Critical Care (Jun 2020)

Significance of body temperature in elderly patients with sepsis

  • Takashi Shimazui,
  • Taka-aki Nakada,
  • Keith R. Walley,
  • Taku Oshima,
  • Toshikazu Abe,
  • Hiroshi Ogura,
  • Atsushi Shiraishi,
  • Shigeki Kushimoto,
  • Daizoh Saitoh,
  • Seitaro Fujishima,
  • Toshihiko Mayumi,
  • Yasukazu Shiino,
  • Takehiko Tarui,
  • Toru Hifumi,
  • Yasuhiro Otomo,
  • Kohji Okamoto,
  • Yutaka Umemura,
  • Joji Kotani,
  • Yuichiro Sakamoto,
  • Junichi Sasaki,
  • Shin-ichiro Shiraishi,
  • Kiyotsugu Takuma,
  • Ryosuke Tsuruta,
  • Akiyoshi Hagiwara,
  • Kazuma Yamakawa,
  • Tomohiko Masuno,
  • Naoshi Takeyama,
  • Norio Yamashita,
  • Hiroto Ikeda,
  • Masashi Ueyama,
  • Satoshi Fujimi,
  • Satoshi Gando,
  • on behalf of the JAAM FORECAST Group

DOI
https://doi.org/10.1186/s13054-020-02976-6
Journal volume & issue
Vol. 24, no. 1
pp. 1 – 9

Abstract

Read online

Abstract Background Elderly patients have a blunted host response, which may influence vital signs and clinical outcomes of sepsis. This study was aimed to investigate whether the associations between the vital signs and mortality are different in elderly and non-elderly patients with sepsis. Methods This was a retrospective observational study. A Japanese multicenter sepsis cohort (FORECAST, n = 1148) was used for the discovery analyses. Significant discovery results were tested for replication using two validation cohorts of sepsis (JAAMSR, Japan, n = 624; SPH, Canada, n = 1004). Patients were categorized into elderly and non-elderly groups (age ≥ 75 or < 75 years). We tested for association between vital signs (body temperature [BT], heart rate, mean arterial pressure, systolic blood pressure, and respiratory rate) and 90-day in-hospital mortality (primary outcome). Results In the discovery cohort, non-elderly patients with BT < 36.0 °C had significantly increased 90-day mortality (P = 0.025, adjusted hazard ratio 1.70, 95% CI 1.07–2.71). In the validation cohorts, non-elderly patients with BT < 36.0 °C had significantly increased mortality (JAAMSR, P = 0.0024, adjusted hazard ratio 2.05, 95% CI 1.29–3.26; SPH, P = 0.029, adjusted hazard ratio 1.36, 95% CI 1.03–1.80). These differences were not observed in elderly patients in the three cohorts. Associations between the other four vital signs and mortality were not different in elderly and non-elderly patients. The interaction of age and hypothermia/fever was significant (P < 0.05). Conclusions In septic patients, we found mortality in non-elderly sepsis patients was increased with hypothermia and decreased with fever. However, mortality in elderly patients was not associated with BT. These results illuminate the difference in the inflammatory response of the elderly compared to non-elderly sepsis patients.

Keywords