Lecture 1:
Introduction Session 1:
Pharmacoepidemiologic questions concerns the next four aspects:
1- Diagnostics: say which disease the patient has.
2- Etiology: the reason of the disease.
3- Prognosis: future of disease.
4- Intervention: best treatment (therapy) or how to prevent the disease (prevention).
Epidemiology: is the science that studies the occurrence of a disease in large population of people as a
function of determinants. It is an essential science to conduct evidence-based medicine.
Etiology and intervention VS diagnosis and prognosis:
Etiology and intervention: their studies are analytical studies, with causal associations. These can be
known after randomized control trials or case control studies. Only one determinant is needed to
determine them and confounding bias can occur. They are represented by explanatory statistical
models.
Diagnosis and prognosis: their studies are descriptive studies, with predictive associations. These can
be determined by getting info from cohorts. No confounding bias can occur. More than one
determinant is needed to know them (age, other comorbidities, bad habits and others. We need all of
these to make, as much as possible, good predictions). They are represented by predictive statistical
models.
Polio story:
In 1932 a polio vaccine was made by Dr. Kolmar. But it turned out that it is unsafe to be given to all
children due to the high percentage of dead and paralyzed children after having the vaccination.
Dr. Salk made the mineral oil vaccine against corona. He made the largest experiment ever. Children
without vaccination have 3.3 higher risk to get polio than vaccinated children.
They decided that the vaccine is safe because: 1- small number of cases in vaccinated persons.
2- lack of correlation between site of paralysis and inoculation. 3- absence in association between
time and vaccination. 4- absence of extensions in the house holds.
After that companies were needed which could make the vaccinations. The protocol given to these
companies was 5 pages instead of the 55 pages protocol of Salk. That leaded to many vaccinations
being unsafe to give. Nowadays we have two valid vaccines against polio: Salk and Sabin vaccines.
So, it takes time to make a safe effective vaccine.
Circular model for drug development:
1- Understand the mechanism of the disease that you want to make a drug for.
2- Drug discovery and finding the target.
3- Characterize the drug.
4- Formulate it, know how to make it in large scales, and pre-marketing phase.
5- Post-marketing phase, pharmacoepidemiologic studies, and economical evaluation.
Field of research in pharmacoepidemiology:
1- What proportion of the population will take the drug.
2- What determinants predict the use of the drug.
3- Do drugs decrease the rate of mortality, and diseased cases or improve the quality of life.
4- The numbers and types of side effects of this drug.
Drugs that have caused severe side effects:
Chloroform, thalidomide or softenon, birth control pill (Diane-35), valproic acid, mefloquine, and
MMRs vaccine.
,Also SSRIs as (venlafaxine) for children, Cox-2 inhibitors, statins, ACE-inhibitors, domperidone,
rosiglitazone, inhaled corticosteroids (in large doses), and anti-depressants.
Withdrawn drugs examples:
Valdecoxib, rofecoxib, Roaccutane, and cerivastatin.
Main reasons for dangerous side effects and withdrawn out of the market:
1- clinical trials are not big as you think.
2- real world use of the drug differs; the trial group is mostly healthy group.
3- the short duration of trials.
4- off-label usage of the drug (e.g. the usage of an adult-drug by a child).
85% of the trails have less than 5000 patients. And some of them have even less than 500 patients.
Some drugs show severe side effects when the DDD is more than 1. Thus, patients take more than the
safe dose. (the side effects then will be a feature of the patient and not of the drug).
Some drugs can be used after repurposing their usage. E.g. finasteride is used for BPH, but also useful
for growing hair. Statins also show goof effects during sepsis and copd exacerbations.
Drugs may also have preventative effects as aspirin that can protect from colorectal cancer.
,Lecture 2:
Introduction session 2:
Phases in drug development process:
Once the pharmaceutical industry gets the license to make the vaccine/drug, we go on to the post
marketing phase where the pharmaceutical product is available for people to buy. This is the post-
marketing phase. During this phase the pharmacoepidemiologic research takes place.
What are the typical pharmacoepidemiologic research questions?
Type 1: what is the use of a drug X in patient Y and what is the variation between age and sex?
In this case; the use of drug X is the outcome. Patient Y is the study domain/hypothetical student.
Variation between age and sex: determinant of the outcome. (We want to know something about the
use of the drug X that is why it is the outcome)
Type 2: What is the effect of drug X in the treatment of disease Z on outcome Y?
Drug X: is the determinant/exposure. Outcome Y: the outcome. Disease Z: the study
domain/hypothetical patient.
(Here is the effect of drug X important for us on the course of disease, that is why drug X is the
determinant that will determine the outcome Y).
To make a proper question for our polio story, we use the second type of question.
What is the effect of vaccine X in preventing the occurrence of polio in children?
Polio vaccination here is the main determinant, the occurrence of polio is the outcome, and the
children is the hypothetical patient/study domain.
Types of epidemiological studies in order to answer the epidemiologic questions:
1- Experimental studies: studies in which the investigator determines what the participant is going to
get (Randomized clinical trials). When talking about community we talk about community
intervention trials e.g. anti-corona measures are working? Go to three UMC’s and install a very
progressive program to get everyone isolated and ensure that everyone is diagnosed as soon as
possible, and compare that to three UMC’s that do not have this program.
Field trials are much bigger. E.g. Malaria trials, like when in Kenia they did bed nets on a lot of
houses and see if it is a helpful manner to get rid of the insect that causes Malaria.
2- Observational studies (post-marketing surveillance): no influence if you get the vaccine or not.
These are descriptive studies (Prognostic and diagnosis). OR case-control studies which gives us
, analytical studies as in interventions and etiology where there is a causal relationship.
General characteristics of randomized clinical trials:
Of the population: 1- narrow clinical indication.
2- homogeneous population (Do not give the drug to people that may develop severe side effects due
to higher risk because of either pregnancy or a bad habit).
Of the exposition: 3- small number.
4- randomized. (computer determines what the patient will get the product or not)
5- controlled. (Everything has to be recorded to make the comparison in the end)
Of the outcome: 6- Double-blind. (Both patient and investigator do not know whether you got the
pharmaceutical or the placebo).
7- Relatively short-term therapy. (because trials are VERY expensive)
RCTs differ from observational studies in the post-marketing phase, where the real-world use of the
drug is shown. Also, there is no limited time where the drug is used in.
RCTs participants make a very small part out of the whole population uses the drug.
What is the important clinical information that has to be obtained from PMS?
Hard-end points are death or hospitalization.
Effect on population level: e.g. when giving a vaccine against corona to all the children, the
transmission may stop. Also, pharmacoeconomic is important to us. Especially cost-effectiveness.