17
stage, histology, and smoking status. A total of 589 patients received isotretinoin
while the remaining patients received placebo.
One of the objectives of the phase III study is to assess the treatment effect
on different types of toxicity and the grade associated with each of them. In this
chapter we focus on inference related to this objective only. The treatment-related
toxicities include: cheilitis, conjunctivitis, arthralgia, hypertriglyceridemia, headache,
and abnormal vision. Cheilitis is dryness, usually associated with an uncomfortable
sensation of the Iips with scaling and cracking and accompanied by a characteristic
burning sensation. Conjunctivitis is one of the most common nontraumatic eye com-
plaints, involving the inflammation of conjunctiva. Arthralgia is pain in the joints.
Hypertriglyceridemia is an excess of triglycerides in the blood. With the exception of
hypertriglyceridemia, toxicity was graded by use of the Common Toxicity Criteria, a
toxicity scale used by the NCI for Adverse Events. Triglyceride toxicity was graded
as follows: grade 1 toxicity was defined as more than 2.5 times but less than or equal
to five times the normal level; grade 2 toxicity was defined as more than five times but
less than or equal to 10 times the normal level; and grade 3 toxicity was defined as
more than 10 times the normal triglyceride level or if a patient experienced complica-
tions (e.g., pancreatitis) at any grade of triglyceride toxicity. If patients experienced
multiple incidents of the same toxicity, only one incident at the highest grade was
counted. When multiple different toxicities are reported for the same patient, the
corresponding observed toxicity levels are expected to correlate. We will introduce
the desired correlation with patient-specific random effects in the model. Reported
toxicity rates are per patient, i.e., toxicity rates are probabilities that a patient re-
ports a certain type of toxicity and grade. A summary of the data is shown in Table
2.1.