Raymond, Elizabeth G. MD, MPH; Dalebout, Sandra M. MPH; Camp, Sharon I. PhD
Emergency contraceptive pills are indicated for use after unprotected sexual intercourse to prevent pregnancy. The levonorgestrel‐only regimen consists of one dose of levonorgestrel 0.75 mg taken as soon as possible after unprotected sex, followed by a second identical dose 12 hours later. A prescription version of this regimen (Plan B, distributed by Women's Capital Corporation, Washington, DC) was approved by the United States Food and Drug Administration in 1999. The distributor of that product now plans to apply to market the regimen over the counter. In considering whether to allow a drug to be marketed over the counter, the Food and Drug Administration often requires data from a label comprehension study to establish that the product label conveys to users the information that is needed for the safe and effective use of the product without the supervision of a health care provider.1 This study was designed to test comprehension of a prototype label for the over‐the‐counter product. The prototype was similar to the approved prescription label but was slightly modified to be consistent with Food and Drug Administration content and formatting requirements for over‐the‐counter drugs (see box on page 347).
MATERIALS AND METHODS
The study design and questionnaire were heavily influenced by the Food and Drug Administration's comments on early drafts of the protocol. The stated goal of the study was to estimate whether subjects understood 11 communication objectives relevant to safe and effective use of the product. These objectives address indications for use of the product, contraindications, instructions, possible side effects, and management of serious complications, as described on the label approved by the Food and Drug Administration for the prescription version of the product.
A convenience sample of study subjects was recruited between June 18, 2001 and July 12, 2001 at malls and family planning clinics in or near eight large United States cities (Denver, CO; Los Angeles, CA; Chicago, IL; San Antonio, TX; Philadelphia, PA; Miami, FL; Phoenix, AZ; and Washington, DC). To be eligible to participate, women had to be 12 to 50 years old; able to read English well enough to read an over‐the‐counter product label; without a health care or marketing background; and without a history of participation in this study. Within these criteria, recruiters at each site were asked to meet or exceed quotas for subgroups of minority women and women felt to be at high risk for poor comprehension. These quotas were based on the proportion of each subgroup in the United States population according to the year 2000 United States census and on advice from expert consultants. The quotas were calculated as percentages of the target sample size of 575 women and were as follows: 20% aged 17 or younger; 11% black; 11% Hispanic; 4% Asian or Pacific Islander; and 20% eighth‐grade reading level or lower.
Interviews were conducted by trained interviewers in a private room within the mall or clinic. Before the survey, a validated test of reading comprehension in adults (the Rapid Estimate of Adult Literacy in Medicine2) was administered to all subjects aged 18 years or older who had not completed college. For this test, the subject was asked to read 66 words aloud, and the number she pronounced correctly was recorded. The developers of the test have published approximate grade levels corresponding to various test scores.
For the survey, each subject was given a prototype over‐the‐counter package. This package, like the prescription version of this product, had text on both the outside and the inside of the sealed four‐panel folder (see box on page 347). The pills were enclosed in blister bubbles embedded in the inside of the folder. The subject was instructed first to look at the outside of the package as if she were thinking about whether to purchase the product. The package was retrieved, and the interviewer asked one question about the product. The package was then returned to the subject, and she was permitted to look at the outside while the interviewer asked five additional questions. The subject was then told to open the package and review it as if she were about to use the product. She was permitted to refer to the entire label as the interviewer asked 24 additional questions. Questions included multiple‐choice and open‐ended formats. Most presented a scenario, and the subject was asked whether or not use of the product would be correct in the situation described. Several questions were included that were not directly addressed by the label; for example, two questions asked whether fever and breathing difficulty (not mentioned on the label) could be adverse effects of the product, and one question required subjects to infer that asthma (also not mentioned) is not a contraindication to use. After completing the product questionnaire, the interviewer recorded demographic information about the subject. Finally, the subject was asked to complete a self‐administered questionnaire about her sexual and contraceptive history. Before initiation of the study, all questions had been pretested on 10 women at a mall not subsequently involved in the study. Other than this pretest, no prior studies of the label or of similar labels had been conducted.
