Date of Award

7-2022

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Program

Health Outcomes and Policy Research

Track

Health Services Research

Research Advisor

Rebecca B. Reynolds, PhD

Committee

Peter R. DeVersa, MD Charisse Madlock-Brown, PhD Angela L. Morey, PhD Simonne S. Nouer, PhD

Keywords

Medical Coding, Sepsis

Abstract

Background: Sepsis is a condition that can be very costly and very deadly. Diagnosing sepsis can be challenging as there is not one specific test that will identify whether a patient has sepsis and there are varying opinions as to the true definition of sepsis. The definition of sepsis used for this research is a combination of System Inflammatory Response Syndrome (SIRS) with an identified infection. Medical Coders must review the documentation provided in a medical record to accurately assign an ICD-10-CM code. Administrative data is then used to provide statistical information for research purposes. When coded data is not accurate, this leads to errors in administrative data and inaccuracies in research. Objectives: The main goal of this study was to identify the accuracy of medical coding for sepsis patients. There were six research questions that guided the research. These included 1) Are cases coded as sepsis that are not clinically supported as sepsis; 2) Are infection cases not coded as sepsis clinically supported as sepsis; 3) Are there any variances for certain physicians; 4) Are there any variances for certain physician specialties; 5) Are there any variances for certain payers; 6) Are there any variances for certain medical coders? Methods: We used a convenience sampling of patient records from 4th quarter 2019 from Erlanger Health Systems that were coded as sepsis and a sampling that were coded as an infection without sepsis. Research Design and Study Procedures: Following Institutional Review Board (IRB) approval from both Erlanger Health Systems and the University of Tennessee Health Science Center (UTHSC), a chart review was conducted. Clinical indicators identified in the created data abstraction tool were abstracted from the patient records. Results: Data analysis concluded that the accuracy rate of medical coding for the sepsis patient records based on the clinical documentation was 98.5%. Physician specialty and payer type had no impact on the accuracy of medical coding on these records. Data analysis concluded the accuracy rate of medical coding for the infection patient records based on clinical documentation was 59%. Logistical regression also identified there were no variances in the coding for the infection patients based on the payer type, medical coder years of inpatient coding experience and the medical coders education level. Analysis determined there was a variance in coding accuracy of the infection patients group based on physician specialty.

Declaration of Authorship

Declaration of Authorship is included in the supplemental files.

ORCID

https://orcid.org/0000-0002-8446-7155

DOI

10.21007/etd.cghs.2022.0599

Additional Files
2022-015-Insco-DOA.pdf (222 kB)
Declaration of Authorship
Download

Share

COinS