Table V.

Implant failure predictive model performance in the Swedish Arthroplasty Register.

Outcome	Model performance: AUC in % (95% CI)
	Random forest	Gradient boosting machine	Ridge regression	Lasso regression	Logistic regression	Classification tree	Classification tree with pruning
Revision due to PPFF
30 days	0.83 (0.79 to 0.88)	0.84 (0.80 to 0.88)	0.83 (0.79 to 0.88)	0.85 (0.81 to 0.89)*	0.83 (0.78 to 0.88)	0.82 (0.78 to 0.85)	0.82 (0.78 to 0.85)
60 days	0.81 (0.76 to 0.86)	0.86 (0.83 to 0.89)*	0.85 (0.81 to 0.89)	0.85 (0.81 to 0.89)	0.85 (0.81 to 0.89)	0.84 (0.80 to 0.87)	0.84 (0.80 to 0.87)
90 days	0.80 (0.75 to 0.85)	0.86 (0.82 to 0.89)*	0.84 (0.81 to 0.88)	0.84 (0.80 to 0.88)	0.82 (0.76 to 0.87)	0.83 (0.80 to 0.86)	0.83 (0.80 to 0.86)
1 year	0.71 (0.65 to 0.77)	0.80 (0.75 to 0.85)*	0.79 (0.74 to 0.84)	0.79 (0.74 to 0.84)	0.79 (0.73 to 0.84)	0.76 (0.71 to 0.82)	0.76 (0.71 to 0.82)
Reoperation due to PPFF
30 days	0.77 (0.69 to 0.84)	0.84 (0.79 to 0.89)*	0.83 (0.78 to 0.88)	0.83 (0.78 to 0.88)	0.83 (0.77 to 0.88)	0.80 (0.74 to 0.85)	0.80 (0.74 to 0.85)
60 days	0.80 (0.75 to 0.86)	0.86 (0.83 to 0.89)*	0.85 (0.81 to 0.89)	0.85 (0.81 to 0.89)	0.85 (0.81 to 0.89)	0.82 (0.78 to 0.86)	0.82 (0.78 to 0.86)
90 days	0.76 (0.69 to 0.82)	0.85 (0.81 to 0.88)*	0.83 (0.80 to 0.87)	0.83 (0.78 to 0.87)	0.81 (0.76 to 0.86)	0.80 (0.76 to 0.84)	0.80 (0.76 to 0.84)
1 year	0.66 (0.60 to 0.73)	0.79 (0.74 to 0.84)*	0.78 (0.73 to 0.83)	0.78 (0.73 to 0.83)	0.77 (0.72 to 0.82)	0.71 (0.66 to 0.76)	0.71 (0.66 to 0.76)

^*Best-performing model for each outcome.

AUC, area under the receiver operating characteristic curve; PPFF, periprosthetic femoral fracture.