The use of natural language processing on pediatric diagnostic radiology reports in the electronic health record to identify deep venous thrombosis in children.

Venous thromboembolism (VTE) is a potentially life-threatening condition that includes both deep vein thrombosis (DVT) and pulmonary embolism. We sought to improve detection and reporting of children with a new diagnosis of VTE by applying natural language processing (NLP) tools to radiologists' reports. We validated an NLP tool, Reveal NLP (Health Fidelity Inc, San Mateo, CA) and inference rules engine's performance in identifying reports with deep venous thrombosis using a curated set of ultrasound reports. We then configured the NLP tool to scan all available radiology reports on a daily basis for studies that met criteria for VTE between July 1, 2015, and March 31, 2016. The NLP tool and inference rules engine correctly identified 140 out of 144 reports with positive DVT findings and 98 out of 106 negative reports in the validation set. The tool's sensitivity was 97.2% (95% CI 93-99.2%), specificity was 92.5% (95% CI 85.7-96.7%). Subsequently, the NLP tool and inference rules engine processed 6373 radiology reports from 3371 hospital encounters. The NLP tool and inference rules engine identified 178 positive reports and 3193 negative reports with a sensitivity of 82.9% (95% CI 74.8-89.2) and specificity of 97.5% (95% CI 96.9-98). The system functions well as a safety net to screen patients for HA-VTE on a daily basis and offers value as an automated, redundant system. To our knowledge, this is the first pediatric study to apply NLP technology in a prospective manner for HA-VTE identification.

Improving Cardiac Surgical Site Infection Reporting and Prevention By Using Registry Data for Case Ascertainment.

BACKGROUND: The use of administrative data for surgical site infection (SSI) surveillance leads to inaccurate reporting of SSI rates [1]. A quality improvement (QI) initiative was conducted linking clinical registry and administrative databases to improve reporting and reduce the incidence of SSI [2]. METHODS: At our institution, The Society of Thoracic Surgeons Congenital Heart Surgery Database (STS-CHSD) and infection surveillance database (ISD) were linked to the enterprise data warehouse containing electronic health record (EHR) billing data. A data visualization tool was created to (1) use the STS-CHSD for case ascertainment, (2) resolve discrepancies between the databases, and (3) assess impact of QI initiatives, including wound alert reports, bedside reviews, prevention bundles, and billing coder education. RESULTS: Over the 24-month study period, 1,715 surgical cases were ascertained according to the STS-CHSD clinical criteria, with 23 SSIs identified through the STS-CHSD, 20 SSIs identified through the ISD, and 32 SSIs identified through the billing database. The rolling 12-month STS-CHSD SSI rate decreased from 2.73% (21 of 769 as of January 2013) to 1.11% (9 of 813 as of December 2014). Thirty reporting discrepancies were reviewed to ensure accuracy. Workflow changes facilitated communication and improved adjudication of suspected SSIs. Billing coder education increased coding accuracy and narrowed variation between the 3 SSI sources. The data visualization tool demonstrated temporal relationships between QI initiatives and SSI rate reductions. CONCLUSIONS: Linkage of registry and infection control surveillance data with the EHR improves SSI surveillance. The visualization tool and workflow changes facilitated communication, SSI adjudication, and assessment of the QI initiatives. Implementation of these initiatives was associated with decreased SSI rates.

Use of administrative data for surgical site infection surveillance after congenital cardiac surgery results in inaccurate reporting of surgical site infection rates.

BACKGROUND: The National Healthcare Safety Network (NHSN) is a safety surveillance system managed by the Centers for Disease Control and Prevention that monitors procedure specific rates of surgical site infections (SSIs). At our institution, SSI data is collected and reported by three different methods: (1) the NHSN database with reporting to the Centers for Disease Control and Prevention; (2) the hospital billing database with reporting to payers; and (3) The Society of Thoracic Surgeons Congenital Heart Surgery Database. A quality improvement initiative was undertaken to better understand issues with SSI reporting and to evaluate the effect of different data sources on annual SSI rates. METHODS: Annual cardiac surgery procedure volumes for all three data sources were compared. All episodes of SSI identified in any data source were reviewed and adjudicated using NHSN SSI criteria, and the effect on SSI rates was evaluated. RESULTS: From January 1, 2008, to December 31, 2011, 2,474 cardiac procedures were performed and reported to The Society of Thoracic Surgeons Congenital Heart Surgery Database. Billing data identified 1,865 cardiac surgery procedures using the 63 CARD International Classification of Diseases-Ninth Revision codes from the NHSN inclusion criteria. Only 1,425 procedures were targeted for NHSN surveillance using the NHSN's CARD operative procedure group in the same period. Procedures identified for NHSN surveillance annually underestimated the number of cardiac operations performed by 17% to 71%. As a result, annual SSI rates potentially differed by 12% to 270%. CONCLUSIONS: The NHSN CARD surveillance guidelines for SSI fail to identify all pediatric cardiac surgical procedures. Failure to target all at-risk procedures leads to inaccurate reporting of SSI rates largely based on identifying the denominator. Inaccurate recording of SSI data has implications for public reporting, benchmarking of outcomes, and denial of payment. Use of The Society of Thoracic Surgeons Congenital Heart Surgery Database as the gold standard to identify procedures for surveillance will lead to more accurate reporting of SSI rates.