Health Information Elements Research Articles

Joining Datasets Without Identifiers: Probabilistic Linkage of Virtual Pediatric Systems and PEDSnet.

To 1) probabilistically link two important pediatric data sources, Virtual Pediatric Systems and PEDSnet, 2) evaluate linkage accuracy overall and in patients with severe sepsis or septic shock, and 3) identify variables important to linkage accuracy. Retrospective linkage of prospectively collected datasets from Virtual Pediatrics Systems, Inc (Los Angeles, CA) and the PEDSnet consortium. Single-center academic PICU. All PICU encounters between January 1, 2012, and December 31, 2017, that were deterministically matched between the two datasets. None. We abstracted records from Virtual Pediatric Systems and PEDSnet corresponding to PICU encounters and probabilistically linked using 44 features shared by the two datasets. We generated a gold standard deterministic linkage using protected health information elements, which were then removed from datasets. We then calculated candidate pair log-likelihood ratios for all pairs of subjects and selected optimal pairs in a two-stage algorithm. A total of 22,051 gold standard PICU encounter pairs were identified over the study period. The optimal linkage model demonstrated excellent discrimination (area under the receiver operating characteristic curve > 0.99); 19,801 cases (89.9%) were matched with 13 false positives. The addition of two protected health information dates (admission month, birth day-of-year) increased to 20,189 (91.6%) the cases matched, with three false positives. Restricting to patients with Virtual Pediatric Systems diagnosis of severe sepsis or septic shock (n = 1,340 [6.1%]) matched 1,250 cases (93.2%) with zero false positives. Increased number of laboratory values present in the first 12 hours of admission significantly increased log-likelihood ratios, suggesting stronger candidate pair matching. We demonstrated the use of probabilistic linkage to accurately join two complementary pediatric critical care datasets at a single academic PICU in the absence of protected health information. Combining datasets with curated diagnoses and granular measurements can validate patient acuity metrics and facilitate multicenter machine learning algorithms. We anticipate these methods will generalize to other common PICU diagnoses.

Art of disaster preparedness in European union: a survey on the health systems.

Introduction: Naturally occurring and man-made disasters have been increasing in the world, including Europe, over the past several decades. Health systems are a key part of any community disaster management system. The success of preparedness and prevention depends on the success of activities such as disaster planning, organization and training. The aim of this study is to evaluate health system preparedness for disasters in the 27 European Union member countries. Method: A cross-sectional analysis study was completed between June-September 2012. The checklist used for this survey was a modified from the World Health Organization toolkit for assessing health-system capacity for crisis management. Three specialists from each of the 27 European Union countries were included in the survey. Responses to each survey question were scored and the range of preparedness level was defined as 0-100%, categorized in three levels as follows: Acceptable; Transitional; or Insufficient. Results: Response rate was 79.1%. The average level of disaster management preparedness in the health systems of 27 European Union member states was 68% (Acceptable). The highest level of preparedness was seen in the United Kingdom, Luxemburg, and Lithuania. Considering the elements of disaster management system, the highest level of preparedness score was at health information elements (86%), and the lowest level was for hospitals, and educational elements (54%). Conclusion: This survey study suggests that preparedness level of European Union countries in 2012 is at an acceptable level but could be improved. Elements such as hospitals and education and training suffer from insufficient levels of preparedness. The European Union health systems need a collective strategic plan, as well as enough resources, to establish a comprehensive and standardized disaster management strategy plan. A competency based training curriculum for managers and first responders is basic to accomplishing this goal. Keywords: Disaster medicine; Disaster preparedness; Disaster epidemiology; Health systems; European Union

Personal Derived Health Information: A Foundation to Preparing the United States for Disasters and Public Health Emergencies

In the days following a disaster/public health emergency, there is great effort to ensure that everyone receives appropriate care and lives are saved. However, evacuees following a disaster/public health emergency often lack access to personal health information that is vital to receive or maintain quality care. Delayed treatment and interruptions of medication regimens often contribute to excess morbidity and mortality following a disaster/public health emergency. This study sought to define a set of minimum health information elements that can be maintained in a personal health record (PHR) and given to first responders/receivers within the first 96 hours of a disaster/public health response to improve clinical health outcomes. A mixed methods approach of qualitative and quantitative data gathering and analyses was completed. Expert panel members (n = 116) and existing health information elements were sampled for this study; 55% (n = 64) of expert panel members had clinical credentials and determined the health information. From an initial set of 6 sources, a step-wise process using a Likert scale survey and thematic data analyses, including interrater reliability and validity checks, produced a set of minimum health information elements. The results identified 30 essential elements from 676 existing health information elements, a reduction of approximately 95%. The elements were grouped into 7 domains: identification, emergency contact, health care contact, health profile -past medical history, medication, major allergies/diet restrictions, and family information. Leading experts in clinical disaster preparedness identified a set of minimum health information elements that first responders/receivers must have to ensure appropriate and timely care. If this set of elements is used as the fundamental information for a PHR, and automatically updated and validated during clinical encounters and medication changes, it is conceivable that following large-scale disasters clinical outcomes may be improved and more lives may be saved.

Large-scale evaluation of automated clinical note de-identification and its impact on information extraction

Objective(1) To evaluate a state-of-the-art natural language processing (NLP)-based approach to automatically de-identify a large set of diverse clinical notes. (2) To measure the impact of de-identification on the performance of information extraction algorithms on the de-identified documents.Material and methodsA cross-sectional study that included 3503 stratified, randomly selected clinical notes (over 22 note types) from five million documents produced at one of the largest US pediatric hospitals. Sensitivity, precision, F value of two automated de-identification systems for removing all 18 HIPAA-defined protected health information elements were computed. Performance was assessed against a manually generated ‘gold standard’. Statistical significance was tested. The automated de-identification performance was also compared with that of two humans on a 10% subsample of the gold standard. The effect of de-identification on the performance of subsequent medication extraction was measured.ResultsThe gold standard included 30 815 protected health information elements and more than one million tokens. The most accurate NLP method had 91.92% sensitivity (R) and 95.08% precision (P) overall. The performance of the system was indistinguishable from that of human annotators (annotators' performance was 92.15%(R)/93.95%(P) and 94.55%(R)/88.45%(P) overall while the best system obtained 92.91%(R)/95.73%(P) on same text). The impact of automated de-identification was minimal on the utility of the narrative notes for subsequent information extraction as measured by the sensitivity and precision of medication name extraction.Discussion and conclusionNLP-based de-identification shows excellent performance that rivals the performance of human annotators. Furthermore, unlike manual de-identification, the automated approach scales up to millions of documents quickly and inexpensively.