RESUMEN
INTRODUCTION: Postoperative fever following elective colectomy is common and often results in comprehensive laboratory and radiographic testing. We hypothesized that risk factors for febrile complications may be identified with data available at the time of initial fever episode and that a significant proportion of patients exists in which additional testing is of minimal utility. METHODS: We performed a retrospective case-control study of patients undergoing elective colorectal resection at a single institution between 2015 and 2020 with postoperative fever ≥38.0°C. Febrile patients with a clinically significant fever source requiring a change in management were compared to patients with fever that resolved spontaneously. An additive risk score from 0 to 3 was generated from selected characteristics with P-value <0.001. Multivariable logistic regression was used to model the odds of infection with the risk score entered as a binary variable (0-1 versus 2-3 risk factors). RESULTS: Of 1036 elective colectomy patients, 143 (13.8%) had postoperative fever ≥38.0°C. Among the febrile patients, 27 had a positive evaluation (18.9%). Active smoking status (P = 0.018), time from surgery to fever, fever ≥38.5°C, other vital sign changes, and documented localizing signs/symptoms (all, P < 0.001) were associated with an identifiable source of fever. On multivariable regression, the presence of 2-3 risk factors and fever onset after postoperative day 1 were associated with a positive fever evaluation. CONCLUSIONS: Few patients with fever after colectomy required a change in clinical management. Those without multiple risk factors elicited from an interval history and physical exam are unlikely to benefit from additional testing.
Asunto(s)
Colectomía , Complicaciones Posoperatorias , Estudios de Casos y Controles , Colectomía/efectos adversos , Colectomía/métodos , Procedimientos Quirúrgicos Electivos/efectos adversos , Procedimientos Quirúrgicos Electivos/métodos , Fiebre/epidemiología , Fiebre/etiología , Humanos , Incidencia , Complicaciones Posoperatorias/diagnóstico , Complicaciones Posoperatorias/epidemiología , Complicaciones Posoperatorias/etiología , Estudios Retrospectivos , Factores de RiesgoRESUMEN
BACKGROUND: We investigated low-grade, hemodynamically normal BHI associated with any of three interventions indicating ICU observation. METHODS: BHI between 01/01/96-6/30/14 were reviewed. Two groups included: Group A (BHI grades 1-3) with normal initial systolic BP and Group B (all other BHI). Interventions necessitating ICU observation were defined with three criteria: PRC transfusion within 24 h, angiography, or laparotomy. Between group and Group A subgroup outcomes were reported. RESULTS: Group A (n = 1088) had lower ISS, LOS, PRCs transfused, and mortality (p < 0.01) than Group B (n = 636). For any of the criteria indicating ICU admission, Group A had a NPV, sensitivity and specificity of 67.9%, 30.3%, and 75.3% respectively; isolated BHI (n = 188) sensitivity, specificity and NPV were 17.8%, 88.1%, and 77.3%. Laparotomy specifically for BHI was 2.0% for Grade I/II, 4.3% for Grade III subgroups. CONCLUSIONS: Hemodynamic stability is insufficient as a sole criterion for safe admission of low-grade BHI to a non-ICU environment.