Therapy Scenario

As a clinical manager of a geriatric unit in a large hospital, you are interested to know whether any benefit can be gained from drugs purporting to improve quality of life for people with Alzheimer’s disease, particularly as new drugs are expensive and may have serious adverse effects. A patient’s wife alerts you to the existence of a new drug she heard about on the Internet called rivastigmine, and you promise her that you will look into it. With interested colleagues, you formulate the question: “Is rivastigmine likely to reduce the effects of Alzheimer’s in people with mild to moderate disease?”

You start up Best Evidence, enter “rivastigmine” and “Alzheimer’s”, and find nothing (yet). You then search PubMed (the Internet version of MEDLINE that is most up to date) for “rivastigmine alzheimer* controlled trial”. (Note there is no need to use connectors such as AND, OR, NOT when doing a quick search on Pub Med). You see that indeed, there is a new trial. Happily it is published in the British Medical Journal – freely available in full text on the Web. BMJ 1999;318:633-40.

Read the article and decide:

  • Is the evidence from this randomised trial valid?
  • If valid, does the evidence address an important issue?
  • If valid and important, can you apply this evidence in caring for your patients and population?

Completed Therapy Worksheet for Evidence-Based Purchasing


Rosler M, Anand R, Cicin-Sain A, Gauthier S et al. Efficacy and safety of rivastigmine in patients with Alzheimer’s Disease: international randomised controlled trial.
BMJ 1999;318:633-40.

Are the results of this single preventive or therapeutic trial valid?

Was the assignment of patients to treatments randomised?
And was the randomisation list concealed?
Yes. Randomisation was generated by a computer. It is not explicitly stated, but reasonable to assume that this code was kept from investigators (p.634).
Were all patients who entered the trial accounted for at its conclusion?
And were they analysed in the groups to which they were randomised?
Yes. Figure 1, page 635. The authors used intention-to-treat analysis.
Were patients and clinicians kept “blind” to which treatment was being received?
Yes. The treatment (low and high dose) and placebo drug capsules were identical, and patients took the same number of capsules, irrespective of dose.
Aside from the experimental treatment, were the groups treated equally?
Were the groups similar at the start of the trial?
Apparently the groups were comparable (results, page 365), although there is no table or published data to demonstrate demographic characteristics here. Table 2 shows that baseline scores were similar for both groups. There were more women than men (59% vs. 41%) in the whole study.

Are the valid results of this randomised trial important?

Your Calculations
4+ point improvement on Alzheimer’s disease assessment scale* using intention-to-treat analysis Relative Benefit Increase
Absolute Benefit Increase
Number Needed to Treat
High dose Low dose Placebo (EER – CER)/CER EER – CER 1/ABI
24% 15% 16% high vs. placebo (24%-16%)/16%
= 50%
24% – 16%
= 8%
= 13 pts
low vs. placebo (15%-16%)/16%
= 63%
= -1%
(or, 1% without the minus sign for an Absolute Risk Increase)
= 100 pts
(i.e. an NNH)
* Note that it is possible to calculate these figures for any of the five outcomes cited in Table 3, using any of the three forms of analysis given, if learners are interested.
** Because the outcome is beneficial, rather than harmful, it makes more sense in this case to use the term “relative benefit increase” rather than “relative risk reduction”.

text{95% Confidence Interval ($CI$) on an $NNT$} &= 1 / (text{limits on the $CI$ of its $ARR$})\
&= pm1.96 sqrt{frac{mathit{CER}times(1-mathit{CER})}{text{# of control pts.}} + frac{mathit{EER} times(1-mathit{EER})}{text{# of exper. pts.}}} \
&= pm1.96 sqrt{{frac{0.096times0.904}{730}} + {frac{0.028times0.972}{711}}} \
&= pm2.4%

Can you apply this valid, important evidence about a treatment in caring for your patient?

Do these results apply to your patient?
Is your patient so different from those in the trial that its results can’t help you?
Maybe. Depends upon patient type – those in this trial were mild to moderately severe patients, including mainly women and white people. Also, it depends upon how clinically relevant you feel the “4+ point improvement” on the Alzheimer’s scale is.
How great would the potential benefit of therapy actually be for your individual patient?
Method I: f
Risk of the outcome (school absence) in your patient, relative to patients in the trial.
Expressed as a decimal:

$$ mathit{NNT}/mathit{f} = underline{qquadqquad} \\ text{($NNT$ for patients like yours)}$$

Method II: 1 / (PEER x RRR)

Your patient’s expected event rate if they received the control treatment:
$$ mathit{PEER}: underline{qquadqquad} \\
1/(mathit{PEER} times mathit{RRR}) \\
= 1/underline{qquadqquad} \\
= underline{qquadqquad} \\
text{($NNT$ for patients like yours)}$$

Are your patient’s values and preferences satisfied by the regimen and its consequences?
Do your patient and you have a clear assessment of their values and preferences?
Need to ascertain from group.
Are they met by this regimen and its consequences?
Need to ascertain from group and decision about clinical relevance of the trial’s findings.

Additional Notes

Alzheimer’s: Rivastigmine improves cognition

Clinical Bottom Line

High dose rivastigmine improves cognitive and functional outcomes in patients with mild to moderate Alzheimer’s disease.


Rosler M, Anand R, Cicin-Sain A, Gauthier S et al. Efficacy and safety of rivastigmine in patients with Alzheimer’s Disease: international randomised controlled trial.
BMJ 1999;318:633-40.

Purchasing Question

Should rivastigmine be considered in patients with Alzheimer’s for improving quality of life and reducing carers’ costs?

Search Terms

MEDLINE search using rivastigmine, AND Alzheimer* AND controlled trial.

The Study

Multicentre, double blind, randomised controlled trial of high and low dose rivastigmine vs. placebo in 725 patients with mild to moderate Alzheimer’s Disease, with a 12 week lead in dosage period and 14 week follow-up, conducted in 45 centres in Europe and North America.

Patients were recruited by their GPs or directly from specialist centres. They were randomly allocated by computer to placebo, low or high dose drug, using identical type and number of dosage capsules. Patients were assessed at baseline and follow-up using the Cognitive subscale of the Alzheimer’s disease assessment scale, the progressive deterioration scale, and clinician interviews. Analysis was conducted in three ways (intention-to-treat, last observation carried forward analysis and observed cases analysis). All patients were accounted for within the intention-to-treat analysis.

The Evidence

Outcome Time to Outcome CER EER RBI ABI NNT (95% CI)
Beneficial outcome (improvement in cognition on Alzheimer’s scale) 26 weeks 16% 24% 50% 8% 13
(7 to 113)


  • In this study, relative risk = RR (here, relative benefit) = 0.24/0.16 = 1.5
  • Similar results were obtained from the other assessment instruments, but the overall clinical validity of the instruments was not clearly presented in the paper.

Appraised By

Anna Donald and Sam Vincent