Table 3. Estimated percent increase in the number of daily deaths related to coronary heart disease for each 1 °C increase in diurnal temperature range^a .

Lag	Method	Percent	95% CI
0	TS	1.12	0.68–1.56
	UCC	1.31	0.43–2.19
	BCC	0.70	0.18–1.22
1	TS	2.12	1.56–2.67
	UCC	2.67	1.41–3.93
	BCC	1.73	0.46–3.00
2	TS	2.46	1.76–3.16
	UCC	3.21	2.23–4.19
	BCC	2.13	1.04–3.22
3	TS	2.63	1.99–3.26
	UCC	3.62	2.09–5.15
	BCC	2.22	1.41–3.03
4	TS	2.60	1.95–3.26
	UCC	3.19	2.03–4.35
	BCC	2.37	1.12–3.62

Abbreviations: TS, time-series analysis; UCC, unidirectional case-crossover analysis; BCC, bidirectional case-crossover analysis.

^aFor UCC analysis, the same weekdays 1, 2, or 3 weeks before the death were selected as the control periods; for BCC analysis, the same weekdays 1, 2, or 3 weeks before, and after, the death were selected as the control periods.