Usability of Graphical User Interfaces With Semiautomatic Delabeling Feature to Improve Drug Allergy Documentation.

BACKGROUND: The quality of allergy documentation in electronic health records is frequently poor. OBJECTIVE: To compare the usability of 3 graphical user interfaces (GUIs) for drug allergy documentation. METHODS: Physicians tested 3 GUIs by means of 5 fictional drug allergy scenarios: the current GUI (GUI 0), using mainly free-text, and 2 new coded versions (GUI 1 and GUI 2) asking information on allergen category, specific allergen, symptom(s), symptom onset, timing of initial reaction, and diagnosis status with a semiautomatic delabeling feature. Satisfaction was measured by the System Usability Scale questionnaire, efficiency by time to complete the tasks, and effectiveness by a task completion score. Posttest interviews provided more in-depth qualitative feedback. RESULTS: Thirty physicians from 7 different medical specialties and with varying degrees of experience participated. The mean System Usability Scale scores for GUI 1 (77.25, adjective rating "Good") and GUI 2 (78.42, adjective rating "Good") were significantly higher than for GUI 0 (56.58, adjective rating "OK") (Z, 6.27, Padj < .001 and Z, 6.62, Padj < .001, respectively). There was no significant difference in task time between GUIs. Task completion scores of GUI 1 and GUI 2 were higher than for GUI 0 (Z, 9.59, Padj < .001 and Z, 11.87, Padj < .001, respectively). Quantitative and qualitative findings were combined to propose a GUI 3 with high usability. CONCLUSIONS: The usability and quality of allergy documentation was higher for the newly developed coded GUIs with a semiautomatic delabeling feature without being more time-consuming.

Added value of patient- and drug-related factors to stratify drug-drug interaction alerts for risk of QT prolongation: Development and validation of a risk prediction model.

AIMS: Many clinical decision support systems trigger warning alerts for drug-drug interactions potentially leading to QT prolongation and torsades de pointes (QT-DDIs). Unfortunately, there is overalerting and underalerting because stratification is only based on a fixed QT-DDI severity level. We aimed to improve QT-DDI alerting by developing and validating a risk prediction model considering patient- and drug-related factors. METHODS: We fitted 31 predictor candidates to a stepwise linear regression for 1000 bootstrap samples and selected the predictors present in 95% of the 1000 models. A final linear regression model with those variables was fitted on the original development sample (350 QT-DDIs). This model was validated on an external dataset (143 QT-DDIs). Both true QTc and predicted QTc were stratified into three risk levels (low, moderate and high). Stratification of QT-DDIs could be appropriate (predicted risk = true risk), acceptable (one risk level difference) or inappropriate (two risk levels difference). RESULTS: The final model included 11 predictors with the three most important being use of antiarrhythmics, age and baseline QTc. Comparing current practice to the prediction model, appropriate stratification increased significantly from 37% to 54% appropriate QT-DDIs (increase of 17.5% on average [95% CI +5.4% to +29.6%], padj = 0.006) and inappropriate stratification decreased significantly from 13% to 1% inappropriate QT-DDIs (decrease of 11.2% on average [95% CI -17.7% to -4.7%], padj ≤ 0.001). CONCLUSION: The prediction model including patient- and drug-related factors outperformed QT alerting based on QT-DDI severity alone and therefore is a promising strategy to improve DDI alerting.

Development and validation of an audit tool for fluid management in non-critically ill adults in the emergency department.

Especially in the emergency department (ED), it is critical to identify weaknesses in prescribing behavior of IV maintenance fluids to ensure a qualitative 24-h fluid management plan. The primary aim of this study was to develop an audit instrument to assess the pitfalls in documentation and prescribing habits of IV fluid therapy for non-critically ill patients admitted to the ED. In this study, an expert panel initially designed the tool. During different phases, adaptations were made to optimize inter-rater agreement (Fleiss' kappa, κ) and validity was determined based on the application of the tool on randomly selected electronic ED patient records. Second, the IV fluid prescription's appropriateness was addressed. The final tool consists of three parts: fluid status assessment, evaluation of maintenance fluid needs and, if present, a limited appropriateness check of the fluid prescription. A manual enhanced inter-rater agreement. κ-values achieved the target value (0.40) after three adaptation rounds, except for the category of 'clinical observations' in the assessment part (κ = 0.531, 95% CI 0.528-0.534), which was acceptable. A check of the IV fluid's indication and volume was only possible with an additional expert evaluation to conclude on prescription appropriateness. Criterion related validity of the final version was high (93.4%). To conclude, the instrument is considered reliable and can be used in clinical practice to evaluate ED fluid management. Thorough documentation is essential to evaluate the appropriateness of the IV fluid prescription, to improve information transfer on IV fluid therapy to the ward and to facilitate retrospective chart review of ED prescribing behavior.

