A systematic review of the performance characteristics of clinical event monitor signals used to detect adverse drug events in the hospital setting

Navigating Drug-Induced Adversities: A Python-Based Console Application for Causality Assessment Using the Naranjo Algorithm

The timely and accurate adverse drug reactions (ADR) assessment is vital for effective patient management and healthcare delivery. The Naranjo Algorithm is a widely recognized tool for determining the probability that a drug induces a given ADR. However, the process can be time-consuming and susceptible to human error. This study introduces a Python-based console application (Python Software Foundation, Wilmington, Delaware, United States) designed to automate the Naranjo Algorithm for ADR causality assessment. The application was developed using Python 3.11.4 on a Windows 11 system (Microsoft Corporation, Redmond, Washington, United States) and compiled in Notepad (Microsoft Corporation), a basic text editor, highlighting its accessibility and ease of use in various settings. User input is solicited for each question in the Naranjo Algorithm, validated for acceptable entries, and subsequently scored. The final score categorizes the reaction into Doubtful, Possible, Probable, or Definite ADR, facilitating rapid clinical decision-making. Preliminary validation shows promising reliability and effectiveness, making it a valuable asset in research and clinical settings for assessment.

Generalizability of machine learning methods in detecting adverse drug events from clinical narratives in electronic medical records

We assessed the generalizability of machine learning methods using natural language processing (NLP) techniques to detect adverse drug events (ADEs) from clinical narratives in electronic medical records (EMRs). We constructed a new corpus correlating drugs with adverse drug events using 1,394 clinical notes of 47 randomly selected patients who received immune checkpoint inhibitors (ICIs) from 2011 to 2018 at The Ohio State University James Cancer Hospital, annotating 189 drug-ADE relations in single sentences within the medical records. We also used data from Harvard's publicly available 2018 National Clinical Challenge (n2c2), which includes 505 discharge summaries with annotations of 1,355 single-sentence drug-ADE relations. We applied classical machine learning (support vector machine (SVM)), deep learning (convolutional neural network (CNN) and bidirectional long short-term memory (BiLSTM)), and state-of-the-art transformer-based (bidirectional encoder representations from transformers (BERT) and ClinicalBERT) methods trained and tested in the two different corpora and compared performance among them to detect drug-ADE relationships. ClinicalBERT detected drug-ADE relationships better than the other methods when trained using our dataset and tested in n2c2 (ClinicalBERT F-score, 0.78; other methods, F-scores, 0.61-0.73) and when trained using the n2c2 dataset and tested in ours (ClinicalBERT F-score, 0.74; other methods, F-scores, 0.55-0.72). Comparison among several machine learning methods demonstrated the superior performance and, therefore, the greatest generalizability of findings of ClinicalBERT for the detection of drug-ADE relations from clinical narratives in electronic medical records.

Classification of Health Information Technology Safety Events in a Pediatric Tertiary Care Hospital

OBJECTIVE

State agencies have developed reporting systems of safety events that include events related to health information technology (HIT). These data come from hospital reporting systems where staff submit safety reports and nurses, in the role of safety managers, review, and code events. Safety managers may have varying degrees of experience with identifying events related to HIT. Our objective was to review events potentially involving HIT and compare those with what was reported to the state.

METHODS

We performed a structured review of 1 year of safety events from an academic pediatric healthcare system. We reviewed the free-text description of each event and applied a classification scheme derived from the AHRQ Health IT Hazard Manager and compared the results with events reported to the state as involving HIT.

RESULTS

Of 33,218 safety events for a 1-year period, 1247 included key words related to HIT and/or were indicated by safety managers as involving HIT. Of the 1247 events, the structured review identified 769 as involving HIT. In comparison, safety managers only identified 194 of the 769 events (25%) as involving HIT. Most events, 353 (46%), not identified by safety managers were documentation issues. Of the 1247 events, the structured review identified 478 as not involving HIT while safety managers identified and reported 81 of these 478 events (17%) as involving HIT.

CONCLUSIONS

The current process of reporting safety events lacks standardization in identifying health technology contributions to safety events, which may minimize the effectiveness of safety initiatives.

Clinical decision support systems-based interventions to improve medication outcomes: A systematic literature review on features and effects

Background: Clinical decision support systems (CDSSs) interventions were used to improve the life quality and safety in patients and also to improve practitioner performance, especially in the field of medication. Therefore, the aim of the paper was to summarize the available evidence on the impact, outcomes and significant factors on the implementation of CDSS in the field of medicine. Methods: This study is a systematic literature review. PubMed, Cochrane Library, Web of Science, Scopus, EMBASE, and ProQuest were investigated by 15 February 2017. The inclusion requirements were met by 98 papers, from which 13 had described important factors in the implementation of CDSS, and 86 were medicated-related. We categorized the system in terms of its correlation with medication in which a system was implemented, and our intended results were examined. In this study, the process outcomes (such as; prescription, drug-drug interaction, drug adherence, etc.), patient outcomes, and significant factors affecting the implementation of CDSS were reviewed. Results: We found evidence that the use of medication-related CDSS improves clinical outcomes. Also, significant results were obtained regarding the reduction of prescription errors, and the improvement in quality and safety of medication prescribed. Conclusion: The results of this study show that, although computer systems such as CDSS may cause errors, in most cases, it has helped to improve prescribing, reduce side effects and drug interactions, and improve patient safety. Although these systems have improved the performance of practitioners and processes, there has not been much research on the impact of these systems on patient outcomes.

Towards Personalized Antithrombotic Treatments: Focus on P2Y12 Inhibitors and Direct Oral Anticoagulants

Oral anticoagulants and antiplatelet drugs are commonly prescribed to lower the risk of cardiovascular diseases, such as venous and arterial thrombosis, which represent the leading causes of mortality worldwide. A significant percentage of patients taking antithrombotics will nevertheless experience bleeding or recurrent ischemic events, and this represents a major public health issue. Cardiovascular medicine is now questioning the one-size-fits-all policy, and more personalized approaches are increasingly being considered. However, the available tools are currently limited and they are only moderately able to predict clinical events or have a significant impact on clinical outcomes. Predicting concentrations of antithrombotics in blood could be an effective means of personalization as they have been associated with bleeding and recurrent ischemia. Target concentration interventions could take advantage of physiologically based pharmacokinetic (PBPK) and population-based pharmacokinetic (POPPK) models, which are increasingly used in clinical settings and have attracted the interest of governmental regulatory agencies, to propose dosages adapted to specific population characteristics. These models have the benefit of combining parameters from different sources, such as experimental in vitro data and patients' demographic, genetic, and physiological in vivo data, to characterize the dose-concentration relationships of compounds of interest. As such, they can be used to predict individual drug exposure. In the near future, these models could therefore be a valuable means of predicting personalized antithrombotic blood concentrations and, hopefully, of preventing clinical non-response or bleeding in a given patient. Existing approaches for personalization of antithrombotic prescriptions will be reviewed using practical examples for P2Y₁₂ inhibitors and direct oral anticoagulants. The review will additionally focus on the existing PBPK and POPPK models for these two categories of drugs. Lastly, we address potential scenarios for their implementation in clinics, along with the main limitations and challenges.

Extraction of Information Related to Adverse Drug Events from Electronic Health Record Notes: Design of an End-to-End Model Based on Deep Learning

Background

Pharmacovigilance and drug-safety surveillance are crucial for monitoring adverse drug events (ADEs), but the main ADE-reporting systems such as Food and Drug Administration Adverse Event Reporting System face challenges such as underreporting. Therefore, as complementary surveillance, data on ADEs are extracted from electronic health record (EHR) notes via natural language processing (NLP). As NLP develops, many up-to-date machine-learning techniques are introduced in this field, such as deep learning and multi-task learning (MTL). However, only a few studies have focused on employing such techniques to extract ADEs.

Objective

We aimed to design a deep learning model for extracting ADEs and related information such as medications and indications. Since extraction of ADE-related information includes two steps—named entity recognition and relation extraction—our second objective was to improve the deep learning model using multi-task learning between the two steps.

Methods

We employed the dataset from the Medication, Indication and Adverse Drug Events (MADE) 1.0 challenge to train and test our models. This dataset consists of 1089 EHR notes of cancer patients and includes 9 entity types such as Medication, Indication, and ADE and 7 types of relations between these entities. To extract information from the dataset, we proposed a deep-learning model that uses a bidirectional long short-term memory (BiLSTM) conditional random field network to recognize entities and a BiLSTM-Attention network to extract relations. To further improve the deep-learning model, we employed three typical MTL methods, namely, hard parameter sharing, parameter regularization, and task relation learning, to build three MTL models, called HardMTL, RegMTL, and LearnMTL, respectively.

