Electronic Healthcare Data Research Articles

Use of Patient Health Records to Quantify Drug-Related Pro-arrhythmic Risk

SummaryThere is an increasing expectation that computational approaches may supplement existing human decision-making. Frontloading of models for cardiac safety prediction is no exception to this trend, and ongoing regulatory initiatives propose use of high-throughput in vitro data combined with computational models for calculating proarrhythmic risk. Evaluation of these models requires robust assessment of the outcomes. Using FDA Adverse Event Reporting System reports and electronic healthcare claims data from the Truven-MarketScan US claims database, we quantify the incidence rate of arrhythmia in patients and how this changes depending on patient characteristics. First, we propose that such datasets are a complementary resource for determining relative drug risk and assessing the performance of cardiac safety models for regulatory use. Second, the results suggest important determinants for appropriate stratification of patients and evaluation of additional drug risk in prescribing and clinical support algorithms and for precision health.

Use of telephone helpline data for syndromic surveillance of adverse events following immunization in Australia: A retrospective study, 2009 to 2017

BackgroundThe increasing availability of electronic healthcare data offers an opportunity to enhance adverse events following immunisation (AEFI) signal monitoring in near real-time. AimTo evaluate the potential use of telephone helpline data to augment the existing AEFI surveillance system in Victoria, Australia. MethodsAnonymised telephone helpline call data were extracted between 2009 and 2017. For comparison, we included AEFI reports to the Victorian enhanced passive surveillance system, SAEFVIC-“Surveillance of Adverse Events Following Vaccination In the Community”. The temporal pattern cross-correlation coefficient at different time lags was estimated as a measure of timeliness evaluation. Historically known AEFI signals in 2010 and 2015 were examined using the Farrington statistical signal detection algorithm. ResultDuring the study period, overall, the telephone helpline centre received 2,005,226 calls. Of these, 0.68% (13,719) were AEFI-related. In the same period, SAEFVIC received 10,367 AEFI related reports. Cross-correlation analysis, generally, showed that the two datasets were moderately correlated (r = 0.4) at a negative lag of 1 week. For individual years, the cross-correlation coefficient was highest (r = 0.66) in 2010 with the telephone helpline data leading by 2 weeks. Our analysis indicated the 2010 reported incidence of febrile convulsions and the 2015 reported increased allergic-related reactions following seasonal influenza vaccination three weeks and one week earlier respectively. ConclusionTelephone helpline data was able to detect an increased rate of AEFI earlier than the enhanced passive AEFI surveillance system. This dataset offers a valuable and near real-time component of an integrated AEFI early signal detection system in Australia.

Validation of a type 1 diabetes algorithm using electronic medical records and administrative healthcare data to study the population incidence and prevalence of type 1 diabetes in Ontario, Canada

IntroductionWe aimed to develop algorithms distinguishing type 1 diabetes (T1D) from type 2 diabetes in adults ≥18 years old using primary care electronic medical record (EMRPC) and administrative healthcare data...

Psychotropic Medication After Intensive Care Unit–Treated Pediatric Traumatic Brain Injury

BackgroundOur aim was to assess the occurrence and risk factors for psychotropic medication use after pediatric traumatic brain injury treated in the intensive care unit. MethodsWe combined data from the Finnish Intensive Care Consortium database, data on reimbursed medications from the Social Insurance Institute, and individual electronic health care data. We analyzed data on children aged five to 17 years treated for traumatic brain injury in intensive care units of four university hospitals in Finland during 2003 to 2013 and being alive six months after injury with no history of psychotropic medication use before traumatic brain injury. ResultsWe identified 248 patients of whom 46 (19%) were prescribed a new psychotropic medication after traumatic brain injury. In multivariable logistic regression, a higher age associated with a higher probability for use of any psychotropic medication. Subgroup analyses showed that higher age associated with an increased risk of antidepressant and antipsychotic use but with a decreased risk of stimulant use. Apart from age, we found no other clinical, radiological, or treatment-related factors that significantly associated with subsequent use of psychotropics. Psychotropic medication was most common (45%) in children aged 12 to 17 years and had moderate disability at six-month follow-up. ConclusionsOne fifth of children treated in the intensive care unit for traumatic brain injury were prescribed a new psychotropic medication during a median follow-up of three years and five months. Psychotropic medication was most common among teenagers with moderate post-traumatic disability. The need and use of psychotropics postinjury seem multifactorial and not related to any traumatic brain injury type.

