Serial Measurement of Pancreatic Stone Protein (PSP) for the Early Recognition and Diagnosis of Sepsis in Intensive Care Unit (ICU) Patients in China: A Multicenter, Prospective, Biomarker-result-blinded Observational Study

In this study, 250 patients with high-risk of sepsis will be enrolled, including ≥ 50 subjects with confirmed sepsis. Diagnostic criteria for sepsis should meet the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3), i.e., simultaneously meet the following two conditions: a) confirmed or suspected infection; b) the SOFA score has increased by ≥ 2 points compared with the baseline. Subjects who meet the inclusion criteria and do not meet the exclusion criteria will be collected 3 ml of peripheral venous blood samples daily (together with blood tests such as complete blood count for normal diagnosis and treatment purposes) until the diagnosis of sepsis or other diseases was confirmed.

During the blood collection period, relevant clinical information, laboratory examination results, treatment information and SOFA scores of the subjects will be collected. This study does not produce any intervention in the normal clinical diagnosis and treatment of the subjects. After the subjects were diagnosed with sepsis, a 28-day follow-up will be performed to record the number of days of ICU treatment and the survival of the subjects 7 and 28 days after the diagnosis of sepsis. If the subjects were diagnosed with septic shock during the follow-up period, they will be recorded as "Ds". If the subjects were still hospitalized in the study center, 3 ml of peripheral venous blood samples will be collected. The completion of the 28-day follow-up will be considered the end of the study.

This study aims to:

1. Comparing to reference method, the clinical diagnosis of sepsis (sepsis-3), and reference reagents, CE-marked IVD PSP capsule on the point-of-care abioSCOPE® device (Abionic SA), C-reactive protein (CRP) assay kit, procalcitonin (PCT) assay kit, etc., to verify and evaluate the comprehensive performance of PSP as a biomarker in the early recognition and diagnosis of sepsis manifesting within the first 3 days after testing. Diagnostic accuracy analysis [including sensitivity, specificity, positive predicted value, negative predicted value, positive likelihood ratio, negative likelihood ratio, area under ROC curve (AUC), etc.], and consistency analysis (including positive coincidence rate, negative coincidence rate, total coincidence rate, Kappa value, etc.) will be performed.

2. The clinical performance of PSP for the diagnosis of sepsis will be evaluated based on the cross-sectional study data (on the day of confirmed visit).

3. The clinical performance of PSP for recognition of sepsis manifesting within the first 3 days after testing will be evaluated based on the longitudinal study data (multiple visits data).

4. Comparing to the reference method and the reference reagent test results, try to use machine learning & deep learning methods to select a subset of relevant features (variable screening) from clinical information and biomarker data (CRP, PCT, IL-6, NT-proBNP, hs-cTnI, SAA, Cys C, CAL, etc.), to construct an innovative combined diagnostic model with PSP, "PSP+ X" model. The comprehensive performance of "PSP+X" model for the early recognition and diagnosis of sepsis manifesting within the first 3 days after testing will be evaluated. Diagnostic

5. The clinical performance of "PSP+X" mode for the diagnosis of sepsis will be evaluated based on the cross-sectional study data (on the day of confirmed visit).

6. The clinical performance of "PSP+X" mode for early recognition of sepsis manifesting within the first 3 days after testing will be evaluated based on the longitudinal study data (multiple visits data).