Table 9 Number of red alarms anticipated when selecting the top “n” patients with higher score versus “n” random patients.

From: Using machine learning to predict deterioration of symptoms in COPD patients within a telemonitoring program

 

Top 2

Top 5

Top 10

Prioritizing by model score

70

107

126

Random prioritization

9.2 ± 1.45

23.5 ± 1.30

46.6 ± 1.41

  1. Highest values are in bold.