The sample size was calculated to allow estimation of the proportion of women understanding each communication objective with a 95% confidence interval of ten percentage points, conservatively assuming that the proportion would be 50%. No adjustments were made for multiple confidence interval estimations. Calculations indicated that the minimum required sample size was 385 women. Based on advice from the Food and Drug Administration and consultants, the minimum desired sample size was increased to 575.
The focus of the analysis was to determine the proportion of subjects who demonstrated “understanding” of each of the communication objectives. Understanding for each objective was operationally defined by a previously specified combination of acceptable responses to designated questions about the product. For example, a woman was considered to have understood objective 1 (the product is indicated for prevention of pregnancy after unprotected sex) if she gave acceptable answers to two of the following four questions:
* Without looking at the label, tell me what Plan B is used for.
* A woman's partner used a condom during sex with her, but the condom broke. The next morning, she used Plan B to prevent pregnancy. Was this a correct use of Plan B?
* A woman with asthma had unprotected sex. The next day, she took Plan B to prevent pregnancy. Was this a correct use of Plan B?
* A woman had unprotected sex 2 days ago and then used Plan B to prevent pregnancy. Was this a correct use of Plan B?
Percentages were truncated rather than rounded. Confidence limits were calculated using binomial standard errors and P values were calculated using χ2 or Fishers exact tests.
The study protocol was approved by the Protection of Human Subjects Committee of Family Health International. Signed parental consent was required for subjects aged 17 or younger who were recruited in mall settings. Parents and subjects were separated before the consent was obtained to ensure that the subjects did not have any advance information about the product when they completed the interview.
A total of 663 interviews were conducted (72–111 in each city). Seven women interviewed did not meet the inclusion criteria. The other 656 women were included in the analysis. They had a median age of 21 years (range 12–50 years) and represented a broad spectrum of demographic characteristics (Table 1). Of 393 women aged 18 years or older who had not completed college, 35.4% scored at the eighth grade level or below on the literacy test. At least 474 of the 599 women who responded to the sexual history questionnaire had previously had sex (79%), and of those, at least 314 (66%) had had intercourse without contraception despite a desire not to become pregnant. At least 390 (82%) of the sexually experienced subjects had had some sort of pregnancy scare: missed oral contraceptive pills, broken condom, unprotected intercourse despite a desire not to be pregnant, or worry that they might be pregnant. Forty‐one (6%) of the eligible women reported ever having used emergency contraceptive pills.
Nineteen of the 30 questions about the package were answered acceptably by at least 85% of subjects, and only four were answered acceptably by fewer than 75% of subjects (Table 2). Incorrect answers were not concentrated among a few women; the median number of questions answered acceptably was 27, and only 42 women (6.4%) gave acceptable answers to all questions. More than 85% of subjects demonstrated understanding of seven of the 11 communication objectives, according to the previously specified criteria (Table 3). The proportion who understood each of the 11 communication objectives was compared in subgroups defined by the following 11 characteristics: age, race, ethnicity, city of interview, interview site (mall or clinic), income category, highest level of schooling completed, literacy level (among subjects aged 18 years or older who had not completed college), previous sexual experience, experience with pregnancy scare (condom break, missed oral contraceptive pills, unprotected intercourse, or worry about unwanted pregnancy), and experience using emergency contraceptive pills. Among these 121 comparisons, 39 statistically significant differences were found (P < .05; data not shown). However, the only apparent pattern was that women of lower literacy were significantly less likely to understand almost all objectives than more literate women. However, eight of the 11 objectives were each understood by more than 80% of women with low literacy.
After reading the prototype over‐the‐counter label, a high proportion of women were able to understand 11 key messages necessary for safe and effective over‐the‐counter use of this product. These included messages about indications for use, contraindications, instructions, possible adverse effects, and management of serious complications of use. Three particularly important objectives were understood by more than 93% of subjects: that the product is indicated for prevention of pregnancy after unprotected sex, that it does not prevent human immunodeficiency virus or acquired immune deficiency syndrome, and that it should not be used by women who are already pregnant. In addition, 97% of subjects understood the most important instruction for use: that the first pill should be taken within 72 hours or as soon as possible after intercourse.
