Table 4. Odd Ratios (OR) for the Association of Frailty Domain Scores (Good versus Fair/Poor) and Number of Frailty Domains Rated as Fair/Poor with Judgment of the Absence or Presence of Frailty According to Family Physicians (FP) and Geriatrician

Predictor	FP	Geriatrician	P Value for Interaction
Predictor	OR [95%-CI] for judging person as frail	OR [95%-CI] for judging person as frail	P Value for Interaction
Domain reported as fair or poor functioning versus reported as good^*
Physical	10.4 [4.4–24.4]	11.5 [4.9–26.8]	.86
Medication	2.5 [1.13–5.6]	2.4 [1.19–4.8]	.93
Cognition	9.3 [3.9–22.0]	79.3 [18.4–341.2]	.005
Sensory	2.0 [1.1–3.6]	7.6 [3.5–16.5]	.007
(Instrumental) activities of daily living	5.9 [2.0–16.8]	Not estimable	.98
Mobility	3.0 [1.5–6.0]	25.0 [8.9–70.3]	<.001
Mental	2.2 [1.1–4.4]	18.0 [7.1–45.9]	<.001
Social	4.1 [1.9–8.7]	3.1 [1.0–9.3]	.67
Increase in number of health domains reported as fair or poor; +1^†	3.8 [2.9–5.0]	7.7 [5.1–11.5]	<.001

↵* Results (OR [95% CI]) taken from a logistic mixed-model regressing profession of evaluator (FP versus geriatrician), all frailty domain scores (fair/poor versus good) simultaneously, and terms for the interaction between domain scores and profession with a random intercept to allow the dependence at participant level.
↵† Results (OR [95% CI]) taken from a logistic mixed-model regressing profession of evaluator (FP versus geriatrician), the number of frailty domains (0–8) rated as fair/poor and terms for the interaction between the number of impaired domains and profession with a random intercept to allow the dependence at participant level.