Table 2. Summary of predictive model outputs for postoperative outcomes.

This table outlines the key predictors, effect sizes, and statistical significance from three regression models evaluating postoperative complications, healing time, and functional outcomes. Implant type and vitamin D levels consistently emerged as significant predictors across all models. Model performance metrics, including accuracy and R² values, indicate moderate predictive strength. All reported predictors showed statistically significant associations (p < 0.05).

Predictive model	Key predictor	Effect size (OR/β)	Model performance	p-value
Binary logistic regression (complications)	Implant type	OR = 2.13	Accuracy = 71.6%	0.005
	Vitamin D level	OR = 1.74		0.01
	Diabetes	OR = 1.65		0.02
	ASA score	Not specified		Not specified
Multiple linear regression (healing time)	Implant type	β = –0.41	R² = 0.38	<0.001
	Vitamin D level	β = –0.28		<0.001
	ASA score	β = 0.25		<0.001
Ordinal logistic regression (functional outcome)	Implant type	OR = 2.73	Nagelkerke R² = 0.41	<0.01
	ASA score	OR = 1.93		<0.01
	Vitamin D status	OR = 1.61		<0.01