Two communication objectives were understood by fewer than 80% of the study population. Although more than 85% of subjects indicated correctly that the product should not be used as regular birth control when they were asked directly, only 67% met the definition of understanding the corresponding objective. One possible explanation for this relatively poor result is that several of the four questions used in the definition may in retrospect have required an excessively strict or unrealistic interpretation. For example, one question stated: “A woman is planning to have sex tonight. She usually uses condoms to prevent pregnancy. This time she plans to use Plan B instead because her husband complains about using condoms. Is this a correct use of Plan B?” Only 46% of subjects gave the acceptable answer to this question, which was “no.” The intent of the question was that if a woman knows in advance that her husband will refuse to use condoms, she should either find another contraceptive method or refuse to have sex with him, rather than plan to use the product. However, this choice may be improbable in the minds of many women. Subjects may have assumed that unprotected sex was inevitable in the specific situation described, after which use of the product would in fact be appropriate. Alternatively, women may not consider that a decision to use the product instead of another method on a single, isolated occasion actually violates the intent of the label warnings; instead, they may interpret the label as advising only against consistent or repeated use of this product as a primary contraceptive method. Such an interpretation is unlikely to be highly detrimental to a woman's health; because pregnancy after a single coital act is statistically unlikely during most of the menstrual cycle,3 occasional use of this product as a primary method will not substantially increase the risk of unwanted pregnancy.
Only three‐quarters of subjects understood that the product should not be used by women with unexplained vaginal bleeding. The questionnaire included just one question to test this objective, which in retrospect would have been too few to test an important concept. However, the rationale for the Food and Drug Administration's requirement that this contraindication be included on the prescription label is unknown. To our knowledge no published medical guidelines for use of emergency contraceptive pills include unexplained bleeding as a contraindication.4,5
Two objectives were understood by between 80% and 85% of subjects: that the first pill should be taken as soon as possible after intercourse, and that the user should seek medical care immediately if severe abdominal pain develops. However, the first of these objectives should not be considered in isolation; the more important message is that almost all women (97%) understood the overall time frame for product use (either within 72 hours or as soon as possible after sex). Furthermore, research has shown that this regimen is highly effective even if delayed up to 72 hours after intercourse, and a similar regimen is effective even if use is delayed longer than that.6 Failure to appreciate the written warning about abdominal pain in the label itself is not of great clinical concern, because women do not need written instructions to know that they should see a doctor if severe pain develops. Moreover, no evidence exists to indicate that emergency contraceptive pills increase the risk of ectopic pregnancy, which was implied in the warning.7
The study population was diverse and included large numbers of minority women, young women, and less‐educated or less‐literate women. Although less‐educated and less‐literate women were less likely to understand the objectives than more‐educated and more‐literate women, understanding was generally high in all subgroups examined. Most of the study subjects had experienced a contraceptive accident or pregnancy scare and thus might have been candidates for using emergency contraception. However, only a small number had ever used emergency contraceptive pills, and therefore it is unlikely that our results were substantially influenced by previous experience with this or a similar product.
At the request of the Food and Drug Administration, the study included questions that were asked when the subject had differing degrees of access to the label. The first question was asked after the package had been taken away from the subject, and five additional questions were asked while the subject was looking only at the outside of the package. These restrictions probably resulted in a lower likelihood of answering correctly than would occur in the real over‐the‐counter setting, in which women would be able to look at the entire package before using the product. Nevertheless, the proportion of subjects who gave acceptable answers to each of these six questions was at least 80%. Because these questions covered the key information that a woman would need to decide whether to purchase the product (indication for treatment, the appropriate timing of use, and the main contraindication, pregnancy), the finding that a high proportion of women answered these questions correctly after reading only the outside of the package is noteworthy.
This study had several limitations. The study population was a convenience sample of women recruited in shopping malls and family planning clinics and thus may not be representative of all women who might ultimately use the product. Whether this selection process may have led to an overestimate or an underestimate of the degree of understanding of the label is unknown, however. A more serious concern is that requirement for parental consent for minor subjects recruited in malls may have resulted in an atypical sample of young women who were comfortable talking about sex and contraception and who therefore may have been more likely than most teens to answer questions correctly. Clearly, our results cannot be extended to the many women in the United States who do not read English.
