Objectives To evaluate the clinical relevance of pharmacist intervention on patient care in emergencies, to determine the severity of detected errors. Second, to analyse the most frequent types of interventions and type of drugs involved and to evaluate the clinical pharmacist's activity.
Methods A 6-month observational prospective study of pharmacist intervention in the Emergency Department (ED) at a 400-bed hospital in Spain was performed to record interventions carried out by the clinical pharmacists. We determined whether the intervention occurred in the process of medication reconciliation or another activity, and whether the drug involved belonged to the High-Alert Medications Institute for Safe Medication Practices (ISMP) list. To evaluate the severity of the errors detected and clinical relevance of the pharmacist intervention, a modified assessment scale of Overhage and Lukes was used. Relationship between clinical relevance of pharmacist intervention and the severity of medication errors was assessed using ORs and Spearman's correlation coefficient.
Results During the observation period, pharmacists reviewed the pharmacotherapy history and medication orders of 2984 patients. A total of 991 interventions were recorded in 557 patients; 67.2% of the errors were detected during medication reconciliation. Medication errors were considered severe in 57.2% of cases and 64.9% of pharmacist intervention were considered relevant. About 10.9% of the drugs involved are in the High-Alert Medications ISMP list. The severity of the medication error and the clinical significance of the pharmacist intervention were correlated (Spearman's ρ=0.728/p<0.001).
Conclusions In this single centre study, the clinical pharmacists identified and intervened on a high number of severe medication errors. This suggests that emergency services will benefit from pharmacist-provided drug therapy services.
Statistics from Altmetric.com
What this study adds?
The potential impact of pharmacist intervention on quality and safety in the ED has not been well-described.
In this prospective observational study, a high number of potentially serious medication errors were detected by clinical pharmacists during medication reconciliation and review of prescribing in the ED. Pharmacist intervention occurred in 19% of patients and 65% of these were considered clinically relevant.
In this paper, we pretend to evaluate the clinical relevance of the pharmacist intervention in patients who go to the ED care and to analyse the most frequent types of interventions.
What is already known on this subject?
As far as the authors know, in the published literature, not all authors measure the potential impact of pharmacist intervention on health.
A medication error can be defined as ‘a failure in the treatment process that leads to, or has the potential to lead to, harming the patient’.1 Errors will always occur in any system, but it is essential to identify causes and attempt to minimise risks2 of these preventable adverse drug events. High-Alert Medications are drugs that bear a greater risk of causing significant patient harm when used by mistake. Although errors may or may not be more common with these drugs, the consequences of a mistake are clearly more devastating to patients.3
The ED is very prone to medication errors due to multiple factors: workload, lack of coordination between different services, inadequate identification of patients, duties, overloaded working shifts, constant interruptions, miscommunications between professionals and a frequent use of intravenous medication.4–6 Prior studies suggest that up to 10% of patients admitted to ED experience medication errors during their ED stay. Even 70% of these mistakes could have been prevented.7 ,8
The European Society of Clinical Pharmacy defines clinical pharmacy as all the services performed by pharmacists who practice in hospitals, community pharmacies, clinics and any other setting where medicines are prescribed and used.9 The overall goal of clinical pharmacy activities is to promote the correct and appropriate use of medicinal products and devices and the clinical pharmacist has proven to be a key part of the multidisciplinary team that improves the quality, safety and efficiency of patient's treatments through the pharmaceutical care. The Society recommends that all prescriptions be reviewed and validated as soon as possible by a hospital pharmacist, that hospital pharmacists should play a full part in decision making including advising, implementing and monitoring medication changes in full partnership with patients and other professionals and pharmacists' clinical interventions should be analysed to inform quality improvement interventions.
Pharmacists have been involved in ED for decades. In 2009, a systematic literature review was conducted to ascertain the scope of involvement of clinical pharmacists in ED; to summarise economic, humanistic and clinical outcomes data.10 Services provided included traditional clinical pharmacy services, responding to medical emergencies, providing consultations on medication issues, identifying and reducing medication errors and conducting medication histories on admission to hospital. However, most studies included in the 2009 review were performed in the USA, where the activity of the emergency pharmacist is more established through American Society of Health-System Pharmacists.11 In Europe, this profession is not well established, and the description of pharmacist activity is reduced to isolated publications.12–14
One of the most common activities of clinical pharmacist is medication reconciliation. This process involves comparing the patient's current medications against the physician's admission, transfer and/or discharge orders. Reconciliation accurately identiﬁes all the medications that patients are taking—including name of drug, dosage, frequency and route, and the pharmacist uses this list to provide correct medications for patients anywhere within the healthcare system.15
The importance of clinical pharmacists in reconciliation has been recognised by WHO.16 The American Society of Critical Care recommends that a pharmacist should be responsible for the pharmacological treatment optimisation in all hospitals which have at least one unit of critical patients.17
In published literature, most studies describe the severity of the detected errors in prescriptions regarding the patient's health, but not record the potential impact of pharmacist intervention.
