Preliminary Analysis of Structured Reporting in the HiGHmed Use Case Cardiology: Challenges and Measures.

The HiGHmed consortium aims to create a shared information governance framework to integrate clinical routine data. One challenge is the replacement of unstructured reporting (e.g. doctoral letters) with structured reporting in clinical routine. The Heidelberg cardiology department evaluates dynamic PDF forms for structured data reporting of heart failure (HF) patients. In this use case, we aim to identify potential caveats or shortcomings in data processing at an early stage. We employed data mining strategies to detect patterns related to incomplete or false data, which we found to be present among all data types. We then discuss the characteristics of the baseline patient cohort in Heidelberg to find out about specific peculiarities and potential biases, which may be site-specific. Briefly, our patient population is predominantly male (67%), NYHA I & II are the most common severity classes, NYHA IV is missing entirely. Most patients have a dilated cardiomyopathy (DCM) or coronary heart disease (CHD) diagnosed as their cause of HF. Finally, we also analyzed how comorbidities and risk factors relate to specific disease entities of heart failure patients. Family anamnesis was more frequent among cardiomyopathy patients than among CHD patients, who show a more dominating presence of dyslipidemia instead. Generally, the most dominant risk factor was arterial hypertension, while at the other end of the scale alcoholism appears to be underreported.

Analysis of long-term survival in multiple myeloma after first-line autologous stem cell transplantation: impact of clinical risk factors and sustained response.

The widespread use of high-dose therapy and autologous stem cell transplantation (ASCT) as well as the introduction of novel agents have significantly improved outcomes in multiple myeloma (MM) enabling long-term survival. We here analyze factors influencing survival in 865 newly diagnosed MM patients who underwent first-line ASCT at our center between 1993 and 2014. Relative survival and conditional survival were assessed to further characterize long-term survivors. Achievement of complete response (CR) post-ASCT was associated with prolonged progression-free survival (PFS) in the whole cohort and with significantly superior overall survival (OS) in the subgroup of patients receiving novel agent-based induction therapy. Landmark analyses performed at 1, 3, and 5 years post-ASCT revealed that sustainment of any response had a highly significant influence on survival with no significant differences between sustained CR and sustained inferior responses. Furthermore, outcome was independently improved by administration of maintenance therapy. A subset of patients did experience long-term survival >15 years. However, conditional survival demonstrated a persistent risk of myeloma-associated death and cumulative relative survival curves did not show development of a clear plateau, even in prognostically advantageous groups. In conclusion, in this large retrospective study, sustained response after first-line ASCT was found to be a major prognostic factor for OS independent of depth of sustained response. Administration of maintenance therapy further improved outcome, supporting the hypothesis that interventions to prolong responses achieved post-ASCT may be essential to reach long-term survival, especially in the setting of persisting residual disease.

Assessment of Automatically Exported Clinical Data from a Hospital Information System for Clinical Research in Multiple Myeloma.

PURPOSE: An important part of the electronic information available in Hospital Information System (HIS) has the potential to be automatically exported to Electronic Data Capture (EDC) platforms for improving clinical research. This automation has the advantage of reducing manual data transcription, a time consuming and prone to errors process. However, quantitative evaluations of the process of exporting data from a HIS to an EDC system have not been reported extensively, in particular comparing with manual transcription. In this work an assessment to study the quality of an automatic export process, focused in laboratory data from a HIS is presented. METHODS: Quality of the laboratory data was assessed in two types of processes: (1) a manual process of data transcription, and (2) an automatic process of data transference. The automatic transference was implemented as an Extract, Transform and Load (ETL) process. Then, a comparison was carried out between manual and automatic data collection methods. The criteria to measure data quality were correctness and completeness. RESULTS: The manual process had a general error rate of 2.6% to 7.1%, obtaining the lowest error rate if data fields with a not clear definition were removed from the analysis (p < 10E-3). In the case of automatic process, the general error rate was 1.9% to 12.1%, where lowest error rate is obtained when excluding information missing in the HIS but transcribed to the EDC from other physical sources. CONCLUSION: The automatic ETL process can be used to collect laboratory data for clinical research if data in the HIS as well as physical documentation not included in HIS, are identified previously and follows a standardized data collection protocol.

Analysis of Existing Guidelines for the Systematic Planning Process of Clinical Registries.

Clinical registries are a powerful method to observe the clinical practice and natural disease history. In contrast to clinical trials, where guidelines and standardized methods exist and are mandatory, only a few initiatives have published methodological guidelines for clinical registries. The objective of this paper was to review these guidelines and systematically assess their completeness, usability and feasibility according to a SWOT analysis. The results show that each guideline has its own strengths and weaknesses. While one supports the systematic planning process, the other discusses clinical registries in great detail. However, the feasibility was mostly limited and the special requirements of clinical registries, their flexible, expandable and adaptable technological structure was not addressed consistently.

The Use of Tools, Modelling Methods, Data Types, and Endpoints in Systems Medicine: A Survey on Projects of the German e:Med-Programme.