Surprisingly little published data are available on how well the public understands other medication package labels. One study of three prototype labels for a proposed over‐the‐counter cholestyramine product found results similar to those presented here: the proportion of subjects who correctly answered questions about key messages on the label ranged from 78% to 99% for various questions, and more‐educated subjects were significantly more likely than less‐educated subjects to answer correctly.8 Another study of two labels for an over‐the‐counter ranitidine product found that 86–90% of people surveyed understood directions for use. Unlike our study, neither of these prior studies included scenario questions that required subjects to apply the information on the label to specific situations.9 Our finding that most subjects could answer such questions correctly suggests that the understanding of the emergency contraception product label may be translated into appropriate action when women use the product.
The characteristics of this product clearly justify consideration for over‐the‐counter distribution. Levonorgestrel is safe—even if taken in pregnancy, evidence suggests that it will not harm the pregnancy or the fetus—and it has no potential for addiction. The regimen has few contraindications, all of which should be easy for the consumer to identify for herself without needing an examination. The instructions for use are simple, and the same dose is used for all patients.10,11 Furthermore, recent data indicate that the efficacy of treatment is substantially greater the sooner the drug is taken; in the trial cited above, the chance of pregnancy was almost seven times lower if treatment was initiated within the first 24 hours after intercourse than if treatment was delayed more than 48 hours.12 Therefore, women should be able to obtain the product promptly whenever they need it. Distribution over the counter may be the most effective means of accomplishing this goal and has been endorsed by many expert groups, including the American College of Obstetricians and Gynecologists (American College of Obstetricians and Gynecologists News Release, February 14, 2001), the American Academy of Pediatrics (Citizen's Petition submitted by Center for Reproductive Law and Policy to USFDA, February 14, 2001), and the American Medical Association (AMA Eyes Over‐the‐Counter Access of Emergency Contraception. American Medical News December 25, 2000; 43:9).
With the favorable findings of this study, the next research step in the process to obtain Food and Drug Administration approval for over‐the‐counter distribution of the product is to proceed with an “actual use study” to assess whether women who obtain this product in a simulated over‐the‐counter setting will use it safely and effectively. Recruitment for such a study began in November 2001.
1. Soller RW. OTCness. Drug Inf J 1998;32:555–60.
2. Davis TC, Long SW, Jackson RH, Mayeaux EJ, George RB, Murphy PW, et al. Rapid estimate of adult literacy in medicine: A shortened screening instrument. Fam Med 1993;25:391–5.
3. Wilcox AJ, Weinberg CR, Baird DD. Timing of sexual intercourse in relation to ovulation. Effects on the probability of conception, survival of the pregnancy, and sex of the baby. N Engl J Med 1995;333:1517–21.
4. American College of Obstetricians and Gynecologists. ACOG practice patterns. Emergency oral contraception. Number 25. Washington, DC: American College of Obstetricians and Gynecologists, 2001.
5. World Health Organization. Improving access to quality care in family planning. Medical eligibility criteria for contraceptive use. 2nd edition. May 11, 2001.
6. Rodrigues I, Grou F, Joly J. Effectiveness of emergency contraceptive pills between 72 and 120 hours after unprotected sexual intercourse. Am J Obstet Gynecol 2001;184:531–7.
7. Nielsen CL, Miller L. Ectopic gestation following emergency contraceptive pill administration. Contraception 2000;62:275–6.
8. Friedman CP, Romeo D, Hinton SS. Healthcare decisions and product labeling: Results of a consumer comprehension study of prototype labeling for proposed over-the-counter cholestyramine. Am J Med 1997;102:50–6.
9. Ciociola AA, Sirgo MA, Pappa KA, McGuire JA, Fung K. A study of the nonprescription drug consumer's understanding of the ranitidine product label and actual product usage patterns in the treatment of episodic heartburn. Am J Ther 2001;8:387–98.
10. Grimes DA, Raymond EG, Scott Jones B. Emergency contraception over-the-counter: The medical and legal imperatives. Obstet Gynecol 2001;98:151–5.
11. Ellertson C, Trussell J, Stewart FH, Winikoff B. Should emergency contraceptive pills be available without prescription? J Am Med Womens Assoc 1998;53:226–9,232.
12. Task Force on Postovulatory Methods of Fertility Regulation. Randomised controlled trial of levonorgestrel versus the Yuzpe regimen of combined oral contraceptives for emergency contraception. Lancet 1998;352:428–33.