- Email: [email protected]
Contraceptive usage, knowledge and correlates of usage among female emergency department patients
Contraception 74 (2006) 201 – 207
Original research article
Contraceptive usage, knowledge and correlates of usage among female emergency department patients Roland C. Merchant a,b,4, Jennifer A. Damergisa, Erin M. Geea, Beth C. Bock c, Bruce M. Becker a,b, Melissa A. Clark b a
Department of Emergency Medicine, Brown Medical School, Rhode Island Hospital, Providence, RI 02903, USA b Department of Community Health, Brown Medical School, Providence, RI 02912, USA c Department of Psychiatry and Human Behavior, Center for Behavioral and Preventive Medicine, Brown University, Providence, RI 02903, USA Received 23 February 2006; revised 13 March 2006; accepted 17 March 2006
Abstract Objectives: For female emergency department (ED) patients, we sought to assess the prevalence of contraceptive usage as well as the extent of contraceptive knowledge and to determine if demographic and sexual health history factors, comprehension of contraceptive methods and moral/religious opinions on contraception were associated with current usage of birth control pills (BCPs), prior usage of emergency contraception (EC) and frequency of condom usage. Methods: English-speaking female ED patients aged between 18 and 55 years at a northeastern United States urban ED were surveyed on their usage, comprehension and opinions regarding BCPs, EC and condoms. Results: Of the 539 respondents (64.6% were aged V 35 years), most were White (63.1%), single (42.5%), Catholic (48.4%) and privately insured (55.3%). Among the 223 women at pregnancy risk [not currently pregnant, not using any form of nonsurgical birth control (except condoms) and with no prior tubal ligation or hysterectomy], about 25% were using BCPs, fewer than 10% had used EC and almost 40% never used condoms. Most women displayed good knowledge about BCPs and condoms but poor understanding about EC. In multivariate logistic regression analyses, current BCP usage among women at risk of pregnancy was associated with younger age [odds ratio (OR) = 0.54; 95% confidence interval (CI) = 0.37–0.79], private insurance (OR = 2.52; 95% CI = 1.30–4.86) and recent intercourse (OR = 1.61; 95% CI = 1.19–2.18). Among women at risk of pregnancy, those who had an abortion (OR = 2.56; 95% CI = 1.17–5.61) and those who displayed greater EC knowledge (OR = 3.23; 95% CI = 1.50–6.95) had greater odds of having used EC. Among all women, more frequent condom usage was associated with being younger (OR = 0.57; 95% CI = 0.46–0.70), having never been married (OR = 0.44; 95% CI = 0.28–0.68) and not having intercourse recently (OR = 0.79; 95% CI = 0.64–0.98). Conclusions: A high percentage of female ED patients (41.4%) were at risk of pregnancy. Demographic and sexual history factors can help identify women who might benefit from receiving referrals or education on contraceptive measures. D 2006 Elsevier Inc. All rights reserved. Keywords: Oral contraceptives; Emergency contraception; Emergency medicine; Contraception; Condoms; Emergency services; Women’s health
1. Introduction The percentage of women in the United States at risk of unintended pregnancy may be increasing. According to estimates from the 1995 National Surveys of Family Growth (NSFG), 49% of pregnancies in the United States during The results of this study were presented at the 2005 American College of Emergency Physicians Scientific Assembly in Washington, DC, on September 27, 2005. 4 Corresponding author. Department of Emergency Medicine, Brown Medical School, Rhode Island Hospital, Providence, RI 02903, USA. Tel.: +1 401 444 5109; fax: +1 401 444 4307. E-mail address: [email protected] (R.C. Merchant ). 0010-7824/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.contraception.2006.03.012
that year were unintended and 54% ended in an induced abortion [1]. According to the 1995 and 2002 NSFGs, the percentage of women aged between 15 and 44 years having sexual intercourse who were not using any form of contraception in fact increased from 5.4% in 1995 to 7.4% in 2002 [2]. If contraception was more widely used, rates of unintended pregnancy would decrease, as would the societal costs of these pregnancies from their adverse health and economic effects [3,4]. The Public Health and Education Task Force (PHTF) of the Society for Academic Emergency Medicine (SAEM) evaluated preventive interventions for the emergency department (ED) in 2000 [5,6]. Its review appraised the
202
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207
applicability of preventive health interventions that were recommended by the U.S. Preventive Services Task Force (USPSTF) for other health care settings to ED practice. Contraceptive counseling was considered as a candidate preventive health measure by the USPSTF but was not reviewed by the SAEM PHTF. This absence was due in large part to the paucity of studies on the usage and effectiveness of reproductive health interventions in the ED. In fact, published studies from EDs on patient contraceptive usage and knowledge have concerned only the use of emergency contraception (EC) [7]. To our knowledge, there has been no published study on birth control pill (BCP) and condom usage among patients in this setting. Nevertheless, the ED is an appropriate place to consider preventive reproductive health interventions for women because it may be the only source of primary health care services to women who do not have to access to them elsewhere. In the United States, there were 61 million patient visits to EDs in 2003, of which 46.7% were by women aged between 15 and 44 years [8]. For many women, the ED is their principal source of medical care [9 –11]. As a consequence, the ED may serve as the only means by which women can access or receive referrals for preventive reproductive health care. Knowing the patterns of and factors related to contraceptive use may assist emergency medicine clinicians in identifying patients who could benefit from using contraception, addressing knowledge deficits and providing referrals or systems to facilitate referrals for contraceptive use. We sought to determine how frequently women visiting the ED are using contraceptive methods (BCPs, EC and condoms) and, consequently, how many are at risk of pregnancy. Determining the frequency of contraceptive usage and the prevalence of women visiting the ED at risk of pregnancy would help highlight the needs of this group. Furthermore, along with pregnancy testing, knowing the prevalence and demographics of women at risk would help clinicians identify these women so as to avoid prescribing or using potentially dangerous or harmful medications, diagnostic tests or procedures. In addition, we aimed to determine how well women know the purpose, applicability and suggested usage of BCPs, EC and condoms. We believe that understanding the extent of patient knowledge on these topics would help clinicians plan for relevant educational interventions when needed. We further sought to examine if contraceptive use (usage of BCPs, EC and condoms) was associated with demographic or sexual health history factors; with comprehension of the purpose, applicability and recommended usage of these contraceptive measures; and with moral and religious opinions on contraception. Our goal was to find potentially modifiable factors that may lead to ED-based preventive health programs to improve contraceptive use among ED patients. For BCPs and EC, we were particularly interested in women who have sex with men and who are at risk of pregnancy [those not currently pregnant, not using a
nonsurgical form of birth control (except condoms) and with no prior history of tubal ligation or hysterectomy]. For condom usage, we considered all respondents who reported having sex with men. It is our hope that the study results may contribute to future evaluations of preventive health measures for the ED.