Results

Since extraction of ADE-related information is a two-step task, the result of the second step (ie, relation extraction) was used to compare all models. We used microaveraged precision, recall, and F1 as evaluation metrics. Our deep learning model achieved state-of-the-art results (F1=65.9%), which is significantly higher than that (F1=61.7%) of the best system in the MADE1.0 challenge. HardMTL further improved the F1 by 0.8%, boosting the F1 to 66.7%, whereas RegMTL and LearnMTL failed to boost the performance.

Conclusions

Deep learning models can significantly improve the performance of ADE-related information extraction. MTL may be effective for named entity recognition and relation extraction, but it depends on the methods, data, and other factors. Our results can facilitate research on ADE detection, NLP, and machine learning.

Evaluation of a Novel System to Enhance Clinicians' Recognition of Preadmission Adverse Drug Reactions

BACKGROUND

Often unrecognized by providers, adverse drug reactions (ADRs) diminish patients' quality of life, cause preventable admissions and emergency department visits, and increase health care costs.

OBJECTIVE

This article evaluates whether an automated system, the Adverse Drug Effect Recognizer (ADER), could assist clinicians in detecting and addressing inpatients' ongoing preadmission ADRs.

METHODS

ADER uses natural language processing to extract patients' medications, findings, and past diagnoses from admission notes. It compares excerpted information to a database of known medication adverse effects and promptly warns clinicians about potential ongoing ADRs and potential confounders via alerts placed in patients' electronic health records (EHRs). A 3-month intervention trial evaluated ADER's impact on antihypertensive medication ordering behaviors. At the time of patient admission, ADER warned providers on the Internal Medicine wards of Vanderbilt University Hospital about potential ongoing preadmission antihypertensive medication ADRs. A retrospective control group, comprised similar physicians from a period prior to the intervention, received no alerts. The evaluation compared ordering behaviors for each group to determine if preadmission medications changed during hospitalization or at discharge. The study also analyzed intervention group participants' survey responses and user comments.

RESULTS

ADER identified potential preadmission ADRs for 30% of both groups. Compared with controls, intervention providers more often withheld or discontinued suspected ADR-causing medications during the inpatient stay (p < 0.001). Intervention providers who responded to alert-related surveys held or discontinued suspected ADR-causing medications more often at discharge (p < 0.001).

CONCLUSION

Results indicate that ADER helped physicians recognize ADRs and reduced ordering of suspected ADR-causing medications. In hospitals using EHRs, ADER-like systems could improve clinicians' recognition and elimination of ongoing ADRs.

Clinical Relation Extraction Toward Drug Safety Surveillance Using Electronic Health Record Narratives: Classical Learning Versus Deep Learning

BACKGROUND

Medication and adverse drug event (ADE) information extracted from electronic health record (EHR) notes can be a rich resource for drug safety surveillance. Existing observational studies have mainly relied on structured EHR data to obtain ADE information; however, ADEs are often buried in the EHR narratives and not recorded in structured data.

OBJECTIVE

To unlock ADE-related information from EHR narratives, there is a need to extract relevant entities and identify relations among them. In this study, we focus on relation identification. This study aimed to evaluate natural language processing and machine learning approaches using the expert-annotated medical entities and relations in the context of drug safety surveillance, and investigate how different learning approaches perform under different configurations.

METHODS

We have manually annotated 791 EHR notes with 9 named entities (eg, medication, indication, severity, and ADEs) and 7 different types of relations (eg, medication-dosage, medication-ADE, and severity-ADE). Then, we explored 3 supervised machine learning systems for relation identification: (1) a support vector machines (SVM) system, (2) an end-to-end deep neural network system, and (3) a supervised descriptive rule induction baseline system. For the neural network system, we exploited the state-of-the-art recurrent neural network (RNN) and attention models. We report the performance by macro-averaged precision, recall, and F1-score across the relation types.

RESULTS

Our results show that the SVM model achieved the best average F1-score of 89.1% on test data, outperforming the long short-term memory (LSTM) model with attention (F1-score of 65.72%) as well as the rule induction baseline system (F1-score of 7.47%) by a large margin. The bidirectional LSTM model with attention achieved the best performance among different RNN models. With the inclusion of additional features in the LSTM model, its performance can be boosted to an average F1-score of 77.35%.

CONCLUSIONS

It shows that classical learning models (SVM) remains advantageous over deep learning models (RNN variants) for clinical relation identification, especially for long-distance intersentential relations. However, RNNs demonstrate a great potential of significant improvement if more training data become available. Our work is an important step toward mining EHRs to improve the efficacy of drug safety surveillance. Most importantly, the annotated data used in this study will be made publicly available, which will further promote drug safety research in the community.

Quality of Decision Support in Computerized Provider Order Entry: Systematic Literature Review

Background

Computerized decision support systems have raised a lot of hopes and expectations in the field of order entry. Although there are numerous studies reporting positive impacts, concerns are increasingly high about alert fatigue and effective impacts of these systems. One of the root causes of fatigue alert reported is the low clinical relevance of these alerts.

Objective

The objective of this systematic review was to assess the reported positive predictive value (PPV), as a proxy to clinical relevance, of decision support systems in computerized provider order entry (CPOE).

Methods

A systematic search of the scientific literature published between February 2009 and March 2015 on CPOE, clinical decision support systems, and the predictive value associated with alert fatigue was conducted using PubMed database. Inclusion criteria were as follows: English language, full text available (free or pay for access), assessed medication, direct or indirect level of predictive value, sensitivity, or specificity. When possible with the information provided, PPV was calculated or evaluated.

Results

Additive queries on PubMed retrieved 928 candidate papers. Of these, 376 were eligible based on abstract. Finally, 26 studies qualified for a full-text review, and 17 provided enough information for the study objectives. An additional 4 papers were added from the references of the reviewed papers. The results demonstrate massive variations in PPVs ranging from 8% to 83% according to the object of the decision support, with most results between 20% and 40%. The best results were observed when patients’ characteristics, such as comorbidity or laboratory test results, were taken into account. There was also an important variation in sensitivity, ranging from 38% to 91%.

Conclusions

There is increasing reporting of alerts override in CPOE decision support. Several causes are discussed in the literature, the most important one being the clinical relevance of alerts. In this paper, we tried to assess formally the clinical relevance of alerts, using a near-strong proxy, which is the PPV of alerts, or any way to express it such as the rate of true and false positive alerts. In doing this literature review, three inferences were drawn. First, very few papers report direct or enough indirect elements that support the use or the computation of PPV, which is a gold standard for all diagnostic tools in medicine and should be systematically reported for decision support. Second, the PPV varies a lot according to the typology of decision support, so that overall rates are not useful, but must be reported by the type of alert. Finally, in general, the PPVs are below or near 50%, which can be considered as very low.

Identifying, Understanding, and Managing Patient Safety and Clinical Risks in the Clinical Research Environment

Meticulous study design, exacting scientific review, scrupulous data management, rigorous human subjects' protection, and effective recruitment strategies are essential components of all clinical research studies that must be in place prior to the conduct of the study. Equally important to the success of the study and to the well-being and protection of study participants is the assurance that safe clinical environments are provided for the care of research participants. This chapter describes approaches to managing patient care and clinical research quality, as well as the application of established health care-based clinical quality methods and tools to the complex processes of clinical research. Techniques for identifying clinical risks, tools for collecting data about the quality of clinical care processes, strategies for mitigating risk and patient harm, and techniques for measuring the participants' perceptions of the research process are discussed in detail.

Measurement of selected preventable adverse drug events in electronic health records: Toward developing a complexity score

PURPOSE

The defining of a select number of high-priority preventable adverse drug events (pADEs) for measurement in the electronic health record (EHR) and the estimation of pADE incidences in two tertiary care facilities are described.

METHODS

This study was part of a larger effort aimed at developing an automated electronic health record (EHR)-based complexity-score (C-score) that ranks hospitalized patients according to their risk for pADEs for clinical intervention. We developed measures for 16 high-priority pADEs often deemed preventable using discrete clinical and administrative EHR data. For each pADE we specified inclusion and exclusion criteria that were used to define risk populations for each specific pADE. The incidence of each type of pADE was then measured during a designated follow-up period considering all adult admissions to 2 large academic tertiary care hospitals, who were eligible for the pADE-specific risk populations during any of their first 5 hospital days.

