Performance of prescriptive algorithms
. | . | kNN . | LASSO . | Random forest . | |||
---|---|---|---|---|---|---|---|
All patient visits (N = 48,140) | |||||||
Mean HbA1c benefit relative to standard of care (SE) | |||||||
Percent | −0.14 (0.01)* | −0.13 (0.01)* | −0.07 (0.01)* | ||||
mmol/mol | −1.5 (0.1)* | −1.4 (0.1)* | −0.8 (0.1)* | ||||
Visits for which algorithm’s recommendation differed from observed standard of care | |||||||
Number of visits (%) | 15,323 (31.8) | 12,684 (26.3) | 14,302 (29.7) | ||||
Mean HbA1c benefit relative to standard of care (SE) | |||||||
Percent | −0.44 (0.03)* | −0.45 (0.03)* | −0.26 (0.03)* | ||||
mmol/mol | −4.8 (0.3)* | −4.9 (0.3)* | −2.8 (0.3)* |
. | . | kNN . | LASSO . | Random forest . | |||
---|---|---|---|---|---|---|---|
All patient visits (N = 48,140) | |||||||
Mean HbA1c benefit relative to standard of care (SE) | |||||||
Percent | −0.14 (0.01)* | −0.13 (0.01)* | −0.07 (0.01)* | ||||
mmol/mol | −1.5 (0.1)* | −1.4 (0.1)* | −0.8 (0.1)* | ||||
Visits for which algorithm’s recommendation differed from observed standard of care | |||||||
Number of visits (%) | 15,323 (31.8) | 12,684 (26.3) | 14,302 (29.7) | ||||
Mean HbA1c benefit relative to standard of care (SE) | |||||||
Percent | −0.44 (0.03)* | −0.45 (0.03)* | −0.26 (0.03)* | ||||
mmol/mol | −4.8 (0.3)* | −4.9 (0.3)* | −2.8 (0.3)* |
*P < 0.001.