Early recognition of sepsis is a key factor to improve survival to this disease in surgical patients, since it allows prompt control of the infectious source. Combining pro-inflammatory and immunosupression biomarkers could represent a good strategy to improve sepsis detection. Here we evaluated the combination of procalcitonin (PCT) with gene expression levels of HLA-DRA to detect sepsis in a cohort of 154 surgical patients (101 with sepsis and 53 with no infection). HLA-DRA expression was quantified using droplet digital PCR, a next-generation PCR technology. Area under the receiver operating curve analysis (AUROC) showed that the PCT/HLA-DRA ratio outperformed PCT to detect sepsis (AUROC [CI95%], p): PCT: 0.80 [0.73-0.88], <0.001; PCT/HLA-DRA: 0.85 [0.78-0.91], <0.001. In the multivariate analysis, the ratio showed a superior ability to predict sepsis compared to that of PCT (OR [CI 95%], p): PCT/HLA-DRA: 7.66 [1.82-32.29], 0.006; PCT: 4.21 [1.15-15.43] 0.030. Multivariate analysis was confirmed using a new surgical cohort with 74 sepsis patients and 21 controls: PCT/HLA-DRA: 34.86 [1.22-995.08], 0.038; PCT: 5.52 [0.40-75.78], 0.201. In conclusion, the combination of PCT with HLA-DRA is a promising strategy for improving sepsis detection in surgical patients.