Systems medicine is the consequent continuation of research efforts on the road to an individualized medicine. Thereby, systems medicine tries to offer a holistic view on the patient by combining different data sources to highlight different perspectives on the patient's health. Our research question was to identify the main data types, modelling methods, analysis tools, and endpoints currently used and studied in systems medicine. Therefore, we conducted a survey on projects with a systems medicine background. Fifty participants completed this survey. The results of the survey were analyzed using histograms and cross tables, and finally compared to results of a former literature review with the same research focus. The data types reported in this survey were widely diversified. As expected, genomic and phenotype data were used most frequently. In contrast, environmental and behavioral data were rarely used in the projects. Overall, the cross tables of the data types in the survey and the literature review showed overlapping results.

Automated Classification of Selected Data Elements from Free-text Diagnostic Reports for Clinical Research.

OBJECTIVES: In the Multiple Myeloma clinical registry at Heidelberg University Hospital, most data are extracted from discharge letters. Our aim was to analyze if it is possible to make the manual documentation process more efficient by using methods of natural language processing for multiclass classification of free-text diagnostic reports to automatically document the diagnosis and state of disease of myeloma patients. The first objective was to create a corpus consisting of free-text diagnosis paragraphs of patients with multiple myeloma from German diagnostic reports, and its manual annotation of relevant data elements by documentation specialists. The second objective was to construct and evaluate a framework using different NLP methods to enable automatic multiclass classification of relevant data elements from free-text diagnostic reports. METHODS: The main diagnoses paragraph was extracted from the clinical report of one third randomly selected patients of the multiple myeloma research database from Heidelberg University Hospital (in total 737 selected patients). An EDC system was setup and two data entry specialists performed independently a manual documentation of at least nine specific data elements for multiple myeloma characterization. Both data entries were compared and assessed by a third specialist and an annotated text corpus was created. A framework was constructed, consisting of a self-developed package to split multiple diagnosis sequences into several subsequences, four different preprocessing steps to normalize the input data and two classifiers: a maximum entropy classifier (MEC) and a support vector machine (SVM). In total 15 different pipelines were examined and assessed by a ten-fold cross-validation, reiterated 100 times. For quality indication the average error rate and the average F1-score were conducted. For significance testing the approximate randomization test was used. RESULTS: The created annotated corpus consists of 737 different diagnoses paragraphs with a total number of 865 coded diagnosis. The dataset is publicly available in the supplementary online files for training and testing of further NLP methods. Both classifiers showed low average error rates (MEC: 1.05; SVM: 0.84) and high F1-scores (MEC: 0.89; SVM: 0.92). However the results varied widely depending on the classified data element. Preprocessing methods increased this effect and had significant impact on the classification, both positive and negative. The automatic diagnosis splitter increased the average error rate significantly, even if the F1-score decreased only slightly. CONCLUSIONS: The low average error rates and high average F1-scores of each pipeline demonstrate the suitability of the investigated NPL methods. However, it was also shown that there is no best practice for an automatic classification of data elements from free-text diagnostic reports.

A reference data model of a metadata registry preserving semantics and representations of data elements.

Integration and analysis of clinical data collected in multiple data sources over a long period of time is a major challenge even when data warehouses and metadata registries are used. Since most metadata registries focus on describing data elements to establish domain consistent data definition and providing item libraries, hierarchical and temporal dependencies cannot be mapped. Therefore we developed and validated a reference data model, based on ISO/IEC 11179, which allows revision and branching control of conceptually similar data elements with heterogeneous definitions and representations.

Comparison of the developmental tests Bayley-III and Bayley-II in 7-month-old infants born preterm.

The study aims on comparing Bayley Scales of infant development third (Bayley-III) and Bayley second (Bayley-II) edition with special focus on patterns in the first year of life. Fifty-five premature infants (43 with low birth weight/LBW >1,499 g and 12 with very/extremely low birth weight/VLBW/ELBW <1,500 g) aged 7 months (corrected for prematurity) were assessed with the complete Bayley-III. From this assessment, Bayley-II results were retrospectively estimated. Bayley-III results were compared to the expected mean with one-sample t-tests. The mean scores of both editions were compared with the aid of paired-sample t-tests. Pearson correlations between subscales and editions were analysed. The Bayley-III cognitive score of the study group was significantly higher than the expected mean of the standardization sample. VLBW/ELBW had significantly lower motor scores than LBW in both editions. When compared to estimated Bayley-II scores, all relevant Bayley-III scores were significantly higher (all p < .01) with highest difference (ten points) between the motor scales of both editions. There were significant correlations not only between Bayley-III cognitive and language scales but also between language and motor scales. Given the strong association between motor and cognitive behaviour in early infancy, this age-specific pattern is heightening the risk of failure to identify infants at risk for both cognitive and motor delay. Therefore, assessment of infants should comprise all subscales. Since Bayley-III probably overestimates especially motor performance in young infants, when interpreting Bayley-III scores in this age, comparison groups are highly recommended until further validation of normative data are outstanding.