COVID-19 challenges to Pakistan: Is GIS analysis useful to draw solutions?

The outbreak of novel coronavirus (COVID-19) is a public health emergency that had caused disastrous results in more than 100 countries. The ability to detect disease outbreaks in the early stages is a key component of efficient disease control and prevention. With the increased availability of electronic health-care data and spatial analysis techniques, there is great potential to develop algorithms to enable more effective disease surveillance. The research focuses to develop a transparent user-friendly method to simulate the outbreak data. The paper describes the GIS tools to identify and define the field of investigation which requires consideration of the strengths and limitations of data collection instruments, facility of locational data collection, accuracy of locational data, and pertinent attributes for understanding disease risk. Using such information, it is quite easy for authorities to locate the highly effected area and take appropriate actions in that particular areas. However, GIS techniques, resources, and methods can be used in Pakistan for more effective investigation of vulnerable geographical locations.

Analysis of antiseizure drug‐related adverse reactions from the electronic health record using the common data model

Antiseizure drugs (ASDs) are known to cause a wide range of adverse drug reactions (ADRs). Recently, electronic health care data using the common data model (CDM) have been introduced and commonly adopted in pharmacovigilance research. We aimed to analyze ASD-related ADRs using CDM and to assess the feasibility of CDM analysis in monitoring ADR in a single tertiary hospital. We selected five ASDs: oxcarbazepine (OXC), lamotrigine (LTG), levetiracetam (LEV), valproic acid (VPA), and topiramate (TPM). Patients diagnosed with epilepsy and exposed to monotherapy with one of the ASDs before age 18years were included. We measured four ADR outcomes: (1) hematologic abnormality, (2) hyponatremia, (3) elevation of liver enzymes, and (4) subclinical hypothyroidism. We performed a subgroup analysis to exclude the effects of concomitant medications. From the database, 1344 patients were included for the study. Of the 1344 patients, 436 were receiving OXC, 293 were receiving LTG, 275 were receiving LEV, 180 were receiving VPA, and 160 were receiving TPM. Thrombocytopenia developed in 14.1% of patients taking VPA. Hyponatremia occurred in 10.5% of patients taking OXC. Variable ranges of liver enzyme elevation were detected in 19.3% of patients taking VPA. Subclinical hypothyroidism occurred in approximately 21.5% to 28% of patients with ASD monotherapy, which did not significantly differ according to the type of ASD. In a subgroup analysis, we observed similar ADR tendencies, but with less thrombocytopenia in the TPM group. The incidence and trends of ADRs that were evaluated by CDM were similar to the previous literature. CDM can be a useful tool for analyzing ASD-related ADRs in a multicenter study. The strengths and limitations of CDM should be carefully addressed.

Developing an FHIR-Based Computational Pipeline for Automatic Population of Case Report Forms for Colorectal Cancer Clinical Trials Using Electronic Health Records.

