Term
|
Definition
Original research
Most current info
Found in journals
Answers "foreground" questions like for patient specific situations and complex clinical decisions |
|
|
Term
|
Definition
|
|
Term
| 1/2 and A/B/C classifications |
|
Definition
1: strong recommendation
2: weaker recommendation
A = good RCTs
B = poor RCTs
C+ = no RCTs but results are
overwhelmingly credible
C = observational studies |
|
|
Term
|
Definition
Meta-analysis
Systematic review
Randomized controlled trials
Cohort
Case-control
Case-series
Case-reports
Expert opinion |
|
|
Term
|
Definition
| Detailed description of a single case |
|
|
Term
|
Definition
| Examining multiple cases with similar characteristics |
|
|
Term
|
Definition
| Examine two populations at a specific moment in time |
|
|
Term
|
Definition
| retrospectively looking back at two populations to determine if a certain risk factor had an effect on a certain outcome |
|
|
Term
|
Definition
Retrospective: seperate by endpoint and look back to see effect of risk factors
Prospective: seperate by risk factors and then see how many people get a positive endpoint |
|
|
Term
|
Definition
an experimental method (rather than observational)
*Gold Standard |
|
|
Term
|
Definition
| Compilation of multiple studies on one subject into a database to make a giant study |
|
|
Term
|
Definition
| The statement in the study which you are looking to disprove |
|
|
Term
|
Definition
acceptable level of significance (usually 5%)
Meaning 5% chance of Type I error (flase + meaning rejecting the null when its actually true) |
|
|
Term
|
Definition
chance of making a Type II error (false - meaning that you didnt reject null hypothesis when its actually not true)
usually 10-20% |
|
|
Term
|
Definition
|
|
Term
|
Definition
says that if the test were repeated multiple times, there is a 95% chance that the value would fall between 2 set numbers
*If 0 is in the 95% CI then there is no statistical difference |
|
|
Term
|
Definition
data thats normally distributed (bell curve)
Seen with Continuous data (means) |
|
|
Term
|
Definition
non-normally distributed data (skewed)
seen with nominal/ordinal data (medians) |
|
|
Term
|
Definition
ordinal data and nonparametric continuous data
NOT nominal data |
|
|
Term
|
Definition
|
|
Term
|
Definition
drug info references, textbooks, review articles
Answers "background" questions like general information or already established knowledge |
|
|
Term
|
Definition
DrugDex
DiseaseDex
PoisonDex
AltMedDex
Reprotox
Trissel's (IV)
*best info for foreign drugs |
|
|
Term
|
Definition
Lexi Drugs
Lexi Pediatric Drugs
Lexi Interact
Lexi Patient Education
Kings Guide (IV) **has references |
|
|
Term
|
Definition
Compare drugs within the same class
Off-Label uses of drugs
Review of Natural Products |
|
|
Term
|
Definition
nominal data
2+ independent groups |
|
|
Term
|
Definition
nominal data
2+ independent
*one cell has an expected frequency of <5 |
|
|
Term
|
Definition
nominal data
2+ independent groups
*want to control for cofounders/strata |
|
|
Term
|
Definition
nominal data
2+ MATCHED groups |
|
|
Term
|
Definition
continuous data (means)
2 independent groups |
|
|
Term
|
Definition
continuous data (means)
3+ independent groups
*tells you there's a difference somewhere between the 3 groups, but doesnt tell between which) |
|
|
Term
|
Definition
continuous data (means)
2+ independent groups
*adjusts for confounders/covariates |
|
|
Term
|
Definition
continuous data
2 MATCHED groups |
|
|
Term
|
Definition
ordinal or nonparametric continuous data (medians)
2 independent groups |
|
|
Term
|
Definition
ordinal or nonparametric continuous data (medians)
2 MATCHED groups |
|
|
Term
|
Definition
ordinal or nonparametric continuous data (medians)
3+ independent groups |
|
|
Term
|
Definition
| database for primary literature |
|
|
Term
| How do you view MESH terms on PubMed? |
|
Definition
|
|
Term
|
Definition
| database for info on Pharmacy Practice |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
| info on Drug Interactions |
|
|
Term
|
Definition
|
|
Term
|
Definition
| info on Pregnancy/Breast Milk interactions |
|
|
Term
|
Definition
| using the patient's data based on randomization, regardless if they were adherent or not |
|
|
Term
| LOCF (last observation carried forward) |
|
Definition
| if a patient drops out, their last recorded measurement is carried forward thru the rest of the trial |
|
|
Term
|
Definition
| only analyze the patients who were adherent |
|
|
Term
| NNT (number needed to treat) calulated by |
|
Definition
| inverse of the absolute risk reduction |
|
|
Term
|
Definition
| used to find info on equivalent drugs from different countries |
|
|
Term
|
Definition
only predicts outcome or the effectiveness of therapy
(does not measure any actual clinical benefit to patient) |
|
|
Term
|
Definition
directly measures clinical benefit to patient
(ie cure rate, measureable reduction of symptoms) |
|
|
Term
|
Definition
people with disease / people who could get the disease
(8 people got sick from fish out of the 10 people who ate fish; 8/10 = 0.8) |
|
|
Term
|
Definition
risk ratio for study group / risk ratio for control group
(8 people got sick out of the 10 people in the control group 8/10 = 0.8)
(3 people got sick out of the 20 people who got treatment; 3/20 = 0.15)
(0.15 / 0.8 = 0.19)
(People who got treatment are 0.19 times AS LIKELY to get sick compared to those who didnt get treatment) |
|
|
Term
|
Definition
risk for exposed group - risk for control group
(8 people got sick out of the 10 people in the control group 8/10 = 0.8)
(3 people got sick out of the 20 people who got treatment; 3/20 = 0.15)
(ARR = 0.15 - 0.8 = -0.65)
(The risk of getting sick was 65% LESS in the treatment group compared to the control group) |
|
|
Term
|
Definition
(risk for control - risk for exposed) / (risk for control)
(8 people got sick out of the 10 people in the control group 8/10 = 0.8)
(3 people got sick out of the 20 people who got treatment; 3/20 = 0.15)
RRR = (0.8 - 0.15) / (0.8) = 0.81
(The treatment reduced the risk of sickness by 81% compared to placebo) |
|
|
Term
|
Definition
The number of people needed to be treated to cure/prevent 1 additional case/event
NNT = (1 / ARR)
(8 people got sick out of the 10 people in the control group 8/10 = 0.8)
(3 people got sick out of the 20 people who got treatment; 3/20 = 0.15)
(ARR = 0.15 - 0.8 = -0.65)
NNT = 1 / 0.65 = 1.6 (rounds to 2) |
|
|
Term
|
Definition
Ratio of (treatment event / treatment no event) and (control event / control no event)
(7 out of 17 people people got sick in control group)
(3 out of 12 people got sick in treatment group)
(7/10 = 0.7)
(3/9 = 0.33)
OR = 0.7 / 0.33 = 2.1
(People without treatment are 2.1 times AS LIKELY to get sick than those who do get treatment) |
|
|