RESULTS

Utilizing the data from 83,787 admissions who were at risk for at least one pADE during at least one of their first five hospital days, we found that 27,193 admissions (32.5%) developed at least one pADE. Uncontrolled postsurgical pain, uncontrolled pneumonia, and drug-associated hypotension had the highest incidences with the following number of days with pADE per number of patients at risk: 13,484 of 19,640; 527 of 1,530; and 13,394 of 43,630, while drug-associated falls (446 of 75,036), drug-associated acute mental status changes (262 of 66,875) and venous thromboembolism (214 of 74,283) had the lowest incidence rates.

CONCLUSION

EHR-based definitions of clinically important pADEs were developed, and the incidence of the pADEs was estimated. These definitions will be advanced for the creation of prediction models to develop a C-score for identifying patients at risk for pADEs to prioritize pharmacist intervention.

Unfractionated heparin versus low molecular weight heparins for avoiding heparin-induced thrombocytopenia in postoperative patients

BACKGROUND

Heparin-induced thrombocytopenia (HIT) is an adverse drug reaction presenting as a prothrombotic disorder related to antibody-mediated platelet activation. It is a paradoxical immune reaction resulting in thrombin generation in vivo, which leads to a hypercoagulable state and the potential to initiate venous or arterial thrombosis. A number of factors are thought to influence the incidence of HIT including the type and preparation of heparin (unfractionated heparin (UFH) or low molecular weight heparin (LMWH)) and the heparin-exposed patient population, with the postoperative patient population at higher risk.Although LMWH has largely replaced UFH as a front-line therapy, there is evidence supporting a lack of superiority of LMWH compared with UFH regarding prevention of deep vein thrombosis and pulmonary embolism following surgery, and similar frequencies of bleeding have been described with LMWH and UFH. The decision as to which of these two preparations of heparin to use may thus be influenced by harmful effects such as HIT. We therefore sought to determine the relative impact of UFH and LMWH on HIT in postoperative patients receiving thromboembolism prophylaxis. This is an update of a review first published in 2012.

OBJECTIVES

The objective of this review was to compare the incidence of heparin-induced thrombocytopenia (HIT) and HIT complicated by venous thromboembolism in postoperative patients exposed to unfractionated heparin (UFH) versus low molecular weight heparin (LMWH).

SEARCH METHODS

For this update, the Cochrane Vascular Information Specialist searched the Specialised Register (May 2016), CENTRAL (2016, Issue 4) and trials registries. The authors searched Lilacs (June 2016) and additional trials were sought from reference lists of relevant publications.

SELECTION CRITERIA

We included randomised controlled trials (RCTs) in which participants were postoperative patients allocated to receive prophylaxis with UFH or LMWH, in a blinded or unblinded fashion. Studies were excluded if they did not use the accepted definition of HIT. This was defined as a relative reduction in the platelet count of 50% or greater from the postoperative peak (even if the platelet count at its lowest remained greater than 150 x 10⁹/L) occurring within five to 14 days after the surgery, with or without a thrombotic event occurring in this timeframe. Additionally, we required circulating antibodies associated with the syndrome to have been investigated through laboratory assays.

DATA COLLECTION AND ANALYSIS

Two review authors independently extracted data and assessed the risk of bias. Disagreements were resolved by consensus with participation of a third author.

MAIN RESULTS

In this update, we included three trials involving 1398 postoperative participants. Participants were submitted to general surgical procedures, minor and major, and the minimum mean age was 49 years. Pooled analysis showed a significant reduction in the risk of HIT with LMWH compared with UFH (risk ratio (RR) 0.23, 95% confidence interval (CI) 0.07 to 0.73); low-quality evidence. The number needed to treat for an additional beneficial outcome (NNTB) was 59. The risk of HIT was consistently reduced comparing participants undergoing major surgical procedures exposed to LMWH or UFH (RR 0.22, 95% CI 0.06 to 0.75); low-quality evidence. The occurrence of HIT complicated by venous thromboembolism was significantly lower in participants receiving LMWH compared with UFH (RR 0.22, 95% CI 0.06 to 0.84); low-quality evidence. The NNTB was 75. Arterial thrombosis occurred in only one participant who received UFH. There were no amputations or deaths documented. Although limited evidence is available, it appears that HIT induced by both types of heparins is common in people undergoing major surgical procedures (incidence greater than 1% and less than 10%).

AUTHORS' CONCLUSIONS

This updated review demonstrated low-quality evidence of a lower incidence of HIT, and HIT complicated by venous thromboembolism, in postoperative patients undergoing thromboprophylaxis with LMWH compared with UFH. Similarily, the risk of HIT in people undergoing major surgical procedures was lower when treated with LMWH compared to UFH (low-quality evidence). The quality of the evidence was downgraded due to concerns about the risk of bias in the included studies and imprecision of the study results. These findings may support current clinical use of LMWH over UFH as front-line heparin therapy. However, our conclusions are limited and there was an unexpected paucity of RCTs including HIT as an outcome. To address the scarcity of clinically-relevant information on HIT, HIT must be included as a core harmful outcome in future RCTs of heparin.

Detection of adverse drug events using an electronic trigger tool

PURPOSE

Implementation and refinement of an integrated electronic "trigger tool" for detecting adverse drug events (ADEs) is described.

METHODS

A three-month prospective study was conducted at a large medical center to test and improve the positive predictive value (PPV) of an electronic health record-based tool for detecting ADEs associated with use of four "trigger drugs": the reversal agents flumazenil, naloxone, phytonadione, and protamine. On administration of a trigger drug to an adult patient, an electronic message was transmitted to two pharmacists, who reviewed cases in near real time (typically, on the same day) to detect actual or potential ADEs. In phase 1 of the study, any use of a trigger drug resulted in an alert message; in subsequent phases, the alerting criteria were narrowed on the basis of clinical criteria and laboratory data with the goal of refining the trigger tool's PPV.

RESULTS

A total of 87 drug administrations were reviewed during the three-month study period, with 27 ADEs detected. PPV values in phases 1, 2, and 3 were 0.33, 0.21, and 0.36, respectively. The relatively low overall PPV of the trigger tool was largely attributable to false-positive trigger messages associated with phytonadione use (such messages were reduced from 35 in phase 1 to 7 in phase 3).

CONCLUSION

Evaluation and refinement of an electronic trigger tool based on detecting the use of the reversal agents flumazenil, naloxone, phytonadione, and protamine found an overall PPV of 0.31 during a three-month study period.

Clinical decision support for drug related events: Moving towards better prevention

Clinical decision support (CDS) systems with automated alerts integrated into electronic medical records demonstrate efficacy for detecting medication errors (ME) and adverse drug events (ADEs). Critically ill patients are at increased risk for ME, ADEs and serious negative outcomes related to these events. Capitalizing on CDS to detect ME and prevent adverse drug related events has the potential to improve patient outcomes. The key to an effective medication safety surveillance system incorporating CDS is advancing the signals for alerts by using trajectory analyses to predict clinical events, instead of waiting for these events to occur. Additionally, incorporating cutting-edge biomarkers into alert knowledge in an effort to identify the need to adjust medication therapy portending harm will advance the current state of CDS. CDS can be taken a step further to identify drug related physiological events, which are less commonly included in surveillance systems. Predictive models for adverse events that combine patient factors with laboratory values and biomarkers are being established and these models can be the foundation for individualized CDS alerts to prevent impending ADEs.

Evaluation of a Computer Application for Retrospective Detection of Vitamin K Antagonist Treatment Imbalance

OBJECTIVE

Management of vitamin K antagonists (VKAs) is difficult, and overdoses can have dramatic hemorrhagic consequences. The adverse drug event (ADE) scorecards is a tool intended for the detection and description of adverse drug reaction/ADE developed during a European computerized medical data processing project. It is used in a quality assurance process. Our objective was to evaluate the performance of the ADE scorecards in the detection of the contributing factors for VKA overdoses, among the cases where a VKA overdose is observed.

METHODS

Twenty-eight rules allow the detection of VKA treatment overdose related to drug or a clinical situation. They were applied on 14,748 electronic medical records from a community hospital. Among 582 records including a VKA prescription, 59 cases of VKA overdoses (international normalized ratio ≥ 5) during the hospital stay have been identified. The ADE scorecards detected 49 of them. We evaluated the positive predictive value and sensitivity of these rules, by an expert review of the cases.

