Infection probability score, APACHE II and KARNOFSKY scoring systems as predictors of infection onset in haematology-oncology patients

J Clin Nurs. 2010 Jun;19(11-12):1560-8. doi: 10.1111/j.1365-2702.2009.03011.x. Epub 2010 Apr 5.


Aim: To assess the predictive power of three systems: Infection Probability Score, APACHE II and KARNOFSKY score to the onset of healthcare-associated infections in haematology-oncology patients.

Background: The high incidence of healthcare-associated infections is a frequent problem in haematology-oncology patients that affects morbidity and mortality of these patients.

Design: A retrospective surveillance survey.

Method: The survey was conducted for seven months in the haematology unit of a general hospital in Greece to assess the predictive power of Infection Probability Score, APACHE II and KARNOFSKY score to the onset of healthcare-associated infections. The sample consisted of 102 hospitalised patients. The diagnosis of healthcare-associated infections was based on the definitions proposed by CDC.

Results: Among the participants, 53 (52%) were males and 49 (48%) were females with a mean age of 53.30 (SD 18.59) years old (range, 17-85 years). The incidence density of healthcare-associated infections (the number of new cases of healthcare-associated infections per 1000 patient-days) was 21.8 infections per 1000 patient-days. Among the 102 patients, healthcare-associated infections occurred in 32 (31.4%) patients who had a total of 48 healthcare-associated infections (47.5%). Among the 38 patients with neutropenia, 26 (68.4%) had more than one healthcare-associated infection. Of the 48 detected healthcare-associated infections, the most frequent type was blood-stream infection (n = 17, 35.4%), followed by Clostridium difficile infection (n = 11, 22.9%) and respiratory tract infection (n = 8, 3.4%). The best cut-off value of Infection Probability Score (IPS) for the prediction of a healthcare-associated infection was 10 with sensitivity of 59.4% and specificity of 74.3%.

Conclusions: Between the three different prognostic scoring systems, IPS had the best sensitivity in predicting healthcare-associated infections.

Relevance to clinical practice: IPS is an effective tool and should be used from nurses for the early detection of haematology-oncology patients who are susceptible to the onset of a healthcare-associated infection.

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Female
  • Hematologic Diseases / complications*
  • Hematologic Diseases / epidemiology
  • Humans
  • Infections / epidemiology*
  • Karnofsky Performance Status*
  • Male
  • Middle Aged
  • Neoplasms / complications*
  • Neoplasms / epidemiology
  • Probability
  • Retrospective Studies
  • Risk Factors
  • Young Adult