Donepezil, rivastigmine, galantamine and memantine for ...

44
Clinical effectiveness
Comparison:
Outcome:
Study
Corey-bloom 1998 riv
Rosler 1999 riv
06 Rivastigmine: ADAS-cog change from baseline
01 Rivastigmine 1–4 mg
Treatment
n
233
243
Mean
(SD)
2.36 (6.00)
1.37 (6.90)
Control
n
234
239
found when combining the studies for the
6–12 mg/day dose group ( 2 test 8.17, df = 1,
p = 0.0043), so the statistically significant
treatment effect seen for the fixed-effect model
combining the two trials with this dosage group is
not valid (WMD –3.08, 95% CI –3.78 to –2.38,
p < 0.00001) (see Figure 12). Exploration of this
statistical heterogeneity by subgroup analysis was
not appropriate owing to the small numbers of
included trials in the meta-analysis.
[Commercial/academic confidential information
removed]
MMSE
Three of the four published studies reported
MMSE as an outcome measure (Table 16). Two of
these reported mean baseline scores of 19–20, 57,58
but baseline values were not reported in the other
study. The first study compared two fixed-dose
groups (4 and 6 mg/day) with placebo, and the
other two studies compared two flexible dose
groups (1–4 and 6–12 mg/day) with placebo. The
Mean
(SD)
4.09 (2.90)
1.34 (7.00)
WMD
(95% CI fixed)
Weight
%
67.8
32.2
WMD
(95% CI fixed)
–1.73 (–2.59 to –0.87)
0.03 (–1.21 to 1.27)
Total (95% CI)
Test for heterogeneity 2 476 473 100.0 –1.16 (–1.87 to –0.46)
= 5.24, df = 1, p = 0.022
Test for overall effect z = 3.24, p = 0.001
FIGURE 11 ADAS-cog change from baseline with rivastigmine 1–4 mg
Comparison:
Outcome:
Study
Corey-bloom 1998 riv
Rosler 1999 riv
06 Rivastigmine: ADAS-cog change from baseline
02 Rivastigmine 6–12 mg
Treatment
n
231
243
Mean
(SD)
0.31 (5.90)
–0.26 (6.80)
Control
n
234
239
–10 –5 0 5 10
Favours treatment Favours control
Mean
(SD)
4.09 (2.90)
1.34 (7.00)
67.9
32.1
–3.78 (–4.63 to –2.93)
–1.60 (–2.83 to –0.37)
Total (95% CI)
Test for heterogeneity 2 474 473 100.0 –3.08 (–3.78 to –2.38)
= 8.17, df = 1, p = 0.0043
Test for overall effect z = 8.65, p < 0.00001
FIGURE 12 ADAS-cog change from baseline with rivastigmine 6–12 mg
WMD
(95% CI fixed)
Weight
%
–10 –5 0 5 10
Favours treatment Favours control
WMD
(95% CI fixed)
fixed-dose study did not report any statistically
significant differences between treatment groups
and placebo for this outcome measure. Both of the
flexible dose studies reported a statistically
significant difference between the high dose
treatment group and placebo. The high-dose
participants in both studies had a mean dose of
~10 mg/day. Corey-Bloom and colleagues 57 used
observed case analysis for this outcome measure,
and found an improvement in the high-dose
group of 0.30 points, compared with a decline in
placebo participants of –0.79 points. The other
flexible dose study 58 used ITT analysis, and
reported an improvement in the high-dose group
of 0.21 points compared with a decline in placebo
participants’ scores of –0.47. In this study, the lowdose
participants also showed a decline in MMSE
score, with a mean change of –0.62, but this was
not statistically significantly different from the
placebo group.
[Commercial/academic confidential information
removed]