RESULTS

The expert review confirmed the contribution of a detected risk factor to the VKA overdose in 11 cases. Therefore, the precision of the rules is 22.4%. The sensitivity is 84.6%. The risk factors were mainly infection and amiodarone introduction. The 4 cases of clinical injury related to a drug were properly designated by the rules.

CONCLUSIONS

Our study shows the great potential of the ADE scorecards for detecting cofactors of VKA overdoses and gives an argument to include complex rules in the knowledge bases used for the detection and identification of ADEs in large medical databases.

Evaluation of an automated surveillance system using trigger alerts to prevent adverse drug events in the intensive care unit and general ward

INTRODUCTION

Adverse events in the intensive care unit (ICU) may be associated with several possible causes, so determining a drug-related causal assessment is more challenging than in general ward patients. Therefore, the hypothesis was that automated trigger alerts may perform differently in various patient care settings. The purpose of this study was to compare the frequency and type of clinically significant automated trigger alerts in critically ill and general ward patients as well as evaluate the performance of alerts for drug-related hazardous conditions (DRHCs).

METHODS

A retrospective cohort study was conducted in adult ICU and general ward patients at three institutions (academic, community, and rural hospital) in a health system. Automated trigger alerts generated during two nonconsecutive months were obtained from a centralized database. Pharmacist responses to alerts and prescriber response to recommendations were evaluated for all alerts. A clinical significant event was defined as an actionable intervention requiring drug therapy changes that the pharmacist determined to be appropriate for patient safety and where the physician accepted the pharmacist's recommendation. The positive predictive value (PPV) was calculated for each trigger alert considered a DRHC (i.e., abnormal laboratory values and suspected drug causes).

RESULTS

A total of 751 alerts were generated in 623 patients during the study period. Pharmacists intervened on 39.8 and 44.8 % alerts generated in the ICU and general ward, respectively. Overall, the physician acceptance rate of approximately 90 % was comparable irrespective of patient care setting. Therefore, the number of clinically significant alerts was 88.9 and 83.4 % for the ICU and non-ICU, respectively. The types of drug therapy changes were similar between settings. The PPV of alerts identifying a DRHC was 0.66 in the ICU and 0.76 in general ward patients.

CONCLUSIONS

The number and type of clinically significant alerts were similar irrespective of patient population, suggesting that the alerts may be equally as beneficial in the ICU population, despite the challenges in drug-related event adjudication. An opportunity exists to improve the performance of alerts in both settings, so quality improvement programs for measuring alert performance and making refinements is needed.

Detection of adverse drug reactions by medication antidote signals and comparison of their sensitivity with common methods of ADR detection

OBJECTIVE

To determine the PPVs of selected ten medication antidote signals in recognizing potential ADRs and comparison of their sensitivity with manual chart analysis, and voluntary reporting recognizing the same ADRs.

METHOD

The inpatient EMR database of internal medicine department was utilized for a period of one year, adult patients prescribed at least one of the ten signals, were included in the study, recipient patients of antidote signals were assessed for the occurrence of an ADR by Naranjo's tool of ADR evaluation. PPVs of each antidote signal were verified.

RESULT

PPV of Methylprednisolone and Phytonadione was 0.28, Metoclopramide and Potassium Chloride - 0.29, Dextrose 50%, Promethazine, Sodium Polystyrene and Loperamide - 0.30, Protamine and Acetylcysteine - 0.33. In comparison of confirmed ADRs of antidote signals with other methods, Dextrose 50%, Metoclopramide, Sodium Polystyrene, Potassium Chloride, Methylprednisolone and Promethazine seem to be extremely significant (P value > 0.0001), while ADRs of Phytonadione, Protamine, Acetylcysteine and Loperamide were insignificant.

CONCLUSION

Antidote medication signals have definitive discerning evaluation value of ADRs over routine methods of ADR detection with a high detection rate with a minimum cost; Their integration with hospital EMR database and routine patient safety surveillance enhances transparency, time-saving and facilitates ADR detection.

Computerised clinical decision support systems to improve medication safety in long-term care homes: a systematic review

OBJECTIVES

Computerised clinical decision support systems (CCDSS) are used to improve the quality of care in various healthcare settings. This systematic review evaluated the impact of CCDSS on improving medication safety in long-term care homes (LTC). Medication safety in older populations is an important health concern as inappropriate medication use can elevate the risk of potentially severe outcomes (ie, adverse drug reactions, ADR). With an increasing ageing population, greater use of LTC by the growing ageing population and increasing number of medication-related health issues in LTC, strategies to improve medication safety are essential.

METHODS

Databases searched included MEDLINE, EMBASE, Scopus and Cochrane Library. Three groups of keywords were combined: those relating to LTC, medication safety and CCDSS. One reviewer undertook screening and quality assessment.

RESULTS

Overall findings suggest that CCDSS in LTC improved the quality of prescribing decisions (ie, appropriate medication orders), detected ADR, triggered warning messages (ie, related to central nervous system side effects, drug-associated constipation, renal insufficiency) and reduced injury risk among older adults.

CONCLUSIONS

CCDSS have received little attention in LTC, as attested by the limited published literature. With an increasing ageing population, greater use of LTC by the ageing population and increased workload for health professionals, merely relying on physicians' judgement on medication safety would not be sufficient. CCDSS to improve medication safety and enhance the quality of prescribing decisions are essential. Analysis of review findings indicates that CCDSS are beneficial, effective and have potential to improve medication safety in LTC; however, the use of CCDSS in LTC is scarce. Careful assessment on the impact of CCDSS on medication safety and further modifications to existing CCDSS are recommended for wider acceptance. Due to scant evidence in the current literature, further research on implementation and effectiveness of CCDSS is required.

Automated Communication Tools and Computer-Based Medication Reconciliation to Decrease Hospital Discharge Medication Errors

This study sought to determine the effects of automated primary care physician (PCP) communication and patient safety tools, including computerized discharge medication reconciliation, on discharge medication errors and posthospitalization patient outcomes, using a pre-post quasi-experimental study design, in hospitalized medical patients with ≥2 comorbidities and ≥5 chronic medications, at a single center. The primary outcome was discharge medication errors, compared before and after rollout of these tools. Secondary outcomes were 30-day rehospitalization, emergency department visit, and PCP follow-up visit rates. This study found that discharge medication errors were lower post intervention (odds ratio = 0.57; 95% confidence interval = 0.44-0.74; P < .001). Clinically important errors, with the potential for serious or life-threatening harm, and 30-day patient outcomes were not significantly different between study periods. Thus, automated health system-based communication and patient safety tools, including computerized discharge medication reconciliation, decreased hospital discharge medication errors in medically complex patients.

Adverse drug events identified by triggers at a teaching hospital in Brazil

Background

Adverse drug events (ADEs) are one of the most frequent causes of patient harm resulting from medical interventions, especially among inpatients. This study aimed to evaluate the incidence of ADEs and characterise them in terms of degree of harm, medication implicated and patient symptoms, at a Brazilian university hospital.

Methods

This is a retrospective study of chart review. The method, developed by the Institute for Healthcare Improvement, uses triggers to identify possible ADEs. The study population comprised adult inpatients at least 15 years old. Obstetric patients and those hospitalised for less than 48 hours were excluded. Time spent in the intensive care unit was not considered for the purposes of this study. Patients were selected on the basis of simple random sampling of records of patients discharged from January to July 2008. The records selected were reviewed by a multidisciplinary team. The indicators of ADE incidence were patients with ADEs and ADE rate per 100 patients. Patients with and without ADE were compared in the bivariate analysis. To identify the drugs classes most often associated with events, the number of prescriptions of each class of drug was related to the number of events assigned to it.

Results

The 240 inpatients studied were of mean age 50.8 (SD = 20.0) years, and mostly male (63.8%). A total of 44 ADEs were identified in 35 patient records, with 14.6% of patients presenting ADE and a rate of 18.3% ADEs per 100 patients. The most frequent were skin rash and nausea and vomiting, but severe ADEs were also identified. In the bivariate analysis long hospital stay and use of 10 or more drugs were associated with the occurrence of ADEs (p-value < 0.01). The drug classes associated with the highest number of events were anti-infective.

Conclusion

About 1/6 of the hospitalized patients in a teaching hospital showed adverse events what is, by itself, cause for concern. Increased number of prescribed drugs and greater period of hospitalization appear to favour the occurrence of these events. In the future studies with higher number of patients may offer evidences of the association.

