Table 3.

Studies comparing CBT-I to pharmacological therapies: follow-up results

Study	Group	Sleep latency	Total sleep time	Total wake time	Efficiency	Other	Notes
CBT-I vs. zopiclone						SWS
Sievertsen 2006 [29]	CBT-I: 18	Not reported	–5.0 min	–60.7 min	+8.7%	+21.1 min
	Zopiclone :16		–56.2 min	–9.9 min	–0.4%	–17.6 min
6 months			p = NS	p = 0.01	p = 0.008	p = 0.001
Polysomnography
Sleep diary	CBT-I: 18	Not reported	+42.4 min	–73.3 min	+14.2%
	Zopiclone :16		+40.5 min	–42.2 min	+10.7%
			p = NS	p = 0.03	p = NS
CBT-I vs. zolpidem
Jacobs 2004 [30]	CBT-I: 8						No long-term follow-up for zolpidem group
12 months	Zolpidem: none
Sleep diary
CBT-I vs. temazepam						WASO
Wu 2006 [31]	CBT-I: 19	–32.8 min	+30.3 min	Not reported	+10.2%		p values based on post-intervention differences
8 months	Temazepam: 17	–17.2 min	–13.0 min		–1.9%
Polysomnography		p < 0.004	p < 0.05		p < 0.01
Sleep diary	CBT-I: 19	–41.8 min	45.5 min	Not reported	+16.8%
	Temazepam: 17	–20.5 min	–6.0 min		+3.9%
		p < 0.003	p < 0.01		p < 0.05
Morin 1999 [31]	CBT-I: 13	Not reported	+65.2 min	Not reported	+16.4%	–16.5 min	All measurements in temazepam group significantly worsened from end of treatment to end of follow-up.
24 months
Sleep diary	Temazepam: 12		+11.5 min		+2.9%	–4.6 min
			p = NR		p = NR	p = NR
CBT-I vs. triazolam
McCluskey 1991 [33]	CBT-I: 15	–45 min	+51 min	Not reported	Not reported
8 weeks	Triazolam :15	–21 min	+14 min
Sleep diary		p < 0.01	p = NR

SWS Slow wave sleep, WASO Wake after sleep onset.