PURPOSEThe Fast Healthcare Interoperability Resources (FHIR) is emerging as a next-generation standards framework developed by HL7 for exchanging electronic health care data. The modeling capability of FHIR in standardizing cancer data has been gaining increasing attention by the cancer research informatics community. However, few studies have been conducted to examine the capability of FHIR in electronic data capture (EDC) applications for effective cancer clinical trials. The objective of this study was to design, develop, and evaluate an FHIR-based method that enables the automation of the case report forms (CRFs) population for cancer clinical trials using real-world electronic health records (EHRs).MATERIALS AND METHODSWe developed an FHIR-based computational pipeline of EDC with a case study for modeling colorectal cancer trials. We first leveraged an existing FHIR-based cancer profile to represent EHR data of patients with colorectal cancer, and then we used the FHIR Questionnaire and QuestionnaireResponse resources to represent the CRFs and their data population. To test the accuracy of and overall quality of the computational pipeline, we used synoptic reports of 287 Mayo Clinic patients with colorectal cancer from 2013 to 2019 with standard measures of precision, recall, and F1 score.RESULTSUsing the computational pipeline, a total of 1,037 synoptic reports were successfully converted as the instances of the FHIR-based cancer profile. The average accuracy for converting all data elements (excluding tumor perforation) of the cancer profile was 0.99, using 200 randomly selected records. The average F1 score for populating nine questions of the CRFs in a real-world colorectal cancer trial was 0.95, using 100 randomly selected records.CONCLUSIONWe demonstrated that it is feasible to populate CRFs with EHR data in an automated manner with satisfactory performance. The outcome of the study provides helpful insight into future directions in implementing FHIR-based EDC applications for modern cancer clinical trials.

Natural language processing for structuring clinical text data on depression using UK-CRIS.

Utilisation of routinely collected electronic health records from secondary care offers unprecedented possibilities for medical science research but can also present difficulties. One key issue is that medical information is presented as free-form text and, therefore, requires time commitment from clinicians to manually extract salient information. Natural language processing (NLP) methods can be used to automatically extract clinically relevant information. Our aim is to use natural language processing (NLP) to capture real-world data on individuals with depression from the Clinical Record Interactive Search (CRIS) clinical text to foster the use of electronic healthcare data in mental health research. We used a combination of methods to extract salient information from electronic health records. First, clinical experts define the information of interest and subsequently build the training and testing corpora for statistical models. Second, we built and fine-tuned the statistical models using active learning procedures. Results show a high degree of accuracy in the extraction of drug-related information. Contrastingly, a much lower degree of accuracy is demonstrated in relation to auxiliary variables. In combination with state-of-the-art active learning paradigms, the performance of the model increases considerably. This study illustrates the feasibility of using the natural language processing models and proposes a research pipeline to be used for accurately extracting information from electronic health records. Real-world, individual patient data are an invaluable source of information, which can be used to better personalise treatment.

Securing and authenticating healthcare records through blockchain technology

Healthcare data is important in making critical policy decisions, patients care, and medical diagnostics to name a few. Due to the importance and market demand, healthcare data is also vulnerable to cyber attacks. The centralized record keeping systems expose a single node for the attackers to attack. A decentralized system is computationally expensive but has the ability to be revolutionary by keeping the patient at the core and providing security, transparency, privacy, and interoperability of the electronic healthcare data. A blockchain is such an implementation of a distributed and decentralized system using reliable cryptographic algorithms. This paper proposes a secure blockchain based architecture tailored specifically to cater to the needs of e-healthcare systems.

Risk of herpes zoster after exposure to varicella to explore the exogenous boosting hypothesis: self controlled case series study using UK electronic healthcare data

ObjectiveTo assess the magnitude and duration of any hypothesised protective effect of household exposure to a child with varicella on the relative incidence of herpes zoster in adults.DesignSelf controlled case...

Implementing healthcare services on a large scale: Challenges and remedies based on blockchain technology

