Evaluating Interoperability in German Critical Incident Reporting Systems.

INTRODUCTION: In industrialised countries, one in ten patients suffers harm during hospitalization. Critical Incident Reporting Systems (CIRS) aim to minimize this by learning from errors and identifying potential risks. However, a lack of interoperability among the 16 CIRS in Germany hampers their effectiveness. METHODS: This study investigates reports' syntactic and semantic interoperability across seven different reporting systems. Syntactic interoperability was examined using WHO's Minimal Information Models (MIM), while semantic interoperability was evaluated with SNOMED concepts. RESULTS: The findings reveal a low structural overlap, with only two terms correctly represented in the SNOMED CT terminology. In addition, most systems showed no syntactic interoperability. CONCLUSION: Improving interoperability is essential for increasing the effectiveness and usability of CIRS. The study suggests a unified data model such as MIM or using Health Level 7 Fast Healthcare Interoperability Resources (HL7 FHIR) resources and expanding SNOMED CT with patient safety-relevant terms for semantic interoperability. Given the current lack of both syntactic and semantic interoperability in CIRS, developing a patient safety ontology is recommended for efficient critical incident analysis too.

[The analysis of CIRSmedical.de using Natural Language Processing]. / Die Analyse des CIRSmedical.de mittels Natural Language Processing.

BACKGROUND: CIRSmedical.de is a publicly accessible, cross-institutional reporting and learning system, which is organized by the German Agency for Quality in Medicine (ÄZQ). CIRSmedical.de has existed since 2005 and has published more than 6,000 event reports. Up to now it has been common practice to analyse these reports in detail or carry out systematic evaluations focusing on specific topics. A systematic evaluation of all case reports has not yet been conducted. Natural Language Processing (NLP) is an analysis strategy from the field of Artificial Intelligence for indexing texts. The examination of case reports using NLP was carried out to describe the characteristics of event reports and comments. MATERIALS AND METHODS: For this analysis 6,480 case reports from CIRSmedical.de (as of December 10, 2019) were provided by the ÄZQ as Excel files. Several free text fields were included in the analysis as well as the feedback of the CIRS team (expert commentary). Text lengths, reporting behaviour, sentiment values and keywords were examined. The algorithms for the analysis were developed with the programming language Python and the corresponding libraries NLTK and SpaCy. RESULTS: The comparison of report lengths depending on the different subject groups presented a heterogeneous picture, in terms of both the number of reports and the number of words. There are more than 4,000 reports from the field of anaesthesiology, whereby text lengths vary particularly strongly with a right-skewed distribution. There are only a few reports from the field of psychotherapy, and these are also very short. The different professional groups (nurses, doctors, other staff) write reports of about the same length. Reports and expert commentaries also differ in terms of sentiment values. Due to the length of the comments, they are more negative in terms of sentiment. Keywords can be identified but show a high heterogeneity. DISCUSSION: Systematic analysis using NLP allows for the description of text properties in event reports and comments. It is now possible to draw a conclusion about the reporters' intention, focus and mood when they report in CIRS. The sentiment analysis is an indication of the mood which the texts convey, both as a report and as a commentary. Text length analysis draws attention to different problems and tendencies: event reports are usually much shorter. Texts that are too short, however, run the risk that the information will not be readily usable for analysis. Comments are often longer, but here one faces the opposite problem: texts that are too long may not be read. The examination of texts by means of NLP helps to rethink the reason for and the form of input, both when reporting and when commenting. It is a first step in the automatic, supportive classification of texts and an improvement of the interaction between reporters and the system.

[The similarity of drug names as a possible cause of confusion: Analysis of data from outpatient care]. / Die Ähnlichkeit von Medikamentennamen als mögliche Ursache von Verwechslungen ­ eine Untersuchung von Daten aus der ambulanten Versorgung.

