A data visualisation method for assessing exposure misclassification in case-crossover studies: the example of tricyclic antidepressants and the risk of hip fracture in older people

Table 3 TCA exposure classification and effect estimates with alternative control windows (n = 8828)

Control Window	TCA Exposure Classification				OR^c (95% CI)
Control Window	Continuous Users^a	Non-Users^a	Intermittent Users^b	Recent Stoppers^b	OR^c (95% CI)
1^d	349	8244	127	108	1.18 (0.91–1.52)
2^e	326	8247	150	105	1.43 (1.11–1.83)

Adapted from [7]
TCA Tricyclic antidepressant, OR Odds ratio, CI Confidence interval
^aExcluded from analysis due to concordant exposure between the case window and the control window
^bIncluded in analysis due to discordant exposure between the case window and the control window
^cOR for intermittent exposure compared with recent stopping
^dA control window of 101–150 days pre-fracture
^eA control window of 151–200 days pre-fracture

ISSN: 1471-2288