A survey of nursing home physicians to determine laboratory monitoring adverse drug event alert preferences

OBJECTIVE

We conducted a survey of nursing home physicians to learn about (1) the laboratory value thresholds that clinical event monitors should use to generate alerts about potential adverse drug events (ADEs); (2) the specific information to be included in the alerts; and (3) the communication modality that should be used for communicating them.

METHODS

Nursing home physician attendees of the 2010 Conference of AMDA: The Society for Post-Acute and Long-Term Care Medicine.

RESULTS

A total of 800 surveys were distributed; 565 completed surveys were returned and seven surveys were excluded due to inability to verify that the respondents were physicians (a 70% net valid response rate). Alerting threshold preferences were identified for eight laboratory tests. For example, the majority of respondents selected thresholds of ≥5.5 mEq/L for hyperkalemia (63%) and ≤3.5 without symptoms for hypokalemia (54%). The majority of surveyed physicians thought alerts should include the complete active medication list, current vital signs, previous value of the triggering lab, medication change in the past 30 days, and medication allergies. Most surveyed physicians felt the best way to communicate an ADE alert was by direct phone/voice communication (64%), followed by email to a mobile device (59%).

CONCLUSIONS

This survey of nursing home physicians suggests that the majority prefer alerting thresholds that would generally lead to fewer alerts than if widely accepted standardized laboratory ranges were used. It also suggests a subset of information items to include in alerts, and the physicians' preferred communication modalities. This information might improve the acceptance of clinical event monitoring systems to detect ADEs in the nursing home setting.

[Adverse drug events in hospital: pilot study with trigger tool]

OBJECTIVE

To estimate the frequency of and to characterize the adverse drug events at a terciary care hospital.

METHODS

A retrospective review was carried out of 128 medical records from a hospital in Rio de Janeiro in 2007, representing 2,092 patients. The instrument used was a list of triggers, such as antidotes, abnormal laboratory analysis results and sudden suspension of treatment, among others. A simple random sample of patients aged 15 and over was extracted. Oncologic and obstetric patients were excluded as were those hospitalized for less than 48 hours or in the emergency room. Social and demographic characteristics and those of the disease of patients who underwent adverse events were compared with those of patients who did not in order to test for differences between the groups.

RESULTS

Around 70.0% of the medical records assessed showed at least one trigger. Adverse drug events triggers had an overall positive predictive value of 14.4%. The incidence of adverse drug events was 26.6 per 100 patients and 15.6% patients suffered one or more event. The median length of stay for patients suffering an adverse drug event was 35.2 days as against 10.7 days for those who did not (p < 0.01). The pharmacological classes most commonly associated with an adverse drug event were related to the cardiovascular system, nervous system and alimentary tract and metabolism. The most common active substances associated with an adverse drug event were tramadol, dypirone, glibenclamide and furosemide. Over 80.0% of events provoked or contributed to temporary harm to the patient and required intervention and 6.0% may have contributed to the death of the patient. It was estimated that in the hospital, 131 events involving drowsiness or fainting 33 involving falls, and 33 episodes of hemorrhage related to adverse drug effects occur annually.

CONCLUSIONS

Almost one-sixth of in-patients (16,0%) suffered an adverse drug event. The instrument used may prove useful as a technique for monitoring and evaluating patient care results. Psycothropic therapy should be critically appraised given the frequency of associated events, such as excessive sedation, lethargy, and hypotension.

Adverse drug events with hyperkalaemia during inpatient stays: evaluation of an automated method for retrospective detection in hospital databases

BACKGROUND

Adverse drug reactions and adverse drug events (ADEs) are major public health issues. Many different prospective tools for the automated detection of ADEs in hospital databases have been developed and evaluated. The objective of the present study was to evaluate an automated method for the retrospective detection of ADEs with hyperkalaemia during inpatient stays.

METHODS

We used a set of complex detection rules to take account of the patient's clinical and biological context and the chronological relationship between the causes and the expected outcome. The dataset consisted of 3,444 inpatient stays in a French general hospital. An automated review was performed for all data and the results were compared with those of an expert chart review. The complex detection rules' analytical quality was evaluated for ADEs.

RESULTS

In terms of recall, 89.5% of ADEs with hyperkalaemia "with or without an abnormal symptom" were automatically identified (including all three serious ADEs). In terms of precision, 63.7% of the automatically identified ADEs with hyperkalaemia were true ADEs.

CONCLUSIONS

The use of context-sensitive rules appears to improve the automated detection of ADEs with hyperkalaemia. This type of tool may have an important role in pharmacoepidemiology via the routine analysis of large inter-hospital databases.

Medication safety and knowledge-based functions: a stepwise approach against information overload

AIMS

The aim was to improve medication safety in an emergency department (ED) by enhancing the integration and presentation of safety information for drug therapy.

METHODS

Based on an evaluation of safety of drug therapy issues in the ED and a review of computer-assisted intervention technologies we redesigned an electronic case sheet and implemented computer-assisted interventions into the routine work flow. We devised a four step system of alerts, and facilitated access to different levels of drug information. System use was analyzed over a period of 6 months. In addition, physicians answered a survey based on the technology acceptance model TAM2.

RESULTS

The new application was implemented in an informal manner to avoid work flow disruption. Log files demonstrated that step I, 'valid indication' was utilized for 3% of the recorded drugs and step II 'tooltip for well-known drug risks' for 48% of the drugs. In the questionnaire, the computer-assisted interventions were rated better than previous paper based measures (checklists, posters) with regard to usefulness, support of work and information quality.

CONCLUSION

A stepwise assisting intervention received positive user acceptance. Some intervention steps have been seldom used, others quite often. We think that we were able to avoid over-alerting and work flow intrusion in a critical ED environment.

Use of BioSense for Rapid Assessment of the Safety of Medical Countermeasures

Objective

To conduct an initial examination of the potential use of BioSense data to monitor and rapidly assess the safety of medical countermeasures (MCM) used for prevention or treatment of adverse health effects of biological, chemical, and radiation exposures during a public health emergency.

Introduction

BioSense is a national human health surveillance system for disease detection, monitoring, and situation awareness through near real-time access to existing electronic healthcare encounter information, including information from hospital emergency departments (EDs). MCM include antibiotics, antivirals, antidotes, antitoxins, vaccinations, nuclide-binding agents, and other medications. Although some MCM have been extensively evaluated and have FDA approval, many do not (1). Current FDA and CDC systems that monitor drug and vaccine safety have limited ability to monitor MCM safety, and in particular to conduct rapid assessments during an emergency (1).

Methods

To provide a preliminary assessment of the use of BioSense for this purpose, we reviewed selected publications evaluating the use of electronic health records (EHRs) to monitor safety of drugs and vaccinations (medications), focusing particularly on systematic reviews, reviewed BioSense data elements, and consulted with a number of subject matter experts.

Results

More than 40 studies have examined use of EHR data to monitor adverse effects (AEs) of medications using administrative, laboratory, and pharmacy records from inpatient- and out-patient settings, including EDs (2–4). To identify AEs, investigators have used diagnostic codes; administration of antidotes, laboratory measures of drug levels and of biologic response, text searches of unstructured clinical notes, and combinations of those data elements. BioSense ED data include chief complaint text, triage notes, text diagnosis, as well as diagnostic and medical procedure codes.

Investigations used a variety of study designs in various populations and settings; examined a wide range of medications, vaccinations, and AEs; and developed a diverse set of analytic algorithms to search EHR data to detect and signal AEs (2–4). Most research has been done on FDA-approved medications. Most studies used EHR data to identify individuals using specific medications and then searched for potential AEs identified from previous research. None of the studies investigated use of EHR data to monitor safety when records of an individual’s medication use could not be linked to that individual’s records of AEs. BioSense data could be used for AE detection, but linking AEs to MCM use would require follow-back investigation. Since there is limited research on AEs of some MCM, there would be limited information to guide identification of potential AEs.

Performance characteristics of the AE monitoring systems have been mixed with reported sensitivities ranging from 40–90%; specificities from 1% to 90%, and positive predictive values from < 1% to 64%, depending on the medication, AE and other characteristics of the study (2, 4). However, the small numbers of studies with common characteristics has limited the ability of reviewers to determine which types of systems have better performance for different medications and AEs.

