|
N-of-1
trials are trials in which patients are repeatedly treated with
the same treatments in order to compare their utility on an individual
basis.
Example
4: March et al. report a series of n-of-1 trials in osteoarthritis
(March
et al., 1994). Diclofenac (D) was compared to paracetamol
(P) in 20 patients. Patients suffering from osteoarthritis were
treated in six periods which, for the purpose of randomization,
were handled in pairs. (In each pair of periods the patients either
received D followed by P or P followed by D.) Pain scores were
used as the main outcome variable.
The object
of the series of trials was to establish what proportion of patients
were better treated with paracetamol. To this end, a matched pairs
t-test was performed for each patient, matching on the period
pairs. (Since there were three pairs of periods for the test,
this was a t-test with two degrees of freedom.)
The figure
below is a graphical illustration of the results as analysed by
March et al. The mean pain difference paracetamol-diclofenac is
illustrated for each patient. Positive values favor diclofenac.
The three differences marked with triangles are "significant"
at the 5% level.
| Figure
5.1: Pain differences in mm VAS from March et al.
|
|---|
|
|
The authors
then concluded that "in osteoarthritis many patients currently
receiving or being considered for non-steroidal anti-inflammmatory
drugs may achieve adequate control with paracetamol".
|