2. Materials and methods 2.1. Study design This was a prospective survey of women aged between 18 and 55 years seeking treatment in an ED. Participants completed a self-administered, written as well as anonymous multiple-choice and best-answer questionnaire that we drafted. The hospital institutional review board approved the study as exempt from review and written informed consent. 2.2. Study setting We conducted the study at an urban academic ED that serves as a tertiary referral center for a catchment area of approximately 1.5 million people. In fiscal year 2002, there were 73,672 adult patient visits to this ED, 36,102 (49.0%) of which were by women. Of these women, 24,087 (66.7%) were aged between 18 and 55 years. Table 1 shows the demographic profile of the women included in our survey. 2.3. Survey content and development We chose three topics for our survey — BCPs, EC and condoms — and created questions patterned on the styles used in other surveys on similar topics in other settings that we reviewed. Our survey focused on the survey participants’ demographics and health history as well as their use, knowledge and opinions regarding contraceptive usage. The written multiple-choice, best-answer questions asked participants to identify the applicability, purpose and recommended usage of birth control, EC and condoms (Appendix A). The reading level of the survey was at a Flesch–Kincaid grade of 5.4 (Microsoft Word, Redmond, WA, USA). We created a draft questionnaire and pilot tested it in June 2002 on a convenience sample of 20 women from the ambulatory care section of the ED. Using a standardized script, we interviewed these women after they completed the questionnaire. Our interview consisted of a brief cognitive assessment of randomly selected questions from the questionnaire, a review of potentially sensitive questions and an analysis of mismarked responses. We also solicited feedback on participants’ opinions, impressions and reactions to the questionnaire. Each of these women received $20 for participating in this pilot test. We revised our survey based on their comments and our observations. The final survey used in this analysis included 28 multiple-choice questions: 6 demographic questions; 11 health history questions; 3 questions on BCP knowledge, 3 on condom knowledge and 1 on EC knowledge; and 4 agree-or-disagree
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207 Table 1 Demographic profile of the survey respondents (N = 539)
Age group (years) 18–25 26–35 36–45 46–55 Race American Indian/ Native American Asian/Pacific Islander Black/African American Hispanic/Latina White Biracial/Multiracial Other Marital status Single Married Divorced/Separated Male partner Widowed Religious identity Buddhist Catholic Jewish Muslim Protestant None Other Health insurance None Government Private
All respondents (N = 539)
Women at pregnancy risk (n = 316)
Women not at pregnancy risk (n = 223)
36.4 28.2 21.5 13.9
48.4 27.5 16.2 7.9
19.3 29.2 29.2 22.4
4.7
4.4
4.9
0.9 9.8 15.0 63.1 5.6 0.9
1.6 8.9 14.5 62.7 6.6 1.3
0.0 11.2 15.7 63.7 4.0 0.5
42.5 28.4 14.6 13.4 1.1
51.9 25.3 9.2 13.3 0.3
29.1 32.7 22.4 13.4 2.2
1.3 48.4 1.1 1.3 26.7 18.1 3.0
1.3 48.7 1.6 1.0 25.0 19.9 2.5
1.3 48.0 0.5 1.8 29.1 15.7 3.6
13.9 30.8 55.3
15.5 21.8 62.7
11.7 43.5 44.8
All values are expressed as percentages.
questions on moral/religious opinions regarding BCP and EC usage. 2.4. Selection of participants We limited the sample to women aged between 18 and 55 years because we hypothesized that women older than 55 years might have different reproductive health concerns (e.g., menopause) and be less likely to use the contraceptive measures of interest. Women were excluded if they could not read or write in English; were being evaluated in the critical care, psychiatric section or alcohol abuse holding areas of the ED; were not awake; or could not physically complete the form. All other women seeking treatment in the ED during the study hours were asked to complete a questionnaire. Each participant who completed the questionnaire received a $2 gift card to a local pharmacy. Research assistants and trained volunteers administered the survey in three 2-month blocks that included sixty-six 8-h shifts in July–August 2002, thirty-two 4 -h shifts in October–November 2002 and thirty-two 4-h shifts in April– May 2003. We purposely sampled three seasons to avoid unknown time-dependent factors that might influence the
203
pattern of ED visits and affect the results. The data collection shifts were scheduled to reflect the time-dependent influx of patients to our ED. Data collection was structured such that 25% of the shifts were between 7 and 11 a.m., 50% were between 11 a.m. and 11 p.m. and 25% were between 11 p.m. and 7 a.m. The ED’s daily volume was essentially the same each day of the week. To reflect this pattern of patient visits, we sampled participants so that each day of the week was represented equally. 2.5. Methods of measurement The primary outcome was use of BCPs, EC and condoms. Independent variables included demographic and sexual health history factors as well as knowledge and opinions about the contraceptive methods. 2.6. Primary data analysis Two trained personnel independently entered the data into separate Epi Info 2002 (Centers for Disease Control and Prevention, Atlanta, GA, USA) databases. These two databases were compared for every entry, and errors were corrected to create a final database. We transferred the final database into Stata 8.2 (Stata, College Station, TX, USA) using Stat/Transfer 6 (Circle Systems, Seattle, WA, USA). We tabulated the frequency of responses to each of the questions and calculated 95% confidence intervals (CIs) for point estimates. We conducted our analyses on respondents for whom we had no missing data on the variables of interest; we then repeated our analyses for all respondents (whether they answered all questions or not) and observed similar results. For BCP and EC usage, we restricted our analyses to women at risk of pregnancy. For condom usage, we conducted the analysis on all women who have sex with men. To examine the correlates of contraceptive usage, we performed univariate logistic regression analyses using the demographic and sexual health history variables as well as responses to the contraceptive knowledge and opinion questions as the variables of interest. Correlates statistically significant at the a = .05 level were placed in multivariate logistic regression models to assess their association with contraceptive usage. For condom usage, we performed ordinal logistic regression in which increasing frequency of condom usage was the outcome. Odds ratios (ORs) and corresponding 95% CIs were calculated from these models.
3. Results 3.1. Respondent demography As shown in Table 1, of the 539 respondents, most were aged 35 years or younger (64.6%), White (63.1%), single (42.5%), Catholic (48.4%) and privately insured (55.3%). Of the 539 respondents, 316 (53.6%) were considered to be at risk of pregnancy [not currently pregnant, not using any nonsurgical form of birth control (except for condoms) and
204
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207
sexual and pregnancy history except for history of having been pregnant.
Table 2 Sexual and contraceptive history All Women at Women not at respondents pregnancy risk pregnancy risk (N = 539) (n = 316) (n = 223) Contraceptive history Currently using BCPs 18.9 Using other birth 11.1 control methods Prior bilateral 23.4 tubal ligation Prior hysterectomy 9.7 Ever used EC 8.4 Ever had abortion 25.1 Sexual history Last intercourse with a man V 1 week 49.0 V 1 month 22.8 V 6 months 9.8 V 1 year 4.8 N 1 year 9.1 Never 4.5 Frequency of condom use No sex/no sex with men 7.2 Never 45.8 Sometimes 16.9 Often 13.9 Always 16.1 Pregnancy history Currently pregnanta 2.4 Ever pregnant 71.1 No. of times pregnant 0 28.9 1 13.9 2 17.4 3 18.0 z4 21.7
25.9 0.0
9.0 26.9
0.0
56.5
0.0 9.8 23.4
23.3 6.3 27.4
48.4 21.5 9.8 4.8 8.5 7.0
49.8 24.7 9.9 4.9 9.9 0.9
8.9 38.9 17.4 16.5 18.3
4.9 55.6 16.1 10.3 13.0
0.0 56.7
5.8 91.5
43.3 14.9 14.2 13.9 13.6
8.5 12.6 22.0 23.8 33.2
All values are expressed as percentages. a Of all respondents, 3.9% were unsure if they were pregnant; of women at pregnancy risk, 6.0% were pregnant.