The accessibility of electronic healthcare data is necessary for effective treatment, policy decisions, and healthcare information exchange. Due to the intangibility of digital data, healthcare information is also prone to privacy-breach and security attacks. Further, the importance of immutability and privacy of healthcare data becomes colossal when a nationwide healthcare and wellness scheme is planned to be implemented. Providing quality healthcare services to such an enormous population size is challenging and requires proper technological infrastructure. The cooperation from the society is equally important to lay such a copious architecture on which the healthcare services should seamlessly run. ObjectiveTo assess the social and technical challenges that lie ahead in implementing large-scale comprehensive healthcare services and suggest a technology-intervening solution to serve the society at large. MethodThis study considers India's National Health Policy (2017) initiatives. The social and technical hurdles in implementation of the schemes are discussed, and AarogyaChain, a Blockchain technology-based solution is proposed to eliminate the health policy implementation hiccups. ResultWe find that the scalability is a primary concern in implementing healthcare services on blockchain at such a large scale. We experimented by creating a blockchain and found that the system throughput is a function of the number of special nodes called ordering nodes, and a trade-off is required to balance between time-to-commit and system's fault tolerance. ConclusionBlockchain provides a secure and transparent system of integrated healthcare services that keeps patients at the center and provides for corruption intolerant and efficient implementation of nationwide health-insurance programs.

CRcoder: An Interactive Web Application and SAS Macro to Support Personalized Clinical Decisions.

Electronic health care data offer an opportunity to improve clinical decision making through advanced statistical analyses of longitudinal observations. To describe a Web application and SAS/STAT macro (SAS Institute Inc, Cary, NC) for computing joint models to estimate the typical and personalized risk of 2 concurrent binary outcomes. Features of the Web application design include uploading longitudinal files formatted with constant or time-varying covariates, specification of 2 binary outcomes, specification of a propensity model for treatment, and joint and separate models of the outcomes. In addition we designed an SAS macro for conducting the analysis. Fitting of joint and separate statistical models was implemented using a model specified in the Web application, with subsequent processing by the SAS macro. To illustrate the fitting of models, a sample of older adults with comorbid hypertension and chronic obstructive pulmonary disease from the Medical Expenditure Panel Survey was created to examine the association between polypharmacy (use of ≥ 5 medication classes) and limitations in social activities and mobility. Relative to separate models, the joint models typically estimated attenuated associations between explanatory variables and the 2 outcomes with smaller standard errors. These joint models yielded estimates of personalized concurrent risk and typical concurrent risk. Clinical decision making based on electronic health data can be improved using joint modeling to generate an individual's probability of concurrent risk. This user-friendly software performs the advanced statistical analyses needed to estimate typical and personalized concurrent risks.

Comparison of Self-controlled Designs for Evaluating Outcomes of Drug–Drug Interactions

Self-controlled designs, both case-crossover and self-controlled case series, are well suited for evaluating outcomes of drug-drug interactions in electronic healthcare data. Their comparative performance in this context, however, is unknown. We simulated cohorts of patients exposed to two drugs: a chronic drug (object) and a short-term drug (precipitant) with an associated interaction of 2.0 on the odds ratio scale. We analyzed cohorts using case-crossover and self-controlled case series designs evaluating exposure to the precipitant drug within person-time exposed to the object drug. Scenarios evaluated violations of key design assumptions: (1) time-varying, within-person confounding; (2) time trend in precipitant drug exposure prevalence; (3) nontransient precipitant exposure; and (4) event-dependent object drug discontinuation. Case-crossover analysis produced biased estimates when 30% of patients persisted on the precipitant drug (estimated OR 2.85) and when the use of the precipitant drug was increasing in simulated cohorts (estimated OR 2.56). Self-controlled case series produced biased estimates when patients discontinued the object drug following the occurrence of an outcome (estimated incidence ratio [IR] of 2.09 [50% of patients stopping therapy] and 2.22 [90%]). Both designs yielded similarly biased estimates in the presence of time-varying, within-person confounding. In settings with independent or rare outcomes and no substantial event-dependent censoring (<50%), self-controlled case series may be preferable to case-crossover design for evaluating outcomes of drug-drug interactions. With frequent event-dependent drug discontinuation, a case-crossover design may be preferable provided there are no time-related trends in drug exposure.