BACKGROUND: The incidence of adverse drug events (ADE) described in the literature varies between 6.5 and 20 %. Furthermore, it is assumed that up to 29 % of ADE are due to medication errors as a result of confusion because of similarities in spelling (sound alike) or in name, physical appearance or packaging (look alike). Studies dealing with the so-called "LASA" issue were mostly carried out in inpatient care. As far as we know, no systematic investigations into this subject have been carried out for the outpatient sector where patients themselves take care of the application of their medication. In addition, there is no documentation about medication errors in the home setting. The aim of the present study is to describe the importance of the LASA issue in the home setting where medication errors are likely to occur due to similarity of drug names. METHODS: In this context, the similarity of names of prescription drugs was systematically analyzed. We examined in detail how often prescription drug pairings showing orthographic and phonetic similarity were dispensed in the investigation period to an individual patient at the same time. Orthographic similarity was defined as relevant at a Levenshtein index value of ≤ 0.4. This corresponds to the similarity measures of the drugs listed in the LASA public lists and means that the similarity in the lettering of two drug names amounts to at least 60 %. Phonetic similarity was analysed using the Cologne Phonetic ("Kölner Phonetik") for the German language. RESULTS: A total of 255,770 prescriptions were included in the analysis. In 11.4 %, drug pairings were detected that fall below the critical orthographic similarity threshold (Levenshtein index value ≤ 0.4), which represents an increased likelihood of medication error due to the critical similarity of drug names in this fraction. Within this group of "LASA drugs" different degrees of similarity were identified. Even drug pairings with very high orthographic similarity (Levenshtein index value from ≤ 0.1 and 0.1 to ≤ 0.2, 12.4 % and 3.6 % of the drug pairings, respectively) were detected. These drug pairings were mostly different in strength while active ingredients, manufacturer name and pharmaceutical form were the same. For the majority of drug pairings (84 %), the orthographic similarity was lower and showed a Levenshtein index value of ≥ 0.2 to 0.4. Despite different active ingredients, there is a degree of similarity resulting from both identical manufacturer name and pharmaceutical form appearing as part of the drug name. At the phonetic level, the analysis shows comparable frequency of similarity of drug pairings that are subject to potential medication error. DISCUSSION: For the first time, a study was carried out in the outpatient setting recording the incidence of drug pairings that carry a risk for medication errors resulting from patients' confusion over too similar drug names. In the light of the age structure of the patients to whom these look- or sound-alike drugs are prescribed, we can assume that there is a considerable risk of ADE. The conceivable consequences of such medication errors on a pharmacological level range from relatively harmless to potentially highly dangerous. CONCLUSION: There is a major need to fully inform patients about this risk of confusion and subsequent medication errors with certain drug combinations. The similarity structures of drug pairings identified in this study could serve as a basis for developing an appropriate information routine.

[A structured case analysis from the Critical Incident Reporting System of the German Medical Association and the National Association of Statutory Health Insurance Physicians]. / Eine strukturierte Fallanalyse aus dem Critical Incident Reporting System der Bundesärztekammer und Kassenärztlichen Bundesvereinigung.

BACKGROUND: Reporting systems for near misses are necessary to improve patient safety. In Germany, different systems are publicly available on both a national and regional level or as systems related to various medical domains. In contrast with the British Registry, our reporting systems still lack systematic evaluation. Using the Open-Task-Process Model (OPT model) one case of CIRSmedical (www.cirsmedical.de) was selected for a systematic analysis. METHOD: Case 148384 reports on a patient with a tentative diagnosis of pulmonary embolism with an oxygen saturation of 71 %. The attending physician was ordered to leave the patient to participate in the daily team meeting. After 40minutes, the nurses transferred the patient from the emergency department to the ICU. The OPT model systematically checks the properties of all tasks in a given process and matches them to requirements or solving capacities of the task. RESULTS: The analysis manifests some structural problems: Although the case was not very difficult (high priority, but a frequent problem), the solving capacities were not adequate in order to avoid errors. Since the physician left the patient, the loyalty toward medical standards and the team error correction activity were low. The team did not intervene to prevent the doctor from leaving his patient. CONCLUSION: The OPT model allows for the analysis of both single cases and complete data sets of CIR systems and is able to disclose structural problems of clinical management.