- Email: [email protected]
Personalized contraceptive assistance and uptake of long-acting, reversible contraceptives by postpartum women: a randomized, controlled trial
Contraception 88 (2013) 45 – 51
Original research article
Personalized contraceptive assistance and uptake of long-acting, reversible contraceptives by postpartum women: a randomized, controlled trial☆ Katharine B. Simmons a,⁎, Alison B. Edelman a , Hong Li b , Keenan E. Yanit a , Jeffrey T. Jensen a b
a Department of Obstetrics & Gynecology, Oregon Health & Science University, Portland, OR 97239, USA Department of Biostatistics and Bioinformatics Resource, School of Public Health & Preventive Medicine, Oregon Health & Science University, Portland, OR 97239, USA Received 23 August 2012; revised 26 October 2012; accepted 31 October 2012
Abstract Background: Many women who intend to use long-acting, reversible contraceptives (LARCs) postpartum do not follow through with initiating use. The objectives of this study were to determine whether support from a contraceptive personal assistant could increase the uptake of LARCs by 3 months postpartum, and to identify risk factors for nonuptake of LARCs among women who planned LARC use. Study Design: This is a randomized, controlled trial of 50 low-income postpartum women who desired LARC. The intervention group received telephone contact from a personal assistant who provided contraception education, facilitation of insurance coverage, appointment scheduling and assistance with childcare and transportation. The control group received routine follow up. Women were surveyed immediately and 3 months postpartum regarding contraceptive use and anticipated barriers to LARC use. Results: A similar proportion of women in both groups received LARC [control 16/24 (67%), intervention 18/25 (72%), p= .76]. More primiparous (86.4%) than multiparous (55.5%) women obtained LARC (p=.04). In addition, women with more prenatal visits were more likely to have initiated LARC (odds ratio, 95% confidence interval for each increased visit: 1.50, 1.15–1.96). No other demographic factors were related to LARC uptake. Conclusions: Providing telephone assistance to help navigate barriers did not increase postpartum uptake of LARCs. A personal history of clinic visit no-shows and/or infrequent prenatal visits were related to poor uptake of LARCs postpartum. © 2013 Elsevier Inc. All rights reserved. Keywords: Postpartum contraception; Long-acting reversible contraception; Intrauterine device; Family planning
1. Introduction New parenthood involves shifts in lifestyle and priorities, and many women are motivated to avoid a repeat pregnancy during this time. The prenatal and postpartum periods are thus an important time to counsel women on contraception. Unintended pregnancies commonly occur after an interpregnancy interval of less than 12 months [1], with 20% of deliveries to multiparous women occurring within a year of a prior delivery [2]. Unintended pregnancies occur disproportionately in poor women [3], and young women who carry unintended pregnancies to term have decreased ☆ Financial support: The authors acknowledge grant support from the Society of Family Planning. ⁎ Corresponding author. Fax: +1 503 494 5680. E-mail address: [email protected] (K.B. Simmons).
0010-7824/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.contraception.2012.10.037
educational achievement and lower socioeconomic status than women who delay childbearing [4]. Likewise, interpregnancy interval under 1 year is an independent risk factor for preterm delivery and neonatal death [5]. These data provide important motivation for health care providers and society to ensure that women are given appropriate resources to avoid an unwanted or a sooner-than-wanted pregnancy. More effective use of contraception in the postpartum period could reduce unintended pregnancies and improve obstetric outcomes. To be effective in preventing pregnancy, contraception must be used correctly and reliably. An Emory University survey of 250 randomly selected postpartum women from around the United States found that they desired a contraceptive method that is easy to use, provides longterm protection, does not require frequent trips to a pharmacy or physician, and has no impact on breastfeeding [6]. Long-
46
K.B. Simmons et al. / Contraception 88 (2013) 45–51
acting reversible contraceptives (LARCs) including intrauterine devices (IUDs) and contraceptive implants meet these criteria and are ideal to reduce unintended pregnancies [7,8]. LARCs offer the most effective form of reversible contraception, with failure rates over the first year of 0.001%, 0.14% and 0.7%, respectively, for the etonogestrel contraceptive implant, levonorgestrel intrauterine system (IUS) and copper IUD [9]. Unfortunately, women who choose LARC methods for postpartum contraception often face barriers to initiation, including inadequate or incorrect counseling, financial hurdles, and difficulty getting to postpartum clinic visits for placement of the LARC device. Because of these and other unforeseen barriers, studies estimate that as many as 40–75% of women who are planning to have a postpartum IUD will not have their IUD placed [10–12]. Chen et al. [12] examined whether immediate postplacental IUD placement could lead to increased rates of IUD use by 6 months postpartum, compared to routine 6-week postpartum placement. Of the women assigned to receive a postplacental IUD, 98% had the device placed as planned, and 84% of these women were still using the device at 6 months postpartum. A similarly high number (90%) of women randomized to 6-week postpartum placement successfully had the device placed, and 76% of these women were using the device at 6 months postpartum [12]. Although usage rates at 6 months were not significantly different for these two groups, a third group emerged. Those women who were initially enrolled but who did not receive an IUD through the study (due to cesarean delivery, infection, hemorrhage or precipitous delivery) and were told to follow up with their provider for an IUD had a much lower rate of use (26.8%) at 6 months postpartum [12]. A possible explanation for this marked discrepancy is that participation in a clinical trial provides additional support that reduces the barriers to uptake present with usual care. If so, providing comprehensive contraception support services without a research component might assist patients to follow through on their expressed desire to use a LARC method during the postpartum period. The primary objective of the present project was to increase the use of LARCs among low-income postpartum women at an academic hospital. We hypothesized that logistical barriers prevent use of LARCs among women who specifically express a desire for these methods at hospital discharge, and that offering personalized assistance designed to anticipate and address these barriers, using currently available social and medical assistance, will increase the rates of LARC usage at 3 months postpartum. This project tested the effectiveness of a contraceptive “personal assistant” (PA), who provided personalized contraception education, navigated insurance coverage, provided referral to agencies with low-cost or sliding scale services, scheduled appointments for LARC placement, and helped arrange childcare and transportation. The secondary objective was to identify
risk factors for nonuptake of LARCs within the postpartum period.