Some experts suggest that data in BioSense, might contribute to safety surveillance of MCM. They also caution that poor predictive values and high rates of false positives reported in the literature raise concerns about burden to those conducting investigations in response to AE alerts, particularly in the context of a public health emergency.

Conclusions

These findings suggest that BioSense data could potentially contribute to rapid identification of safety issues for MCM and that some methods from published research could be applicable to the use of BioSense for this purpose. However, such use would require careful development and evaluation.

Unfractionated heparin versus low molecular weight heparin for avoiding heparin-induced thrombocytopenia in postoperative patients

BACKGROUND

Heparin-induced thrombocytopenia (HIT) is an adverse drug reaction presenting as a prothrombotic disorder related to antibody-mediated platelet activation. It is a poorly understood paradoxical immune reaction resulting in thrombin generation in vivo, which leads to a hypercoagulable state and the potential to initiate venous or arterial thrombosis. A number of factors are thought to influence the incidence of HIT including the type and preparation of heparin (unfractionated heparin (UFH) or low molecular weight heparin (LMWH)) and the heparin-exposed patient population, with the postoperative patient population presenting a higher risk.Although LMWH has largely replaced UFH as a front-line therapy, there is evidence supporting a lack of superiority of LMWH compared with UFH regarding prevention of deep vein thrombosis and pulmonary embolism following surgery, and similar frequencies of bleeding have been described with LMWH and UFH. The decision as to which of these two preparations of heparin to use may thus be influenced by adverse reactions such as HIT. We therefore sought to determine the relative impact of UFH and LMWH specifically on HIT in postoperative patients receiving thromboembolism prophylaxis.

OBJECTIVES

The objective of this review was to compare the incidence of HIT and HIT complicated by thrombosis in patients exposed to UFH versus LMWH in randomised controlled trials (RCTs) of postoperative heparin therapy.

SEARCH METHODS

The Cochrane Peripheral Vascular Diseases Group searched their Specialised Register (March 2012) and CENTRAL (2012, Issue 2). In addition, the authors searched LILACS (March 2012) and additional trials were sought from reference lists of relevant publications.

SELECTION CRITERIA

We were interested in comparing the incidence of HIT occurring during exposure to UFH or LMWH after any surgical intervention. Therefore, we studied RCTs in which participants were postoperative patients allocated to receive UFH or LMWH, in a blinded or unblinded fashion. Eligible studies were required to have as an outcome clinically diagnosed HIT, defined as a relative reduction in the platelet count of 50% or greater from the postoperative peak (even if the platelet count at its lowest remained > 150 x 10(9)/L) occurring within five to 14 days after the surgery, with or without a thrombotic event occurring in this timeframe. Additionally, circulating antibodies associated with the syndrome were required to have been investigated through laboratory assays.

DATA COLLECTION AND ANALYSIS

Two review authors independently extracted data and assessed the risk of bias. Disagreements were resolved by consensus with participation of a third author.

MAIN RESULTS

In total two studies involving 923 participants met all the inclusion criteria and were included in the review. Pooled analysis showed a statistically significant reduction in the risk of HIT with LMWH compared with UFH (risk ratio (RR) 0.24, 95% confidence interval (CI) 0.07 to 0.82; P = 0.02). This result suggests that patients treated with LMWH would have a relative risk reduction (RRR) of 76% in the probability of developing HIT compared with patients treated with UFH.Venous thromboembolism (VTE) complicating HIT occurred in 12 of 17 patients who developed HIT. Pooled analysis showed a statistically significant reduction in HIT complicated by VTE with LMWH compared with UFH (RR 0.20, 95% CI 0.04 to 0.90; P = 0.04). This result indicates that patients using LMWH would have a RRR of 80% for developing HIT complicated by VTE compared with patients using UFH. Arterial thrombosis occurred in only one patient who received UFH and there were no amputations or deaths documented.

AUTHORS' CONCLUSIONS

There was a lower incidence of HIT and HIT complicated by VTE in postoperative patients undergoing thromboprophylaxis with LMWH compared with UFH. This is consistent with the current clinical use of LMWH over UFH as front-line heparin therapy. However, conclusions are limited by a scarcity of high quality evidence. We did not expect the paucity of RCTs including HIT as an outcome as heparin is one of the most commonly used drugs worldwide and HIT is a life-threatening adverse drug reaction. To address the scarcity of clinically-relevant information on the topic of HIT as a whole, HIT should be included as an outcome in future RCTs of heparin, and HIT as an adverse drug reaction should be considered in clinical recommendations regarding monitoring of the platelet count for HIT.

Vigilância de eventos adversos a medicamentos em hospitais: aplicação e desempenho de rastreadores

Os eventos adversos a medicamentos (EAMs) são causa importante de comprometimento da qualidade da atenção ao paciente hospitalizado e, por isso, devem ser identificados e caracterizados. Para tanto surgiram listas de rastreadores, entre elas a proposta pelo Institute for Healthcare Improvement. Aqui é apresentado o processo da aplicação dos rastreadores e o seu desempenho em um hospital de ensino. As informações sobre os rastreadores e os EAM foram coletadas por meio de revisão retrospectiva dos prontuários de pacientes com alta hospitalar de janeiro a junho de 2008. Foram identificados 497 rastreadores em 177 prontuários, onde cada prontuário apresentou, em média, 2,33 (DP = 2,7) rastreadores. Os encontrados com mais frequência foram: "antiemético" (72,1/100 prontuários), "interrupção abrupta da medicação" (70,0/100 prontuários) e "sedação excessiva, sonolência, torpor, letargia, queda e hipotensão" (34,6/100 prontuários). Os mais eficientes na captura de EAM (rendimento), isto é, aqueles que uma vez identificados sinalizaram possíveis eventos foram "antagonista de benzodiazepínico", "antidiarréicos" e "rash cutâneo". Os EAM mais encontrados foram relacionados aos rastreadores "interrupção abrupta da medicação" (8,3/100 prontuários), "antiemético" (4,6/100 prontuários) e "rash cutâneo" (2,1/100 prontuários). Essas considerações apontam para a utilidade do emprego da lista de rastreadores e podem contribuir para decidir sobre ajustes na sua aplicação.

Use of an abnormal laboratory value-drug combination alert to detect drug-induced thrombocytopenia in critically Ill patients

PURPOSE

The aim of this study was to assess the performance of a commercially available clinical decision support system (CDSS) drug-laboratory result alert in detecting drug-induced thrombocytopenia in critically ill patients.

MATERIALS AND METHODS

Adult patients admitted to the medical and cardiac intensive care unit during an 8-week period and identified by 1 of 3 signals in the CDSS, TheraDoc, were eligible. Alerts were generated when the patient had a low platelet count and was ordered a potentially causal drug. Patients were evaluated in real time for the occurrence of an adverse drug reaction using 3 causality instruments. Positive predictive values were calculated for the alert.

RESULTS

Sixty-four patients with a mean age of 54 years met the inclusion criteria, generating 350 alerts. Positive predictive values were 0.36, 0.83, and 0.40 for signals 1, 2, and 3, respectively. Overall, there were 137 adverse drug reactions identified in the 350 alerts, with heparin, vancomycin, and famotidine as the 3 most common potential causes.

CONCLUSIONS

A commercial CDSS drug-laboratory alert is effective at identifying drug-induced thrombocytopenia in the intensive care unit and may improve patient safety. Compared with previous studies, the combination alert performs better than alerts based exclusively on laboratory values and should be considered to reduce alert fatigue.

A systematic review to evaluate the accuracy of electronic adverse drug event detection

OBJECTIVE

Adverse drug events (ADEs), defined as adverse patient outcomes caused by medications, are common and difficult to detect. Electronic detection of ADEs is a promising method to identify ADEs. We performed this systematic review to characterize established electronic detection systems and their accuracy.

METHODS

We identified studies evaluating electronic ADE detection from the MEDLINE and EMBASE databases. We included studies if they contained original data and involved detection of electronic triggers using information systems. We abstracted data regarding rule characteristics including type, accuracy, and rationale.

RESULTS

Forty-eight studies met our inclusion criteria. Twenty-four (50%) studies reported rule accuracy but only 9 (18.8%) utilized a proper gold standard (chart review in all patients). Rule accuracy was variable and often poor (range of sensitivity: 40%-94%; specificity: 1.4%-89.8%; positive predictive value: 0.9%-64%). 5 (10.4%) studies derived or used detection rules that were defined by clinical need or the underlying ADE prevalence. Detection rules in 8 (16.7%) studies detected specific types of ADEs.

