Biostatistics/Ophthalmology Basics

Question 1

What are the features of a complete CNIII palsy?

Answer 1

Only lateral rectus and superior oblique will be active

LR = abduction
SO = Downwards and abduction

Result = Down and outward eye turn with mydriasis (dilated pupil)

Kevan Otology Page 25

Question 2

Differential diagnosis of sudden vision loss

Answer 2

1. Vitreous hemorrhage
2. Central retinal (Artery or vein) occlusion (pale fundus + cherry red spots)
3. Retinal detachment
4. Temporal arteritis
5. Optic neuritis (if young, consider MS)
6. TIA
7. Malingering

Question 3

Normal visual fields

Answer 3

• 90 degrees ipsilateral
• 50 degrees contralateral

https://www.vision-and-eye-health.com/images/xVisualFieldHoriz.gif.pagespeed.ic.iHnN9uU-0e.png

Question 4

What is strabismus vs. amblyopia?

Answer 4

Strabismus = misalignment of the eyes

Ambloypia = Unilateral or bilateral reduced vision without physiologic defect of one eye or the other (abnormal central processing)
- Does not improve with glasses
- Greater refractive error so the brain suppresses the weaker image
- Treatment = patching the stronger eye (younger = improved outcome)

Question 5

Differentiate Hordeolum, Chalazion, and Blepharitis

Answer 5

Hordeolum = Infection of lid glands, Rx topical antibiotics

Chalazion = Obstructed meibomian gland, Rx warm compress, tobradex or curettage

Blepharitis = Chronic staph infection of the upper lid, associated with seborrhea

Question 6

What are the different types of conjunctivitis?

Answer 6

1. Virus: Adenovirus (pink eye)
2. Bacterial: Staph, Haemophilus influenzae, diplococcus, N Gonorrhea (if ++ purulent)
3. Allergic

Question 7

What are the main causes of:
1. Keratitis
2. Scleritis
3. Iritis

Answer 7

Keratitis:
- HSV
- VZV

Scleritis:
- 50% systemic disease (e.g. RA)

Iritis:
- TB
- Sarcoid
- Ankylosing Spondylitis
- RA
- Reiters
- Behcet’s
- Gonorrhea

Question 8

What does p < 0.001 mean?

Answer 8

Assuming that H0 is true (ie. there is no difference between the two groups studied), the probability of observing the result obtained in the experiment is 1/1000 (0.1%)

H0 = usually defined as “no difference” between groups, therefore p = probability of H0 being true

P = probability that result occurred by chance alone (probability that H0 is true)

Question 9

What is the precise meaning of a p value < 0.05?

Answer 9

Assuming that the null hypothesis is correct, the probability of obtaining the observed result in a study is < 5% (5/100)

Question 10

Define the meaning of a 95% confidence interval

Answer 10

If a study is repeated infintely on new samples of individuals, the confidence interval will include the true result 95% of the time

Note: It does NOT mean that there is a 95% chance that the true value falls within the reported confidence intervals

Question 11

Define Sensitivity

Answer 11

True positive Rate - The proportion of those who have the condition that received a positive results on the test (TP / (TP + FN))

100% sensitive = negative test rules the condition out (SNout)
- If test positive, can be true positive or false positive, therefore sensitivity doesn’t tell you about positive tests as well. Just tells you that if you DO TRULY have the disease, then you will be positive
- If you test negative, can’t be a false negative because 100% Sensitivity means FN = 0.

Vancouver Page 69

Question 12

Define Specificity

Answer 12

True negative Rate - The Proportion of thsoe who do not have the condition that received a negative result on this test
- TN / (FP + TN)

100% specific = Positive test rules the condition in (SPin)
- If test negative, can be true negative or false negative, therefore specifisity doesn’t tell you about negative tests as well. Just tells you that if you DON’T TRULY have the disease, then you will be negative
- If you test positive, can’t be a false positive because 100% Specificity means FN = 0.

