Table 4.

Odds ratios^a (95% confidence intervals) for the association between antiepileptic refilling and switching and seizure-related outcomes for sensitivity analyses^b

No. days in case/control periods	Refill of the same drug, strength and dosage form from the same manufacturerc	Switch between the same drug, strength and dosage form but from different manufacturersd	Refill-adjusted odds ratio for switching
	Odds ratio (95% CI)	Odds ratio (95% CI)	Odds ratio (95% CI)
Adjusted for use of other medications and visits to general practitioners and neurologists
21	2.31 (1.56–3.44)	2.65 (0.68–10.34)	1.15 (0.28–4.73)
28	2.08 (1.48–2.93)	3.05 (0.84–11.04)	1.46 (0.39–5.55)
Excluding diagnosis code 780.3x
21	2.18 (1.23–3.87)	3.00 (0.81–11.08)	1.38 (0.33–5.74)
28	1.71 (1.08–2.73)	2.20 (0.76–6.33)	1.28 (0.40–4.07)
Excluding clobazam and clonazepam
21	2.23 (1.45–3.44)	3.33 (0.92–12.11)	1.49 (0.38–5.82)
28	2.07 (1.43–3.01)	3.00 (0.97–9.30)	1.45 (0.44–4.76)
Excluding clobazam, clonazepam, gabapentin, topiramate
21	2.00 (1.29–3.10)	3.00 (0.81–11.08)	1.50 (0.38–5.95)
28	1.95 (1.33–2.87)	2.75 (0.88–8.64)	1.41 (0.42–4.72)
Excluding those with one or more visit to a general practitioner nor neurologist
21	2.31 (1.56–3.44)	2.00 (0.18–22.05)	0.86 (0.08–9.85)
28	2.08 (1.48–2.93)	4.00 (0.45–35.79)	1.92 (0.21–17.65)

^aIn case-crossover studies, odds ratios are inherently adjusted for time-invariant patient factors

^bThese sensitivity analyses assume a 1-day induction period and a 2-day grace period consistent with primary analyses

^cIncludes refilling of brand products and refilling of generic products

^dincluded switching between brand and generic products, generic and brand products, and two generic products from different manufacturers