Biostats Flashcards
what is nominal data?
categories are in an arbitrary order - no set order
ex: gender, ethnicity, marital status
what is ordinal data?
catergories are ranked in a logical order, but the difference between categories is not equal- ordered, ranked
ex: NYHA scale, 0-10 pain scale
Independent variable
-manipulated/changed by the investigator to determine effect on the dependent variable (outcome)
–> think study design (inclusion criteria, intervention/study drug)
Dependent variable
the outcomes being measured in a study
What is the null hypothesis?
Ho: states that there is no statistically significant difference between groups
–> this is what the researcher tried to disprove or reject
What is the alternative hypothesis?
Ha: states that there is a statically significant difference between the groups
–> this is what the researcher hopes to prove or accept
what is the alpha value and what does it means?
-researchers select a maximum possible error margin (5% , 0.05), comparing the p-value to alpha determined statistical significance
P < 0.05 = null hypothesis is REJECTED
P >/ 0.05 = null hypothesis ACCEPTED
what is a confidence interval?
provides the same information about significance as the P-value, plus the precision of the results
–> CI = 1 - alpha
Interpreting confidence intervals: comparing difference (means) data
-results statistically significant if the CI range does NOT include 0
ex: Difference (95% CI) : 38 (18-58) = sig
ex: Difference (95% CI) : 0.131 (-0.26- 0.89) = not sig
Interpreting confidence intervals: comparing ratio data (relative risk, odds ration, hazard ratio)
-result is statistically significant if the CI range does NOT include one
ex: Relative Risk : 0.92 (0.61-1.29) = not sig
ex: Relative Risk: 0.85 (0.72-0.99) = sig
Type 1 error
-null hypothesis is rejected in error
–> risk is determined by alpha and is related to the confidence interval
Cl = 1 - alpha (type 1 error)
Type 2 error
-null hypothesis is accepted in error
-probability denoted as beta (B) –> typically 0.1 or 0.2
-related to power (the probability of avoiding a type 2 error)
Power = 1- B
Relative Risk and Risk Ratio
Risk = (# of subjects with the unfavorable event) / (3 of subjects in that arm of the study)
RR = (risk in the treatment group) / (risk in the control group)
RR = 1 –> no difference in risk
RR > 1 –> higher risk in the treatment group
RR < 1 –> lower risk in the treatment group “as likely”
Relative Risk Reduction (RRR)
-how much the risk is reduced in the treatment group vs the control group –> “less likely)
RRR = (% risk control gr - % risk in tx gr) / (% risk in control group)
or
RRR= 1- RR
Absolute risk reduction (ARR)
- the absolute difference in outcome rates between 2 groups - more important to know
ARR = (% risk in the control group) - (% risk in the treatment group)
Interpretation: for every 100 pts treated with __, x fewer pts will have HF progression
Number needed to treat (NNT)
-number of ppl who need to be treated for a certain period of time in order for 1 patient to benefit
NNT= 1 / (risk in control group - risk in treatment group). OR
NNT = 1/ ARR
Rounding rules for NNT and NNH
-NNT: round up anything that is greater than a whole number (NNT = 52.1 –> 53)
-NNH: round down any number greater than a whole number ( NNH = 41.9 –> 41)
Number needed to harm (NNH)
-number of people who need to be treated for a certain period of time in order for 1 patient to experience harm ( an adverse event)
NNH = 1/ARR
Odds Ratio (OR)
-probability of an outcome occurring with an exposure versus without the exposure
OR = AD/BC
A: have the outcome, with exposure
B : dont have the outcome, without exposure
C: have the outcome, without exposure
D: dont have the outcome, without exposure
Hazard Ratio (HR)
-negative outcome measured in clinical trials
-rate at which an unfavorable event occurs within a short period of time
-assumes ratio is constant over time
-used in survival analysis
HR: (hazard rate in the tx group/ hazard rate in the control group)
Odds ratio and Hazard ratio interpretation
OR/HR =1, event rate in similar in both groups
OR/HR > 1: event rate is higher in the treatment group
OR/HR <1: event rate is lower in the treatment group
T-test
used when endpoints have continuous data and the data is normally distributed
-student t-test: used when the study has 2 independent samples: the tx group and the control group
Chi-squared/Fisher’s exact test
-used to determine statistical significance between treatment groups
ANOVA test
used to test for statistical significance when using continuous data with 3 or more samples or groups
Correlation
-used to determine if one variable is related to another variable
r: determines relationship and direction (-1 to 1)
–> Spearman’s rank-order correlation: used for ordinal, ranked data
–>Pearson’s correlation coefficient: used for continuous data
Regression
used to describe the relationship between a dependent variable and one or more independent variables
–> types: linear (continuous data), logistic (categorical data), Cox (categorical in a survival analysis)
Sensitivity
-described how effectively a test identifies pts with the condition
–> higher the sensitivity the better
Specificity
-described how effectively a test identifies patients without the condition
–> higher the specificity the better
–> a “true - negative”
Intention to treat
includes data for all pts orignally allocated to each treatment group (active and control) even if the pt did not complete the trial according to the study protocol (eg. due to non-compliance, protocol or study withdrawal)
Types of medical studies (ranked from most to least reliable)
-systematic reviews and meta-analyses
-randomized controlled trials
-cohort studies
-case-controlled studies
-case series and case reports
-expert opinion
Cost-minimization analysis (CMA)
compares costs of interventions with demonstrated equivalence
Cost benefit analysis (CBA)
compared benefits and costs in monetary units
Cost-effectiveness Analysis (CEA)
most common, costs are monetary but outcomes are in clinical units (easy to quantify)
Cost-utility analysis (CUA)
outcomes based on quality of life assessments
