Browse Prior Art Database

Clinical Content Categorization from Electronic Medical Records (EMRs) without Assignment of Codes

IP.com Disclosure Number: IPCOM000245141D
Publication Date: 2016-Feb-12

Publishing Venue

The IP.com Prior Art Database

Abstract

An Evidence Based Platform (EBP) is a methodology for developing and leveraging applications that categorize the clinical content of health care data (such as APR-DRG and other quality or performance measures) directly from electronic medical records without assigning medical codes. The method described in this paper provides an efficiently automated, cost effective, transparent, and defensible solution for assessing the quality, risk, and outcome of a patient healthcare encounter.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 8% of the total text.

Page 01 of 14

Clinical Content Categorization from Electronic Medical Records (EMRs) without Assignment of Codes

Abstract

An Evidence Based Platform (EBP) is a methodology for developing and leveraging applications that categorize the clinical content of health care data (such as APR-DRG and other quality or performance measures) directly from electronic medical records without assigning medical codes. The method described in this paper provides an efficiently automated, cost effective, transparent, and defensible solution for assessing the quality, risk, and outcome of a patient healthcare encounter.

Introduction

An evidence based platform allows applications to be built that use the EMR directly, rather than requiring that a coder or auto-coding software be used to extract information and render it in coded form (e.g., ICD, CPT, and the like). The method described in this paper is described in detail using the All Patient Refined-Diagnostic Related Groups (APR-DRGs) as an example of an application that could be built to operate on an evidence based platform. The word "evidence" is defined as any information in the medical record that might currently be extracted and used in a "grouper" rather than the word "evidence" as it is used in "evidence-based medicine", where it means "scientifically established clinical relationships as reported in peer-reviewed literature".

A conventional grouper works bottom-up. It begins with a set of codes, dates, and a few patient descriptors like sex, age, and discharge status and builds a structure on top of those to arrive at its objective. Along the way, it may define intermediate concepts and build upon those. The word "grouper" is defined as any program that calculates a categorization (of any kind) that is based, at least in part, on clinical data in the medical record. Quality and performance measurements are included, even though they are not usually called "groupers."

Bottom-up groupers like APR-DRGs and Clinical Risk Groups (CRGs) rely on being presented with comprehensive records composed of diagnosis and procedure codes (e.g., ICD-9 or - 10, CPT, ICD-10-PCS), which are expensively derived from the medical record. An entire industry has matured around these classification systems and the requirement for "correct coding." Yet, groupers typically use only a small fraction of the information that is so assiduously collected. Much of the detail represented by the codes is ignored, as the grouper frequently treats all codes identically in a list of potentially thousands of codes. Further, it is not uncommon for the APR-DRG grouper, for example, to be given fifty codes for one inpatient stay - but it never uses more than five codes. From the standpoint of computing an APR-DRG, wasted time, effort, and money are visible results.

Classification systems' reliance on codes is an artifact of the pre-digital age. In 1974 (and since updated), a set of guidelines (UHDDS - Uniform Hospital Discharge Data Set) was...