Interpretations of a teratogen warning symbol

TERATOLOGY 64:148 –153 (2001)
Interpretations of a Teratogen
Warning Symbol
KATHERINE LYON DANIEL,1* KAREN DENARD GOLDMAN,2 SUE LACHENMAYR,3
J. DAVID ERICKSON,1 AND CYNTHIA MOORE1
1Birth Defects and Pediatric Genetics Branch, National Center on Birth Defects and Developmental
Disabilities, Centers for Disease Control and Prevention, Atlanta, Georgia 30341
2Health Education and Promotion Program, Health Services Department, Lehman College, City
University of New York, New York, New York 10012
3New Jersey Department of Health and Senior Services, Lebanon, New Jersey 08833
ABSTRACT
INTRODUCTION
Background: Warning symbols are used on terato-
Since 1988, a warning symbol has been used to in- genic medications to communicate the message that form prospective patients and consumers that a medi- women should (1) not take that medication if they may cation is a teratogen—a substance known to cause already be pregnant, and (2) not get pregnant while birth defects (Fig. 1). The symbol depicts the silhouette taking that medication. Communications research indi- of a pregnant woman covered by the international sym- cates that people interpret symbols or pictures in dif- bol for “No” (a red circle with a slash across it). The ferent ways. Other studies have shown that patients do phrases “Avoid Pregnancy,” “Do NOT Get Pregnant,” not always receive education materials that are part of or a combination of both messages are superimposed on prescription protocol. Researchers at the Centers for the circle and/or slash, or appear above and below the Disease Control and Prevention (CDC) tested the inter- symbol itself. This symbol has appeared on prescrip- pretation of the teratogen warning symbol and its abil- tion drug labeling, medication packaging, and educa- ity to convey the correct information without accompa- tional materials and has been published on the Inter- net in private industry and federal government web Methods: A teratogen warning symbol currently
pages. The warning symbol has been used for the printed on some medication packaging uses graphics widely used acne medication isotretinoin (Accutane௡).
and text warning the user not to get pregnant. Re- Accutane has been advocated to treat both serious and searchers interviewed women of childbearing age mild acne, and approximately 2,000 pregnancies have about their interpretation of the warning symbol and its occurred in women taking Accutane (Cunliffe et al., ’97; meaning. Ninety-seven women were interviewed in a Dermatologic and Ophthalmic Drugs Advisory Com- variety of locations, including public health clinics, lit- mittee, ’00a). The warning symbol is now being used for eracy and job training offices, health clubs, and malls.
the drug thalidomide (Thalomid௡), which was approvedfor the first time in the United States in 1998 (Lary et Results: Only 21% of women interpreted correctly
al., ’99). Thalidomide was responsible for severe dis- without prompting that they should either not take the abling birth defects in an estimated 10,000 children medication if they are pregnant or not get pregnant during the 1950s and 1960s before it was taken off the while taking the medication. Twenty-seven percent of market. However, applications for thalidomide’s anti- women first thought the symbol meant the package angiogenic and immunomodulatory properties are in- contained birth control medication, and 24% said it creasing, and it is likely to be prescribed more fre- simply indicated the package contained drugs or med- quently (Lary et al., ’99). Among women, the total icine. An additional 7% said they did not know what the number of physician visits for acne that resulted in symbol was supposed to mean; 39% of respondents Accutane prescriptions has grown from an estimated offered circumstances in which prescription medica- 150,000 in 1989 to 406,000 in 1997 (Stern, ’00).
The teratogen warning symbol was used as part of Conclusions: Misinterpretation of warning symbols
the Accutane Pregnancy Prevention Program devel- can result in serious consequences. This research oped by the manufacturer in 1988. The following year, should serve as an urgent call for mandating educationfor all patients receiving drugs with teratogenic prop-erties, and careful pretesting and modification of warn- *Correspondence to: Katherine Lyon Daniel, Birth Defects and Pedi- ing symbols before they are used on medications with atric Genetics Branch, National Center on Birth Defects and Devel-opmental Disabilities, Centers for Disease Control and Prevention, 4770 Buford Hwy, F-45, Atlanta GA 30341.
Received 6 November 2000; Accepted 7 May 2001 Published 2001 WILEY-LISS, INC.
US government work and, as such, is in the public domain inthe United States of America.
INTERPRETATIONS OF A TERATOGEN WARNING SYMBOL
Some teratogens are so potent that just one dose taken in early pregnancy can cause severely disablingor fatal birth defects. For these medications, there maybe “no safe dose, no safe duration of fetal exposure, nosafe period of exposure during pregnancy” (Dai, ’92).
Research on physician–patient interaction by Lipkin(’96) demonstrated that physicians often used overlytechnical explanations and that, even under ideal con-ditions, patients retained only 75% of what they heard.
When patients receive incomplete or inadequate coun-seling, or are not able to read or understand the infor-mation presented to them, labeling and warning sym-bols on medication packaging take on a largereducational burden. To be fully effective in helping toprevent birth defects, a teratogen warning symbolshould effectively convey two related messages: that Fig. 1. Teratogen warning symbol used in conjunction with the ques-
nonpregnant women who are taking the accompanying tionnaire to determine women’s interpretation of this symbol.
medication should not become pregnant, and that preg-nant women (or women who might be pregnant) shouldnot take the medication. The research question of this in 1989, Accutane packaging was changed from loose study was: How do women interpret the teratogen pills in a bottle to a blister pack, and the symbol was warning symbol outside of the context of a patient placed behind every capsule (Mitchell, ’95). The symbol has been used as one component of educational pro-grams on the appropriate and inappropriate uses of medications. These programs are either voluntary ormandatory. When used as part of these materials, the A questionnaire instrument was developed to deter- symbol is intended as a visual aid to remind patients of mine women’s interpretation of the teratogen warning the danger of therapeutic agents they should have symbol, and the instrument was pretested with 24 learned about in an initial counseling or education women in well-baby, Special Supplemental Nutritional session on the medication and contraceptive issues. For Program for Women, Infants, and Children (WIC), and some of these teratogenic drugs, prescription has be- family planning clinics within the New York metropol- come more acceptable over time for a wider range of itan area and surrounding suburbs. After refinement of conditions and severities (Cunliffe, ’97). Patients do not the instrument, 97 ethnographic interviews were con- always receive sufficient information, and some may ducted with women of childbearing age in 10 locations: not receive any counseling at all (MMWR, ’00).
a well-baby clinic, a family planning clinic, an HIV/ In addition, medications may be taken or prescribed AIDS treatment clinic, a health club, a prenatal-care contrary to indications, and warnings about terato- clinic, a WIC clinic, a literacy program office, a home genic potential of medications are unheeded by some health aide training program office, an HIV/AIDS peer patients (Mitchell, ’95). For example, studies of isotreti- training seminar location, and a shopping mall. The noin use to assess prevention program effectiveness questions were open-ended in format, allowing partic- have indicated that patient compliance with pregnancy ipants to contribute information in their own words.
prevention measures during the course of treatment Although the questions on the instrument were stan- ranged from 32% to 38% (Pastusak et al., ’94). Of dardized, the ordering of follow-up questions and women whose pregnancies were reported to the manu- prompts sometimes varied depending on the partici- facturer, only 51% resulted from contraceptive failure pant’s responses. This was designed to achieve stan- (Dermatologic and Ophthalmic Drugs Advisory Com- dardized data collection while still allowing the inter- mittee, ’00b). One survey of 14 women who became viewer to pursue unanticipated responses (Patton, ’87).
pregnant while using isotretinoin found that none of Locations were chosen to include participants of varied the women had been exposed to all elements of the race, ethnicity, age, and education (Table 1), and to multifaceted education program designed by the drug make sure that participants included some women manufacturer to reduce the likelihood of pregnancy with lower literacy skills and others with limited En- during medication use (MMWR, ’00). This exhorts us to glish proficiency. Two interviewers were trained and ensure the effectiveness of the warning symbol in con- checked for consistency, and interviews were con- veying a serious prevention message where other forms ducted in the language of the participant’s choice. In- of education may be lacking. The best way to under- terviews in Spanish were conducted with a bilingual stand this is to pretest the symbol in a sample of the interviewer. Interviewers recorded the answers on a intended target audience that has not been exposed to form for later analysis; most interviews were also tape- previous education materials (Lippin and Fingeret, ’91; recorded. A limited number of interviews were con- DANIEL ET AL.
TABLE 1. First response to question “what is this?” by demographics of respondent, percent (n)
bTotals do not always equal 100 because of rounding, and/or missing data.
PROCEDURE
ANALYSIS
First, participants were shown a computer-gener- The responses were recorded and coded, then ana- lyzed inductively across cases using qualitative analy- ated picture of a medication blister pack with the warn- sis techniques described by Miles and Huberman (’94).
ing symbol as it appears on a medication package and These include (1) grouping or condensing data, (2) cre- were asked what they thought it was. This initial view ating a display of indices in matrix form, (3) partition- of the symbol was the actual size used on packages of ing or dividing the data points in logical ways, and (4) teratogenic drugs (approximately one square inch).
verifying the data clusters so that variables of interest Second, participants were shown slightly enlarged can be labeled and distinctions between them can be versions of the symbol printed in color on a page, with further examined. The partitioning and clustering ac- the accompanying text (“Do NOT Get Pregnant”). The tivities allow for the observation of patterns or themes interviewer explained that the symbol would be that are reported in results below. For some variables, printed on a package of medication. Interviewers re- the data were analyzed and reported quantitatively corded initial responses on conversation guideline sheets without comment and probed for other interpre-tations. In the event that participants still gave inter- pretations that did not include the intended message/s, Ninety-seven women participated in the interview.
the interviewer followed up with a little more informa- Eight percent of the respondents were teens aged 18 or tion, for example, “and if I told you this wasn’t birth 19 years, 56% were aged 20 – 40 years, and 35% were control, what else might this symbol mean?” older than 40 years (Table 1). Thirty-three percent of Third, the interviewer explained the concept of ter- the women identified themselves as black, 33% as atogens, revealed the intended message of the symbol white, 25% as Hispanic, 4% as Asian, and 3% as other.
to participants, and asked for their reaction to that About 23% of the sample had less than a high school information. A full page-sized symbol without the ac- degree; 49% had a high school degree, GED, vocational companying text was shown to women to facilitate school degree, or some college classes; and 28% had a discussion of the symbol itself (apart from the text).
college degree or beyond. Fifty percent of the respon- Women were asked how they thought the symbol might dents reported that they had had a child, 41% reported be improved to more clearly communicate the intended they had not, and 9% declined to answer.
message. Several potential alternative phrases devel- The order of presentation of the materials did not oped using information from the pretest were provided systematically affect the interpretation in the pretest.
to choose from if participants did not have a suggestion.
Eight unique conceptual themes or response categories Questions at the close of the interview included partic- were clustered into categories (Table 2). The themes ipant demographics and, in later interviews, views on are presented in order of the most common first im- INTERPRETATIONS OF A TERATOGEN WARNING SYMBOL
TABLE 2. Response summaries of teratogen warning symbol interpretations
bTotal does not equal 100 because of rounding.
Theme 1—“Birth Control”: Twenty-eight percent of might induce an abortion; only one person gave the women interviewed reported their first inter- pretation was that the symbol indicated the med- Theme 9 —“Other Responses”: Five women gave var- ication was birth control or was used for emer- ious other responses first, including that the drug gency contraception. A total of 78% of women was for HIV/AIDS treatment, or that the symbol reported this interpretation at some point in the indicated that pregnancy is “bad.” About 25% of women gave answers in this category at some Theme 2—“Drugs or medicine”: Twenty-five percent of respondents initially interpreted the symbol Combination Themes 2 and 3—“Complete under- solely to indicate that the package contained drugs standing” (both correct messages): Nine women or medicine, without an associated warning.
(11%) reported both of the “correct” desired mes- Theme 3—“If Pregnant, Don’t Take This Medication” sages during the course of the interview. Only 20% (correct message): Nineteen percent of respondents of participants were able to identify either one of reported as their first interpretation that women the “correct” messages first, without probing.
who were already pregnant should not take themedication. Fifty-one percent reported this inter- Sixty-one women were asked whether there was a pretation at some point in the interview.
situation in which it was okay to share medications: Theme 4 —“Hormones or Vitamins”: Eight percent of 39% said it was never okay, 21% said for nonprescrip- women first thought the symbol meant the pack- tion medications only (e.g., for headaches or pain re- age contained hormones or vitamins; a total of 11% lief), and 39% gave circumstances in which sharing of women interpreted this throughout the inter- prescription medication was acceptable. Reasons for sharing included: it is common practice (everyone does Theme 5—“I Don’t Know”: Seven percent of women it), cost of medication, share with friends and family, to were unable to interpret what the symbol might help others, the same medication for which person has mean. An additional 9% did not offer any addi- a prescription, and illegal/irresponsible use (addic- tional interpretations after their initial thoughts.
Theme 6 —“If Taking This Medication, Don’t Get Pregnant” (correct message): One percent of women DISCUSSION
reported first that the symbol warned nonpregnant A large proportion of the women in this study did not women they should not get pregnant while taking understand the intent of the warning symbol when the medication; 26% reported this at any time dur- they saw the symbol without counseling or education.
By itself, this symbol does not appear to communicate Theme 7—“Danger for Pregnant Women”: In a re- the risk for birth defects. Twenty-eight percent of lated interpretation to Theme 2, 12% of the respon- women interpreted the message with a meaning oppo- dents said they thought it meant the pills were site of its intent: they initially thought this symbol “dangerous” for pregnant women, but without fur- meant the medication prevented pregnancy. This dem- onstrates a serious weakness of this symbol and its use Theme 8 —“Abortion Pill”: A total of 16% said they in helping to prevent teratogenic exposure. In fact, this thought the symbol meant that the pill could or incorrect interpretation may increase teratogenic risk DANIEL ET AL.
because women who have not received counseling may genic drugs are less common or nonexistent. Of partic- be less likely to use other birth control methods while ular concern in this country are patients who have low literacy skills, limited intellectual capacity, or limita- All participants confirmed that once they had been tions in the language of the warning text. Their reli- told what the message was supposed to convey, they ance on visual cues underscores the importance of uni- could easily see and understand the correct warning versally understood symbols. Research results by message(s). This demonstrates the difference between Lippin and Fingeret (’91) demonstrated that picto- “recognition” of a message and initial “interpretation” grams often needed to be explained before low literacy of a message and underscores the importance of initial patients understood their meaning. Presenting the education and counseling. The study also showed that symbol out of context may contribute to some of the many women experienced difficulty seeing beyond vague response themes such as “Drugs or Medicine” their initial interpretation. Once some participants had and “Hormones or Vitamins.” The response “Danger for reached a conclusion they repeatedly stated that the Pregnant Women” might indicate at least a partial medication must be birth control, even when the inter- understanding of the risk. Repeat treatment/use is an- viewer suggested otherwise. This tendency is consis- other important issue because patients may not receive tent with the interpretation of risk information (Arnon the same counseling as they did during the initial visit and Kreitler, ’84; Hogarth, ’87), especially when the and course of treatment, and may forget the contrain- patient has no reason to question her initial interpre- tation. This also underscores the importance of provid-ing complete education for all patients, including test- CONCLUSIONS
ing the patients’ understanding of risks and counselingpatients not to share medications with others.
The existing teratogen warning symbol may be an The nature of the weaknesses of the symbol were effective reminder of educational counseling if the that some women reported they did not understand the health care provider has already conveyed that infor- meaning of the international symbol for “NO,” (the red mation. However, for women who have not received circle with a slash through it), some read the words as this education, the symbol is likely to convey a message “TO not get pregnant” instead of “DO not get preg- counter to its intent and could possibly increase the nant,” and a few said they had not read beyond their likelihood that a patient would not use effective con- initial interpretation because it was a directive state- traception while taking a prescribed teratogenic medi- ment and/or negative in tone. Once the intent of the cation. Although drugs with teratogenic properties message was explained, many participants reported may provide significant and unique benefits, their ef- they wanted more information about why they should fectiveness is coupled with the potential for harmful heed the warning. A few expressed resentment at being effects. It is thus imperative that women patients directed to take action without being given a reason.
clearly understand the teratogenic risks and the need When asked for improvements, a number of partici- to avoid pregnancy. Recommendations include the fol- pants suggested that warning messages that explained lowing: (1) health-care providers who prescribe terato- the outcome would be most effective for them (exam- genic medications should always provide complete ed- ples include “Causes Birth Defects” or “Will Cause ucational counseling about birth defects and pregnancy prevention; and (2) although the warning symbols This study has several limitations. First, partici- should be used to reinforce the educational counseling pants were recruited from a convenience sample and conducted by the referring physician, drug manufac- may not be typical of patients who could be prescribed turers and package designers should assume the re- these medications. The results should be interpreted to sponsibility to carefully test the interpretation of warn- apply to a high-risk group of lower socioeconomic sta- ing symbols outside of the context of education tus and limited literacy skills (although one-half of the programs for their effectiveness in communicating the sample reported some education beyond high school, potential risks of the teratogen and modify symbols as and one quarter had a college degree). Second, as dis- needed. The need for a warning symbol, tested for cussed earlier, the symbol was shown to women com- effectiveness and clarity for both the general popula- pletely outside of a patient education context—they tion and higher risk groups, is imperative to help pre- had not received any information previously about vent birth defects associated with teratogenic medica- birth defects, teratogens, or other related issues.
Therefore, these results may or may not apply towomen who receive brief or limited educational coun- LITERATURE CITED
seling along with their prescription. The conditions Arnon R, Kreitler S. 1984. Effects of meaning training on overcoming under which these findings do apply include: instances functional fixedness. Curr Psychol Res Rev Winter:11–24.
where no education is provided, when medication is Cunliffe WJ, van der Kerkhof PCM, Caputo R, Cavicchini S, Cooper A, shared with others, when medication samples are ob- Fryand OL, Gollnick H, Layton AM, Leyden JJ, Mascaro JM, Or- tained from health-care providers or friends, or when tonne JP, Shalita A. 1997. Roaccutane treatment guidelines: resultsof an international survey. Dermatology 194:351–357.
the medication is obtained in other countries where Dai WS, LaBraico JM, Stern RS. 1992. Epidemiology of isotretinoin pregnancy prevention programs accompanying terato- exposure during pregnancy. J Am Acad Dermatol 26:599 – 606.
INTERPRETATIONS OF A TERATOGEN WARNING SYMBOL
Dermatologic and Ophthalmic Drugs Advisory Committee. Briefing Lippin TM, Fingeret HA. 1991. How adults with limited literacy skills Information, 2000a. Appendix II: overview of Accutane and preg- cope with reading medication labels. Durham, NC: Literacy South.
nancy exposure, September 18 –19, 2000, Rockville MD: US Depart- ment of Health and Human Services, Food and Drug Administration Miles MB, Huberman AM. 1994. Qualitative analysis: an expanded [http://www.fda.gov/ohrms/dockets/ac/00/backgrd/3639b1.htm].
sourcebook. Thousand Oaks, CA: Sage. p 1–11.
Dermatologic and Ophthalmic Drugs Advisory Committee. Briefing Mitchell AA, Van Bennekom CM, Louick C. 1995. A pregnancy pre- Information, 2000b. Pregnancy prevention program: a risk man- vention program in women of childbearing age receiving isotreti- agement program to reduce pregnancies during Accutane treat- ment, September 18 –19, 2000, Rockville MD: US Department Pastuzak A, Koren G, Rieder MJ. 1994. Use of the retinoid pregnancy of Health and Human Services, Food and Drug Administra- prevention program in Canada: patterns of contraceptive use in tion [http://www. fda.gov/ohrms/dockets/ac/00/backgrd/3639b1.
women treated with isotretinoin and etretinate. Reprod Toxicol8:63– 68.
Patton MQ. 1987. How to use qualitative methods in evaluation.
Hogarth R. 1987. Judgement and choice: the psychology of decision.
Newbury Park, CA: Sage. p 108 –143.
2nd ed. New York: John Wiley & Sons. p 114 –119.
Stern RS 2000. Medication and medical service utilization for acne: Lammer EJ, Chen DT, Hoar RM, Agnish ND, Benke PJ, Braun, JT 1995–1998. Report to the Dermatologic and Ophthalmic Drugs Ad- Curry CJ, Fernhoff PM, Grix AW, Lott IT, Richard JM, Sun SC.
visory Committee, September 18 –19, 2000. Rockville, MD: US De- 1985. Retinoic acid embryopathy. N Engl J Med 313:837– 841.
partment of Health and Human Services, Food and Drug Adminis- Lary JM, Daniel KL, Erickson JD, Roberts HE, and Moore CA. 1999.
The return of thalidomide: can birth defects be prevented? Drug System for Thalidomide Education and Prescribing Safety (STEPS).
1998. [prescriber folder]. Warren, NJ: Celgene Corp.
Lipkin M. 1996. Physician–patient interaction in reproductive coun- Theis SL, Johnson JH 1995. Strategies for teaching patients: a meta seling. Obstet Gynecol 88:31S– 40S.
analysis. Clin Nurse Spec 9:100 –105.

Source: http://www.kdgconsulting.net/resources/teratogen_article.pdf