ADVANCE system testing: Can safety studies be conducted using electronic healthcare data? An example using pertussis vaccination

IntroductionThe Accelerated Development of Vaccine benefit-risk Collaboration in Europe (ADVANCE) public-private collaboration, aimed to develop and test a system for rapid benefit-risk monitoring of vaccines using healthcare databases in Europe. The objective of this proof-of-concept (POC) study was to test the feasibility of the ADVANCE system to generate incidence rates (IRs) per 1000 person-years and incidence rate ratios (IRRs) for risks associated with whole cell- (wP) and acellular- (aP) pertussis vaccines, occurring in event-specific risk windows in children prior to their pre-school-entry booster. MethodsThe study population comprised almost 5.1 million children aged 1 month to <6 years vaccinated with wP or aP vaccines during the study period from 1 January 1990 to 31 December 2015. Data from two Danish hospital (H) databases (AUH and SSI) and five primary care (PC) databases from, UK (THIN and RCGP RSC), Spain (SIDIAP and BIFAP) and Italy (Pedianet) were analysed. Database-specific IRRs between risk vs. non-risk periods were estimated in a self-controlled case series study and pooled using random-effects meta-analyses. ResultsThe overall IRs were: fever, 58.2 (95% CI: 58.1; 58.3), 96.9 (96.7; 97.1) for PC DBs and 8.56 (8.5; 8.6) for H DBs; convulsions, 7.6 (95% CI: 7.6; 7.7), 3.55 (3.5; 3.6) for PC and 12.87 (12.8; 13) for H; persistent crying, 3.9 (95% CI: 3.8; 3.9) for PC, injection-site reactions, 2.2 (95% CI 2.1; 2.2) for PC, hypotonic hypo-responsive episode (HHE), 0.4 (95% CI: 0.4; 0.4), 0.6 (0.6; 0.6) for PC and 0.2 (0.2; 0.3) for H; and somnolence: 0.3 (95% CI: 0.3; 0.3) for PC. The pooled IRRs for persistent crying, fever, and ISR, adjusted for age and healthy vaccinee period were higher after wP vs. aP vaccination, and lower for convulsions, for all doses. The IRR for HHE was slightly lower for wP than aP, while wP was associated with somnolence only for dose 1 and dose 3 compared with aP. ConclusionsThe estimated IRs and IRRs were comparable with published data, therefore demonstrating that the ADVANCE system was able to combine several European healthcare databases to assess vaccine safety data for wP and aP vaccination.

2457. Data Science for Outbreak Investigation: Identifying Risk Factors, Tracing Contacts, and Eliciting Transmission Pathways in a Vancomycin-Resistant Enterococci (VRE) Outbreak

BackgroundIn 2018 we experienced a nosocomial outbreak due to vancomycin-resistant enterococci (VRE) in our hospital network. Our goals were to characterize risk factors for VRE acquisition, elicit potential hot spots of transmission, and delineate an optimized approach to tracing contacts.MethodsWe assembled diverse datasets of variable quality and covering different aspects of care from electronic medical records generated during the outbreak period (1/2018–9/2018). Patients who tested VRE-positive during this period were compared with controls with up to 3 negative screenings. First, we identified risk factors for VRE colonization by means of uni- and multivariate analyses. Next, we elicited transmission pathways by detecting commonalities between VRE cases, and determined whether patients with characteristics and connections similar to VRE cases were missed by our current contact tracing strategy.ResultsWe compared 221 VRE patients to 33,624 controls. Independent predictors of VRE colonization were ICU admission (OR 4.9, with 95% confidence interval [3.7–6.5], P < 0.001)], number of records in the database (a proxy for severity-of-illness, OR 1.1 [1.1–1.1], P < 0.001), length of hospital stay (OR 2.7 [2.0–3.5], P < 0.001), age (OR 1.3 [1.2–1.4], P < 0.001), and weeks of antibiotics (OR 1.2 [1.1–1.3], P < 0.001). By using complex network analysis, we were able to establish three main pathways by which the 221 VRE cases are connected: healthcare personnel, medical devices, and patient rooms. This multi-dimensional network extends beyond our current contact tracing strategy, which captures inpatients based on geographical proximity (cf. figure).ConclusionIn this outbreak investigation based on a large electronic healthcare data collection, we found three main risk factors for being a VRE carrier (ICU admission, length of hospital stay, antibiotic exposure), along with three important links between VRE cases (healthcare personnel, medical devices, patient rooms). Data science is likely to provide a better understanding of outbreaks, but interpretations should take data maturity, the scope of included sources, and potential confounding factors into account. Disclosures All authors: No reported disclosures.

Injury Burden in the United States: Accurate, Reliable, and Timely Surveillance Using Electronic Health Care Data.

Current injury surveillance systems in the United States, including the National Electronic Injury Surveillance System (NEISS), are unable to draw reliable subnational and subannual incidence estimates.Compared with the International Classification of Diseases (ICD), the clinical ontology system currently used widely in health care, NEISS's coding structure lacks specificity and consistency. In parallel, the quality of ICD codes depends on accurate and complete documentation by health care providers and skillful translation into ICD codes in electronic health care data. Additionally, there is no national mandate to collect external cause of injury data.Electronic health care data, such as health records and claims, with updated codes and uniform adherence to recommendations for coding external cause of injury, have the potential to be used for a more robust and timely surveillance of injury to accurately and reliably reflect the injury burden in the United States.

Developing a FHIR-based EHR phenotyping framework: A case study for identification of patients with obesity and multiple comorbidities from discharge summaries.

Standards-based clinical data normalization has become a key component of effective data integration and accurate phenotyping for secondary use of electronic healthcare records (EHR) data. HL7 Fast Healthcare Interoperability Resources (FHIR) is an emerging clinical data standard for exchanging electronic healthcare data and has been used in modeling and integrating both structured and unstructured EHR data for a variety of clinical research applications. The overall objective of this study is to develop and evaluate a FHIR-based EHR phenotyping framework for identification of patients with obesity and its multiple comorbidities from semi-structured discharge summaries leveraging a FHIR-based clinical data normalization pipeline (known as NLP2FHIR). We implemented a multi-class and multi-label classification system based on the i2b2 Obesity Challenge task to evaluate the FHIR-based EHR phenotyping framework. Two core parts of the framework are: (a) the conversion of discharge summaries into corresponding FHIR resources - Composition, Condition, MedicationStatement, Procedure and FamilyMemberHistory using the NLP2FHIR pipeline, and (b) the implementation of four machine learning algorithms (logistic regression, support vector machine, decision tree, and random forest) to train classifiers to predict disease state of obesity and 15 comorbidities using features extracted from standard FHIR resources and terminology expansions. We used the macro- and micro-averaged precision (P), recall (R), and F1 score (F1) measures to evaluate the classifier performance. We validated the framework using a second obesity dataset extracted from the MIMIC-III database. Using the NLP2FHIR pipeline, 1237 clinical discharge summaries from the 2008 i2b2 obesity challenge dataset were represented as the instances of the FHIR Composition resource consisting of 5677 records with 16 unique section types. After the NLP processing and FHIR modeling, a set of 244,438 FHIR clinical resource instances were generated. As the results of the four machine learning classifiers, the random forest algorithm performed the best with F1-micro(0.9466)/F1-macro(0.7887) and F1-micro(0.9536)/F1-macro(0.6524) for intuitive classification (reflecting medical professionals' judgments) and textual classification (reflecting the judgments based on explicitly reported information of diseases), respectively. The MIMIC-III obesity dataset was successfully integrated for prediction with minimal configuration of the NLP2FHIR pipeline and machine learning models. The study demonstrated that the FHIR-based EHR phenotyping approach could effectively identify the state of obesity and multiple comorbidities using semi-structured discharge summaries. Our FHIR-based phenotyping approach is a first concrete step towards improving the data aspect of phenotyping portability across EHR systems and enhancing interpretability of the machine learning-based phenotyping algorithms.

Does the frequency of routine follow-up after curative treatment for head-and-neck cancer affect survival?