CONCLUSION

Several factors led to inaccurate ADE detection algorithms, including immature underlying information systems, non-standard event definitions, and variable methods for detection rule validation. Few ADE detection algorithms considered clinical priorities. To enhance the utility of electronic detection systems, there is a need to systematically address these factors.

Predictive value of alert triggers for identification of developing adverse drug events

OBJECTIVE

Adverse drug events (ADEs) are the most common type of iatrogenic injury in hospitalized patients. However; the ability of electronic triggers to identify patients at high risk for inpatient ADEs before they occur has not been well studied. The objective of this study was to assess the positive predictive value of event triggers to detect developing ADEs.

METHODS

We conducted a prospective observational study in patients at a university-based teaching hospital during a 5-month period. Patients were monitored using electronic triggers designed to detect patients at increased risk for 4 types of ADEs: hypoglycemia, hypokalemia, hyperkalemia, and thrombocytopenia. Each patient for whom a trigger fired was followed to determine whether a drug-induced markedly abnormal laboratory result occurred between 1 and 72 hours after the initial trigger firing.

RESULTS

Overall, the triggers fired 611 times on 456 patients. Of the 456 patients, 101 experienced 1 or more related ADEs between 1 and 72 hours after the initial trigger firing. The positive predictive value of the triggers and median time from trigger firing to ADE was 31% and 11.6 hours for hypoglycemia, 4.0% and 17 hours for hypokalemia, 31% and 25.4 hours for hyperkalemia, and 21% and 48.4 hours for thrombocytopenia.

CONCLUSION

Computerized triggers have sufficient predictive value to detect developing ADEs and can help clinicians avert ADEs. More research is required to determine whether real-time, primary-prevention alerts may reduce the incidence of ADEs.

Comparison of a basic and an advanced pharmacotherapy-related clinical decision support system in a hospital care setting in the Netherlands

OBJECTIVE To compare the clinical relevance of medication alerts in a basic and in an advanced clinical decision support system (CDSS).

DESIGN

A prospective observational study.

MATERIALS AND METHODS

We collected 4023 medication orders in a hospital for independent evaluation in two pharmacotherapy-related decision support systems. Only the more advanced system considered patient characteristics and laboratory test results in its algorithms. Two pharmacists assessed the clinical relevance of the medication alerts produced. The alert was considered relevant if the pharmacist would undertake action (eg, contact the physician or the nurse). The primary analysis concerned the positive predictive value (PPV) for clinically relevant medication alerts in both systems.

RESULTS

The PPV was significantly higher in the advanced system (5.8% vs 17.0%; p<0.05). Significant differences were found in the alert categories: drug-(drug) interaction (9.9% vs 14.8%; p<0.05), drug-age interaction (2.9% vs 73.3%; p<0.05), and dosing guidance (5.6% vs 16.9%; p<0.05). Including laboratory values and other patient characteristics resulted in a significantly higher PPV for the advanced CDSS compared to the basic medication alerts (12.2% vs 23.3%; p<0.05).

CONCLUSION

The advanced CDSS produced a higher proportion of clinically relevant medication alerts, but the number of irrelevant alerts remained high. To improve the PPV of the advanced CDSS, the algorithms should be optimized by identifying additional risk modifiers and more data should be made electronically available to improve the performance of the algorithms. Our study illustrates and corroborates the need for cyclic testing of technical improvements in information technology in circumstances representative of daily clinical practice.

Data mining to generate adverse drug events detection rules

Adverse drug events (ADEs) are a public health issue. Their detection usually relies on voluntary reporting or medical chart reviews. The objective of this paper is to automatically detect cases of ADEs by data mining. 115,447 complete past hospital stays are extracted from six French, Danish, and Bulgarian hospitals using a common data model including diagnoses, drug administrations, laboratory results, and free-text records. Different kinds of outcomes are traced, and supervised rule induction methods (decision trees and association rules) are used to discover ADE detection rules, with respect to time constraints. The rules are then filtered, validated, and reorganized by a committee of experts. The rules are described in a rule repository, and several statistics are automatically computed in every medical department, such as the confidence, relative risk, and median delay of outcome appearance. 236 validated ADE-detection rules are discovered; they enable to detect 27 different kinds of outcomes. The rules use a various number of conditions related to laboratory results, diseases, drug administration, and demographics. Some rules involve innovative conditions, such as drug discontinuations.

Computerized detection of adverse drug reactions in the medical intensive care unit

OBJECTIVE

Clinical event monitors are a type of active medication monitoring system that can use signals to alert clinicians to possible adverse drug reactions. The primary goal was to evaluate the positive predictive values of select signals used to automate the detection of ADRs in the medical intensive care unit.

METHOD

This is a prospective, case series of adult patients in the medical intensive care unit during a six-week period who had one of five signals presents: an elevated blood urea nitrogen, vancomycin, or quinidine concentration, or a low sodium or glucose concentration. Alerts were assessed using 3 objective published adverse drug reaction determination instruments. An event was considered an adverse drug reaction when 2 out of 3 instruments had agreement of possible, probable or definite. Positive predictive values were calculated as the proportion of alerts that occurred, divided by the number of times that alerts occurred and adverse drug reactions were confirmed.

RESULTS

145 patients were eligible for evaluation. For the 48 patients (50% male) having an alert, the mean±SD age was 62±19 years. A total of 253 alerts were generated. Positive predictive values were 1.0, 0.55, 0.38 and 0.33 for vancomycin, glucose, sodium, and blood urea nitrogen, respectively. A quinidine alert was not generated during the evaluation.

CONCLUSIONS

Computerized clinical event monitoring systems should be considered when developing methods to detect adverse drug reactions as part of intensive care unit patient safety surveillance systems, since they can automate the detection of these events using signals that have good performance characteristics by processing commonly available laboratory and medication information.

A review on systematic reviews of health information system studies

The purpose of this review is to consolidate existing evidence from published systematic reviews on health information system (HIS) evaluation studies to inform HIS practice and research. Fifty reviews published during 1994-2008 were selected for meta-level synthesis. These reviews covered five areas: medication management, preventive care, health conditions, data quality, and care process/outcome. After reconciliation for duplicates, 1276 HIS studies were arrived at as the non-overlapping corpus. On the basis of a subset of 287 controlled HIS studies, there is some evidence for improved quality of care, but in varying degrees across topic areas. For instance, 31/43 (72%) controlled HIS studies had positive results using preventive care reminders, mostly through guideline adherence such as immunization and health screening. Key factors that influence HIS success included having in-house systems, developers as users, integrated decision support and benchmark practices, and addressing such contextual issues as provider knowledge and perception, incentives, and legislation/policy.

Impact of health information technology interventions to improve medication laboratory monitoring for ambulatory patients: a systematic review

Medication errors are a major source of morbidity and mortality. Inadequate laboratory monitoring of high-risk medications after initial prescription is a medical error that contributes to preventable adverse drug events. Health information technology (HIT)-based clinical decision support may improve patient safety by improving the laboratory monitoring of high-risk medications, but the effectiveness of such interventions is unclear. Therefore, the authors conducted a systematic review to identify studies that evaluate the independent effect of HIT interventions on improving laboratory monitoring for high-risk medications in the ambulatory setting using a Medline search from January 1, 1980 through January 1, 2009 and a manual review of relevant bibliographies. All anticoagulation monitoring studies were excluded. Eight articles met the inclusion criteria, including six randomized controlled trials and two pre-post intervention studies. Six of the studies were conducted in two large, integrated healthcare delivery systems in the USA. Overall, five of the eight studies reported statistically significant, but small, improvements in laboratory monitoring; only half of the randomized controlled trials reported statistically significant improvements. Studies that found no improvement were more likely to have used analytic strategies that addressed clustering and confounding. Whether HIT improves laboratory monitoring of certain high-risk medications for ambulatory patients remains unclear, and further research is needed to clarify this important question.

Do geriatric conditions increase risk of adverse drug reactions in ambulatory elders? Results from the VA GEM Drug Study

BACKGROUND

Many clinicians prescribe cautiously to older adults with common geriatric conditions for fear of causing adverse drug reactions (ADRs). However, little is known about the association between these conditions and risk of ADRs.

