PELS (Procedia of Engineering and Life Science)
Vol. 9 (2025): Proceedings of the 2025 Annual Meeting of APTIRMIKI

Length of Stay Patterns and Their Relation to Coding Accuracy: Polanya Lama Tinggal Pasien dan Hubungannya dengan Akurasi Koding

Muchlis, Husni Abdul (Unknown)
Qomarania, Witri Zuama (Unknown)
Nurmalasari, Mieke (Unknown)
Kurniawati, Anastasia Cyntia Dewi (Unknown)
Lestari, Betri Widya (Unknown)



Article Info

Publish Date
07 Nov 2025

Abstract

Hospitals face efficiency and quality challenges within the Case-Based Groups (CBG's) financing system, where a patient’s Length of Stay (LOS) is critical. Accurate LOS data is crucial for strategic decisions, cost management, and quality care. A study at a Type B Hospital in Bekasi City found significant variation and outliers in LOS, indicating a non-normal distribution. This observational analytic study, involving 3,151 inpatient claims from January 2024, analyzed LOS data and its impact on clinical documentation and coding quality. The analysis compared the Arithmetic Mean Length of Stay (AMLOS) and the Geometric Mean Length of Stay (GMLOS) to identify outliers, followed by a Wilcoxon test. Results showed LOS varied from 1 to 48 days, with an AMLOS of 7.13 and a GMLOS of 6.76 days, indicating positive skewness from outliers. AMLOS was consistently higher than GMLOS in the top 10 CBG's, especially for moderate and severe cases. The Wilcoxon test (p<0.05) confirmed a significant statistical difference, showing GMLOS more accurately represents the appropriate LOS. The presence of outliers (e.g., >30 or 44 days) suggests potential issues with documentation or coding. Therefore, using the more robust GMLOS is crucial for hospitals to optimize management, improve care, and maintain the quality of clinical documentation and coding.

Copyrights © 2025






Journal Info

Abbrev

PELS

Publisher

Subject

Computer Science & IT Control & Systems Engineering Electrical & Electronics Engineering Engineering

Description

PELS (Procedia of Engineering and Life Science) is an international journal published by Faculty of Science and Technology Universitas Muhammadiyah Sidoarjo. The research article submitted to this online journal will be double blind peer-reviewed (Both reviewer and author remain anonymous to each ...