Therefore, the main aim of this study was to determine the severity of detected errors detected by clinical pharmacists and evaluate the clinical relevance of pharmacist intervention in patients who undergo ED care. Secondary objectives were to evaluate the activity of pharmaceutical care (interventions per day and per patient) to determine the most frequent types of interventions and to analyse the interventions conducted within drugs included in High-Alert Medications ISMP list.
An observational prospective study of pharmacist intervention in ED over a non-consecutive period of 6 months was carried out. The study was performed at a 400-bed hospital in Spain that attends a population of 250 000 inhabitants and receives around 100 000 visits per year.
The hospital ED has 74 beds, divided into different zones: 1. triage rooms; 2. hospital stretchers to explore patients classified according to pathology: ophthalmology, otolaryngology, orthopaedic surgery, psychiatry, room negative pressure (boxes); 3. semicritical patients area; 4. basic or minor care unit; 5. chest pain unit; 6. observation unit and 7. short-stay unit.
The clinical pharmacists are present in the ED Monday to Friday 8:00 to 15:00. Their job is to review and monitor the medical orders in patients admitted to any ED area and detect and intervene on medication errors. Medical orders are conducted through a system of electronic prescriptions. Clinical pharmacists also perform reconciliation of home medications in the short-stay unit and observation unit, where patients stay at least 24 hours before discharge, and boxes (hospital stretchers to explore patients classified according to pathology), where patients could be admitted directly or discharged within a few hours. Reconciliation involves review and verification of all home medications and then making sure these are given while patients are in the ED/observation area, that doses and regimen are correct and that no medications are given that would interact. In addition, their task is to support and answer any queries or doubts about pharmacotherapy from physicians and nurses. Pharmacists prioritise the observation unit and the short-stay unit because these are the areas where patients spend more time and are more likely to receive their home medications.
A database (Microsoft Office Access) was designed to record the interventions that were carried out by clinical pharmacists during a period of study of 6 months (October–November 2013 and January–April 2014). December 2013 was excluded because there was not a clinical pharmacist responsible for ED that month. At the end of each day, the pharmacists registered the interventions carried out that day. During the study period, one pharmacist was employed in ED along with hospital pharmacists in training who worked with ED pharmacist. The patient population consisted of all patients for whom the clinical pharmacist performed any medication review.
For each incidence where there was pharmacist intervention, we collected patients’ demographic characteristics (age, gender), the drug involved and whether the identification of the error occurred during reconciliation of home medication or from another type of clinical pharmacist activity. For each medication, we determined the pharmacotherapeutic group to which the medication belonged (according to the Anatomical Therapeutic Chemical (ATC) Classification) and whether the drug was considered a High-Alert Medications according to the classification established by The Institute for Safe Medication Practices (ISMP). For each intervention, we also determined the severity of the prescribing error detected, and the clinical relevance of the pharmacist intervention.
To evaluate the severity of the errors detected and clinical relevance of the pharmacist intervention, modified assessment scales by Overhage and Lukes18 were used. These scales evaluate the potential impact of the error and the interventions on the patient’s health. For the errors, five severity categories are possible: 1. potentially lethal, 2. serious, 3. significant, 4. less significant and 5. no error. The clinical relevance of pharmacist intervention was determined based on the likely consequences in patient care: 1. extremely significant, 2. highly significant, 3. significant, 4. hardly significant, 5. insignificant and 6. injurious intervention (box 1). As an example the order of digoxin treatment every 24 hours instead of 48 hours in patients with nephropathy would be classified as ‘3-significant’ because a double dose is used in a drug with a narrow therapeutic index. The clinical relevance of the pharmacist intervention to change the dosing interval to the correct one would be ‘highly significant’ because damage to a vital organ (and possible lethal arrhythmias) is avoided.
Classification of the different categories of the validated method proposed by Overhage and Lukes18 regarding ‘severity of prescription error’ and ‘impact of service provided by the pharmacist’
Severity of prescription error
1. Potentially lethal
High potential to compromise the life of patients.
Potentially life-saving drug prescribed at a dose too low for the condition for which it is being used.
High dose (>normal dose ×10) of a drug with a narrow therapeutic range.
Route of administration can lead to severe toxicity.
Dose of a medication too low for a serious disease in a patient with distress acute.
High dose (normal dose ×4–10) of a drug with a narrow therapeutic range.