Vancouver Page 69

Question 13

Define Type 1 vs. Type 2 error

Answer 13

Type 1 Error (alpha): Chance of finding a difference when there isn’t one
- Reject the H0 (null hypothesis) when it is in fact true (false rejection)
- False positive rate
- 1 - specificity

Type 2 error (beta): Chance of finding no difference when there is one
- False acceptance of H0 (null hypothesis)
- Dependent on Type 1 error and the sample size
- False negative rate
- 1 - sensitivity
- Generally worse than a Type 1 error (don’t want inaccurate false negatives, if a patient is actually positive!)

Null hypothesis = there is no difference. If you reject this that means there is a difference, ie. you would get a positive result. If you accept this (not reject) that means there is no difference, ie. you would get a negative result.

Think of rejection as a positive result.
Not rejection is a negative result.
If you reject when its actually true, then its a false rejection (ie. false positive)

Vancouver Page 69
https://www.scribbr.com/wp-content/uploads/2021/01/type-i-and-ii-error-2.png

Question 14

Define prevalence calculation

Answer 14

Condition positive / Total # of people in the sample

Question 15

Define power

Answer 15

Power = Sensitivity = 1 - beta (type 2 error)

Probability of rejecting a false null hypothesis

Question 16

How do you calculate positive predictive value and negative predictive value?

Answer 16

Positive predictive value = proportion of subjects with a positive test result who truly have the outcome of interest
- Looking at all those that test positive, who actually has the disease (whereas sensitivity looks at the proportion who ACTUALLY have the disease, who test positive)
- TP / TP + FP

Negative predictive value = proportion of subjects with a negative test result who truly do not have the outcome of interest
- At all those who test negative, who actually does not have the disease
- TN / TN + FN

Vancouver Page 69

Question 17

What is a likelihood ratio? How is it interpreted?

Answer 17

LR = How much we should trust a positive / negative test result

LR > 1 = Indicates an increased probability that the target disorder is present
LR < 1 = indicates a decreased probability that the target disorder is present
LR = 1 = test result does not change the probability of disease at all

Question 18

What is a positive likelihood ratio vs. negative likelihood ratio

Answer 18

LR+ = Probability that a person with the disease tested positive/probability that a person without the disease tested positive.
- = Sensitivity / (1-Specificity)

LR− = Probability that a person with the disease tested negative/probability that a person without the disease tested negative.
- = (1 - Sensitivity) / Specificity

Vancouver Page 69

Question 19

Define relative risk

Answer 19

Ratio of the risks for the disease in the exposure group, to the risks for the disease in the non-exposure group

[Disease with exposure] vs. [Disease with No exposure]

[A/(A+B)] / [C/(C+D)]

A = Exposed + disease
B = Exposed + no disease
C = Not exposed + disease
D = Not exposed + no disease

Question 20

Define Odds ratio

Answer 20

OR represents the odds that an outcome will occur given a particular exposure, compared to the odds of the outcome occurring in the absence of that exposure.

(sick vs. not sick in exposed group) vs. (sick vs not sick in unexposed group)

(A/B) / (C/D)

Question 21

What is absolute or attributable risk (or risk difference)?

Answer 21

How much of the disease can be attributed to the exposure (absolute terms)?
[Disease with exposure] - [Disease with no exposure]

[A/(A+B)] - [C/(C+D)]

Also known as:
Absolute risk reduction = absolute change in risk (e.g. risk reduction from 50% to 25% - the absolute risk reduction is 25%. The relative risk reduction is 50%)

Question 22

How do you calculate the Number needed to treat?

Answer 22

Number needed to Treat (NNT) = 1/ARR
- Number of people needed to treat to prevent one bad outcome

Vancouver Page 69

Question 23

What does “spread about mean” mean in statistical terms?

Answer 23

Variance (or dispersion) –> ie. wider standard deviation

Question 24

What does skew mean in statistical terms?

Answer 24

Long tail (bulk of values lie on opposite side of mean)

Question 25

What does validity mean in statistical terms? What is internal vs. external validity?

Answer 25

Validity = Test correctly measures/detects what it’s supposed to measure/detect
- Internal validity = accuracy
- External validity = generalizability

Question 26

What does reliability mean in statistical terms?

Answer 26

Consistency (reproducibility)

Question 27

What is the PICO method for article review?

Answer 27

P = Population being studied
I = Intervention applied
C = Comparison to what intervention (or placebo)
O = Outcome analyzed