Platelet Indices as Predictive Markers for Sepsis

Abstract

Abstract Background: Sepsis continues to represent a significant global health challenge which is defined as an aberrant host reaction to infection that results in potentially fatal organ damage. Although recent strides have been made to improve early detection, prompt diagnosis still remains a big challenge. This study assessed the predictability of Platelet Distribution Width (PDW), Platelet count (PLT), and Mean Platelet Volume (MPV) in sepsis. Methodology: This case-control study was conducted from June to September 2024 at the Intensive Care Unit of Tamale Teaching Hospital in Tamale, Ghana. Clinical records and Full Blood Counts (FBC) of 300 participants; including 150 participants diagnosed with sepsis and 150 participants without sepsis were assessed. The sample size of 300 participants was chosen to reflect equal representation between the two groups while maintaining statistical power for meaningful comparison. The smaller sample size, however, constrains subgroup analyses and wider generalizability. Further research with bigger sample sizes is recommended to confirm these results and explore other subgroups. Data was analysed using SPSS version 27.0, and statistical significance was set at p < 0.05. Results: platelet count was found to be significantly lower (p = 0.001) in participants with sepsis (79.0[56.50 - 111.00] x 109g /L) than participants without sepsis (98.50[76.00 - 123.75]). PDW was significantly higher in participants with sepsis (12.10[8.45-14.35] fL) than their non sepsis counterparts (9.70[7.9 - 14.2] fL). Also, PCT was significantly higher in sepsis patients (0.21[0.14 - 0.32] %) than their counterparts without sepsis (0.19[0.13 - 0.30] %). PLT significantly predicted sepsis (AUC: 0.919, sensitivity: 72.8%, specificity: 96.6%, p < 0.001) in sepsis patients compared to MPV, PDW, and PCT. Conclusion: Sepsis was significantly predicted by platelet parameters, with PLT being the most powerful predictor.