An Assessment of the Accutane (Isotretinoin) Pregnancy Prevention Program Allen A. Mitchell, M.D. Carla Van Bennekom, R.N., M.P.H. Carol Louik, Sc.D. Slone Epidemiology Unit Boston University School of Public Health
Pregnancy Prevention Program (PPP): Introduced in fall, 1988 Multi-component program Aimed at female patients and their physicians Unprecedented and novel approach
Does the PPP work?
Survey of Accutane Use in Women Sponsored by: Hoffmann-La Roche, Inc. SEU Responsible for: - Design/Protocol - Data Collection - Data Processing - Data Analysis Guidance provided by: Independent Advisory Committee
Slone Epidemiology Unit Accutane Survey Advisory Committee 1988 - 2000 Paul D. Stolley, M.D., Chair* University of Maryland School of Medicine Baltimore, MD Edward L. Decker, Pharm.D. New England Medical Center Boston, MA Karen McKoy, M.D. Lahey Clinic Burlington, MA Observers: Charlotte Catz, M.D., NICHD John Melski, M.D. Marshfield Clinic Marshfield, WI Peter Pochi, M.D. Boston University School of Medicine Boston, MA Robert S. Stern, M.D. Harvard Medical School José F. Corderó, M.D., CDC *Dr. Stolley resigned as Committee Chair on July 19, 2000 because he has joined the FDA as a Senior Consultant.
Objectives To assess compliance with the Pregnancy Prevention Program Awareness of the teratogenic risk Patient and physician behaviors Pregnancy rate Pregnancy outcome Risk factors for pregnancy
Limitations No pilot testing No pre/post comparison What is “success”? Survey itself is an intervention Voluntary enrollment/? representativeness
Accutane Survey Design
Accutane Survey TIME PERIOD OF INTEREST (Duration of Follow-up)
To Encourage Enrollment Multiple/different opportunities Keep it simple Provide payment
Enrollment Opportunities 1) Physician-generated All prescriptions originate with physician Logistically feasible Strong incentive
Enrollment Opportunities 2) Toll-free telephone call Simple Quick Familiar
Enrollment Opportunities 3) Package-generated Unique Direct to patient Bypasses physician May preferentially target women who were not encouraged by their physician may be relatively noncompliant are attracted by payment
Follow-Up Approaches Frequent: During and After Treatment (DAT) Advantage - Avoid recall bias Disadvantage - ? Affect outcome beyond PPP
Follow-Up Approaches Infrequent: After Treatment only (AT) Advantage - Avoids intervention beyond PPP Disadvantage - Potential recall bias
Information Obtained at Follow-Up Patient characteristics Education + - Doctor specialty Pregnancy risk Patient knowledge Risk of birth defects Avoid pregnancy Compliance Pregnancy testing Contraception Wait for next menstrual period Pregnancy occurrence
Accutane Survey Results
Accutane Survey Enrollments: 1989-1999 Cumulative enrollments through 6/30/00: >494,000
Method of Enrollment
Follow-Up Rates During and After Treatment (“DAT”) Number to date (6/00): 61,659 1/89 - 12/94 (telephone) 32,301 Completed Follow-Up: Tx start 98% Tx middle 98% 6 mos.`p Tx 93%
Follow-Up Rates During and After Treatment (“DAT”) 1/95 - 6/00 (mail) 29,358 Completed Follow-Up: Tx start 98% Tx middle 98% 6 mos.`p Tx 97%
Follow-Up Rates After Treatment (“AT”) Number to date (6/00): 433,256 Completed Follow-Up: 6 mos.`p Tx 80-86%
Results: Demographics Accutane Survey Results: Demographics
Age 1/89 – 6/00 (n=494,915) Mean (years) 25.6 Median (years) 24
Age of Enrollees by Enrollment Year
Highest Level of Education
Type of Prescriber
Past Treatment for Acne
Pregnancy Risk Categories (at DAT 1) 1/95 – 6/00 (n=28,016) 3% Hysterectomy/postmenopausal 57% Not sexually active 40% Sexually active
Pregnancy Risk Categories (at DAT 1) 1/95 – 6/00 (n=28,016) 57% Not sexually active: Using birth control 44% Not using birth control 56% 100% 40% Sexually active: Using birth control 98% Not using birth control 2%
Results: Women’s Knowledge Accutane Survey Results: Women’s Knowledge
Receipt of PPP Information 1/95 – 6/00 (n=28,386) Patient brochure 92% True/false test 42% Birth Control brochure 53%
Knowledge of Accutane Risk 1/95 – 6/00 (n=28,386) May cause miscarriage 43% May cause birth defects 99%
Accutane Survey Results: Compliance
