[Prognosis in meningococcal disease: methodology and practice]

Klin Med (Mosk). 1999;77(4):32-8.
[Article in Russian]

Abstract

78 patients with systemic meningococcal disease admitted to the Intensive Care Unit of the 2nd Moscow Hospital for Infectious Diseases were studied and the composite prognostic score was developed to estimate the risk of lethal outcome. The stepwise variable selection procedure for the multiple logistic regression was applied to 30 potential clinical and laboratory risk factors and markers. Five factors were selected for the score, namely the platelet count (< 150 x 10(6)/ml), the presence of hemorrhages into the eye or mucosal tissue, the interval from the last urination before admission (> 4 h), the respiration rate (> 170% of age-adjusted normal value) and age (< 2 or > 65 years) with regression coefficients 0.3, 0.2, 0.2 and 0.1, respectively. Both for the source clinical group and for the additional test group (64 patients), the scale was able to classify correctly 95% of cases using the data collected at admission. Ten prognostic scores proposed previously by foreign investigators were tested in the same patients and the best four scores were selected (GMSPS, Gedde-Dahl, Niklasson, Kahn); the scores classified correctly 85-90% of cases. This study is an example of methodological approaches to prognostic score construction in medicine.

Publication types

  • Comparative Study
  • English Abstract

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Child
  • Child, Preschool
  • Disease Progression
  • Female
  • Follow-Up Studies
  • Humans
  • Infant
  • Male
  • Meningococcal Infections / diagnosis*
  • Middle Aged
  • Prognosis
  • Retrospective Studies
  • Risk Assessment / methods