Cost-Effective Enhancement of Claims Data to Improve Comparisons of Patient Safety

Publisher Website
Google Scholar
Abstract
Tools that support screening for medical errors can help to identify potential patient safety events for further investigation and can provide benchmarks against which providers, localities, and states can compare themselves. The Agency for Healthcare Research and Quality Patient Safety Indicators, which are based solely on hospital administrative or claims data, represent one such tool. Without sufficient clinical detail, measures based on claims data may not accurately reflect hospital quality of care. To construct risk-adjustment models, we used hospital discharge data from July 2000 to June 2003 from 188 Pennsylvania hospitals supplied by the Pennsylvania Health Care Cost Containment Council. We augmented the hospital claims data with clinical data (also supplied by the Pennsylvania Health Care Cost Containment Council) abstracted from medical records using MediQual's proprietary Atlas™ (MediQual, Westborough, MA, a subsidiary of CardinalHealth) clinical information system. Clinical data elements included such items as patient history, laboratory results, vital signs, and other clinical findings. Our cost-effectiveness analyses strongly support the value of enhancing administrative claims data with a present-on-admission code and adding a limited set of numerical laboratory values. Reasonable additional benefit may be gained by adding vital signs to this data set, but the trade-off between effectiveness and cost is not as clear. Also, more accurate International Classification of Diseases, Ninth Revision, Clinical Modification coding of specific secondary diagnoses that are currently undercoded could improve the validity of risk-adjustment equations without the added cost of abstracting clinical findings from medical records. There seems to be little justification for secondary abstraction of medical records to obtain data for risk-adjusting the Agency for Healthcare Research and Quality Patient Safety Indicators.