Observational Studies

Question 1

What are observational studies?

Answer 1

Observe the effect of an exposure (or intervention) on an outcome without “interfering” with anything

ex. Survey all PHAR 271 students about Vitamin D intake. Compare exam grades between those who declared using Vitamin D in the past and those who have not.

Question 2

Are observational studies looking for causation over associations?

Answer 2

No, Observational Studies are looking for associations. They are unable to truly find causation

Question 3

Why are observational studies weaker than randomized controlled trials in terms of scientific rigour?

Answer 3

Lack of randomization in observational studies increases the effect of confounding on results

Question 4

Are randomized controlled trials more common than observational studies in terms of published research?

Answer 4

No, despite RCTs having more significant findings, observational studies form a significant proportion of published research

Observations Studies are more practical (all individuals using a drug can be tracked), ethical, and the findings are more generalizable (due to a larger sample size and less sterile conditions)

Question 5

Can the findings of a RCT overrule the results of multiple observational studies?

Answer 5

Yes, RCTs have greater scientific rigour and have more substantial internal validity

Question 6

What is the main challenge faced by the findings of an observational study?

Answer 6

Is the association found by the study real… or due to a confounding variable

Question 7

What are the main types of observational studies?

Answer 7

1. Case reports (Case series)
2. Cross-sectional
3. Case-control
4. Cohort

Question 8

What are some qualities of a case report?

Answer 8

They describe and interpret an individual case

1. Usually written as a narrative (no predefined methods or outcomes)
2. Patient described in detail
3. Observations described in detail (exposure and outcomes)

Case reports often describe unexpected events, adverse effects, variations of known disease, weird observations that do not fall within a known disease

Question 9

What are some advantages of case reports?

Answer 9

1. Useful when identifying new or unusual trends or diseases
2. Can identify new drug effects (can be good or bad)
3. Describes novel interactions (treatment and diagnostic procedure)
4. Suggest areas for areas of further research

Question 10

What are some disadvantages of case reports?

Answer 10

1. Difficult to interpret (spontaneous appearance and disappearance of symptoms, placebo effect, other influences)
2. Cannot confirm or prove anything

Question 11

What are case series in the context of observational studies?

Answer 11

They are a collection of individual reports which typically occur within a fairly short period of time

ex. CDC received many reports of the rising prevalence across the US during 1985. Turns out these reports were the first detection of HIV/AIDs virus in North America

Question 12

What are cross-sectional studies?

Answer 12

They look at population or study sample at one point in time (snapshot of what is happening at that time)

Collect information on: (Exposure, Outcome, Health status, behaviours, etc.)

Collect information by: (Surveys/Questionnaires, Chart reviews, administrative agencies like eHealth)

Question 13

What are some benefits of cross-sectional studies?

Answer 13

They are useful for prevalence (number of patients with condition at a given point in time)

Good for quick examination of potential associations to be taken further with other study designs like RCTs in a separate research study

They are also inexpensive vs. RCTs

Question 14

What are some disadvantages of cross-sectional studies?

Answer 14

They are limited to what we know about the potential associations (lack of before and after study information)

Transient effects (unusual circumstances) can affect the baseline of some measurements.

Question 15

What are case-control studies?

Answer 15

They are retrospective studies. Researchers look at outcomes and attempt to determine what factors caused a given outcome.

There are two groups in a case-control study (cases and controls)

Cases = people with the outcome

Controls = people without outcome

Question 16

When selecting an outcome for a case-control studies, what are some requirements?

Answer 16

1. Need to have a very specific, validated outcome (use standardized accurate measurements/tools, criteria)
2. Ideally select incidence cases over prevalent cases (allows us to study when outcome occurred and the exposure prior)

Question 17

Review slide 18 for a illustrated description of a case-control study

Question 18

When selecting the controls in a case-control study, what are some requirements?

Answer 18

The control and case groups should be as similar as possible (except for the outcome)

The standard number of controls is 4 for every case (increases the power of the study)

Question 19

What strategies can increase the similarities between case and control groups in case-control studies?

Answer 19

Matching cases and controls increase the similarity between the two groups and decrease confounding.

Question 20

What are some advantages of case-control studies?

Answer 20

1. Good for studying rare diseases or outcomes (we start from rare outcome and go retroactive to study potential causes)
2. Relatively quick to conduct (data has already been collected, just needs analysis)
3. Good for establishing an association that can be tested via further research

Question 21

What are some disadvantages of case-control studies?

Answer 21

1. Not randomized (susceptible to confounding)
2. Susceptible to bias
3. Only suggest association (cannot prove cause and effect)

Question 22

What are cohort studies?

Answer 22

Groups (or cohorts) of subjects are created and compared according to exposure (exposed vs. unexposed group over time)

Similar to RCTs, but subjects are not randomized into groups or assigned the exposure

This type of observational study produces the highest level of evidence, and it is the best type for finding a valid association

Question 23

What are the two types of cohort studies?

Answer 23

Prospective (At the start of study, the outcome has not occurred)

Retrospective (Outcome occurred before the study started)

Question 24

Review slide 24 for an illustration of a prospective cohort study

Question 25

Review slide 25 for an illustration of a retrospective cohort stufy

Question 26

What are the differences between a prospective and retrospective cohort studies?

Answer 26

Prospective:
1. Conducted in real-time
2. Researchers have some control over what data is collected

Retrospective:
1. Good for long follow-up studies (observation is already complete)
2. Use retrospective data collected prospectively
3. Data limitations (questionable accuracy and missed information)

Question 27

What are the advantages of a cohort study?

Answer 27

1. Provides highest level of evidence for observational studies
2. Can study multiple outcomes after a single exposure

Question 28

What are the disadvantages of a cohort study?

Answer 28

1. Susceptible to bias
2. Not randomized, so susceptible to confounding
3. More resource intensive compared to other observational study designs