The effect of slightly increasing the correlation of the first lag but running the simulation as a stationary series is to slightly elongate
the typical length scale of the simulated process (Isaaks and Srivastava, p196 et seq show this very elegantly).
While second - stage results showed no significant differences between participants continuing to receive weekly doses of the active drug, those receiving biweekly doses and those shifted to placebo, Kimball notes that, since it is
typical for the severity of HS symptoms to increase and decrease and because the study protocol required treatment discontinuation for participants whose symptoms stopped responding to the drug, larger
scale studies will be required to better define the ideal
length and frequency of treatment.