The dose may result in potentially toxic concentrations.
The drug may exacerbate the condition of the patient (side effects or contraindications).
Spelling or interpretation errors of medications that could lead to dispensing the wrong drug.
Documented allergy to a drug.
High dose (>normal dose×10) of a drug with normal therapeutic range.
Omission of a pretest for a drug that could cause hypersensitivity.
High dose (normal dose ×1.5–4) of drugs with narrow therapeutic range.
Dose of a medication too low for the condition of the patient.
High dose (normal dose ×1–1.5) of a drug with normal therapeutic range.
Inappropriate dosage interval.
Omission of a drug in the medical prescription.
Incomplete information in the medical prescription.
Inappropriate or unavailable pharmaceutical form.
Prescription of drugs not included in the pharmacotherapeutic guide of the hospital.
Illegible, ambiguous or non-standard abbreviations.
5. No error
Medical prescription required additional information or clarification.
Impact of service provided by the pharmacist
1. Extremely significant
The recommendation avoids a situation that endangers the life of patients.
2. Highly significant
The recommendation prevents damage for a vital organ.
Recommendation prevents serious adverse reactions resulting from a drug interaction or a contraindication for use.
Recommendation improves the quality of life of the patient (standard practices described in the hospital).
4. Little significant
The benefit of the recommendation is neutral depending on the interpretation of the professional.
Generalised recommendations not specific of a patient.
6. Injurious intervention
Inappropriate recommendations that can lead to a worsening of the condition patient.
The classification of the severity of errors and relevance of the pharmacist intervention were performed independently by two pharmacists; disagreements were resolved by a third pharmacist.
Statistical analysis was performed using absolute frequencies (percentages) for qualitative variables and central measures (mean) with dispersion measures (SD, range) for quantitative variables. We summarised the main reasons for pharmacist intervention and the most common drugs involved. We performed comparative statistics for those patients with and without relevant interventions according to the gender, age and number of interventions using χ2 test, independent sample Student's t-test and independent sample Mann-Whitney U test, respectively.
The relationship between clinical relevance of the pharmacist intervention and the severity of medication errors detected was assessed using Spearman’s correlation coefficient. Errors were defined as ‘significant’ if they corresponded to categories 1–3 of the Overhage and Lukes scale for severity and interventions relevant if they corresponded to categories 1–3 in the relevance scale. The association of pharmacist intervention was considered as ‘relevant’ (categories: 1–3) and ‘non-relevant’ (categories: 4–5).
Association of the type of activity, the severity of the error, relevance of the intervention and the involvement of High-Alert Medications was assessed by OR method using IBM SPSS V.15 version statistics, while considering a two-sided significance level of 0.05 as statistically significant. Interventions as injurious intervention for Overhage and Lukes18 were excluded from this analysis as they did not correspond to any error.
As the study was considered a service evaluation, ethics committee approval was not required. The relevant legislation concerning data protection was fulfilled, as only patient data relating to gender and age were handled.
During the study period, there were 30855 ED visits. ED pharmacists reviewed the pharmacotherapy history of 2984 (9.7%) patients of the total. Six hundred and sixty-six patients (22.32%) were in either the short-stay unit or observation unit, of whom 395 patients (13.24%) were finally admitted into a hospitalisation unit. The remaining patients were discharged directly from the ED or admitted into hospitalisation units without passing through the short-stay unit or observation.
Of all patients reviewed, 557 patients (18.9%) had interventions, for a total of 991 interventions (1.8 interventions per patient); 49.6% were male; median age=69 years (14–103). On average, there were eight interventions per day and 67.2% occurred during reconciliation. The severity of medication errors was considered significant in 572 (57.7%) corresponding to categories 1–3 from Overhage and Lukes18 scale. Out of the total number of interventions carried out, 64.9% were considered relevant corresponding to categories 1–3 from Overhage and Lukes18 scale. There was no association between age or gender with whether or not the pharmacist intervention was relevant, although patients receiving greater number of interventions were more likely to have a relevant pharmacist intervention. The relationship between the relevance of pharmacist intervention and the baseline characteristic of patients and the number of interventions by patients is shown in table 1.
The number of procedures performed for each of the categories belonging to the two scales are shown in figure 1. There was one injurious intervention (0.10%), which consisted of the validation of a prescription of urapidil in a patient with aortic stenosis, and the patient suffered severe hypotension. (The treatment was stopped and finally, the patient was stabilised.) A statistically significant association between the detection of a significant error and a relevant intervention was detected and correlation between the different levels for pharmacist intervention and the severity of medication errors detected was demonstrated (Spearman's ρ=0.728./p=0.<001) (figure 1).
