1 Facing Facts (and Lack of Facts) about Effectiveness of ECPs Elizabeth Raymond, MD, MPH James Trussell, PhD Chelsea Polis Obstet Gynecol 109(1):181-188,

1

Facing Facts(and Lack of Facts)

about Effectiveness of ECPs

Elizabeth Raymond, MD, MPH

James Trussell, PhD

Chelsea Polis

Obstet Gynecol 109(1):181-188, January 2007

2

The Hope

Widespread use of ECPscould prevent HALF of all

unintended pregnancies and abortionsin the US each year

(Trussell et al., 1992)

3

The Reality

Eleven studies have examined the impact of increased access to ECPs

on pregnancy and abortion rates

4

Fourteen Years Later

Eleven studies have examined the impact of increased access to ECPs

on pregnancy and abortion rates

None have shown any benefit

5

Agenda Today

• Review these studies

• Discuss possible explanations for disappointing findings

6

The Eleven Studies

• Conducted 1998-2006

• 4 countries

• Yuzpe, LNG, mifepristone

7

The Eleven Studies (continued)

• 9 randomized trials, 1 cohort study – Total of 11,830 women enrolled

• 1 demonstration project – >17,831 women given ECPs

• Followed women up to one year• Compared increased access to standard

access

8

Selected Studies

N Regimen1 Yr Preg. Rates

intervention control

Glasier 1083 Yuzpe 5% 6%

Lo 1030 LNG 1% 2%

Hu 2000 mife 4% 3%

Raine 2117 LNG 8% 9%

Raymond 1490 LNG 9% 10%

9

Explanations

1. Flaws in studies

10

Problems with the Studies

• Small size (160-2868 women)• Huge loss to follow-up (1-62%)• Weak intervention• Good access in comparison group• Low baseline risk of pregnancy – little

room for improvement with EC• Not randomized

11

However…

None of the 11 studies had all ofthese problems ―

Indeed, some were very good!

Consistency of findings hard to ignore

12

Explanations

1. Flaws in studies

2. Increased risk taking

13

ECPs and Risk Taking

• Lots of data• None show (almost) any effect of enhanced

ECP access on unprotected sex or use of regular contraception

• Some suggest “improved” behavior• Most data self-reported• 2 studies showed no effect on STIs

14

Explanations

1. Flaws in studies

2. Increased risk taking

3. Low ECP efficacy

15

Current Estimates

• ECP efficacy conveys the reduction in pregnancy risk after a single coital act

• Plan B package (LNG regimen): 89%

• ICEC Service Delivery Guidelines:– LNG regimen 60% - 94%– Yuzpe regimen 56% - 89% (74%)

16

Methodology

In a group of ECP users, compare:

• observed number of pregnancies

• expected number of pregnancies (number that would have occurred without ECPs)

Calculate the amount of reduction due tothe ECPs

17

Example

WHO 1998 trial of LNG vs. Yuzpe regimen

• 1001 women using LNG regimen• Pregnancies observed: 11• Pregnancies expected without EC: 75.3• Pregnancies prevented: 75.3 - 11 = 64.3 • Efficacy: 64.3 = 85%

75.3

18

Example

WHO 1998 trial of LNG vs. Yuzpe regimen

• 1001 women using LNG regimen• Pregnancies observed: 11• Pregnancies expected without EC: 75.3• Pregnancies prevented: 75.3 - 11 = 64.3 • Efficacy: 64.3 = 85%

75.3

19

Expected Pregnancies

• Collect data on:– date of sex act– date of last menstrual period– usual menstrual cycle length

• Determine the day of the menstrual cycle when the coital act occurred

• Use published data to estimate expected pregnancies

20

Pregnancy Risk by Cycle Day

0%

5%

10%

15%

20%

25%

30%

35%

40%

-8 -7 -6 -5 -4 -3 -2 -1 0 1 2

Menstrual cycle day

Pre

gn

an

cy

ris

k Wilcox 1998

21

Pregnancy Risk by Cycle Day

0%

5%

10%

15%

20%

25%

30%

35%

40%

-8 -7 -6 -5 -4 -3 -2 -1 0 1 2

Menstrual cycle day

Pre

gn

an

cy

ris

k

Schw artz 1979Schw artz 1980Bremme 1991

Weinberg 1998Wilcox 1998Colombo 2002

22

Do Any of the Charts Apply?

Women in the charts wanted to be pregnant

ECP users wanted NOT to be pregnant

• Used broken condom, withdrawal

• Different fertility?

• Different amount of unprotected sex?

• Different accuracy in reporting data?

23

Results

Numbers of expected pregnancies reportedby studies are probably too high

Most published efficacy figures are probably overestimates

24

Don’t Give Up…

ECPs do work!

• Physiology studies show effects incompatible with pregnancy

• LNG regimen proven to be more effective than Yuzpe it must be more effective than nothing

25

Efficacy of LNG Regimen

0%

25%

50%

75%

100%

0% 25% 50% 75%

Efficacy of Yuzpe

Eff

ica

cy

of

LN

G

26

Efficacy of LNG Regimen

0%

25%

50%

75%

100%

0% 25% 50% 75%

Efficacy of Yuzpe

Eff

ica

cy

of

LN

G

27

Explanations

1. Flaws in studies

2. Increased risk taking

3. Low ECP efficacy

4. Insufficient use

28

Effects of Increased Access

• In all studies so far (except one), increased access resulted in substantially increased use

• But…

29

But…

• Repeated use was uncommon

• Many unprotected acts remained uncovered by ECPs

• No ECPs were used in most pregnancy cycles

30

Reasons for Non-Use

Reasons given by women:

• Failure to perceive pregnancy risk

• Forgetting• Lack of motivation to use EC (“I’m lazy”)

• Inconvenience

31

Why Didn’t You Take Them?

“Now, my experience is that most of the time, people have no idea why they're doing what they're doing. They have no idea, so they're going to try to make up something that makes sense.”

- Clotaire Rapaille, Chairman, Archetype Discoveries Worldwide

32

Limits on ECP Use

• Expense

• Side effects

• High dose of hormones

• Psychologic stress

33

Ineffective Use

• Many uses are “wasted” – ECPs taken at times when risk is really low

• Does “wasted” use limit “truly needed” use?

34

Eleven Studies, No Benefit

1. Flaws in studies - but consistency compelling

2. Increased risk taking - no evidence

3. Low ECP efficacy - precise efficacy unknown

4. Insufficient use - definitely a problem

35

What To Do Now?

• Be honest: do not oversell by implying Plan B will reduce unintended pregnancy

• Everyone deserves a second chance to prevent an unintended pregnancy