Research and EBP Flashcards
Discuss descriptive statistics and distributions.
If data are distributed normally, the three measures of central tendency are equal: mean, median, and mode. In this case, mean is the most commonly used. If there are are extreme scores, the median is considered the best measure of central tendency.The standard deviation is most descriptive of the variability when mean is used, while min-max is probably best when median is used. A normal distribution is also called a bell curve and gaussian distribution. A skewed distribution occurs to the right when there are abnormally high scores, and to the let with ow scores.
Describe the best statistical tests.
If you want to learn about the association between two variables, calculate the correlation coefficient. If prediction is required, a regression analysis is appropriate. If you want to determine if a treatment has an effect, a chi-square, ANOVA, or t-test is appropriate.
What is the difference between ordinal, nominal, ratio, and interval info?
Nominal is categorical, ordinal is rank order without set differences between values. For statistical purposes, there is no real difference between ratio and interval info.
What tests are used for ordinal data?
For correlation studies, Spearman rho is used (rho is rank order). For interval data, a correlation cooeficient is used. values of -1.00 to +1.00 is used.
What are statistical tests for differences?
A chi-square is used for nominal data, ANOVA is typically used for interval data of normal distribution.
Define sensitivity, specificity, PPV, and NPV.
Sensitivity: probability of having a true-positive test (proportion of people with a condition who also have a positive test)
Specificity: probability of having a true negative test (proportion of people without a condition who also have a negative test)
PPV is the proportion of patients with a positive clinical tests who actually have the condition
NPV is the proportion of patients with a negative test who really don’t have the condition.
In other words, specificity and sensitivity deal with reality, whereas predictive values deal with the results of a clinical test (more common clinical situation). The problem with predictive values is they change based on prevalence, where as sensitivity and specificity are properties of the tests themselves. PPV/NPV will change based on the population, sensitivity/specificity/LR will not.
Is there a way to combine the best characteristics of sensitivity, specificity, PPV, and NPV?
Yes. Likelihood ratios. A LR+ is sensitivity/complement of specificity (1-specificity). A LR- is complement of sensitivity/specificity.
A LR of 1.00 means you could just flip a coin.
What are SnNouts and SpPins?
Used to help remember the most useful aspects of a test with moderate to high values.
SnNout- A test with a high sensitivity value that when negative helps to rule out a disorder
SpPin - A test with a high specificity value when positive helps to rule in a disease.
What is the Ottawa Ankle Rule?
You only need to do an ankle x ray if there is pain in the malleolar zone and either 1) bone tenderness in the posterior aspect of either the anterior or posterior malleoli or 2) inability to bear weight both immediately and in the ER
You only need to do a foot x-ray if there is pain in the midfoot and either 1) bone tenderness in the naviculum or base of the 5th metatarsal or 2) inability to bear weight both immediately and in the ER
Sensitivity is about 98%, specificity is about 50%
What is the Ottawa knee rule?
An x-ray for the knee is only required if one of the following is met:
1) Age 55 or older
2) ISOLATED tenderness of patella (no where else)
3) Fibular head tenderness
4) inability to flex to 90*
5) inability to WB both immediately and in the ER for 4 steps
Sensitivity is about 98%, specificity about 50&
What is the Canadian C-spine rule? How does it compare to the Nexus cervical spine study?
Only applicable for alert patients with trauma:
A need for an x-ray is determined by:
1) Any high risk category is met (older than 65, parasthesias in limb, dangerous mechanism)
2) If the patient has a low risk factor that allows for movement assessment (pt was in a small MVA, has no midline pain, is sitting or walking, or has delayed onset pain) AND they can rotate their head 45*, an x-ray is not needed.
Dangerous mechanisms include fall from higher than 3 feet, axial loading, high speed MVA, high speed collision.
Sensitivity is 100%
Nexus cervical spine study basically says an x-ray is not needed if there is no midline tenderness and the patient is alert and has no neurological symptoms. It only has a sensitivity of 90%, so it isn’t as good.
What is the CPR for patients with low back pain who are most likely to benefit from spinal manipulations.
Flynn et al.
If 4 of the 5 findings are positive, they would most likely benefit from spinal manipulation (LR+ = 23)
1) Episode of pain < 16 days duration
2) FABQ work score <19
3) no radicular symptoms below knee
4) hypomobility in at least one lumbar segment
5) at least one hip with 35* IR
