Table 1.
Datasets included in the analysis
| Dataset accession | First author | Cohort description | Timing of sepsis diagnosis | Percent bacterial infection | Age | Sex (% male) | Severity | Country | No. survived | No. died |
|---|---|---|---|---|---|---|---|---|---|---|
| 1a: Discovery Cohorts | ||||||||||
| E-MEXP-3567 | Irwin | Children with meningococcal sepsis +/− HIV co-infection | Admission to ED | 100 | 2.0 (IQR 0.6–6.9) | 55 | unk. | Malawi | 6 | 6 |
| E-MEXP-3850 | Kwan | Children w/ meningococcal sepsis | Admission to hospital; sampled at multiple times 0–48 h | 100 | 1.3 (range 0.8–2.0) | 40 | PELOD; 29.2 (range 11–61) | UK | 19 | 5 |
| E-MTAB-1548 | Almansa | Adult surgical patients with sepsis (EXPRESS study) | Average post-operation day 4 (hospital acquired) | 100 | 69.7 (std. dev. 13.1) | 67 | APACHE II 17.0 (std. dev. 5.4) | Spain | 50 | 24 |
| GSE10474 | Howrylak | Adults in MICU with sepsis +/− ALI | Admission to ICU | 75+ | 57 (std. dev. 4.3) | 45 | APACHE II 20.7 (std. dev. 1.6) | USA | 22 | 11 |
| GSE13015a | Pankla | Adults with sepsis, many from burkholderia | Within 48 h of diagnosis; both community-acquired and hospital-acquired | 100 | 54.7 (std. dev. 11.7) | 54 | unk. | Thailand | 35 | 13 |
| GSE13015b | 8 | 7 | ||||||||
| GSE27131 | Berdal | Adults with severe H1N1 influenza requiring mechanical ventilation | Admission to ICU | 0 | unk. | unk. | SAPS II 29.3 (std. dev. 10.3) | Norway | 5 | 2 |
| GSE32707 | Dolinay | Adults in MICU with sepsis+/− ARDS | Admission to ICU | unk. | 57.1 (std. dev. 14.9) | 53 | APACHE II 26.7 (std. dev. 8.5) | USA | 31 | 17 |
| GSE40586 | Lill | Infants, children, and adults with bacterial meningitis | Within 48 h of hospital admission | 100 | 43.4 (range 17 days –70 years) | unk. | unk. | Estonia | 19 | 2 |
| GSE63042 | Langley | Adults with sepsis (CAPSOD study) | Admission to ED | 80+ | 59.1 (std. dev. 18.3) | 59 | APACHE II 16.5 (std. dev. 7.3) | USA | 76 | 28 |
| GSE66099 | Wong | Children in ICU with sepsis/septic shock | Admission to ICU | 72 | 3.7 | 58 | PRISM 15.7 | USA | 171 | 28 |
| GSE66890 | Kangelaris | Adults in ICU with sepsis +/− ARDS | Admission to ICU | 63 (std. dev 19) | 56 | APACHE III 100 (std. dev. 35) | USA | 43 | 14 | |
| 1b: Validation cohorts | ||||||||||
| GSE21802 | Bermejo-Martin | Adults in ICU with severe H1N1 influenza | Within 48 h of admission to ICU | 0 | 43 (std. dev. 11) | 47 | SOFA 4.1 (std. dev. 3.5) | Spain | 7 | 4 |
| GSE33341 | Ahn | Adults with 2+ SIRS criteria and bacteremia | Within 24 h of admission to hospital | 100 | 58 (range 24–91) | 61 | unk. | USA | 49 | 2 |
| GSE54514 | Parnell | Adults in ICU with sepsis | Admission to ICU | unk. | 61 (std. dev. 16) | 40 | APACHE II 21 (std. dev. 6) | Australia | 26 | 9 |
| GSE63990 | Tsalik | Adults with bacterial infection plus 2 + SIRS criteria | Admission to ED | 100 | 49 (range 14–88) | 50 | unk. | USA | 64 | 6 |
| E-MTAB-4421.51 | Davenport | Adults with sepsis (GAinS study) | Day of hospital admission | 92 | 64.2 (std. dev. 15.2 | 55 | APACHE II 18.6 (std. dev. 9.7) | UK | 15 | 7 |
| 1c: Hospital-acquired infection cohorts | ||||||||||
| Duke HAI | Tsalik (unpublished) | Adults who developed ventilator-associated pneumonia (VAP) | Hospital days 1–30 | unk. | 58.0 (std. dev. 17.9) | 75 | unk. | USA | 60 | 10 |
| Glue Grant Burns | Glue Grant authors | Adults with severe burns (whole blood) | Hospital days 1–30 | 100 | 14.1 (std. dev. 16.2) | 64 | Denver Score 1.5 (S 1.7) | USA | 84 | 8 |
| Glue Grant Trauma | Glue Grant authors | Adults with severe traumatic injuries (buffy coat) | Hospital days 1–30 | 100 | 33.2 (std. dev. 10.2) | 74 | MODS 6.4 (std. dev. 3.3) | USA | 48 | 1 |
| UF P50 12H | Moldawer (unpublished) | Adults with hospital-acquired sepsis | Hospital days 1–30 | 100 | unk. | unk. | SOFA 5.5 (std. dev. 3.9) | USA | 66 | 5 |
Unk, unknown data or not available; IQR, inter-quartile range; std. dev., standard deviation; ED, emergency department; ICU, intensive care unit; MICU, medical ICU; ARDS, acute respiratory distress syndrome; SIRS, systemic inflammatory response syndrome; VAP, ventilator-associated pneumonia