METHODS

Using data from the VA Geriatric Evaluation and Management Drug Study, we determined any, preventable, and serious ADRs in 808 elders for 12 months after hospital discharge using a validated process involving patient self-report and chart review adjudicated by two health care professionals. Eight common geriatric conditions (activities of daily living, dementia, incontinence, falls, difficulty ambulating, malnourishment, depression, and prolonged bed rest) were evaluated at study baseline through self-report and structured assessments. We used Poisson regression to model the relationship between these geriatric conditions and ADRs.

RESULTS

Participants had a mean of 2.9 ± 1.2 geriatric conditions. Over the 12-month follow-up period, 497 ADRs occurred in 269 participants, including 187 ADRs considered preventable and 127 considered severe. On multivariable analyses, participants with dependency in one or more activities of daily living were less likely to suffer ADRs than those who were fully independent (incidence rate ratio: 0.78, 95% confidence interval = 0.62-1.00). None of the other seven geriatric conditions assessed were associated with ADR risk. Results were similar for preventable and serious ADRs, although participants with a history of falls were more likely to develop serious ADRs (incidence rate ratio: 1.49, 95% confidence interval = 1.00-2.21).

CONCLUSIONS

Many geriatric conditions were not associated with risk of ADRs. Although it is prudent to prescribe judiciously in patients with these conditions, excessive caution may not be warranted.

Culture counts--sustainable inpatient computerized surveillance across Duke University Health System

PURPOSE

The authors report on the managerial and logistical details of deploying a computerized adverse drug event surveillance system that was at first a grant-funded research project and ultimately was changed to a sustained safety-monitoring application serving 3 different hospitals.

METHODS

Surveillance was deployed in 3 phases to 2 community-based hospitals and an academic medical center. A logic-based rules engine surveyed electronic records for laboratory, medication, and demographic information indicative of safety concerns. Potential adverse events triggered manual chart review by pharmacists to verify patient harm.

RESULTS

During Phase 1, the research team created trigger rules for each hospital. In Phase 2, the trigger review was transitioned to hospital personnel and rule sets were reshaped for specific hospital needs. In Phase 3, surveillance was integrated into daily work flows and organizational balanced scorecards where it was accepted as a quantitative measure of medication safety performance.

DISCUSSION AND CONCLUSION

Computerized surveillance helps detect potentially harmful events regardless of hospital size. Active leadership, change-tolerant culture, and hospital pharmacy practice models significantly impact successful adoption. Entrenched cultural issues impeded sustainability at the academic center but not at the 2 community hospitals. Tailoring surveillance to the needs of different inpatient settings is crucial to developing a sustainable model.

The impact of eHealth on the quality and safety of health care: a systematic overview

BACKGROUND

There is considerable international interest in exploiting the potential of digital solutions to enhance the quality and safety of health care. Implementations of transformative eHealth technologies are underway globally, often at very considerable cost. In order to assess the impact of eHealth solutions on the quality and safety of health care, and to inform policy decisions on eHealth deployments, we undertook a systematic review of systematic reviews assessing the effectiveness and consequences of various eHealth technologies on the quality and safety of care.

METHODS AND FINDINGS

We developed novel search strategies, conceptual maps of health care quality, safety, and eHealth interventions, and then systematically identified, scrutinised, and synthesised the systematic review literature. Major biomedical databases were searched to identify systematic reviews published between 1997 and 2010. Related theoretical, methodological, and technical material was also reviewed. We identified 53 systematic reviews that focused on assessing the impact of eHealth interventions on the quality and/or safety of health care and 55 supplementary systematic reviews providing relevant supportive information. This systematic review literature was found to be generally of substandard quality with regards to methodology, reporting, and utility. We thematically categorised eHealth technologies into three main areas: (1) storing, managing, and transmission of data; (2) clinical decision support; and (3) facilitating care from a distance. We found that despite support from policymakers, there was relatively little empirical evidence to substantiate many of the claims made in relation to these technologies. Whether the success of those relatively few solutions identified to improve quality and safety would continue if these were deployed beyond the contexts in which they were originally developed, has yet to be established. Importantly, best practice guidelines in effective development and deployment strategies are lacking.

CONCLUSIONS

There is a large gap between the postulated and empirically demonstrated benefits of eHealth technologies. In addition, there is a lack of robust research on the risks of implementing these technologies and their cost-effectiveness has yet to be demonstrated, despite being frequently promoted by policymakers and "techno-enthusiasts" as if this was a given. In the light of the paucity of evidence in relation to improvements in patient outcomes, as well as the lack of evidence on their cost-effectiveness, it is vital that future eHealth technologies are evaluated against a comprehensive set of measures, ideally throughout all stages of the technology's life cycle. Such evaluation should be characterised by careful attention to socio-technical factors to maximise the likelihood of successful implementation and adoption.

Testing a trigger tool as a method of detecting harm from medication errors in a UK hospital: a pilot study

Objectives

The aim was to adapt a US adverse drug event (ADE) trigger tool for UK use, and to establish its positive predictive value (PPV) and sensitivity in comparison to retrospective health record review for the identification of preventable ADEs, in a pilot study on one hospital ward.

Methods

An established US trigger tool was adapted for UK use. We applied it retrospectively to 207 patients' health records, following up positive triggers to identify any ADEs (both preventable and non-preventable). We compared the preventable ADEs to those identified using full health record review.

Key findings

We identified 168 positive triggers in 127 (61%) of 207 patients. Seven ADEs were identified, representing an ADE in 3.4% of patients or 0.7 ADEs per 100 patient days. Five were non-preventable adverse drug reactions and two were due to preventable errors. The prevalence of preventable ADEs was 1.0% of patients, or 0.2 per 100 patient days. The overall PPV was 0.04 for all ADEs, and 0.01 for preventable ADEs. PPVs for individual triggers varied widely. Five preventable ADEs were identified using health record review. The sensitivity of the trigger tool for identifying preventable ADEs was 0.40, when compared to health record review.

Conclusions

Although we identified some ADEs using the trigger tool, more work is needed to further refine the trigger tool to reduce the false positives and increase sensitivity. To comprehensively identify preventable ADEs, retrospective health record review remains the gold standard and we found no efficiency gain in using the trigger tool.

Developing a patient safety surveillance system to identify adverse events in the intensive care unit

Aggregation of adverse drug event data has evolved in the last decade. Several approaches are available to augment the standard voluntary incident reporting system. Most of these methods are applicable to nonmedication adverse events as well. To identify appropriately system trends as well as process failures, intensive care units should participate in various collection methods. Several different methods are available for robust adverse drug event data collection, such as target chart review, nontargeted chart review, and direct observation. As the various methods usually capture different types of events, employing more than one technique will improve the assessment of intensive care unit care. Some of these surveillance methods offer real-time or near real-time identification of adverse drug events and potentially afford the practitioner time for intervention. Continued development of adverse drug event detection will allow for further quality improvement efforts and preventive strategies to be utilized.

Evaluating the positive predictive values of antidote signals to detect potential adverse drug reactions (ADRs) in the medical intensive care unit (ICU)

PURPOSE

Signals are used to alert clinicians of potential ADRs. Positive predictive values (PPVs) of antidote signals in ICUs are unknown. The primary purpose was to determine PPVs of six signals. The secondary objective was to determine the sensitivity of various ADR detection strategies including manual chart review, administrative data review, and voluntary reporting at identifying the same ADRs discovered using antidotes as a signal.

METHODS

Adult patients admitted to a medical ICU from July 1, 2005 to June 30, 2006 who were prescribed select signals were eligible. Evaluated antidote signals included injectable diphenhydramine, protamine, phytonadione, dextrose 50%, injectable methylprednisolone, and sodium polystyrene. For each signal, a random sample of 50 patients was evaluated for the presence of an ADR. ADR occurrences were determined using two objective causality instruments through retrospective chart review. Agreement between the instruments was required for ADR consideration. PPVs were determined for each signal.

RESULTS

Two hundred and twenty three patients (52% male) were evaluated, with a mean +/- SD age of 60 +/- 17 years, and a median simplified acute physiology score (SAPSII) of 48. PPVs were 0.64, 0.50, 0.38, 0.26, 0.24, and 0.02 for protamine, sodium polystyrene, dextrose 50%, diphenhydramine, phytonadione, and methylprednisolone, respectively. Sensitivity of other detection strategies from highest to lowest was chart review for explicit documentation, administrative database review, and voluntary reporting.

CONCLUSIONS

Protamine and sodium polystyrene performed the best by detecting ADRs in at least one out of two evaluations. Detection strategies other than signals were not as sensitive at identifying the same ADRs as antidote signals.