Compliance with PPP – I 1/95 – 6/00 (n=28,386) Told to avoid pregnancy 99% Signed consent form 77% Postpone Accutane until pregnancy test 67% Postpone Accutane until next menstrual period 57%
Compliance with Selected Measures of the PPP
Compliance with PPP – II Pregnancy Testing 1/95 – 6/00 (n=27,073) Serum 56% Urine 9% Both serum and urine 8% Any 76% None 24%
Pregnancy Testing Before Starting Accutane by Enrollment Year
Knowledge and Compliance at Onset of Therapy 1989 -1990 (n=9,365)
Original Medication Package
Revised Medication Package
Knowledge and Compliance at Onset of Therapy
Compliance with PPP - III Contraceptive Status, By Age
Contraceptive Status Noncontraceptors by Age Group
Contraceptive Status Surgical Sterility by Age Group
Contraceptive Status Nonsurgical Contraception by Age Group
Contraceptive Status Nonsurgical Contraception by Enrollment Year
Among Nonsurgical Contraceptors, % Reporting Oral Contraceptive Use by Enrollment Year
Among Contraceptors % Reporting Use of 2+ Methods by Age Group
Among Contraceptors % Reporting Use of 2+ Methods by Enrollment Year
Proportion of the Most “At Risk” Women Who Are Noncompliant with Pregnancy Testing and Waiting to Start Accutane Among the 27,439 women in DAT1, 6,549 (24%) reported no pregnancy test before starting Accutane Of those, 3,958 (60%) did not wait until their next menstrual period Of those, 1,554 (39%) were sexually active and 39 (1%) were sexually active, not using birth control Thus, among 27,439 women, 39 (0.1%) were sexually active, not using birth control, had no pregnancy test, and did not wait until their menstrual period before starting Accutane.
Accutane Survey Results: Pregnancies
Pregnancies During Accutane Treatment 1989 – 1999 Cohorts Number of women completing follow-up 347,857 Number of pregnancies reported 1,019
Number of Pregnancies Reported by Enrollment Year
Accutane Survey Characteristics of Pregnancies (n=1,019) Pregnant at start of treatment 10% No contraceptive use 23% Contraceptive failure 66%
Outcomes of Accutane Exposed Pregnancies 1989 – 1999 cohorts (n=1,019) No. % Therapeutic abortion 681 67 Spontaneous abortion 177 17 Liveborn 117 11 Ectopic 29 3 Unknown 15 1
Pregnancy Rate During Accutane Treatment 1989 – 1999 Cohorts (Duration of Use <1 Year) Number of women completing follow-up 339,994 Number of pregnancies reported 992 Rate per 1000 Accutane courses 2.8 Rate per 1000 person years 7.4
Pregnancy Rates During Accutane Treatment by Age 1989 – 1999 Cohorts Age (years) No. of women Pregnancy rate/1000 courses <15 12,290 1.0 15-24 138,497 2.8 25-34 89,040 3.7 35-44 44,574 1.7 45+ 7,970 0.5
Pregnancy Rate/1000 Person Courses by Enrollment Year
Pregnancy rate/1000 courses Pregnancy Rate During Accutane Treatment by Contraceptive Method 1989 – 1999 Cohorts Method No. of Women Pregnancy rate/1000 courses None 113,247 1.7 Oral contraceptive 97,895 2.5 Tubal ligation 23,194 0.5 Condom 17,475 10.6 Vasectomy 15,965 0.3 Other 12,318 3.1 Unknown 4,622 Diaphragm 4,519 5.9 IUD 2,212 3.6 Rhythm 855 15.2
Accutane Survey Validity
Accutane Survey Validity Follow-up rates are high
Accutane Survey Validity Follow-up rates are high Responses consistent (AT vs. DAT) Knowledge Behaviors Pregnancy rates
Accutane Survey vs. U.S. Female Population Pregnancy Rates (per 1,000 women/years) U.S. Population* 7.4 105 *NCHS, 2000; women ages 15-44
Accutane Survey vs. U.S. Female Population Proportion Sexually Active and Not Using Contraception (n=26,074)
Accutane Survey vs. U.S. Female Population Among Contraceptors: Proportions using birth control pill
Major Factors Associated with Successful Contraception* Age (coital frequency, fecundity) Education Duration of method use Motivation *Trussell J and Kost K: Studies in Family Planning 1987;18:237-183
% of Pregnancies by Contraceptive Method
Accutane Survey Validity Follow-up rates are high Responses consistent Knowledge Behaviors Pregnancy rates Pregnancy rates “rebound” after stopping Accutane
Pregnancy Rates and Outcomes During and After Therapy with Isotretinoin in 122,582 Women, 1989-1993
Accutane Survey Representativeness
Accutane Survey Representativeness 1) What proportion of eligible women enroll in the Survey? 2) Do women in the Survey represent the larger population of Accutane users?
