Date Article Title Citation

1
DateArticle TitleCitation

• Resident Name
• Faculty Name

2
Background

• Type of Study
• Question being addressed Not the study
  hypothesis, but the issue that lead to the study
  being done.
• Clinical Context the information that goes into
  the introduction section of the article

3
Framwork Overview MAARIE

• Method
• Assignment
• Assessment
• Results
• Interpretation
• Extrapolation

4
Method Null Hypothesis

• Study hypothesis What is the study question?
• What type of investigation will be used to
  address this hypothesis?

5
Method Population

• Study population What are the inclusion and
  exclusion criteria, or how was the study
  population chosen?
• How were controls chosen (if applicable)?
• Were the study group and the control group
  similar at baseline?

6
Method Sample Size and Power

• Sample size and statistical power How many
  subjects are there? Are the numbers sufficient to
  demonstrate statistical significance if the
  hypothesis is true?
• Was a power analysis performed to determine the
  enrollment goal of the study?

7
Assignment

• How were participants assigned to study and
  control arms of the experiment?

8
Assignment Questions

• 1. Process of assignment What method is used to
  identify and assign individuals to study and
  control groups?
• 2. Confounding Variables Are there differences
  between the study and control groups, other than
  the characteristic being studied, which may
  affect outcomes?
• 3. Masking or blinding methods

9
Assessment

• Measurements of outcomes or endpoints in the
  study and control groups
• Does the measurement of an outcome address the
  study question?
• Is the measurement of an outcome an accurate and
  precise measure of the phenomenon that the
  investigation seeks to measure?

10
Assessment, continued

• Thoroughness and Biases of Assessment
• How efficient was follow-up?
• Why were subjects lost to follow-up?
• Are the incomplete data likely to influence the
  data in support, or against support, of the study
  hypothesis?
• Was the completeness of follow-up affected by the
  participant or investigators knowledge of study
  vs. control group assignments?

11
Results

• What are the outcomes in the study group?
• What are the outcomes in the control group?
• How do they compare?

12
Results Key Questions

• 1. Estimation What is the magnitude or strength
  of the association observed between an outcome
  and an exposure?
• 2. Inference What statistical techniques are
  used for statistical significance testing?
• 3. Adjustment What techniques have been used to
  control for differences between study groups and
  control groups, if any?

13
Interpretation

• Does the factor under study alter the probability
  of the outcome of interest?
• Are there harms associated with the intervention?
• Are outcomes observed in subgroups of the study
  population different from outcomes observed in
  the overall study?

14
Extrapolation

• Do the investigators extend the conclusions of
  the study to a population similar to the study
  population? Is this warranted?
• Do the investigators extrapolate their findings
  beyond the scope of their data?
• Can the results be applied to our patient
  population (or a subset of our patients)?

15
Applying MAARIE to case-control studies

• Method case-control
• Assignment The study group have had the outcome
  of interest the controls are otherwise similar
  subjects who have not had the outcome of interest.

• Assessment Determination of whether cases and
  controls have had the same exposure history
  preceding the development of the outcome of
  interest in the study group.

16
MAARIE for case-control study

• The previous presence or absence of a risk factor
  or exposure of interest in the study group
  compared to the control group is the outcome.
  Since the outcome has already occurred,
  independently of the investigation, the
  allocation to study and control groups is
  pre-defined. This assignment is called observed
  assignment.

17
Case-control Results

• Calculation of the chance that the study
  population had the exposure of interest, compared
  to the chance that the control population had the
  exposure of interest.
• Presented as an odds ratio using discordant pairs.

18
Case-Control Interpretation and Extrapolation

• What conclusions can be drawn with respect to the
  differential effects of exposures of interest in
  the populations of the cases and controls?
• Can these conclusions be extended to categories
  of subjects or types of risk exposure different
  from those in the present study?

19
MAARIE for a Cohort Study

• The cohort study design differs from the
  case-control study design in that the
  identification of individuals for study and
  control groups is made before the investigator
  knows which subjects have developed the disease
  of interest. The cohort is a group of subjects
  who share a similar environment.

20
Cohort Methodology

• The investigational cohort possesses
  characteristics (occupational, environmental,
  geographic, etc.) which are hypothesized to
  contribute to disease pathogenesis.

• The control cohort is similar to the
  investigational cohort, except with respect to
  the exposures of interest.

21
Cohort Study Design, MAARIE

• Assignment A group with the circumstance of
  interest is selected as the study cohort. A
  similar group, except in the circumstance of
  scientific interest, serves as the control cohort.

• Assessment Who in the study group, and who in
  the control group, develops the prospectively
  defined outcome?

22
Cohort Study Design MAARIE

• Results The chance of developing the endpoint in
  the study cohort is compared with the chance of
  developing th endpoint in the control cohort.

• Interpretation What are the implications of the
  incidence of the endpoint in the two cohorts?
• Extrapolation Can the same conclusions be
  extended to a broader group of subjects, such as
  our patients?

23
Randomized Clinical Trial

• Method assign individuals to a study or control
  group in random order
• Assignment randomized, prospective distribution
  to the control or to the intervention arm
• Assessment observe which of the subjects sustain
  the outcome.

24
Randomized Clinical Trial

• Result calculate the chance of the outcome for
  each group (study subjects and control subjects).
• Interpret how the intervention influenced the
  incidence of the outcome.
• Extrapolate whether the inference between
  intervention and outcome can be extended to
  subjects who differ from the population enrolled
  in the trial.

25
Compare and Contrast Study Designs

• 1. Case-Control
• 2. Cohort Study
• 3. Randomized Clinical Trial

26
Case-Control Design

• Useful for study of a rare disorder
• Fast, as outcome has already developed
• Multiple hypotheses can be tested at once (diet,
  drugs, lifestyle, family history)

• Cannot establish causation
• Subject to the bias of the investigator in
  selection of the controls and the features
  analyzed

27
Cohort Study

• Stronger cause-and-effect evidence, as the
  outcome has not occurred at the time of the
  assigment
• Multiple outcomes can be followed in one cohort
  (MI, CVA, Heart Failure, Renal Failure)

• Can be expensive and time consuming, depending on
  size of cohort, length of time it is observed,
  and depth of investigation

28
Randomized Clinical Trials

• Randomization helps to assure that the study
  intervention, not an underlying disposition,
  causes the outcome of interest.
• Critical for interventions to prevent, treat, or
  palliate disease

• Capable of establishing efficacy of an
  intervention
• Expensive
• Require Consent
• Require subjects to take risks