Statistical models for occurrence and impact of nosocomial infections

Prof. Dr. M. Schumacher IMBI/FDM
Dr. G. Schulgen IMBI
Dipl. Stat. G. Schwarzer FDM/IMBI
Prof. Dr. F. Daschner Institute of Environmental Medicine and Hospital Epidemiology, Freiburg
Dr. I. Kappstein Institute of Environmental Medicine and Hospital Epidemiology, Freiburg
Dr. A. Kropec Institute of Environmental Medicine and Hospital Epidemiology, Freiburg
Dr. U. Frank Institute of Environmental Medicine and Hospital Epidemiology, Freiburg

• Summary of the project
  

In this project statistical methods were developed for evaluation of risk factors and construction of risk scores for occurrence of hospital acquired infections. Furthermore, statistical methods were proposed for the analysis of the impact of nosocomial infections on duration and outcome of hospital treatment. Statistical models based on multivariate counting processes were used for modelling transitions among several health states of individual patients admitted to hospital. As these methods allow deeper insight into the disease process, factors could be identified which increase risk for acquiring nosocomial infections. These risk factors have been summarised into risk-scores applicable in daily routine and in clinical and epidemiological studies to identify high-risk patients.

In addition, statistical methods have been derived to analyse impact of hospital acquired infections and other complications occurring during medical treatment. It has been demonstrated that statistical methods traditionally used result in estimates of the extra stay in hospital attributable to nosocomial infections that are biased upwards. New statistical methods have been developed which provide a valid basis for economic analyses. Data from prospective studies in intensive care patients of the university hospital in Freiburg, data from a nationwide prevalence survey and studies on HIV-Patients in Freiburg and San Francisco were used to evaluate the performance of the proposed methods.

The potential use of the methods in other areas of medical research has been demonstrated in the analysis of outcome of psychotherapeutic treatment, estimation of local recurrence rates in breast cancer patients and in the analysis of failure time of hip prostheses. The project was supported by the DFG (Schu 756/2-3).

• Publications

(1)

Schulgen G, Schumacher M: Estimation of prolongation of hospital stay attributable to nosocomial infections: new approaches based on multistate models, Lifetime Data Analysis, 2, 219-40, 1996.

(2)

Kropec A, Schulgen G, Schumacher M, Just H, Geiger K, Daschner F: Entwicklung eines Risiko Scores für nosokomiale Sepsis bei Intensivpflegepatienten, Intensivmedizin, 33, 126-32, 1996.

(3)

Kropec A, Schulgen G, Schumacher M, Just H, Geiger K, Daschner F: Scoring system for nosocomial pneumonia in ICUs, Intensiv Care Medicine, 22, 1155-61, 1996.

(4)

Frank U, Daschner F, Schulgen G, Mills J: Incidence and epidemiology of nosocomial infections in patients infected with human immunodeficiency virus, Clinical Infectious Diseases, 25, 318-20, 1997.

(5)

Hartmann A, Schulgen G, Olschewski M, Herzog T: Modeling psychotherapy outcome as event in time: an application of multistate analysis, Journal of Consulting and Clinical Psychology, 65, 262-8, 1997.

(6)

Schulgen G, Schmoor C, Sauerbrei W, Schumacher M: A note on estimating local recurrence rates in clinical trials on the treatment of breast cancer, Breast Cancer Research and Treatment, 49, 87-91, 1998.

(7)

Kampf G, Wischnewski N, Schulgen G, Schumacher M, Daschner F: Prevalence and risk factors for nosocomial lower respiratory tract infections in German hospitals, Journal of Clinical Epidemiology, 51, 495-502, 1998.

(8)

Schulgen G, Kropec A, Kappstein I, Daschner F, Schumacher M: Estimation of extra hospital stay attributable to nosocomial infections: on heterogeneity and timing of events, submitted, 1999.

(9)

Schulgen G, Schumacher M: Scores to predict intermediate events and outcome of intensive care: methods for construction and assessment, submitted, 1999.

(10)

Schwarzer G, Schumacher M, Maurer T, Ochsner PE: Statistical analysis of failure times in total joint replacement, submitted, 2000.