Table 2.

Multivariable analysis of risk factors for the presence of drug-drug interactions at baseline (index hospitalization)

Variables at baselinea	OR (95% CI)	P value
Age (reference: 70–79 years) - 80–89 years - > 90 years	0.79 [0.64–0.97] 0.65 [0.44–0.96]	.02 .03
Sex (reference: female) - male	0.81 [0.66–0.98]	.03
Medical history of depression	2.88 [1.99–4.26]	< .001
Medical history of coronary artery disease	1.25 [1.02–1.54]	.03
Medical history of heart failure	1.60 [1.26–2.02]	< .001
Medical history of atrial fibrillation	1.55 [1.26–1.90]	< .001
Medical history of chronic renal failure	1.09 [0.87–1.36]	.46
Medical history of chronic hepatic failure	0.49 [0.31–0.77]	.002
Medical history of COPD	1.91 [1.48–2.47]	< .001
Hospitalizations during the previous year	1.17 [0.96–1.42]	.10
Hyperpolypharmacy	2.21 [1.77–2.77]	< .001
Non-independent livingc	1.26 [0.98–1.63]	.06
Main reason for hospital admission (reference: surgical)d - Medical	1.09 [0.82–1.45]	.55

N = 1950; AIC = 2490

C statistic: 0.69 IC95% (0.68–0.72).

^a Baseline = Index hospitalization

^b ≥ 10 drugs par day at admission.

^c Non-independent living was defined as living in a nursing home (at least 3 months in the 6 months before the index admission) or being housebound.

^dClassification coded from the main reasons for hospitalization (free text in the original database)

Abbreviations: AIC Akaike’s criterion, COPD chronic obstructive pulmonary disease