Clinical prediction rules for bacteremia and in-hospital death based on clinical data at the time of blood withdrawal for culture: An evaluation of their development and use

J Eval Clin Pract. 2006 Dec;12(6):692-703. doi: 10.1111/j.1365-2753.2006.00637.x.

Abstract

Rationale, aims and objectives: To develop clinical prediction rules for true bacteremia, blood culture positive for gram-negative rods, and in-hospital death using the data at the time of blood withdrawal for culture.

Methods: Data on all hospitalized adults who underwent blood cultures at a tertiary care hospital in Japan were collected from an integrated medical computing system. Logistic regression was used for developing prediction rules followed by the jackknife cross validation.

Results: Among 739 patients, 144 (19.5%) developed true bacteremia, 66 (8.9) were positive for gram-negative rods, and 203 (27.5%) died during hospitalization. Prediction rule based on the data at the time of blood withdrawal for culture stratified them into five groups with probabilities of true bacteremia 6.5, 9.6, 21.9, 30.1, and 59.6%. For blood culture positive for gram-negative rods, the probabilities were 0.6, 4.7, 8.6, and 31.7%, and for in-hospital death, those were 6.7, 15.5, 26.0, 35.5, and 56.1%. The area of receiver operating characteristic for true bacteremia, blood culture positive for gram-negative rods, and in-hospital death were 0.73, 0.64, and 0.64, respectively, in original cohort and 0.72, 0.64, and 0.64 in validation respectively.

Conclusions: The clinical prediction rules are helpful for improved clinical decision making for bacteremia patients.

Publication types

  • Evaluation Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Bacteremia / diagnosis*
  • Bacteremia / mortality*
  • Blood / microbiology*
  • Chi-Square Distribution
  • Female
  • Gram-Negative Bacterial Infections / diagnosis*
  • Gram-Negative Bacterial Infections / mortality*
  • Hospital Mortality*
  • Humans
  • Japan / epidemiology
  • Logistic Models
  • Male
  • Microbiological Techniques
  • Predictive Value of Tests
  • ROC Curve