Sampling rate causes bias in APACHE II and SAPS II scores

Area of Science:

• Critical Care Medicine
• Health Informatics
• Biostatistics

Background:

• Accurate patient severity scoring is crucial for intensive care unit (ICU) management and research.
• Existing scoring systems (APACHE II, SAPS II) rely on specific data points, but sampling rates can vary.
• Variations in data collection frequency may introduce bias in severity assessments.

Purpose of the Study:

• To investigate the impact of different sampling rates for laboratory and hemodynamic data on patient severity scores.
• To determine how these sampling rate variations affect the predicted risk of hospital death.
• To assess the influence of data sampling frequency on the standardized mortality ratio (SMR).

Main Methods:

• A prospective study was conducted in a 23-bed medical-surgical ICU.
• Sixty-nine emergency admission patients were enrolled.
• Laboratory and hemodynamic data were collected at varying intervals (2-hourly, hourly, 2-min medians) to calculate traditional, CIMS, and high-rate severity scores.

Main Results:

• Increasing the sampling rate for hemodynamic data by 2-min intervals raised APACHE II and SAPS II scores by 7.8% and 11.5% respectively.
• Combined increases in sampling rates for both data types led to 14.4% and 14.5% higher APACHE II and SAPS II scores.
• The predicted probability of hospital death increased, while the standardized mortality ratio decreased with higher sampling rates.

Conclusions:

• Elevated sampling rates for critical data result in higher severity scores and lower standardized mortality ratios.
• Discrepancies in data sampling frequencies between institutions can significantly bias comparisons of severity scores.
• Standardization of data collection protocols is essential for reliable inter-hospital comparisons.