Routine follow-up is a cornerstone of oncology practice, but evidence to support most aspects of follow-up is lacking. Our objective was to investigate the relationship between frequency of routine follow-up and survival. This population-based study used electronic health care data relating to 5310 patients from Ontario diagnosed with squamous-cell head-and-neck cancer during 2007-2012. Treatments included surgery (24.6%), radiotherapy with or without chemotherapy (52.4%), and combined surgery and radiotherapy (23%). We determined the oncologist who was following each patient after treatment; calculated the average follow-up visits to the oncologist during the subsequent 2.5 years for all patients who were doing well; and used Kaplan-Meier and multiple variable regression analysis to compare, by treatment, overall survival for patients in the high, typical, and low follow-up oncologist groups. Many oncologists saw patients 40%-80% more often than other oncologists did. No relationship of appointment frequency with survival was observed for patients in any treatment group. The practice of routine follow-up varies and is costly both to a health care system and to patients. Without evidence about the effectiveness of current policies, further research is required to investigate new or optimal practices.

Identification of potential drug name confusion errors in the Sentinel System.

In July 2015, the US Food and Drug Administration (FDA) published a drug safety communication regarding errors in prescribing and dispensing of the antidepressant Brintellix (vortioxetine) and the antiplatelet Brilinta (ticagrelor) that arose due to proprietary drug name confusion. Brintellix is indicated for major depressive disorder; Brilinta is indicated to reduce cardiovascular death, myocardial infarction, and stroke in patients with acute coronary syndrome or history of myocardial infarction. Brintellix was renamed to Trintellix in May 2016. Using Brilinta and Brintellix as a proof-of-concept feasibility use case, we assessed whether drug name confusion errors between the pair could be identified in electronic health care data via the combination of a claims-based algorithm and limited manual claims data review. Using data from the Sentinel System, we defined potential errors as Brintellix users without an on- or off-label indication for Brintellix, without a dispensing for a drug with the same indications as Brintellix, and with an on- or off-label indication for Brilinta between -365 and +30 days after index Brintellix dispensing; the reverse was done for Brilinta. We manually reviewed claims profiles of potential cases. We identified 27 (0.1%) potential errors among 21 208 Brintellix users; 16 appeared to be likely errors based on claims profile review. Fifty-one (0.3%) of the 16 779 Brilinta users were identified as potential errors, and four appeared to be likely errors. A claims-based algorithm combined with manual review of claims profiles could identify potential drug name confusion errors, and narrow down likely errors that warrant further investigation.

Did advice on the prescription of sodium valproate reduce prescriptions to women? An observational study in three European countries between 2007 and 2016.

In November 2014, the CMDh (a regulatory body representing EU Member States) advised doctors not to prescribe sodium valproate for epilepsy or bipolar disorder in preg nant women, in women who can become pregnant, or in girls unless other treatments are ineffective or not tolerated. This study aimed to determine if this warning led to changes in prescription patterns. Cohort of 5.4 million women aged between 10 and 50 years identified in electronic health care data from United Kingdom, France, and Italy (2007-2016). Anti-epileptic drug (AED) prescriptions. The prevalence of women receiving AED prescriptions in 2016 varied from 12.2 per 1000 to 29 per 1000 in the four regions. The incidence of prescribing any AED (excluding clonazepam, gabapentin, and pregabalin) fell each year on average by 7.5% (95% CI, 7.0%-8.0%; Emilia Romagna), 9.6% (8.3%-11.0%; France), 7.1% (6.7%-7.6%; Tuscany), and 0.4% (0.2%-1.0%; United Kingdom). The relative odds of prescribing sodium valproate rather than any other AED decreased more after 2014 compared with before the end of 2014 in France (OR = 0.77; 95% CI, 0.60-0.98), Tuscany (0.81; 0.76-0.86), Emilia Romagna (0.83; 0.76-0.90), and the United Kingdom (0.92; 0.80-1.06; not statistically significant). There is evidence that the CMDh warning did lead to changes in prescription patterns of sodium valproate in women of childbearing age. There were considerable differences in prescribing practice amongst regions of Europe.