McNeil's notes - stats/epi, coagulation cascade, trauma, cancer (gen, uro only), ICU, fluids, anesthesia, Neurology, SCI/nerve injuries Flashcards
A new test is able to identify true positives in 350 patients, true negatives in 1200 patients and false negatives in 150 patients. What is the sensitivity of the test? A. 60% B. 70% C. 80% D. 90%
Answer: 70%
Sen = TP/TP + FN
When the test has 95% sensitivity, this means:
a. 95% of the patient will be positive
b. Patients who are tested positive, 95% of them will have the disease
c. Patients who are tested positive, 5% will have the disease
d. Of the patients who have the condition, the test will detect 95% of them
d. Of the patients who have the condition, the test will detect 95% of them.
sensitivity is the ability of the test to identify correctly those who have the disease (SNOUT-highly sensitive test, a negative result helps rule out disease); specificity is the ability of the test to identify correctly those who do not have the disease (SPIN-a highly specific test, a positive result helps rule in the disease).
A new blood test is available to diagnose pulmonary embolus. The data from a trial of 1000 post operative patients is shown in the 2x2 table below:
PE Present PE Absent
Test Positive 95 100
Test Negative 5 800
1) The sensitivity of the above test is calculated by the equation
a. 800/(100+800)
b. 800/(800+5)
c. 95/(95+5)
d. 95/(95+100)
e. none of the above
c. 95/(95+5)
Sensitivity = true positive (TP)/true positive + false negative
A new blood test is available to diagnose pulmonary embolus. The data from a trial of 1000 post operative patients is shown in the 2x2 table below: PE Present PE Absent Test Positive 95 100 Test Negative 5 800 2) The specificity of the above test is calculated by the equation a. 800/(100+800) b. 800/(800+5) c. 95/(95+5) d. 95/(95+100) e. none of the above
a . 800/(100+800)
Specificity = true neg / (true neg + false positive)
A new blood test is available to diagnose pulmonary embolus. The data from a trial of 1000 post operative patients is shown in the 2x2 table below: PE Present PE Absent Test Positive 95 100 Test Negative 5 800 3) The positive predictive value of the above test is calculated by the equation: a. 800/(100+800) b. 800/(800+5) c. 95/(95+5) d. 95/(95+100) e. none of the above
d. 95/(95+100)
PPV = true positive/(true positive + false positive)
A test that detects the number of people who actrally have the disease measures …
a) specificity
b) sensitivity
c) positive predictive value
d) negative predictive value
Answer given: sensitivity. Disagree however - this sounds more like PPV.
sensitivity – the proportion (%age) of truly diseased people identified as diseased by a screening test
specificity - the proportion (%age) of truly non-diseased prople identified as non-diseased by the screening test
PPV - the proportion of true positives in all positive tests (the number of cases that truly have disease among all those who test positive)
NPV – the proportion of true negatives in all the negative tests
Which of the following is the MOST helpful in establishing a causal relationship between exposure and disease? A. Positive predictive value B. Sensitivity C. Odds ratio D. T-test
Odds ratio = retrospective studies
Risk = prosp[ective studies
Answer: Odds ratio
A world health organization epidemiologist is studying esophageal cancer in females. In Canada, 5,000,000 females over the age of 25 years have been followed from January 1, 1980 to January 1, 2000. Within this population, a group of 1,000,000 women chronically exposed to sulphur dioxide fumes are found to have an increased incidence of developing esophageal cancer as compared to the 4,000,000 that were not chronically exposed. The data is shown below:
Group Esophageal Cancer Incidence (per 1,000,000)
Exposed 100
Not exposed 20
Calculate the relative risk of developing esophageal cancer in those women chronically exposed to sulphur dioxide fumes
a) 100
b) 20
c) 5
d) 0.2
e) none of the above
Answer: 5
Relative Risk = A/ A+B
C/ C+D
RR = 100/1,000,000
80/4,000,000
= 100/1,000,000 20/1,000,000 = 5
A new treatment changes the mortality of acute MI from 26 % in the placebo group to 16 % in the treatment group. The number needed to treat is:
a. 10
b. 100
c. 200
d. 1000
a. 10
ARR = |CER - EER| = |26% - 16%| =10% NNT = 1/ARR = 1/.1 = 10
If β blockers decrease risk of MI by 25% and the mortality of MI is 1% in one year, what is the absolute risk reduction and the number of patients needed to be treated to decrease mortality by one patient:
a. 0.25% - 400 patients.
b. 2.5% - 40 patients.
c. 25% - 4 patients.
d. 1% - 10 patients
Answer: 0.25% - 400 patients
ARR = |CER - EER| where CER = control group event rate (1%) EER = experimental group event rate (1% * [1- 0.25] = 0.75%) ARR = 1% - 0.75% = 0.25%
Power is:
a. Probability to detect statistically significance if one exists
b. A calculation of sample size
c. A calculation of validity
d. Positive predictive value
e. 1 – sensitivity
f. is not related to specificity
g. a stastistic that is not dependent on the prevalence
h. Calculation of the sample size needed to determine if a difference exists
Answer: Probability to detect statistically significance if one exists
The power of a trial is the probability of detecting a treatment effect of a given size, if one truly exists.[65] [66] [67] [68] Studies are usually designed to have a power of 0.80 or greater. Because the power of the trial is the chance of finding a true treatment effect, the quantity (1 - power) is the chance of missing a true treatment effect (i.e., risk of committing a type II error).[65] [67] The value of (1?β), or the power, and the magnitude of the treatment effect the clinical trial is designed to detect (defined by the alternative hypothesis) determine the sample size required for the study.[68]
Regarding statistical errors all are true EXCEPT:
a. The probability that the null hypothesis is considered false when it is true is called the beta
b. In a fixed sample population α error is inversely proportional to the β error.