not having undergone a hysterectomy or bilateral tubal ligation]. Among women at risk of pregnancy, most were aged 35 years or younger (75.9%), White (62.7%), single (51.9%), Catholic (48.7%) and privately insured (62.7%). 3.2. Respondent health history/contraceptive use history Fewer than 35% of all respondents (Table 2) were currently using any type of nonsurgical birth control method (except for condoms). Of all respondents, 33.1% had undergone surgical sterilization (hysterectomy or bilateral tubal ligation). Three times more women had undergone an abortion (25.1%) than had ever used EC (8.4%). Although almost half of all respondents (49%) had had intercourse within the past week, most did not use condoms during intercourse (16.1% reported always using them). More than 70% of the women reported being pregnant at some time in the past. Although few women were currently pregnant by self-report, 3.9% of all participants and 6.0% of women at risk of pregnancy were unsure if they were currently pregnant. The women at risk of pregnancy were similar to all other participants in their contraceptive,
3.3. Contraception knowledge and opinions Overall knowledge regarding BCPs and condoms was high (Table 3). A higher percentage of participants correctly answered all condom questions than the birth control questions. Participants were 3.2 and 3.0 times more likely to correctly answer all of the condom and BCP knowledge questions, respectively, than the EC knowledge question (although there was only one EC knowledge question). Fewer than 10% of all respondents believed that BCPs (whether as EC or as daily usage) caused abortions, were morally wrong or were against their personal religious beliefs. The subset of women at risk of pregnancy answered the knowledge and opinions questions similarly to the entire sample of participants but tended to answer the knowledge questions correctly more often than women who were not at risk of pregnancy. 3.4. Correlates of contraceptive usage The outcomes of the logistic regression analyses that compared factors associated with current usage of BCPs, ever having used EC and frequency of condom usage among women at risk of pregnancy are depicted in Table 4. In a univariate analysis of current BCP use among women at risk of pregnancy, contraceptive use was not associated with current or prior marriage or having a male partner (partner status), religious preference, increasing number of pregnancies, ever having had an abortion or believing that BCPs are against the respondent’s religious or moral beliefs or cause an abortion. Current usage of BCPs was associated with being younger, being White, having private health insurance, having more recent intercourse, never having been Table 3 Correct responses to survey questions
BCP knowledge Who Why How often All correct EC knowledge Condom knowledge Who Why How often All correct BCP opinions Cause abortions Morally wrong Against religion EC opinions Use after intercourse prevents pregnancy
All respondents (N = 539)
Women at pregnancy risk (n = 316)
Women not at pregnancy risk (n = 223)
89.6 98.3 95.0 85.9 28.3
91.8 99.1 96.2 88.9 32.3
86.6 97.3 93.3 81.6 22.9
95.9 98.3 95.9 91.8
96.2 98.7 96.8 93.4
95.5 97.8 94.6 89.7
8.5 6.1 9.7
9.2 5.1 8.9
7.6 7.6 10.8
8.9
11.1
5.8
All values are expressed as percentages.
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207
205
Table 4 Factors associated with contraceptive usage Women at risk of pregnancy BCP usage (n = 316)
Age White race Partner status Divorced/Widowed Married/Male partner Religion Protestant Catholic Other Religious worship frequency Protestant Catholic Private health insurance Recent intercourse Ever pregnant No. of pregnancies Prior abortion Contraceptive knowledge BCP beliefs Causes abortions Morally wrong Against religious beliefs EC beliefs Causes abortions
All respondents Condom usage (n = 500)a
EC usage (n = 316)
Univariate
Multivariate
Univariate
Univariate
Multivariate
0.58 (0.42–0.79) 2.06 (1.17–3.60)
0.54 (0.37–0.79) 1.67 (0.90–3.10)
0.89 (0.62–1.29) 0.52 (0.25–1.10)
0.51 (0.43–0.61) 0.86 (0.61–1.20)
0.57 (0.46–0.70)
0.59 (0.23–1.52) 0.67 (0.39–1.15)
1.73 (0.59–5.11) 0.69 (0.30–1.60)
0.48 (0.30–0.78) 0.23 (0.16–0.33)
1.17 (0.66–2.06) 0.44 (0.28–0.68)
0.34 0.67 0.31 0.78 0.75 0.78 2.23 1.52 1.88 0.84 1.18 2.83
2.14 1.59 1.64 1.19 1.20 1.14 0.61 0.98 0.69 1.10 2.65 3.31
0.57 0.57 0.56 0.96 0.97 1.07 0.83 0.78 3.12 0.99 1.55 1.03
0.74 (0.35–1.55) 0.86 (0.51–1.44) 0.81 (0.51–1.29)
(0.16–0.75) (0.36–1.26) (0.08–1.16) (0.62–0.98) (0.60–0.94) (0.65–0.96) (1.27–3.94) (1.14–2.02) (1.13–3.12) (0.63–1.11) (0.66–2.10) (1.02–7.83)
0.84 0.83 0.80 2.52 1.61 1.14
(0.65–1.06) (0.64–1.06) (0.64–1.01) (1.30–4.86) (1.19–2.18) (0.60–2.16)
2.40 (0.80–7.16)
(0.64–7.17) (0.51–5.0) (0.28–9.69) (0.85–1.65) (0.95–1.52) (0.90–1.45) (0.29–1.27) (0.67–1.43) (0.32–1.50) (0.78–1.56) (1.23–5.71) (1.55–7.07)
1.10 (0.47–2.58) 0.18 (0.02–1.39) 0.95 (0.39–2.32)
1.54 (0.50–4.76) 2.24 (0.60–8.34) 1.11 (0.32–3.93)
1.46 (0.61–3.47)
0.61 (0.22–1.71)
Multivariate
2.56 (1.17–5.61) 3.23 (1.50–6.95)
(0.35–0.92) (0.37–0.88) (0.28–1.15) (0.82–1.11) (0.86–1.10) (0.95–1.20) (0.60–1.16) (0.64–0.94) (2.16–4.50) (0.85–1.14) (1.06–2.26) (0.63–1.70)
0.79 (0.64–0.98) 1.52 (0.96–2.41) 1.05 (0.69–1.63)
All values are expressed as OR (95% CI). a Excludes women who reported not having sexual intercourse.
pregnant, attending worship less frequently (for all worship groups and Catholics and Protestants separately) and having greater BCP knowledge. In multivariate analyses, current BCP usage was associated with younger age, private health insurance and more recent intercourse. Stated in terms of preventive health needs, lack of current BCP usage was associated with being older, not having private health insurance and not having had sex recently. Current BCP usage was not associated with currently or ever having been married, having a male partner (partner status), religious preference, increasing number of pregnancies or ever having had an abortion. BCP usage was also not associated in either univariate or multivariate analysis with believing that BCPs are against the respondent’s religious/moral beliefs or cause an abortion. For prior usage of EC among women at risk of pregnancy, in the univariate and multivariate analyses, ever having had an abortion and answering the EC knowledge question correctly were associated with ever having used EC. Conversely, stated in preventive health needs terms, not having ever used EC was associated with not understanding its function. No other factor was associated with EC use in either the univariate or multivariate analysis. Among all respondents who reported having intercourse with men, more frequent condom usage was associated in the univariate analysis with being younger, never having been married, not having a male partner, being Catholic or
Protestant, having had intercourse less recently, never having been pregnant and having had an abortion. In the univariate analysis, more frequent condom usage was not associated with race, private health insurance and condom usage knowledge. In the multivariate analysis, more frequent condom usage was associated with being younger, not being married or having a male partner and not having intercourse recently. Again, in preventive health needs terms, less frequent condom usage was associated with being older, being married or not having a male partner and having intercourse recently.
4. Discussion Contraceptive usage is a relevant topic to explore among ED patients given that 67% of women visiting our ED were of childbearing age (i.e., 18–55 years). A high percentage (59%) of ED patients eligible to participate were at risk of pregnancy and had recent sexual intercourse. This risk of pregnancy is highlighted by the relatively low percentages of women using BCPs, EC or condoms. Among women at risk of pregnancy, approximately 25% were currently using BCPs, fewer than 10% had ever used EC and almost 40% never used condoms during intercourse. Of course, some of these women may want to become pregnant, but given the high rates of unwanted pregnancy in the United States and
206
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207
increasing rates of failure to use contraception, these results verify that many women visiting the ED are at risk of pregnancy, whether desired or not. These low percentages of contraceptive usage serve as a reminder to emergency medicine practitioners that a high percentage of women of childbearing age are at risk of pregnancy. This has important implications on practitioners’ choice of medications (e.g., teratogens) and diagnostic procedures (e.g., radiography). The results of this study also suggest that an ED visit may be an excellent opportunity to provide women seeking treatment with contraceptive services or referrals. EC definitely appears to be underused in this population. Women were three times more likely to have had an abortion than to have ever used EC. This finding was true for women at risk and those not at risk of pregnancy. Despite their known efficacy in preventing pregnancy and sexually transmitted diseases, condoms were not frequently used by most women at risk of pregnancy. Less than one fifth of the respondents reported always using condoms. This was of particular concern given that nearly 70% of women at risk of pregnancy had sexual intercourse with a man at least once within the prior month. Again, some of these women may have desired pregnancy, but given the high rates of infrequent condom usage among all women sampled, it is likely that women who did not desire pregnancy were not using condoms. Participants had reasonably good knowledge of BCPs and condom use but exhibited a significant deficit in understanding the concept of EC. The results were essentially the same for women regardless of pregnancy risk status. Few women expressed moral or religious objections to BCP and EC usage, which suggests that one of the real challenges in improving women’s understanding about reproductive preventive health choices (at least in the population of women seeking treatment in the ED) is increasing their understanding about the purpose and usage of EC. We identified several important factors associated with usage of BCPs, EC and condoms in our sample. Younger women with private health insurance who had had sexual intercourse with a man recently were more likely to be currently using BCPs. Women who ever had an abortion and exhibited a greater understanding of the concept EC were more likely to have ever used EC. More frequent condom use was observed among younger unmarried/no male partner women who had not had sexual intercourse recently. Surprisingly, women who had had sexual intercourse recently were less likely to have used condoms recently. If these women were not seeking to be pregnant, then they are at substantial risk of pregnancy and sexually transmitted diseases. It is possible that women who have sexual intercourse less often are more likely to protect their reproductive health status and to prevent pregnancy. It is important to emphasize that contraceptive knowledge, race, religious identity, pregnancy history, frequency of religious worship and beliefs and opinions regarding
contraception were not associated with contraceptive usage, with the exception of EC knowledge and usage. As such, except for EC, lack of knowledge does not appear to be the primary barrier to contraceptive use for women seeking treatment in an ED. The lack of association of a history of prior pregnancy with current contraceptive use highlights an increased probability of subsequent pregnancies, whether desired or not. In summary, our study results indicate that a large proportion of women visiting the ED are at risk of pregnancy. Many of these women, whether desiring pregnancy or not, are having intercourse and do not use contraception and therefore might not know they are pregnant, which has implications on the selection of medications and diagnostic studies. Knowledge of BCPs and condoms is high; however, EC knowledge among these women is poor. Our study results pinpoint groups of women who might benefit from expanded educational and outreach programs in an ED that might serve to improve their use of reproductive health choices. For example, women without health insurance could receive educational pamphlets outlining their reproductive health options (e.g., BCPs) and be provided with a list of local resources that offer these services; older women could receive targeted educational messages informing them of the option of using BCPs and condoms; and women who have sexual intercourse frequently could be provided with educational pamphlets emphasizing the risks of pregnancy and sexually transmitted diseases. It is our hope that this study can contribute to efforts regarding the assessment of women’s reproductive preventive health needs in the ED. 4.1. Limitations Given our sampling techniques (ED volume matched, capture of different periods and equal weighting of days of the week), we believe that we have a representative sample of those eligible to participate in the study. However, our sample is from a single ED and involved English-speaking women who were predominantly from certain demographic groups (White, Catholic, single, privately insured). Therefore, the findings may not be applicable to other settings with more diverse populations. We are hopeful that our questionnaire will be adapted for other settings, cultures and languages to corroborate our results; however, we also recognize that with larger samples or different populations, the results may change. Likewise, other modeling techniques may produce different results and ORs that did not reach significance may become significant. It is also possible that, despite our use of cognitive assessments and screening of questions, some subjects may not have understood certain questions, did not accurately recall their health history or were hesitant to answer some questions (or answer truthfully). However, we believe that, on average, since the questionnaire was answered privately and anonymously, the responses are likely to be correct.
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207
Acknowledgments Dr. Merchant was supported by a National Institutes of Health training grant through the Division of Infectious Diseases, Brown Medical School, The Miriam Hospital, from the National Institute on Drug Abuse, 5 T32 DA13911-02, and a Resident Research Training Award from the SAEM. We thank Lyn Robillard, Kristina Casadei, and Robert Andreozzi for their data entry expertise, volunteers from the Society for Clinical Research for Undergraduates at the Brown University for their help in survey administration and CVS for providing a $500 donation of gift cards. Appendix A. Contraceptive knowledge questions Who may take birth control pills? 1. 2. 3. 4. 5. 6.
5 Women and men 5 Only pregnant women 5 Only women whose family members have had cancer 5 Only women who have cancer 5 Only women 5 I am not really sure
Why do you think someone might use birth control pills? 1. 2. 3. 4.
5 5 5 5
To prevent infection To prevent cancer To prevent pregnancy I am not really sure
How often should someone take birth control pills? 1. 2. 3. 4. 5.
5 5 5 5 5
Once a day Once a week Once a month Once every 6 months I am not really sure
If a woman has had vaginal sexual intercourse with a man (without using birth control), can she take birth control pills AFTERWARDS to prevent pregnancy? 1. 2. 3.
5 Yes 5 No 5 I am not really sure
Who should use condoms during sexual intercourse? 1. 2. 3. 4. 5.
5 Only men who have an infection/sexually transmitted disease 5 Only women who have an infection/sexually transmitted disease 5 Only pregnant women 5 Anyone may use condoms during sexual intercourse 5 I am not really sure
207
Why do you think someone should use a condom during sexual intercourse? 1. 2. 3. 4. 5.
5 To remain a virgin 5 To keep from getting breast cancer 5 To prevent someone from having sexual intercourse 5 To prevent pregnancy and/or infection (sexually transmitted disease) 5 I am not really sure
How often should most people use condoms during sexual intercourse in order to prevent pregnancy and infection? 1. 2. 3. 4. 5.
5 5 5 5 5
Never Sometimes/occasionally Most of the time All the time I am not really sure
References [1] Henshaw SK, Unintended pregnancy in the United States. Fam Plann Perspect 1998;30(1):24–29, 46. [2] Mosher WD, Martinez GM, Chandra A, Abma JC, Willson SJ. Use of contraception and use of family planning services in the United States: 1982–2002. Adv Data 2004;350:1 – 36. [3] Trussell J, Leveque JA, Koenig JD, et al. The economic value of contraception: a comparison of 15 methods. Am J Public Health 1995;85:494 – 503. [4] Sonnenberg FA, Burkman RT, Hagerty CG, Speroff L, Speroff T. Costs and net health effects of contraceptive methods. Contraception 2004;69:447 – 59. [5] Rhodes KV, Gordon JA, Lowe RA. Preventive care in the emergency department: Part I. Clinical preventive services — are they relevant to emergency medicine? Society for Academic Emergency Medicine Public Health and Education Task Force Preventive Services Work Group. Acad Emerg Med 2000;7:1036 – 41. [6] Babcock Irvin C, Wyer PC, Gerson LW. Preventive care in the emergency department: Part II. Clinical preventive services—an emergency medicine evidence-based review. Society for Academic Emergency Medicine Public Health and Education Task Force Preventive Services Work Group. Acad Emerg Med 2000;7(9): 1042 – 54. [7] Ranney ML, Gee EM, Merchant RC. Non-prescription availability of emergency contraception in the United States: history, current status, controversies, and impact on emergency medicine practice. Ann Emerg Med 2006;47(5):461 – 71. [8] McCaig LF, Burt CW. National Hospital Ambulatory Medical Care Survey: 2003 emergency department summary. Adv Data 2005;358: 1 – 38. [9] Cunningham PJ, Clancy CM, Cohen JW, Wilets M. The use of hospital emergency departments for nonurgent health problems: a national perspective. Med Care Res Rev 1995;52:453 – 74. [10] O’Brien GM, Stein MD, Zierler S, Shapiro M, O’sullivan P, Woolard R. Use of the ED as a regular source of care: associated factors beyond lack of health insurance. Ann Emerg Med 1997; 30(3):286 – 91. [11] Liu T, Sayre MR, Carleton SC. Emergency medical care: types, trends, and factors related to nonurgent visits. Acad Emerg Med 1999;6: 1147 – 52.
Original research article
Contraceptive usage, knowledge and correlates of usage among female emergency department patients Roland C. Merchant a,b,4, Jennifer A. Damergisa, Erin M. Geea, Beth C. Bock c, Bruce M. Becker a,b, Melissa A. Clark b a
Department of Emergency Medicine, Brown Medical School, Rhode Island Hospital, Providence, RI 02903, USA b Department of Community Health, Brown Medical School, Providence, RI 02912, USA c Department of Psychiatry and Human Behavior, Center for Behavioral and Preventive Medicine, Brown University, Providence, RI 02903, USA Received 23 February 2006; revised 13 March 2006; accepted 17 March 2006
Abstract Objectives: For female emergency department (ED) patients, we sought to assess the prevalence of contraceptive usage as well as the extent of contraceptive knowledge and to determine if demographic and sexual health history factors, comprehension of contraceptive methods and moral/religious opinions on contraception were associated with current usage of birth control pills (BCPs), prior usage of emergency contraception (EC) and frequency of condom usage. Methods: English-speaking female ED patients aged between 18 and 55 years at a northeastern United States urban ED were surveyed on their usage, comprehension and opinions regarding BCPs, EC and condoms. Results: Of the 539 respondents (64.6% were aged V 35 years), most were White (63.1%), single (42.5%), Catholic (48.4%) and privately insured (55.3%). Among the 223 women at pregnancy risk [not currently pregnant, not using any form of nonsurgical birth control (except condoms) and with no prior tubal ligation or hysterectomy], about 25% were using BCPs, fewer than 10% had used EC and almost 40% never used condoms. Most women displayed good knowledge about BCPs and condoms but poor understanding about EC. In multivariate logistic regression analyses, current BCP usage among women at risk of pregnancy was associated with younger age [odds ratio (OR) = 0.54; 95% confidence interval (CI) = 0.37–0.79], private insurance (OR = 2.52; 95% CI = 1.30–4.86) and recent intercourse (OR = 1.61; 95% CI = 1.19–2.18). Among women at risk of pregnancy, those who had an abortion (OR = 2.56; 95% CI = 1.17–5.61) and those who displayed greater EC knowledge (OR = 3.23; 95% CI = 1.50–6.95) had greater odds of having used EC. Among all women, more frequent condom usage was associated with being younger (OR = 0.57; 95% CI = 0.46–0.70), having never been married (OR = 0.44; 95% CI = 0.28–0.68) and not having intercourse recently (OR = 0.79; 95% CI = 0.64–0.98). Conclusions: A high percentage of female ED patients (41.4%) were at risk of pregnancy. Demographic and sexual history factors can help identify women who might benefit from receiving referrals or education on contraceptive measures. D 2006 Elsevier Inc. All rights reserved. Keywords: Oral contraceptives; Emergency contraception; Emergency medicine; Contraception; Condoms; Emergency services; Women’s health
1. Introduction The percentage of women in the United States at risk of unintended pregnancy may be increasing. According to estimates from the 1995 National Surveys of Family Growth (NSFG), 49% of pregnancies in the United States during The results of this study were presented at the 2005 American College of Emergency Physicians Scientific Assembly in Washington, DC, on September 27, 2005. 4 Corresponding author. Department of Emergency Medicine, Brown Medical School, Rhode Island Hospital, Providence, RI 02903, USA. Tel.: +1 401 444 5109; fax: +1 401 444 4307. E-mail address: [email protected] (R.C. Merchant ). 0010-7824/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.contraception.2006.03.012
that year were unintended and 54% ended in an induced abortion [1]. According to the 1995 and 2002 NSFGs, the percentage of women aged between 15 and 44 years having sexual intercourse who were not using any form of contraception in fact increased from 5.4% in 1995 to 7.4% in 2002 [2]. If contraception was more widely used, rates of unintended pregnancy would decrease, as would the societal costs of these pregnancies from their adverse health and economic effects [3,4]. The Public Health and Education Task Force (PHTF) of the Society for Academic Emergency Medicine (SAEM) evaluated preventive interventions for the emergency department (ED) in 2000 [5,6]. Its review appraised the
202
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207
applicability of preventive health interventions that were recommended by the U.S. Preventive Services Task Force (USPSTF) for other health care settings to ED practice. Contraceptive counseling was considered as a candidate preventive health measure by the USPSTF but was not reviewed by the SAEM PHTF. This absence was due in large part to the paucity of studies on the usage and effectiveness of reproductive health interventions in the ED. In fact, published studies from EDs on patient contraceptive usage and knowledge have concerned only the use of emergency contraception (EC) [7]. To our knowledge, there has been no published study on birth control pill (BCP) and condom usage among patients in this setting. Nevertheless, the ED is an appropriate place to consider preventive reproductive health interventions for women because it may be the only source of primary health care services to women who do not have to access to them elsewhere. In the United States, there were 61 million patient visits to EDs in 2003, of which 46.7% were by women aged between 15 and 44 years [8]. For many women, the ED is their principal source of medical care [9 –11]. As a consequence, the ED may serve as the only means by which women can access or receive referrals for preventive reproductive health care. Knowing the patterns of and factors related to contraceptive use may assist emergency medicine clinicians in identifying patients who could benefit from using contraception, addressing knowledge deficits and providing referrals or systems to facilitate referrals for contraceptive use. We sought to determine how frequently women visiting the ED are using contraceptive methods (BCPs, EC and condoms) and, consequently, how many are at risk of pregnancy. Determining the frequency of contraceptive usage and the prevalence of women visiting the ED at risk of pregnancy would help highlight the needs of this group. Furthermore, along with pregnancy testing, knowing the prevalence and demographics of women at risk would help clinicians identify these women so as to avoid prescribing or using potentially dangerous or harmful medications, diagnostic tests or procedures. In addition, we aimed to determine how well women know the purpose, applicability and suggested usage of BCPs, EC and condoms. We believe that understanding the extent of patient knowledge on these topics would help clinicians plan for relevant educational interventions when needed. We further sought to examine if contraceptive use (usage of BCPs, EC and condoms) was associated with demographic or sexual health history factors; with comprehension of the purpose, applicability and recommended usage of these contraceptive measures; and with moral and religious opinions on contraception. Our goal was to find potentially modifiable factors that may lead to ED-based preventive health programs to improve contraceptive use among ED patients. For BCPs and EC, we were particularly interested in women who have sex with men and who are at risk of pregnancy [those not currently pregnant, not using a
nonsurgical form of birth control (except condoms) and with no prior history of tubal ligation or hysterectomy]. For condom usage, we considered all respondents who reported having sex with men. It is our hope that the study results may contribute to future evaluations of preventive health measures for the ED.
2. Materials and methods 2.1. Study design This was a prospective survey of women aged between 18 and 55 years seeking treatment in an ED. Participants completed a self-administered, written as well as anonymous multiple-choice and best-answer questionnaire that we drafted. The hospital institutional review board approved the study as exempt from review and written informed consent. 2.2. Study setting We conducted the study at an urban academic ED that serves as a tertiary referral center for a catchment area of approximately 1.5 million people. In fiscal year 2002, there were 73,672 adult patient visits to this ED, 36,102 (49.0%) of which were by women. Of these women, 24,087 (66.7%) were aged between 18 and 55 years. Table 1 shows the demographic profile of the women included in our survey. 2.3. Survey content and development We chose three topics for our survey — BCPs, EC and condoms — and created questions patterned on the styles used in other surveys on similar topics in other settings that we reviewed. Our survey focused on the survey participants’ demographics and health history as well as their use, knowledge and opinions regarding contraceptive usage. The written multiple-choice, best-answer questions asked participants to identify the applicability, purpose and recommended usage of birth control, EC and condoms (Appendix A). The reading level of the survey was at a Flesch–Kincaid grade of 5.4 (Microsoft Word, Redmond, WA, USA). We created a draft questionnaire and pilot tested it in June 2002 on a convenience sample of 20 women from the ambulatory care section of the ED. Using a standardized script, we interviewed these women after they completed the questionnaire. Our interview consisted of a brief cognitive assessment of randomly selected questions from the questionnaire, a review of potentially sensitive questions and an analysis of mismarked responses. We also solicited feedback on participants’ opinions, impressions and reactions to the questionnaire. Each of these women received $20 for participating in this pilot test. We revised our survey based on their comments and our observations. The final survey used in this analysis included 28 multiple-choice questions: 6 demographic questions; 11 health history questions; 3 questions on BCP knowledge, 3 on condom knowledge and 1 on EC knowledge; and 4 agree-or-disagree
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207 Table 1 Demographic profile of the survey respondents (N = 539)
Age group (years) 18–25 26–35 36–45 46–55 Race American Indian/ Native American Asian/Pacific Islander Black/African American Hispanic/Latina White Biracial/Multiracial Other Marital status Single Married Divorced/Separated Male partner Widowed Religious identity Buddhist Catholic Jewish Muslim Protestant None Other Health insurance None Government Private
All respondents (N = 539)
Women at pregnancy risk (n = 316)
Women not at pregnancy risk (n = 223)
36.4 28.2 21.5 13.9
48.4 27.5 16.2 7.9
19.3 29.2 29.2 22.4
4.7
4.4
4.9
0.9 9.8 15.0 63.1 5.6 0.9
1.6 8.9 14.5 62.7 6.6 1.3
0.0 11.2 15.7 63.7 4.0 0.5
42.5 28.4 14.6 13.4 1.1
51.9 25.3 9.2 13.3 0.3
29.1 32.7 22.4 13.4 2.2
1.3 48.4 1.1 1.3 26.7 18.1 3.0
1.3 48.7 1.6 1.0 25.0 19.9 2.5
1.3 48.0 0.5 1.8 29.1 15.7 3.6
13.9 30.8 55.3
15.5 21.8 62.7
11.7 43.5 44.8
All values are expressed as percentages.
questions on moral/religious opinions regarding BCP and EC usage. 2.4. Selection of participants We limited the sample to women aged between 18 and 55 years because we hypothesized that women older than 55 years might have different reproductive health concerns (e.g., menopause) and be less likely to use the contraceptive measures of interest. Women were excluded if they could not read or write in English; were being evaluated in the critical care, psychiatric section or alcohol abuse holding areas of the ED; were not awake; or could not physically complete the form. All other women seeking treatment in the ED during the study hours were asked to complete a questionnaire. Each participant who completed the questionnaire received a $2 gift card to a local pharmacy. Research assistants and trained volunteers administered the survey in three 2-month blocks that included sixty-six 8-h shifts in July–August 2002, thirty-two 4 -h shifts in October–November 2002 and thirty-two 4-h shifts in April– May 2003. We purposely sampled three seasons to avoid unknown time-dependent factors that might influence the
203
pattern of ED visits and affect the results. The data collection shifts were scheduled to reflect the time-dependent influx of patients to our ED. Data collection was structured such that 25% of the shifts were between 7 and 11 a.m., 50% were between 11 a.m. and 11 p.m. and 25% were between 11 p.m. and 7 a.m. The ED’s daily volume was essentially the same each day of the week. To reflect this pattern of patient visits, we sampled participants so that each day of the week was represented equally. 2.5. Methods of measurement The primary outcome was use of BCPs, EC and condoms. Independent variables included demographic and sexual health history factors as well as knowledge and opinions about the contraceptive methods. 2.6. Primary data analysis Two trained personnel independently entered the data into separate Epi Info 2002 (Centers for Disease Control and Prevention, Atlanta, GA, USA) databases. These two databases were compared for every entry, and errors were corrected to create a final database. We transferred the final database into Stata 8.2 (Stata, College Station, TX, USA) using Stat/Transfer 6 (Circle Systems, Seattle, WA, USA). We tabulated the frequency of responses to each of the questions and calculated 95% confidence intervals (CIs) for point estimates. We conducted our analyses on respondents for whom we had no missing data on the variables of interest; we then repeated our analyses for all respondents (whether they answered all questions or not) and observed similar results. For BCP and EC usage, we restricted our analyses to women at risk of pregnancy. For condom usage, we conducted the analysis on all women who have sex with men. To examine the correlates of contraceptive usage, we performed univariate logistic regression analyses using the demographic and sexual health history variables as well as responses to the contraceptive knowledge and opinion questions as the variables of interest. Correlates statistically significant at the a = .05 level were placed in multivariate logistic regression models to assess their association with contraceptive usage. For condom usage, we performed ordinal logistic regression in which increasing frequency of condom usage was the outcome. Odds ratios (ORs) and corresponding 95% CIs were calculated from these models.
3. Results 3.1. Respondent demography As shown in Table 1, of the 539 respondents, most were aged 35 years or younger (64.6%), White (63.1%), single (42.5%), Catholic (48.4%) and privately insured (55.3%). Of the 539 respondents, 316 (53.6%) were considered to be at risk of pregnancy [not currently pregnant, not using any nonsurgical form of birth control (except for condoms) and
204
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207
sexual and pregnancy history except for history of having been pregnant.
Table 2 Sexual and contraceptive history All Women at Women not at respondents pregnancy risk pregnancy risk (N = 539) (n = 316) (n = 223) Contraceptive history Currently using BCPs 18.9 Using other birth 11.1 control methods Prior bilateral 23.4 tubal ligation Prior hysterectomy 9.7 Ever used EC 8.4 Ever had abortion 25.1 Sexual history Last intercourse with a man V 1 week 49.0 V 1 month 22.8 V 6 months 9.8 V 1 year 4.8 N 1 year 9.1 Never 4.5 Frequency of condom use No sex/no sex with men 7.2 Never 45.8 Sometimes 16.9 Often 13.9 Always 16.1 Pregnancy history Currently pregnanta 2.4 Ever pregnant 71.1 No. of times pregnant 0 28.9 1 13.9 2 17.4 3 18.0 z4 21.7
25.9 0.0
9.0 26.9
0.0
56.5
0.0 9.8 23.4
23.3 6.3 27.4
48.4 21.5 9.8 4.8 8.5 7.0
49.8 24.7 9.9 4.9 9.9 0.9
8.9 38.9 17.4 16.5 18.3
4.9 55.6 16.1 10.3 13.0
0.0 56.7
5.8 91.5
43.3 14.9 14.2 13.9 13.6
8.5 12.6 22.0 23.8 33.2
All values are expressed as percentages. a Of all respondents, 3.9% were unsure if they were pregnant; of women at pregnancy risk, 6.0% were pregnant.
not having undergone a hysterectomy or bilateral tubal ligation]. Among women at risk of pregnancy, most were aged 35 years or younger (75.9%), White (62.7%), single (51.9%), Catholic (48.7%) and privately insured (62.7%). 3.2. Respondent health history/contraceptive use history Fewer than 35% of all respondents (Table 2) were currently using any type of nonsurgical birth control method (except for condoms). Of all respondents, 33.1% had undergone surgical sterilization (hysterectomy or bilateral tubal ligation). Three times more women had undergone an abortion (25.1%) than had ever used EC (8.4%). Although almost half of all respondents (49%) had had intercourse within the past week, most did not use condoms during intercourse (16.1% reported always using them). More than 70% of the women reported being pregnant at some time in the past. Although few women were currently pregnant by self-report, 3.9% of all participants and 6.0% of women at risk of pregnancy were unsure if they were currently pregnant. The women at risk of pregnancy were similar to all other participants in their contraceptive,
3.3. Contraception knowledge and opinions Overall knowledge regarding BCPs and condoms was high (Table 3). A higher percentage of participants correctly answered all condom questions than the birth control questions. Participants were 3.2 and 3.0 times more likely to correctly answer all of the condom and BCP knowledge questions, respectively, than the EC knowledge question (although there was only one EC knowledge question). Fewer than 10% of all respondents believed that BCPs (whether as EC or as daily usage) caused abortions, were morally wrong or were against their personal religious beliefs. The subset of women at risk of pregnancy answered the knowledge and opinions questions similarly to the entire sample of participants but tended to answer the knowledge questions correctly more often than women who were not at risk of pregnancy. 3.4. Correlates of contraceptive usage The outcomes of the logistic regression analyses that compared factors associated with current usage of BCPs, ever having used EC and frequency of condom usage among women at risk of pregnancy are depicted in Table 4. In a univariate analysis of current BCP use among women at risk of pregnancy, contraceptive use was not associated with current or prior marriage or having a male partner (partner status), religious preference, increasing number of pregnancies, ever having had an abortion or believing that BCPs are against the respondent’s religious or moral beliefs or cause an abortion. Current usage of BCPs was associated with being younger, being White, having private health insurance, having more recent intercourse, never having been Table 3 Correct responses to survey questions
BCP knowledge Who Why How often All correct EC knowledge Condom knowledge Who Why How often All correct BCP opinions Cause abortions Morally wrong Against religion EC opinions Use after intercourse prevents pregnancy
All respondents (N = 539)
Women at pregnancy risk (n = 316)
Women not at pregnancy risk (n = 223)
89.6 98.3 95.0 85.9 28.3
91.8 99.1 96.2 88.9 32.3
86.6 97.3 93.3 81.6 22.9
95.9 98.3 95.9 91.8
96.2 98.7 96.8 93.4
95.5 97.8 94.6 89.7
8.5 6.1 9.7
9.2 5.1 8.9
7.6 7.6 10.8
8.9
11.1
5.8
All values are expressed as percentages.
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207
205
Table 4 Factors associated with contraceptive usage Women at risk of pregnancy BCP usage (n = 316)
Age White race Partner status Divorced/Widowed Married/Male partner Religion Protestant Catholic Other Religious worship frequency Protestant Catholic Private health insurance Recent intercourse Ever pregnant No. of pregnancies Prior abortion Contraceptive knowledge BCP beliefs Causes abortions Morally wrong Against religious beliefs EC beliefs Causes abortions
All respondents Condom usage (n = 500)a
EC usage (n = 316)
Univariate
Multivariate
Univariate
Univariate
Multivariate
0.58 (0.42–0.79) 2.06 (1.17–3.60)
0.54 (0.37–0.79) 1.67 (0.90–3.10)
0.89 (0.62–1.29) 0.52 (0.25–1.10)
0.51 (0.43–0.61) 0.86 (0.61–1.20)
0.57 (0.46–0.70)
0.59 (0.23–1.52) 0.67 (0.39–1.15)
1.73 (0.59–5.11) 0.69 (0.30–1.60)
0.48 (0.30–0.78) 0.23 (0.16–0.33)
1.17 (0.66–2.06) 0.44 (0.28–0.68)
0.34 0.67 0.31 0.78 0.75 0.78 2.23 1.52 1.88 0.84 1.18 2.83
2.14 1.59 1.64 1.19 1.20 1.14 0.61 0.98 0.69 1.10 2.65 3.31
0.57 0.57 0.56 0.96 0.97 1.07 0.83 0.78 3.12 0.99 1.55 1.03
0.74 (0.35–1.55) 0.86 (0.51–1.44) 0.81 (0.51–1.29)
(0.16–0.75) (0.36–1.26) (0.08–1.16) (0.62–0.98) (0.60–0.94) (0.65–0.96) (1.27–3.94) (1.14–2.02) (1.13–3.12) (0.63–1.11) (0.66–2.10) (1.02–7.83)
0.84 0.83 0.80 2.52 1.61 1.14
(0.65–1.06) (0.64–1.06) (0.64–1.01) (1.30–4.86) (1.19–2.18) (0.60–2.16)
2.40 (0.80–7.16)
(0.64–7.17) (0.51–5.0) (0.28–9.69) (0.85–1.65) (0.95–1.52) (0.90–1.45) (0.29–1.27) (0.67–1.43) (0.32–1.50) (0.78–1.56) (1.23–5.71) (1.55–7.07)
1.10 (0.47–2.58) 0.18 (0.02–1.39) 0.95 (0.39–2.32)
1.54 (0.50–4.76) 2.24 (0.60–8.34) 1.11 (0.32–3.93)
1.46 (0.61–3.47)
0.61 (0.22–1.71)
Multivariate
2.56 (1.17–5.61) 3.23 (1.50–6.95)
(0.35–0.92) (0.37–0.88) (0.28–1.15) (0.82–1.11) (0.86–1.10) (0.95–1.20) (0.60–1.16) (0.64–0.94) (2.16–4.50) (0.85–1.14) (1.06–2.26) (0.63–1.70)
0.79 (0.64–0.98) 1.52 (0.96–2.41) 1.05 (0.69–1.63)
All values are expressed as OR (95% CI). a Excludes women who reported not having sexual intercourse.
pregnant, attending worship less frequently (for all worship groups and Catholics and Protestants separately) and having greater BCP knowledge. In multivariate analyses, current BCP usage was associated with younger age, private health insurance and more recent intercourse. Stated in terms of preventive health needs, lack of current BCP usage was associated with being older, not having private health insurance and not having had sex recently. Current BCP usage was not associated with currently or ever having been married, having a male partner (partner status), religious preference, increasing number of pregnancies or ever having had an abortion. BCP usage was also not associated in either univariate or multivariate analysis with believing that BCPs are against the respondent’s religious/moral beliefs or cause an abortion. For prior usage of EC among women at risk of pregnancy, in the univariate and multivariate analyses, ever having had an abortion and answering the EC knowledge question correctly were associated with ever having used EC. Conversely, stated in preventive health needs terms, not having ever used EC was associated with not understanding its function. No other factor was associated with EC use in either the univariate or multivariate analysis. Among all respondents who reported having intercourse with men, more frequent condom usage was associated in the univariate analysis with being younger, never having been married, not having a male partner, being Catholic or
Protestant, having had intercourse less recently, never having been pregnant and having had an abortion. In the univariate analysis, more frequent condom usage was not associated with race, private health insurance and condom usage knowledge. In the multivariate analysis, more frequent condom usage was associated with being younger, not being married or having a male partner and not having intercourse recently. Again, in preventive health needs terms, less frequent condom usage was associated with being older, being married or not having a male partner and having intercourse recently.
4. Discussion Contraceptive usage is a relevant topic to explore among ED patients given that 67% of women visiting our ED were of childbearing age (i.e., 18–55 years). A high percentage (59%) of ED patients eligible to participate were at risk of pregnancy and had recent sexual intercourse. This risk of pregnancy is highlighted by the relatively low percentages of women using BCPs, EC or condoms. Among women at risk of pregnancy, approximately 25% were currently using BCPs, fewer than 10% had ever used EC and almost 40% never used condoms during intercourse. Of course, some of these women may want to become pregnant, but given the high rates of unwanted pregnancy in the United States and
206
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207
increasing rates of failure to use contraception, these results verify that many women visiting the ED are at risk of pregnancy, whether desired or not. These low percentages of contraceptive usage serve as a reminder to emergency medicine practitioners that a high percentage of women of childbearing age are at risk of pregnancy. This has important implications on practitioners’ choice of medications (e.g., teratogens) and diagnostic procedures (e.g., radiography). The results of this study also suggest that an ED visit may be an excellent opportunity to provide women seeking treatment with contraceptive services or referrals. EC definitely appears to be underused in this population. Women were three times more likely to have had an abortion than to have ever used EC. This finding was true for women at risk and those not at risk of pregnancy. Despite their known efficacy in preventing pregnancy and sexually transmitted diseases, condoms were not frequently used by most women at risk of pregnancy. Less than one fifth of the respondents reported always using condoms. This was of particular concern given that nearly 70% of women at risk of pregnancy had sexual intercourse with a man at least once within the prior month. Again, some of these women may have desired pregnancy, but given the high rates of infrequent condom usage among all women sampled, it is likely that women who did not desire pregnancy were not using condoms. Participants had reasonably good knowledge of BCPs and condom use but exhibited a significant deficit in understanding the concept of EC. The results were essentially the same for women regardless of pregnancy risk status. Few women expressed moral or religious objections to BCP and EC usage, which suggests that one of the real challenges in improving women’s understanding about reproductive preventive health choices (at least in the population of women seeking treatment in the ED) is increasing their understanding about the purpose and usage of EC. We identified several important factors associated with usage of BCPs, EC and condoms in our sample. Younger women with private health insurance who had had sexual intercourse with a man recently were more likely to be currently using BCPs. Women who ever had an abortion and exhibited a greater understanding of the concept EC were more likely to have ever used EC. More frequent condom use was observed among younger unmarried/no male partner women who had not had sexual intercourse recently. Surprisingly, women who had had sexual intercourse recently were less likely to have used condoms recently. If these women were not seeking to be pregnant, then they are at substantial risk of pregnancy and sexually transmitted diseases. It is possible that women who have sexual intercourse less often are more likely to protect their reproductive health status and to prevent pregnancy. It is important to emphasize that contraceptive knowledge, race, religious identity, pregnancy history, frequency of religious worship and beliefs and opinions regarding
contraception were not associated with contraceptive usage, with the exception of EC knowledge and usage. As such, except for EC, lack of knowledge does not appear to be the primary barrier to contraceptive use for women seeking treatment in an ED. The lack of association of a history of prior pregnancy with current contraceptive use highlights an increased probability of subsequent pregnancies, whether desired or not. In summary, our study results indicate that a large proportion of women visiting the ED are at risk of pregnancy. Many of these women, whether desiring pregnancy or not, are having intercourse and do not use contraception and therefore might not know they are pregnant, which has implications on the selection of medications and diagnostic studies. Knowledge of BCPs and condoms is high; however, EC knowledge among these women is poor. Our study results pinpoint groups of women who might benefit from expanded educational and outreach programs in an ED that might serve to improve their use of reproductive health choices. For example, women without health insurance could receive educational pamphlets outlining their reproductive health options (e.g., BCPs) and be provided with a list of local resources that offer these services; older women could receive targeted educational messages informing them of the option of using BCPs and condoms; and women who have sexual intercourse frequently could be provided with educational pamphlets emphasizing the risks of pregnancy and sexually transmitted diseases. It is our hope that this study can contribute to efforts regarding the assessment of women’s reproductive preventive health needs in the ED. 4.1. Limitations Given our sampling techniques (ED volume matched, capture of different periods and equal weighting of days of the week), we believe that we have a representative sample of those eligible to participate in the study. However, our sample is from a single ED and involved English-speaking women who were predominantly from certain demographic groups (White, Catholic, single, privately insured). Therefore, the findings may not be applicable to other settings with more diverse populations. We are hopeful that our questionnaire will be adapted for other settings, cultures and languages to corroborate our results; however, we also recognize that with larger samples or different populations, the results may change. Likewise, other modeling techniques may produce different results and ORs that did not reach significance may become significant. It is also possible that, despite our use of cognitive assessments and screening of questions, some subjects may not have understood certain questions, did not accurately recall their health history or were hesitant to answer some questions (or answer truthfully). However, we believe that, on average, since the questionnaire was answered privately and anonymously, the responses are likely to be correct.
R.C. Merchant et al. / Contraception 74 (2006) 201 – 207
Acknowledgments Dr. Merchant was supported by a National Institutes of Health training grant through the Division of Infectious Diseases, Brown Medical School, The Miriam Hospital, from the National Institute on Drug Abuse, 5 T32 DA13911-02, and a Resident Research Training Award from the SAEM. We thank Lyn Robillard, Kristina Casadei, and Robert Andreozzi for their data entry expertise, volunteers from the Society for Clinical Research for Undergraduates at the Brown University for their help in survey administration and CVS for providing a $500 donation of gift cards. Appendix A. Contraceptive knowledge questions Who may take birth control pills? 1. 2. 3. 4. 5. 6.
5 Women and men 5 Only pregnant women 5 Only women whose family members have had cancer 5 Only women who have cancer 5 Only women 5 I am not really sure
Why do you think someone might use birth control pills? 1. 2. 3. 4.
5 5 5 5
To prevent infection To prevent cancer To prevent pregnancy I am not really sure
How often should someone take birth control pills? 1. 2. 3. 4. 5.
5 5 5 5 5
Once a day Once a week Once a month Once every 6 months I am not really sure
If a woman has had vaginal sexual intercourse with a man (without using birth control), can she take birth control pills AFTERWARDS to prevent pregnancy? 1. 2. 3.
5 Yes 5 No 5 I am not really sure
Who should use condoms during sexual intercourse? 1. 2. 3. 4. 5.
5 Only men who have an infection/sexually transmitted disease 5 Only women who have an infection/sexually transmitted disease 5 Only pregnant women 5 Anyone may use condoms during sexual intercourse 5 I am not really sure
207
Why do you think someone should use a condom during sexual intercourse? 1. 2. 3. 4. 5.
5 To remain a virgin 5 To keep from getting breast cancer 5 To prevent someone from having sexual intercourse 5 To prevent pregnancy and/or infection (sexually transmitted disease) 5 I am not really sure
How often should most people use condoms during sexual intercourse in order to prevent pregnancy and infection? 1. 2. 3. 4. 5.
5 5 5 5 5
Never Sometimes/occasionally Most of the time All the time I am not really sure
References [1] Henshaw SK, Unintended pregnancy in the United States. Fam Plann Perspect 1998;30(1):24–29, 46. [2] Mosher WD, Martinez GM, Chandra A, Abma JC, Willson SJ. Use of contraception and use of family planning services in the United States: 1982–2002. Adv Data 2004;350:1 – 36. [3] Trussell J, Leveque JA, Koenig JD, et al. The economic value of contraception: a comparison of 15 methods. Am J Public Health 1995;85:494 – 503. [4] Sonnenberg FA, Burkman RT, Hagerty CG, Speroff L, Speroff T. Costs and net health effects of contraceptive methods. Contraception 2004;69:447 – 59. [5] Rhodes KV, Gordon JA, Lowe RA. Preventive care in the emergency department: Part I. Clinical preventive services — are they relevant to emergency medicine? Society for Academic Emergency Medicine Public Health and Education Task Force Preventive Services Work Group. Acad Emerg Med 2000;7:1036 – 41. [6] Babcock Irvin C, Wyer PC, Gerson LW. Preventive care in the emergency department: Part II. Clinical preventive services—an emergency medicine evidence-based review. Society for Academic Emergency Medicine Public Health and Education Task Force Preventive Services Work Group. Acad Emerg Med 2000;7(9): 1042 – 54. [7] Ranney ML, Gee EM, Merchant RC. Non-prescription availability of emergency contraception in the United States: history, current status, controversies, and impact on emergency medicine practice. Ann Emerg Med 2006;47(5):461 – 71. [8] McCaig LF, Burt CW. National Hospital Ambulatory Medical Care Survey: 2003 emergency department summary. Adv Data 2005;358: 1 – 38. [9] Cunningham PJ, Clancy CM, Cohen JW, Wilets M. The use of hospital emergency departments for nonurgent health problems: a national perspective. Med Care Res Rev 1995;52:453 – 74. [10] O’Brien GM, Stein MD, Zierler S, Shapiro M, O’sullivan P, Woolard R. Use of the ED as a regular source of care: associated factors beyond lack of health insurance. Ann Emerg Med 1997; 30(3):286 – 91. [11] Liu T, Sayre MR, Carleton SC. Emergency medical care: types, trends, and factors related to nonurgent visits. Acad Emerg Med 1999;6: 1147 – 52.