Machine Learning Techniques Outperform Conventional Statistical Methods in the Prediction of High Risk QTc Prolongation Related to a Drug-Drug Interaction.

In clinical practice, many drug therapies are associated with prolongation of the QT interval. In literature, estimation of the risk of prescribing drug-induced QT prolongation is mainly executed by means of logistic regression; only one paper reported the use of machine learning techniques. In this paper, we compare the performance of both techniques on the same dataset. High risk for QT prolongation was defined as having a corrected QT interval (QTc) ≥ 450 ms or ≥ 470 ms for respectively male and female patients. Both conventional statistical methods (CSM) and machine learning techniques (MLT) were used. All algorithms were validated internally and with a hold-out dataset of respectively 512 and 102 drug-drug interactions with possible drug-induced QTc prolongation. MLT outperformed the best CSM in both internal and hold-out validation. Random forest and Adaboost classification performed best in the hold-out set with an equal harmonic mean of sensitivity and specificity (HMSS) of 81.2% and an equal accuracy of 82.4% in a hold-out dataset. Sensitivity and specificity were both high (respectively 75.6% and 87.7%). The most important features were baseline QTc value, C-reactive protein level, heart rate at baseline, age, calcium level, renal function, serum potassium level and the atrial fibrillation status. All CSM performed similarly with HMSS varying between 60.3% and 66.3%. The overall performance of logistic regression was 62.0%. MLT (bagging and boosting) outperform CSM in predicting drug-induced QTc prolongation. Additionally, 19.2% was gained in terms of performance by random forest and Adaboost classification compared to logistic regression (the most used technique in literature in estimating the risk for QTc prolongation). Future research should focus on testing the classification on fully external data, further exploring potential of other (new) machine and deep learning models and on generating data pipelines to automatically feed the data to the classifier used.

Usability of Three Graphical User Interfaces for Drug Allergy Documentation.

The current drug allergy documentation module in the electronic health record of our institution is in a free-text format. Two versions of a structured and coded drug allergy documentation module were developed. Twenty-five physicians tested the three interfaces via 3x5 test scenarios. The usability was measured for each interface with a system usability scale questionnaire. Both new versions scored significantly better than the current free-text version. User feedback will be used to further optimize the new module.

Optimization of Performance by Combining Most Sensitive and Specific Models in Data Science Results in Majority Voting Ensemble.

Ensemble modeling is an increasingly popular data science technique that combines the knowledge of multiple base learners to enhance predictive performance. In this paper, the idea was to increase predictive performance by holding out three algorithms when testing multiple classifiers: (a) the best overall performing algorithm (based on the harmonic mean of sensitivity and specificity (HMSS) of that algorithm); (b) the most sensitive model; and (c) the most specific model. This approach boils down to majority voting between the predictions of these three base learners. In this exemplary study, a case of identifying a prolonged QT interval after administering a drug-drug interaction with increased risk of QT prolongation (QT-DDI) is presented. Performance measures included accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Overall performance was measured by calculating the HMSS. Results show an increase in all performance measure characteristics compared to the original best performing algorithm, except for specificity where performance remained stable. The presented approach is fairly simple and shows potential to increase predictive performance, even without adjusting the default cut-offs to differentiate between high and low risk cases. Future research should look at a way of combining all tested algorithms, instead of using only three. Similarly, this approach should be tested on a multiclass prediction problem.

Interdisciplinary knowledge gaps on intravenous fluid management in adult patients: Survey among physicians and nurses of a university hospital.