c. Increasing the sample size decreases α error but doesn’t change the β error.
d. Power equals (1-beta)
Power: The probability of detecting an effect in the treatment vs. control group if a difference actually exists. Must also specify the size of the difference. For example, a paper describing a clinical trial with a new hypertension medication may contain the following statement - “The study had a power of 80% to detect a difference of 5 mm Hg in diastolic blood pressure between the treatment and control groups.” Typical power probabilities are 80% or greater. Power = 1 - ß
Type I Error: Mistakenly rejecting the null hypothesis when it is actually true. The maximum probability of making a Type I error that the researcher is willing to accept is call alpha (a). Alpha is determined before the study begins. False positive conclusion. Studies commonly set alpha to 1 in 20 (=0.05).
Type II Error: Mistakenly accepting (not rejecting) the null hypothesis when it is false. The probability of making a Type II error is called beta (b). Power = 1 - b (see above). False negative conclusion. For trials the probability of a b error is usually set at 0.20 or 20% probability. A 20% chance of missing a true difference.
Which is the most appropriate test to compare the means of 2 normal distributions?
a. Chi square
b. t-test
c. ANOVA
d. Fischer exact test
e. Variance
Answer: t-test
Student’s t test for independent samples is used to determine whether two samples were drawn from populations with different means.
Needs to be normally distributed population
A study has been designed to assess the wear properties of two different metal interfaces for total hip arthroplasty. There are 11 patients in one group and 13 in the other. Which test would best determine if a difference exists between these two groups?
a) t-test
b) Fischer exact test
c) Chi-square test
d) Linear regression
? T-test if the outcome is the same and normal distribution. Need more info.
Chi-square test - Used with categoric variables (two or more discrete treatments with two or more discrete outcomes) to test the null hypothesis that there is no effect of treatment on outcome; assumes at least five expected observations of each combination of treatment and outcome under the null hypothesis
Fisher’s exact test - Used similar to chi-square test; may be used even when fewer than five observations are expected in one or more categories of treatment and outcome
- Marx: Rosen’s Emergency Medicine, 7th ed.
Which test used to measure proportion :
a. Chi square
b. t-test
b. Nova
c. None of the above
d. All of the above
Answer: Chi square
To complete a retrospective study with a dichotomous outcome i.e. yes or no, the following tests could be used except:
a) Chi Square
b) Fischer T
c) Students T
d) Odds ratio
T-test
a) Chi Square (yes)
b) Fischer T (used for categorical data)
c) Students T (measure differenceb/w two means)
d) Odds ratio (yes)
Doing a study of femoral head size (26- 28-32 and 36) in conjuntion with rate of dislocation. What is the most appropriate statistical test? a - student t teat b – fisher exact test c – ANOVA d – chi squared
Chi squared?
Z-test/t-test tests differences between two sample means for continuous data.
Chi-square test - Used with categoric variables (two or more discrete treatments with two or more discrete outcomes) to test the null hypothesis that there is no effect of treatment on outcome; assumes at least five expected observations of each combination of treatment and outcome under the null hypothesis
Fisher’s exact test - Used similar to chi-square test; may be used even when fewer than five observations are expected in one or more categories of treatment and outcome
Analysis of variance compares mean values from three or more groups simultaneously
Define standard deviation. A. Difference between mean and median B. Measure of variance and dispersion C. The midpoint in a series of numbers D. Measure of dispersion around the mode
Answer: Measure of variance and dispersion
The standard deviation is the average deviation of scores around the mean of the variable for the set of observations in the sample. It is often designated SD.
- Tasman: Psychiatry, 1st ed
Given a normal distribution 1 standard deviation is:
a. 2/3
b. 99/100
c. 3/4
d. 1/3
e. None of the above
- 1 SD = 68%, 2 SD = 95%, 3 SD = 99%
The standard deviation is the square root of the variance. For a normally distributed population, 68% of values fall within 1 standard deviation of the mean and 95% of values within 1.96 standard deviations.
- Long: Principles and Practice of Pediatric Infectious Diseases, 3rd ed.
Patients are administered a survey with pain scale to check the effectiveness of a new non-steroidal antiinflammatory drug. The mean is 12, the mode is 8, median is 10, and SD is 11. Which of the following is true
a) most common value is 10
b) the average is 8
c) 68% of values will fall between 1 and 23
c.
a) most common value is 10 (8)
b) the average is 8 (12)
c) 68% of values will fall between 1 and 23 (12 +/-11) one SD
What is the variance of 3,6,9,11:
a. 3.00
b. 5.00
c. 7.00
d. 11.00
e. 29.00
Answer is none of the above. (12.25)
Summary of the calculation procedures:
- subtract the mean from each score
- square each result
- sum all the square
- divide the sum of square by N. Now you get variance
All of the following represent quantitative continuous data EXCEPT: A. Age B. BP C. Number of asthma attacks per month D. Height
Answer: Number of asthma attacks per month
Which of the following is an example of ordinal data? (Ordinal = order) A – stages of breast cancer B – ABO blood type C – sex D – Death / Life
Answer: stages of breast cancer
Ordinal data are categorical data where there is a logical ordering to the categories. A good example is the Likert scale that you see on many surveys: 1=Strongly disagree; 2=Disagree; 3=Neutral; 4=Agree; 5=Strongly agree.
- http://www.cmh.edu/stats/definitions/ordinal.htm
What does the Kaplan-Meier Curver measure?
Answer: survival