The proportion of medication errors and interventions in each of the Overhage and Lukes categories, as well as the main reasons that required pharmacist's recommendations (medications errors), and the most common drugs involved are shown in table 2 The most common interventions were therapeutic exchange of a drug not included in the hospital Drug Therapy Guide (189 interventions, 19.1%) and the inclusion of a drug in the treatment that the patient was taking regularly but had not been prescribed (191 interventions, 19.3%). Omeprazole and simvastatin were the most common drugs in this category.
About 10.9% of the drugs involved were included in High-Alert Medications ISMP list. There were 35 pharmacist intervention for antithrombotic agents, 19 for oral hypoglycaemic agents, 16 for insulins (subcutaneous and intravenous), 15 for narcotics/opioids and 6 for immunosuppressive agents (table 2).
High-Alert Medications were involved in a higher frequency of relevant interventions, although without statistical significance (OR=1.201 (0.782–1.844). Analysing the severity of the errors, we observed that the frequency of serious errors detected in High-Alert Medications (59/108) was comparable to the frequency of serious errors in drugs not belonging to this group (513/883) (OR=0.868 (0.581–1.298)). Table 3 represents the relationship between the relevance of pharmacist intervention and reconciliation and High-Alert Medications.
Our results show a high clinical relevance of interventions by pharmacists on medication, with a prominent percentage of significant interventions. The majority of interventions occurred during medication reconciliation.
The proportion of interventions that was derived from medication reconciliation with respect to the total interventions was higher than that found in similar studies conducted in ED.19 This could be due to the fact that other authors only considered pharmacist intervention derived from the start of treatment, while we recorded any intervention. This may also be the reason we have more types of interventions than in other studies.19 ,20 A strong association between the clinical relevance of pharmacist intervention and the severity of the detected error has also been detected. This is expected due to the relationship between the possible consequences of a highly significant error and the influence that pharmacist intervention may have.
The majority of interventions were the result of substitution of a home medication for one on the hospital formulary or failure to prescribe a home medication. The first case may be due to lack of knowledge by clinicians about the equivalence of dose between drugs from the same family. Interventions to prescribe a home medication omitted by the physician were lower than in other published studies,19 ,20 suggesting that ED clinicians are becoming increasingly becoming aware of the importance of continuing chronic treatment of patients and digital medical records allows access to home medications easier.
Main drugs involved in these interventions were omeprazole and simvastatin. Omeprazole is a drug generally overused in our environment.21 Gené et al22 proposed an algorithm for gastroprotection in emergencies based on the presence of different gastrointestinal risk factors (history of gastrointestinal bleeding) and cardiovascular risk factors in patients. This algorithm was used by our pharmacists to determine appropriate initiation or suspension of prophylactic treatment of non-steroidal anti-inflammatory drug gastropathy. Moreover, medication errors in statins may be significant in patients with acute coronary syndrome,23 a frequent pathology in ED. Patient safety improves when the pharmacist monitors these drugs.24
No statistical association was found between High-Alert Medications and an increased frequency of significant errors or interventions. The ISMP defined these drugs as those that could cause greater adverse effects in patients, although this has not been previously studied. In our study, this association could not be proven, but this could be because of the presence of a pharmacist allowing physicians to ask for help before prescribing these drugs.
Much of the published literature on pharmacists in the ED describes the severity of the detected errors in prescriptions but not the potential impact of interventions on the patient's health. Some studies have assessed the quality of the interventions carried out based on the available published evidence,16 concluding that interventions with higher quality were those related to medication reconciliation. Other studies established the quality of pharmacist intervention based on a prior detection of medication errors,25 similar to our study.
A limitation of this study is that not every patient from ED was included. Pharmacists are only in the ED 20% of the time the ED works and thus are only able to intervene in a small percentage of patients. The study was conducted at a single ED, thus making it difficult to generalise the findings. Also, only pharmacists evaluated significance of interventions, which could introduce bias. We prioritised patients for review who were admitted to the short-stay unit and the observation unit, and thus the severity and impact of the errors and interventions may not be representative of what would occur with a general ED population. In conclusion, in this single-centre retrospective study we found that clinical pharmacists identified a high number of significant errors and their interventions were highly relevant. This suggests that patients would substantially benefit from pharmacist-provided drug therapy services in ED.
Contributors MAP-M, JMR-C and BC-H have contributed to the design of the study and database; MAP-M and JMR-C have realised the acquisition and analysis of data and the drafting of the manuscript; BC-H, JT-T and BC-D have revised the manuscript critically for important intellectual content, all of these have approved the final version to be submitted and MAP-M have submitted the manuscript.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.