Studying veteran ED pain care with an electronic pain care abstraction tool
Ula Hwang MD MPH
James J. Peters VA Medical Center, Bronx, NY
Bronx, NY
Funding Period: September 2010 - August 2011

BACKGROUND/RATIONALE:
This is a pilot project assessing the feasibility of collecting Veteran ED pain care data. Information gathered from this pilot project will provide descriptive data on the quality of acute pain care at Veterans Health Administration (VHA) emergency departments (EDs) and preliminary data to support the long-term goal of developing an electronic medical record abstraction technique to support research and quality improvement methods to accurately abstract information on Veteran ED pain care. The development of a pain care informatics abstraction tool is aligned with VHA clinical and research priorities.

OBJECTIVE(S):
Project objectives are: 1.To assess the feasibility of extracting process of pain care measures with HUMAN RECORD REVIEW of VHA ED medical records; 2.To explore sources and types of data required to extract (future) processes of pain care measures using ELECTRONIC ABSTRACTION of data from VHA sources by identifying alternate data sources of Veteran ED pain care from local (X), regional (Y), and national (Z) VHA sources and identifying the data types available (whether as structured or free text information) and how to efficiently and effectively access them; and 3.To provide recommendations on improving/modifying/structuring VHA data for accurate and consistent electronic abstraction of process of pain care data (e.g., implementing templates for missing data).

METHODS:
This will be a retrospective medical record review of ED pain care at three VA medical centers (VAMCs)(Bronx VA, West Haven VA, West LA VA). A total of 450 ED medical records will be randomly selected for medical record review of process of pain care measures in the ED setting during the study period of 1/1/09-12/31/09. The feasibility of collecting ED process of pain care data with human record review will provide pilot study information to support a future proposal to develop an electronic pain care abstraction tool (EPCAT), using these human abstraction data as the gold standard for automated review in a future proposal.

Data sources include local CPRS electronic text records, national access to text records through Vista Capri, and national database access to CDWWORD on VINCI (tables Dim.Location, Dim.OrderableItem, Outpat.Visit, CPRSOrder.CPRSORder, CPRSOrder.OrderedItem). Review of ED visit medical records were completed by a trained research assistant and investigators for chief complaints of painful conditions and for the quality of acute pain care received. This included documentation of pain assessment, recorded pain scores, and administration of analgesic medications.

FINDINGS/RESULTS:
A total of 435 medical records were reviewed of randomly selected ED visits at the three sites: 93% were male, mean age was 61 years (sd 49), 51% were White, 14% were Hispanic. Sixty-one percent of these ED visits involved a chief complaint (cc) of a painful condition. Of those with cc of painful conditions: 80% had pain scores available from national database sources ((on a scale of 0-10, 10 worst pain ever) mean 5.85 (sd 3.50); 38% had physician documentation of a pain assessment in the text note; 45% received ED analgesics (vs. 13% of those with cc of nonpainful cc, p<0.0001); mean time to first analgesic was 101 min (sd 101) (vs. 142 min (sd 139 of those with nonpainful cc), p=0.11). There were no differences in pain documentation nor the administration of analgesics for subjects based on age, gender, or race/ethnicity.

By site, however, there was significant variation as to what portion of patients had any recorded pain scores (database or physician documentation) (p<0.0001) and whether or not patients received ED analgesics (p<0.0001). The physician pain score documentation ranged from 8% to 64% while receiving any ED analgesics ranged from 14% to 41%. These differences remained when adjusted for painful cc, gender, age, and race/ethnicity.

IMPACT:
Based on preliminary findings, unlike national EDs, these three VAMC EDs do not have significant disparities in the quality of ED pain care for veterans by age, race/ethnicity, nor gender. By site, however, there were significant differences in the level of pain score documentation and whether or not analgesics were administered for acute pain. These findings indicate there is wide variation in the quality of acute pain care delivered at VAMC ED's with regards to pain assessment and treatment. Use of electronic medical record abstraction techniques to assess for pain may not be feasible if clinician documentation of pain care remains minimal. Attempts to standardize and improve acute pain care with documented assessment and treatment across different VAMCs are needed before the development and implementation of an EPCAT can be effective.

PUBLICATIONS:

Book Chapters

1. Hwang U. Geriatric Trauma. In: Shah K, Egan D, Quaas J, editors. Essential Emergency Trauma. 1 ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2010. 394-398 p.

DRA: Musculoskeletal Disorders, Neurodegenerative Diseases
DRE: Research Infrastructure, Diagnosis
Keywords: Data Management, Decision Support, Information Management
MeSH Terms: none