Dear Professor Mean, I am volunteering as a data analyst in a post-marketing surveillance to assess the safety and efficacy of a drug. I'm not sure what to measure and how to measure it. Can you help me figure out what really needs to be done?
I do my best thinking on sandy beaches by the ocean. I'm sure you are the same. Can you arrange for your company to fly us to South Padre Island to work out this problem together?
Normally the data analyst will not make the choices of what to measure all by himself/herself. You need to work with people with a strong medical background. It might help to find other published studies of post marketing surveillance through a medline search and see if you can adopt their general approach. The one important issue is the need to balance practical considerations (what can be measured without excessive time and expense) and medical considerations (what is important from the clinician's perspective and from the patient's perspective).
There's an important acronym, POEM, which stands for Patient Oriented Evidence that Matters. The POEM philosophy is that patients don't care about concentrations of various markers in their blood or what their cells look like under a microscope. They only care about three things: mortality (will I die?), morbidity (will I go blind?) or quality of life (will I be able to walk up two flights of stairs without getting winded). All those other things like pathology and blood chemistry are important only if they are tightly linked to morbidity, mortality, or quality of life.
I realize this is a somewhat vague answer. Quite honestly, every project is different so it is hard to come up with a specific answer.