OBGYN JOURNAL CLUB NOTES Martha A Wojtowycz Ph

OB/GYN JOURNAL CLUB NOTES Martha A. Wojtowycz, Ph. D January 11, 2019

LEARNING OBJECTIVES • Distinguish between incidence and prevalence • Calculate Absolute Risk, Relative Risk, and Odds Ratios • Distinguish between Relative Risk and Absolute Risk • Explain relative risk and absolute risk of an outcome to your patients

HOW DO WE MEASURE DISEASE? INCIDENCE • Incidence: # of new cases of disease # of individuals at risk for developing disease Ø Over a given time frame (usually 1 year) Ø Usually expressed per 1, 000 but can express as appropriate (eg per 100, 000 for rare diseases) Ø Incidence of disease in a population is the absolute risk

HOW DO WE MEASURE DISEASE? PREVALENCE • Prevalence: # of cases of disease (old and new) # of individuals at risk for developing disease Ø At a given time (point prevalence v. period prevalence) Ø Usually expressed per 1, 000 but can express as appropriate (e. g. , per 100, 000 for rare diseases)

INCIDENCE AND PREVALENCE: ACUTE, SELF-LIMITED DISEASE Incidence (new cases going in) XXXXXXXXXXXXX Prevalence XXXXXXXXXXXXX Examples: GI illness, Influenza, pancreatic cancer Cases leaving prevalence pool - cure or death

INCIDENCE AND PREVALENCE: CHRONIC DISEASE Incidence (new cases going in) Examples: HIV, DM XXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXX Prevalence XXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXX Cases leaving prevalence pool - cure or death

RELATIVE RISK (RR) • Ratio of the risk of disease (or incidence rate) in the exposed group to the risk of disease (incidence rate) in the nonexposed group. • RR= (Risk in exposed)/(Risk in nonexposed) • RR= (Incidence in exposed)/(Incidence in nonexposed)

RELATIVE RISK INTERPRETATION RR > 1. 0 RR = 1. 0 RR < 1. 0 Risk in exposed group > risk in nonexposed group; positive association; may be causal Equal risks; no association Risk in exposed group < risk in nonexposed group; negative association; may be protective

RELATIVE RISK Disease Does Not Develops Develop Total Incidence Rates Exposed a b a+b a/(a+b) Not exposed c d c+d c/(c+d)

RELATIVE RISK CALCULATION • Incidence in exposed = a/(a+b) • Incidence in nonexposed = c/(c+d) • RR = (a/(a+b)) / (c/(c+d))

ODDS RATIO (OR) • Another measure of strength of association • Used when we do not have information on the incidence in the exposed or nonexposed • Case control study • Approximation of Relative Risk • OR= (Odds that person with disease was exposed)/ (Odds that person without disease was exposed) • OR can be <1. 0, = 1. 0, >1. 0

ODDS RATIO Disease (+) Disease (-) History of Exposure a b No history of exposure c d

ODDS RATIO • Odds that person with disease was exposed = a / c • Odds that person without the disease was exposed = b / d • OR = (a/c) / (b/d) = ad / bc

JORGENSEN ET AL. , “INCIDENCE OF VENOUS THROMBOEMBOLISM AFTER DIFFERENT MODES OF GYNECOLOGIC SURGERY”, OBSTETRICS AND GYNECOLOGY, VOL. 132, NO. 5, NOVEMBER 2018.

ABSOLUTE RISK OF VENOUS THROMBOEMBOLISM • Absolute risk (or incidence) of venous thromboembolism is 0. 2% or 2 per 1, 000 gynecologic surgeries Ø 0. 7% or 7 per 1, 000 hysterectomies Ø 0. 2% or 2 per 1, 000 myomectomies • Venous thromboembolism is a rare event • Patients should be provided with the absolute risk information so that they are aware that this is rare

RELATIVE RISK OF VENOUS THROMBOEMBOLISM • Exposure is minimally invasive gynecologic surgery (laparoscopy or vaginal) • Unexposed group underwent laparotomy • RR for laparoscopy vs laparotomy Ø Incidence for laparoscopy = 0. 19% or 1. 9 per 1, 000 Ø Incidence for laparotomy = 0. 85% or 8. 5 per 1, 000 Ø RR =. 0019/. 0085 = 0. 22 Ø Shows that laparoscopy has a protective effect • Compared with laparotomy, laparoscopic surgery is associated a 78% lower risk of venous thromboembolism

RELATIVE RISK OF VENOUS THROMBOEMBOLISM • RR for vaginal vs laparotomy Ø Incidence for vaginal = 0. 06% or 0. 6 per 1, 000 Ø Incidence for laparotomy = 0. 85% or 8. 5 per 1, 000 Ø RR =. 0006/. 0085 = 0. 07 Ø Shows that vaginal surgery has a protective effect • Compared with laparotomy, vaginal surgery is associated a 93% lower risk of venous thromboembolism

WHAT TO TELL YOUR PATIENT • Relative risk allows for a comparison of the absolute risks but caution patients not to overestimate the absolute risk Ø For example, suppose we have a rare event where: Ø Absolute risk for all surgeries is 1 per 1, 000 surgeries Ø Absolute risk for surgery type A is 3 per 1, 000 surgeries Ø Absolute risk for surgery type B is 0. 3 per 1, 000 surgeries Ø Relative risk =. 003/. 0003 = 10. 0 • Compared with surgery type B, patients who undergo surgery type A are 10 times more likely to experience this rare event

SAMPLE SIZE/STUDY DESIGN ISSUES • Despite the fact that this study included over 40, 000 surgical encounters, the authors were unable to: • Run all of the analyses for myomectomies because of the small numbers of myomectomies • Simultaneously adjust for potential confounders because there were very few cases with a venous thromboembolism • Case control study design often used to study rare events but is more susceptible to bias