Appraisal of clinical studies

Question 1

Nominal -

Answer 1

Categorical data; shows differences in proportions
* ex. Hair color

Question 2

• Ordinal -

Answer 2

Numerical scores to establish a ranking over a set of data points; shows differences
in rank order
* Ex. Pain scale

Question 3

• Continuous -

Answer 3

Numerical scores that are infinite and have meaning; shows differences in
means (if normal distribution)
* Ex. Perio pockets

Question 4

ABSOLUTE RISK REDUCTION (ARR)
(3)

Answer 4

• Prevalence between groups is RISK
• ARR is the difference (subtracting one group from another) to
  show the difference between groups
• Usually control (CER) and experimental (EER) groups
• CER-EER = ARR

Question 5

RISK RATIO
* Usually stated as an estimate of —
* Commonly used to assess

Answer 5

“____ times more likely”
cause, prevention and treatment

Question 6

RELATIVE RISK REDUCTION
(3)

Answer 6

• CER-EER/CER = RRR
• Does not discriminate huge absolute treatment effects from trivial ones
  because the RRR discards the underlying susceptibility (baseline risk) of
  patients entering RCTs
• Therefore, RRR cannot discriminate huge benefits and risks from small
  ones

Question 7

NNT (NUMBER NEEDED TO TREAT)
(4)

Answer 7

• Average # of patients who need to be treated to prevent
  one bad outcome
• Ideal NNT = 1 (everyone gets better)
• The higher the NNT = the less effective
• Used with therapeutic interventions

Question 8

NNH
(NUMBER NEEDED TO HARM)
(3)

Answer 8

• Similar to NNT except NNH is a bad outcome or risk
  factor
• Used for detrimental effect or risk factor
• The lower the number = worse outcomes
• NNH = 1 means every patient is exposed to harm