RATIONALE: Intravenous (IV) fluids are frequently involved in iatrogenic complications in hospitalized patients. Knowledge of IV fluids seems inadequate and is not covered sufficiently in standard medical education. METHODS: Two surveys were developed, based on the 2016 British National Institute for Health and Care Excellence guideline 'IV fluid therapy in adults in hospital', to provide insight on the learning needs and expectations of physicians and nurses. Each survey focused on profession-specific practice and consisted of three parts: demographics, knowledge questions and evaluation of current habits. Physicians and nurses practicing in a Belgian university hospital were invited to complete the survey electronically, respectively, in January and May 2018. RESULTS: A total of 103 physicians (19%) and 259 nurses (24%) participated. Although every indication for fluid therapy may require a specific fluid and electrolyte mixture, and hence, knowledge of their exact composition, most physicians and nurses did not know the composition of commonly prescribed solutions for IV infusion. Senior physicians did not score better than juniors did on questions concerning the daily needs of a nil-by-mouth patient. The availability of an IV fluid on the ward guides physicians to prescribe IV fluids (17%). Nurses (56%) feel they share responsibility in fluid management as they frequently intervene in urgent situations. More than half of participants (70% of physicians, 79% of nurses) indicated a need for additional information. CONCLUSIONS: A clear need for more structured information on IV fluids was identified. Both physicians and nurses struggle with fluid therapy. Continuing education on IV fluid management, emphasizing multidisciplinary collaboration, and monitoring evidence-based practice is essential to support the clinical decision process in daily practice.

Determinants for under- and overdosing of direct oral anticoagulants and physicians' implementation of clinical pharmacists' recommendations.

AIMS: To analyse the appropriateness of direct oral anticoagulant (DOAC) dosing and determinants for under-and overdosing as well as acceptance and implementation rates of pharmacists' interventions. METHODS: Cross-sectional study in a tertiary hospital in hospitalized patients with atrial fibrillation on DOACs in 2019 (n = 1688). Primary outcome was the proportion of patients with inappropriate DOAC prescribing with identification of determinants for under-and overdosing. Secondary outcomes included acceptance and implementation rates of pharmacists' recommendations and determination of reasons for nonacceptance/nonimplementation. RESULTS: Inappropriate prescribing was observed in 16.9% of patients (n = 286) with underdosing (9.7%) being more prevalent than overdosing (6.9%). For all DOACs considered together, body weight<60 kg (odds ratio [OR] 0.46 [0.27-0.77]), edoxaban use (OR 0.42 [0.24-0.74]), undergoing surgery (OR 0.57 [0.37-0.87]) and being DOAC naïve (OR 0.45 [0.29-0.71]) were associated with significantly lower odds of underdosing. Bleeding history (OR 1.86 [1.24-2.80]) and narcotic use (OR 1.67 [1.13-2.46]) were associated with significantly higher odds for underdosing. Determinants with a significantly higher odds of overdosing were renal impairment (OR 11.29 [6.23-20.45]) and body weight<60 kg (OR 2.34 [1.42-3.85]), whereas dabigatran use (OR 0.24 [0.08-0.71]) and apixaban (OR 0.18 [0.10-0.32]) were associated with a significantly lower odds of overdosing compared to rivaroxaban. Physicians accepted the pharmacists' advice in 179 cases (79.2%) consisting of 92 (51.4%) recommendations for underdosing, 82 (45.8%) for overdosing and 5 (2.8%) for contraindications. CONCLUSION: Inappropriate DOAC prescribing remains common, although there is a slight improvement compared to our study of 2016. Clinical services led by pharmacists help physicians to reduce the number of inadequate prescriptions for high-risk medications such as DOACs.

Assessment of adherence, treatment satisfaction and knowledge of direct oral anticoagulants in atrial fibrillation patients.

AIMS: Direct oral anticoagulants (DOACs) are increasingly used for stroke prevention in atrial fibrillation. However, little is known about the association between medication adherence, patient satisfaction and treatment knowledge. The objective was to determine patients' DOAC adherence and their treatment satisfaction over time. Furthermore, we respectively investigated possible associations of treatment satisfaction and treatment knowledge in relation to adherence. METHODS: Longitudinal study conducted in atrial fibrillation patients hospitalized in 2019 in a tertiary university hospital. DOAC adherence, treatment satisfaction and knowledge were assessed with validated questionnaires. Mixed effects logistic regression was modelled to investigate the effect of both treatment satisfaction and knowledge on DOAC adherence over time. RESULTS: In total, 164 patients participated of whom 128 and 101 patients could be recontacted after a period of, respectively, 3 (first contact) and 6 months (second contact) to assess adherence and treatment satisfaction. Suboptimal adherence was observed in 40.6% of the patients after 3 months and in 42.6% after 6 months (P = .78). There was no significant difference (P = .29) in the total score for treatment satisfaction between the first (79.2%) and the second contact (80.6%). DOAC adherence was not affected by time (P = .71) nor by total knowledge score (P = .61) or treatment satisfaction score (P = .34). Nonetheless, a strong correlation between treatment satisfaction and knowledge was found (P = .004). CONCLUSION: DOAC adherence was suboptimal. Treatment satisfaction and knowledge were not associated with DOAC adherence over a 6-month period. Knowledge gaps were identified that could be remediated through patient education and follow-up.

