review of the literature

Objectives To identify, review and evaluate the published literature on the incidence, type and causes of dispensing errors in community and hospital pharmacy. Method Electronic databases were searched from 1966 to February 2008. This was supplemented by hand‐searching the bibliographies of retrieved articles. Analysis of the findings explored the research methods, operational definitions, incidence, type and causes of dispensing errors. Key findings Sixty papers were identified investigating dispensing errors in the UK, US, Australia, Spain and Brazil. In general, the incidence of dispensing errors varied depending on the study setting, dispensing system, research method and operational definitions. The most common dispensing errors identified by community and hospital pharmacies were dispensing the wrong drug, strength, form or quantity, or labelling medication with the incorrect directions. 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. Conclusions Comparison of the reviewed studies was confounded by differences in study setting, research method and operational definitions for dispensing errors, error rate and classification of error types. The World Health Organization is currently developing global patient safety taxonomy. Such a standardized taxonomy for dispensing errors would facilitate consistent data collection and assist the development of error‐reduction strategies.     

A study of unprevented dispensing incidents in Welsh NHS hospitals

Objective To monitor unprevented dispensing incidents in NHS hospitals by identifying incident types, drugs involved and factors that may have contributed to the occurrence of incidents. Setting All 20 Welsh NHS hospitals (15 district general; 2 teaching; 2 psychiatric and 1 other specialist hospital). Method Unprevented dispensing incidents that occurred between January 2003 and December 2004 were reported and analysed using a validated method. Incident rate was calculated for those hospitals that provided both incident and issue data. Incident rate was compared with previous research using a two‐sample t‐test. Reported incident types and contributory factors were compared with previous research using chi‐square analysis. Key findings A total of 1005 unprevented dispensing incidents were reported by 20 hospitals. The overall incident rate, based on data from 17 hospitals, was 16 incidents per 100000 items dispensed (range 0.2–46 incidents per 100000 items dispensed). The overall dispensing incident rate was less than previously reported (18 incidents per 100000 items dispensed).This finding was not statistically significant (t = 0.421, P = 0.676). The most common incidents reported were dispensing the wrong strength of drug (n = 241, 24%), wrong drug (n = 168, 17%), wrong form (n = 134, 13%) and printing the wrong warnings/directions on the label (n = 112, 11%). A statistically significant finding was that fewer incidents involving dispensing the wrong drug were reported by Welsh hospitals (17%) compared to previous research in the UK (23%; P = 0.01). Drugs most commonly involved in incidents were insulin (n = 34; 6 incidents per 10000 issues of insulin), nifedipine (n = 16; 10 incidents per 10000 issues of nifedipine) and carbamazepine (n = 10; 5 incidents per 10000 issues of carbamazepine). Conclusion The overall unprevented dispensing incident rate was less than previously reported. Dispensing the wrong strength of the correct drug is a problem. Staff should be aware of the risk of dispensing incidents involving insulin, nifedipine and carbamazepine. Strategies for minimising dispensing incidents include using shelf labels to highlight different strengths or formulations of the same drug, and educating staff about easily confused drugs.