Roche Consumer Survey Female Accutane Users 1990 – 1991 (n=400) Proportion reporting enrollment in survey: 239/400; 60%
Consumer Survey Accutane Survey Enrolled Not enrolled Age (mean) 29 31 Education (median) 14 Dermatologist prescribed 91% 88% Sexually active 64% 54% - Using contraception (96%) (92%)
Consumer Survey Contraceptive Methods (among 205 contraceptors; in %) Accutane Survey Enrolled Not enrolled Birth control pill 40 16 Condom 10 12 Other barrier methods 9 8 Surgical sterilization 44 65
United Health Care Population: ³1.2 million women ages 12-59 1/90 – 6/96 5,095 women ages 12-59, who filled Accutane prescription between 1/90 and 6/96
United Health Care Population: ³1.2 million women ages 12-59 1/90 – 6/96 5,095 women ages 12-59, who filled Accutane prescription between 1/90 and 6/96 Proportion enrolled in Accutane Survey Definite: 38.4% Possible: 7.5% Total: 45.9%
United Health Care Accutane Survey: Enrolled vs. Not Enrolled Age (mean) 26.4 28.2 *“Definite” matches only.
United Health Care Accutane Survey Enrollment Rates by Age Enrolled <20 38% 20-29 44% 30-39 41% 40-49 26% 50-59 14%
SEU Accutane Survey Is there selection bias favoring women at low risk?
Characteristics of Survey Participants According to Enrollment Method Doctor-generated Package-generated Age (mean) 24.3 25.9 Prescriber –dermatologist 98% 88% Signed consent 71% Pregnancy test before starting 84% 69% Pregnancy rate (per 1000 courses) 2.5 2.8
Among women who enrolled via the Medication Package, only 13% reported that they did so because of their doctor’s encouragement.
Doctor Encouraged to Enroll in Survey by Age Group
Summary Since 1989, 494,915 women have enrolled in the Accutane survey. Follow-up rates are high. Except for a slight decline in age, demographic characteristics have remained stable. Awareness of teratogenic risk is high. Compliance with pregnancy testing and contraceptive guidelines is variable but not decreasing.
Summary (cont’d) Pregnancy rates are appreciably lower than U.S. population and have declined over time. Data appear to be valid. No evidence that Survey results are not representative.
Future Directions Increase enrollments Increase payment at enrollment Add payment upon completion Modify doctor-patient enrollment procedures Enroll women on multiple courses Increase power and value of DAT follow-up 5000/yr 25,000/yr
Future Directions (cont’d) Modify Survey questions Oral contraceptive type Refine pregnancy risk questions Initiate regular feedback to prescribers to encourage them to value the Survey and enroll patients.
Method of Enrollment by Enrollment Year
Age of Enrollees by Enrollment Year
Region of Residence by Enrollment Year
Highest Level of Education by Enrollment Year
Past Treatment for Acne by Enrollment Year
Pregnancy Risk Category by Enrollment Year
Oral Contraceptive Users by Sexual Activity Status Age 15-24 years
Doctor Encouraged to Enroll in Survey by Enrollment Method
Number of Liveborns by Enrollment Year
% of Pregnancies by Contraceptive Method
Pregnancy Rate During Accutane Treatment by Contraceptive Method by Enrollment Year
Pregnancy Rates According to Follow-Up by Enrollment Year
Contraceptive Failure Rates Typical Use* Perfect Use* Accutane Survey† Method (%) Diaphragm 18 6 1.5 Condom (male) 12 3 2.8 IUD 0.1-2.0 0.1-1.5 0.9 Oral contraceptive pill 0.1-0.5 0.6 Female sterilization 0.4 0.1 Male sterilization 0.15 0.10 0.08 *Percent pregnant in first year of use; Hatcher RA et al.: Contraceptive Technology, 17th Edition, 1994, p113. †Rate per 100 woman-years.
Compliance According to Pregnancy Risk Category (DAT1, 1995 – 2000) Hyst. Not sexually active, using birth control Not sexually active, not using birth control Sexually active, using birth control Sexually active, not using birth control Unk. N 943 6,970 8,821 10,811 195 276 % Told to avoid pregnancy 95 99 Signed consent 51 81 78 76 69 79 Had pregnancy test N/A 82 70 64 Waited for test after starting 74 58 63 53 Waited for next month after starting 48 60 47
Does the more intensive DAT follow-up itself result in enhanced compliance?
Accutane Survey Pregnancy Rates According to Follow-Up Pregnancies Rate/1000 Courses AT 292,955 848 2.8 DAT 46,989 144 3.0
Pregnancy Rate in 30 Days Following Accutane by Enrollment Year (n=338/338,959; Rate=1.0/1000 enrollees)
Pregnancy Rate During Treatment Among Women on Accutane for 1-2 Years N=27/7,862, or 27/9,641 person-years of treatment Rate=2.8/1000 person-years of treatment
United Health Care Accutane Survey Enrollment Rates by Age