Platelet-to-Lymphocyte Ratio as a Predictor of Exacerbation Severity and Hospital Length of Stay in Acute Exacerbation of COPD: A Dual-Center Study in West Sumatra, Indonesia

Abstract

Introduction: Acute exacerbation of chronic obstructive pulmonary disease (AECOPD) is a leading driver of morbidity, mortality and health-care expenditure, and simple admission biomarkers that grade severity are needed in resource-limited settings. The platelet-lymphocyte ratio (PLR) integrates thrombo-inflammation and relative lymphopenia, yet Indonesian evidence linking it to exacerbation severity and length of stay (LOS) remains limited.

Methods: In this dual-center analytical cross-sectional study, 141 hospitalized AECOPD patients at Dr. M. Djamil General Hospital, Padang and RS Madina Bukittinggi (January-December 2024) were analyzed. Admission PLR was related to Anthonisen exacerbation severity and LOS (>7 days) using Kruskal-Wallis and Mann-Whitney tests, Spearman correlation, receiver-operating-characteristic (ROC) analysis and multivariable logistic regression (adjusted odds ratios, aOR).

Results: Most patients were male (86.5%), aged 40-70 years (56.0%) and heavy smokers (87.5%). Mean PLR increased across severity strata (mild 126.8±40.8, moderate 192.3±69.1, severe 444.9±241.7; p<0.001) and correlated moderately with severity (ρ=0.534, p<0.001) but weakly with LOS (ρ=0.295). PLR discriminated severe exacerbation with excellent accuracy (AUC 0.929, 95% CI 0.874-0.985; cut-off ≥216.3, sensitivity 100.0%, specificity 73.2%, accuracy 89.6%) but predicted LOS poorly (AUC 0.566, p=0.273). After adjustment, PLR independently predicted severe exacerbation (aOR 3.73 per 50 units, 95% CI 1.87-7.42, p<0.001; Nagelkerke R2=0.642), whereas prolonged stay was driven by pneumonia (aOR 6.40, 95% CI 2.50-16.37, p<0.001) rather than PLR (aOR 1.15, p=0.139).

Conclusion: Admission PLR is an inexpensive, widely available biomarker that accurately identifies severe AECOPD and may support early risk stratification, although it should not be used alone to predict length of stay, which is governed chiefly by comorbidity.