Use of dynamic microsimulation to predict disease progression in patients with pneumonia-related sepsis
2007

Predicting Disease Progression in Pneumonia-Related Sepsis

Sample size: 1888 publication 10 minutes Evidence: moderate

Author Information

Author(s): Saka Görkem, Kreke Jennifer E, Schaefer Andrew J, Chang Chung-Chou H, Roberts Mark S, Angus Derek C

Primary Institution: University of Pittsburgh

Hypothesis

Can a dynamic microsimulation model accurately predict disease progression in patients with pneumonia-related sepsis?

Conclusion

The simulation model can effectively predict the progression of sepsis by incorporating the duration of illness and the patient's health history.

Supporting Evidence

  • The model predicted discharges and deaths closely matching actual outcomes when including duration of illness.
  • Bootstrap validation confirmed the model's stability and predictive ability.
  • Patients with longer illness duration had different outcomes compared to those with shorter durations.

Takeaway

This study created a computer model to help doctors understand how pneumonia can lead to severe illness and death, showing that knowing how long someone has been sick is really important.

Methodology

The study used a Monte Carlo simulation model based on data from the GenIMS study to predict outcomes in hospitalized patients with pneumonia-related sepsis.

Potential Biases

Potential bias due to the exclusion of patients without complete data and the reliance on historical data for predictions.

Limitations

The model is data-intensive and may struggle with small subgroup sizes due to the criteria used for patient similarity.

Participant Demographics

Most participants were elderly (mean age 67.77 years), predominantly white (80.72%), and about half were female (47.98%).

Statistical Information

P-Value

p=0.91 to 0.98 for discharges; p=0.26 to 0.68 for deaths

Confidence Interval

95% confidence intervals were computed for model predictions.

Statistical Significance

p<0.001

Digital Object Identifier (DOI)

10.1186/cc5942

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication