Dynamic platelet profiles in the emergency department and their predictive value for 28-day all-cause mortality in septic patients.

backgroundSepsis is a significant cause of global mortality and most frequently first presents in the emergency department. Dynamic platelet changes play a key pathophysiological role in sepsis, yet their comprehensive prognostic value-particularly from a multi-dimensional perspective that includes value, magnitude, and timing-remains underexplored.

methodsIn this retrospective cohort study, 363 episodes of sepsis first diagnosed in the emergency department of Beijing Aviation General Hospital between April 2020 and January 2025 were included. We systematically analyzed three dimensions of platelet counts: 1) nadir value; 2) the magnitude of platelet decline; 3) dynamic timing (time to nadir and maximum decline). Multivariable Cox regression models were used to assess their independent associations with 28-day all-cause mortality, adjusting for demographics, comorbidities, infection characteristics, and disease severity.

resultsAmong 363 patients (median age 81 years, 54.5% male), the 28-day mortality rate was 39.9%. Compared to survivors, non-survivors had a lower platelet nadir [70.0 vs. 134 × 10⁹/L, p < 0.001] and a greater magnitude of platelet decline [62.8% vs. 31.5%, p < 0.001].Multivariable analysis revealed dose-response relationships for platelet nadir (moderate [50-99 × 10⁹/L]: aHR = 2.04, p = 0.014; severe [20-49 × 10⁹/L]: aHR = 2.98, p < 0.001; profound [<20 × 10⁹/L]: aHR = 3.46, p = 0.003) and magnitude of decline (moderate [30-50%]: aHR = 1.79, p = 0.044; severe [>50%]: aHR = 2.43, p < 0.001). Timing analysis demonstrated that intermediate and late platelet nadir were associated with higher mortality risk (intermediate: aHR = 4.39-7.28, all p < 0.001; late: aHR = 5.85-9.58, all p < 0.001), while initial and early nadir showed no significant association. Similarly, both intermediate and late maximum platelet decline were associated with increased mortality (intermediate: aHR = 2.70-2.98, all p < 0.001; late: aHR = 6.33-6.87, all p < 0.001).

conclusionThrough stepwise analysis, we demonstrate that dynamic platelet parameters-magnitude and timing of decline-independently predict 28-day mortality in sepsis. The magnitude applies to all patients but is static, while timing adds a critical dimension: the same degree of decline has different prognostic implications depending on when it occurs. Delayed decline and intermediate/late nadir indicate poor prognosis, whereas early nadir does not. These findings underscore the importance of longitudinal platelet monitoring in the emergency setting, as the trajectory-integrating both magnitude and timing-identifies at-risk patients, including those without thrombocytopenia.