Management of drug-disease interactions: a best practice from the Netherlands.

Background Drug-disease interactions are situations where pharmacotherapy may have a negative effect on patients' comorbidities. In these cases, it can be necessary to avoid that drug, adjust its dose or monitor therapy. In the Netherlands, pharmacists have developed a best practice how to systematically evaluate drug-disease interactions based on pharmacological considerations and implement recommendations for specific drug-disease interactions. Aim To describe the development of recommendations for drug-disease interactions and the implementation in prescribing and dispensing practice in the Netherlands. Setting Pharmacies and physicians' practices in primary care and hospitals in the Netherlands. Development A multi-disciplinary expert panel assessed if diseases had clinically relevant drug-disease interactions and evaluated drug-disease interactions by literature review and expert opinion, and subsequently developed practice recommendations. Implementation The recommendations were implemented in all clinical decision support systems in primary care and hospitals throughout the Netherlands. Evaluation Recommendations were developed for 57 diseases and conditions. Cardiovascular diseases have the most drug-disease interactions (n = 12, e.g. long QT-syndrome, heart failure), followed by conditions related to the reproductive system (n = 7, e.g. pregnancy). The number of drugs with recommendations differed between 6 for endometriosis and tympanostomy tubes, and up to 1171 in the case of porphyria or even all drugs for pregnancy. Conclusion Practice recommendations for drug-disease interactions were developed, and implemented in prescribing and dispensing practice. These recommendations support both pharmacists and physicians by signalling clinically relevant drug-disease interactions at point of care, thereby improving medication safety. This practice may be adopted and contribute to safer medication use in other countries as well.

Evaluation of an optimized context-aware clinical decision support system for drug-drug interaction screening.

OBJECTIVE: Evaluation of the effect of six optimization strategies in a clinical decision support system (CDSS) for drug-drug interaction (DDI) screening on alert burden and alert acceptance and description of clinical pharmacist intervention acceptance. METHODS: Optimizations in the new CDSS were the customization of the knowledge base (with addition of 67 extra DDIs and changes in severity classification), a new alert design, required override reasons for the most serious alerts, the creation of DDI-specific screening intervals, patient-specific alerting, and a real-time follow-up system of all alerts by clinical pharmacists with interventions by telephone was introduced. The alert acceptance was evaluated both at the prescription level (i.e. prescription acceptance, was the DDI prescribed?) and at the administration level (i.e. administration acceptance, did the DDI actually take place?). Finally, the new follow-up system was evaluated by assessing the acceptance of clinical pharmacist's interventions. RESULTS: In the pre-intervention period, 1087 alerts (92.0 % level 1 alerts) were triggered, accounting for 19 different DDIs. In the post-intervention period, 2630 alerts (38.4 % level 1 alerts) were triggered, representing 86 different DDIs. The relative risk forprescription acceptance in the post-intervention period compared to the pre-intervention period was 4.02 (95 % confidence interval (CI) 3.17-5.10; 25.5 % versus 6.3 %). The relative risk for administration acceptance was 1.16 (95 % CI 1.08-1.25; 54.4 % versus 46.7 %). Finally, 86.9 % of the clinical pharmacist interventions were accepted. CONCLUSION: Six concurrently implemented CDSS optimization strategies resulted in a high alert acceptance and clinical pharmacist intervention acceptance. Administration acceptance was remarkably higher than prescription acceptance.

A cross-sectional study of the Belgian community pharmacist's satisfaction with the implementation of the electronic prescription.

