Donepezil, rivastigmine, galantamine and memantine for ...

150
Discussion and conclusions
Strengths and limitations of the
review
This review has a number of strengths which lead
to a minimisation of bias. The review is
independent of any vested interest and it brings
together the evidence for the effectiveness of
donepezil, rivastigmine, galantamine and
memantine for AD, by the application of
consistent methods of critical appraisal. It was
guided by the principles for undertaking a
systematic review and prior to undertaking the
review, the methods were set out in a research
protocol (Appendix 2). This protocol defined the
research question, inclusion criteria, quality
criteria, data extraction process and methods
employed to undertake the different stages of the
review. Finally, an advisory group has informed the
review from its initiation, through the
development of the research protocol and
completion of the report.
There were certain limitations placed upon this
review. Despite being guided by the principles for
undertaking a systematic review, owing to time
restrictions placed upon the review authors of
references were not contacted for further details of
their trials where data were lacking. As published
papers are usually limited to 2500–5000 words it
may be that some details of the trials are not
published.
Implications for further research
The review has identified a wide range of
subjective outcome measures among the included
trials. The review has also noted that it is often
unclear how statistically significant changes on
these outcome measures translate into real benefit
for the person with AD and for their carers. In
many cases there is also uncertainty about the
reliability and validity of the outcome measures
used for these populations. Further research is
required to answer a number of related questions
(these may or may not need to be carried out
separately for the population with mild to
moderately severe AD and the population with
moderately-severe to severe AD).
● To review the evidence base of the psychometric
properties of key measures of global, cognitive,
functional, behavioural and mood and QoL
outcomes for the population of people with AD.
In future research can be focused on using the
most reliable and valid measures for this
population.
● To investigate the correlation between statistical
improvements on these key measures with
perceived patient and carer benefit.
● To develop QoL measurement scales that are
reliable and valid for the population with AD
and their carers.
The review found only a few trials with a duration
of follow-up of ≥12 months. The extent to which
benefits on outcomes are maintained in patients
over longer periods cannot be unequivocally
assumed. It would be desirable to assess the effects
of these drugs over long periods of follow-up;
however, placebo-controlled randomised trials are
now unlikely to be ethically acceptable. It may be
possible to undertake a randomised ‘withdrawal’
trial after long-term treatment, randomising to
either continued treatment or placebo.
The review identified only three randomised
comparisons between the different cholinesterase
inhibitors. These comparisons were small scale
and offered very little to the evidence base
regarding which intervention is most beneficial to
patients. Larger, long-term RCTs comparing
cholinesterase inhibitors in those with mild to
moderately severe AD on outcomes such as
cognition, function (ADLs), and behaviour and
mood are required.
Few studies of memantine have been undertaken
on patients with moderately severe to severe AD.
Further RCTs comparing the effects of memantine
with placebo in these populations on measures of
function, behaviour and mood and carer QoL are
required to inform any future up-date of the
present review.
Research is required on the effectiveness of
treatment on patient outcomes, such as healthrelated
QoL, need for institutional care and delay
in disease progression as defined by measures
other than cognitive function alone.
Research on the prediction of disease progression,
using a broad range of AD signs and symptoms (to
include ADL and functional outcomes), is required.
There appears to be an absence of data to model
disease progression using multivariate analysis,
including functional outcomes and measures of
ADL, and initiatives to collect such data in an
unbiased, methodologically rigorous and credible
manner, should be encouraged. The current
methods available to model disease progression
over time are dominated by the use of cognitive
function (MMSE), which is regarded as an
insufficient marker of disease progression for AD.