Table 2 Association of subphenotypes with ICU mortality and 28-day mortality.

From: A machine learning-based prediction model for poor prognosis in sepsis using lymphocyte count: a national, multicenter prospective cohort

28-day mortality

PL

SIL

NL

RDL

HR (95%CI)

p

HR (95%CI)

p

HR (95%CI)

p

(A)

       

Derivation cohort

       

 Unadjusted

Reference

0.61 (0.49–0.75)

 < 0.001

0.55 (0.41–0.73)

 < 0.001

0.86 (0.66–1.13)

0.283

 Adjusted

Reference

0.71 (0.57–0.88)

0.002

0.67 (0.50–0.89)

0.005

0.92(0.70–1.20)

0.520

External validation cohort

       

 Unadjusted

Reference

0.46 (0.34–0.63)

 < 0.001

0.26 (0.17–0.40)

 < 0.001

0.65 (0.38–1.11)

0.116

 Adjusted

Reference

0.52 (0.36–0.74)

 < 0.001

0.34 (0.20–0.57)

 < 0.001

0.61 (0.33–1.13)

0.118

ICU mortality

PL

SIL

NL

RDL

HR (95%CI)

p

HR (95%CI)

p

HR (95%CI)

p

(B)

       

Derivation cohort

       

 Unadjusted

Reference

0.69 (0.56–0.86)

 < 0.001

0.61 (0.46–0.82)

 < 0.001

0.86 (0.65–1.14)

0.313

 Adjusted

Reference

0.75 (0.61–0.94)

0.012

0.76 (0.57–0.99)

0.048

0.90 (0.68–1.20)

0.491

External validation cohort

       

 Unadjusted

Reference

0.62 (0.47–0.89)

0.007

0.50 (0.33–0.77)

0.001

0.70 (0.41–1.19)

0.187

 Adjusted

Reference

0.62 (0.44–0.87)

0.006

0.49 (0.30–0.80)

0.004

0.64 (0.35–1.16)

0.141

  1. Multivariate Cox regression was adjusted by age, sex, SOFA, APACHEII, Immunocompromised status and pulmonary infection. Abbreviations: SOFA: Sequential Organ Failure Assessment score, APACHE II: The Acute Physiology and Chronic Health Evaluation II score, HR: hazard ratio, PL: Persistent Lymphopenia, SIL: Slowly Increasing Lymphocyte, NL: Normal Lymphocyte, RDL: Rapidly Declining Lymphocyte.
Back to article page