BACKGROUND: Several benefits and problems of electronic prescribing (ePrescribing) are described in scientific literature, though problems remain in the implementation. In this study, we evaluated the pharmacist's perception of the ePrescription implementation within the community pharmacy software in Belgium, and the frequency and hindrance of encountered problems. MATERIAL AND METHODS: A cross-sectional study was conducted among community pharmacists in Belgium to measure satisfaction with the ePrescribing implementation and factors influencing this satisfaction. RESULTS: In total 246 pharmacists (3.3 % response rate) rated the implementation in their software with an average score of 6.46 ± 2.16 (SD) on a scale of 10. In Belgium, French-speaking pharmacists gave a significantly higher satisfaction score compared to Dutch-speaking pharmacists (P = 0.032), whereas Dutch-speaking pharmacists perceived to process significantly more ePrescriptions compared to French-speaking pharmacists (P < 0.001). Satisfaction with the implementation of the ePrescription was significantly associated with the software package (P < 0.001), the knowledge of the ePrescribing workflow (P = 0.036), the frequency of slow responses of the software (P < 0.001) and the perception of unavailability of the system (P = 0.003). CONCLUSIONS: The Belgian pharmacist was moderately satisfied with the implementation of the ePrescription. Problems with the availability of Belgian eHealth systems and interoperability issues with national codes used between prescriber and dispenser have to be resolved in the future in order to meet the Belgian community pharmacist's needs.

Evaluation of context-specific alerts for potassium-increasing drug-drug interactions: A pre-post study.

OBJECTIVE: To investigate whether context-specific alerts for potassium-increasing drug-drug interactions (DDIs) in a clinical decision support system reduced the alert burden, increased alert acceptance, and had an effect on the occurrence of hyperkalemia. MATERIALS AND METHODS: In the pre-intervention period all alerts for potassium-increasing DDIs were level 1 alerts advising absolute contraindication, while in the post-intervention period the same drug combinations could trigger a level 1 (absolute contraindication), a level 2 (monitor potassium values), or a level 3 alert (informative, not shown to physicians) based on the patient's recent laboratory value of potassium. Alert acceptance was defined as non-prescription or non-administration of the interacting drug combination for level 1 alerts and as monitoring of the potassium levels for level 2 alerts. RESULTS: The alert burden decreased by 92.8%. The relative risk (RR) for alert acceptance based on prescription rates for level 1 alerts and monitoring rates for level 2 alerts was 15.048 (86.5% vs 5.7%; 95% CI 12.037-18.811; P < 0.001). With alert acceptance for level 1 alerts based on actual administration and for level 2 alerts on monitoring rates, the RR was 3.597 (87.6% vs 24.4%; 95% CI 3.192-4.053; P < 0.001). In the generalized linear mixed model the effect of the intervention on the occurrence of hyperkalemia was not significant (OR 1.091, 95% CI 0.172-6.919). CONCLUSION: The proposed strategy seems effective to get a grip on the delicate balance between over- and under alerting.

Why do Belgian Community Pharmacists Still Treat Electronic Prescriptions as Paper-Based?

Belgium is in a transition phase from paper-based prescriptions to electronic prescriptions (ePrescriptions) during which both the paper and electronic format are valid. Since patients still get a paper proof of the ePrescription, sometimes pharmacists use the ePrescription as paper-based prescription. When the government demands a complete dematerialization, i.e. no more paper-based prescriptions, this will no longer be possible. Therefore, we questioned the frequency and reasons for treating an ePrescription as paper-based. The logged interactions in the national database were used to identify possible reasons. The tarification service Koninklijk Limburgs Apothekers Verbond (KLAV) provided prescriptions of June 2018. KLAV supports tarification for community pharmacies all over Belgium, thereby providing a representative sample for the Belgian community pharmacies. A two-stage cluster random sampling technique was applied to retrieve a subset of 10,000 prescriptions. In this subset we identified 4961 ePrescriptions (49.61%) of which 226 (4.56%, in total 2.26%) were treated as paper-based. Reasons observed for this incorrect handling are (1) non-compliance of the community pharmacist; (2) errors in software or handling of the community pharmacist; (3) errors at the prescriber side or patient tries to fraud; (4) incorrectly revoking the ePrescription; and (5) errors in software of prescriber. The main reasons for treating ePrescriptions as paper-based are non-compliance of the community pharmacist (n = 124, 54.87%) by ignoring its digital nature, and errors in software or handling of the community pharmacist (n = 85, 37.61%). Future research is necessary to investigate user opinions and to measure the impact of introducing ePrescribing in the daily routine.

An Evaluation of the Belgian Community Pharmacist's Satisfaction with the Implementation of the Electronic Prescription Within a Pharmacist's Software.

In a cross-sectional study, we evaluated the perception of the community pharmacist in Belgium about his satisfaction with the implementation of the electronic prescription in his software. 246 community pharmacists scored the implementation in their software with an average of 6.46±2.16(SD) on a score of 10. This satisfaction is associated with the software used (p<0.001), the frequency of slow responses (p<0.001), the perception of unavailability of systems (p=0.003) and the knowledge of the pharmacist (p=0.036).