2. Materials and methods A randomized, controlled trial was conducted at Oregon Health & Science University (OHSU) in Portland, OR, from May 2011 to February 2012. The institutional review board at OHSU approved the study protocol. Fifty postpartum women with public insurance (Oregon Medicaid) were recruited on the OHSU postpartum units and signed an informed, written consent. All subjects meeting inclusion criteria during the enrollment period were identified by daily review of the postpartum unit census. We focused on the Medicaid population because women on Medicaid are more likely than women with private insurance to have gaps in contraceptive use [13]. Subjects were 18–38 years old, English or Spanish speaking, received prenatal care through the OHSU resident clinic and had a stated intention to use a LARC (etonogestrel implant, levonorgestrel IUS and copper IUD). All women received routine contraceptive counseling during prenatal care and had a contraceptive plan in place prior to delivery. Women were excluded if they did not meet inclusion criteria or were incarcerated at the time of delivery due to their inability to return for follow-up visits. LARC devices were not provided as part of the study, in order to ensure that the observed effect was related to the intervention only and not the removal of financial barriers. All participants were asked a baseline questionnaire of 27 multiple-choice questions regarding basic demographics, future pregnancy plans, contraceptive history, work and childcare plans and access to resources like regular medical care and transportation. Medical records were reviewed to obtain information on prenatal care visits. Subjects were randomized into two groups: control (standard care, n= 25) and intervention (standard care plus PA, n= 25). Randomization was generated by a computer-generated randomization scheme in a 1:1 ratio. Women in the control group had routine postpartum follow up including scheduling their own 6-week postpartum appointment. This meant that women desiring LARC had to request it at the time of appointment scheduling in order for the LARC to be available for placement at the 6-week postpartum appointment. This control group was believed to most accurately represent the postpartum care received by women at our center. Women in the intervention group had additional assistance from one of two contraceptive PAs. Both of these assistants had experience with contraception counseling; one was a resident in obstetrics/gynecology (KBS), and the second was a fourth year medical student planning to pursue ob/gyn (KEY). Two weeks after hospital discharge, the PA telephoned each subject to confirm her interest in LARC and to answer questions and provide counseling as needed.
K.B. Simmons et al. / Contraception 88 (2013) 45–51
During this call, insurance coverage and cost were explained. For patients without insurance coverage, the PA referred subjects to assistance programs that provide full coverage for a LARC method. Further assistance was provided by scheduling the postpartum visit for LARC placement. Three days prior to the scheduled appointment, the PA made a second call to the subject to confirm the appointment time and availability of childcare and transportation for the medical appointment. Assistance with transportation to the appointment was arranged as needed using medical or public transportation. Subjects who did not keep their appointments received additional phone calls for rescheduling, up to three times, and additional targeted contraceptive counseling, as needed. Placement of the LARC for all subjects occurred at their routine postpartum clinic visit between 4 and 6 weeks postpartum. This was not a research visit. Providers were not aware of subject involvement in the study and were free to provide additional contraceptive counseling as needed. At 3 months postpartum, all subjects were contacted by phone to answer a follow-up questionnaire regarding current contraceptive method, other methods used since delivery, and ease of obtaining LARC. Subjects who had indicated that they would like to use a LARC but who were not using a LARC at 3 months postpartum were asked additional questions to specifically evaluate why they changed their plan. Subjects were compensated US$50 each at the completion of their follow-up phone call. Based on prior reports, we estimated that approximately 25% of subjects receiving standard care (control group) would have LARC placed by 3 months postpartum [11,12]. Power calculations for group size were based on a goal of
75% LARC use by 3 months in the intervention group as has been previously reported [12]. A sample size of 25 in each group would provide 90% power to detect this 50% difference at an alpha level of .05. All analyses were performed using Statistical Analysis System (SAS) software (SAS version 9.2, SAS Institute). Differences between groups were compared using t test and chi-square/Fisher's Exact test according to an intention-to-treat basis. The associations between LARC placement and prenatal visit history were determined using Cochran–Armitage trend test. Two-sided p values were presented, and p values less than .05 were considered as statistically significant.
3. Results Project enrollment started on 5/15/2011 and was completed on 11/1/2011. Flow of study subjects is documented in Fig. 1. During this time period, there were 56 eligible patients at OHSU out of 302 resident patient deliveries. A total of 51 subjects signed informed consent. One subject was excluded after enrollment but before randomization upon learning that she was incarcerated (exclusion criteria). Fifty subjects were randomized. One subject in the control group was excluded after randomization because she was not planning a LARC device (incorrectly stated in her medical chart). Within the intervention group, 23 of 25 (92%) subjects had at least one of the intervention phone calls completed. All were analyzed as intention-to-treat within the intervention group. Three-month follow-up data on
302 resident patient deliveries 56 eligible patients 1 excluded (incarcerated, exclusion criteria) 2 declined 3 not approached (missed) 50 women randomized Control (n=25)
Routine care LARC placement 4-6 wks PP Excluded for not planning LARC (n=1)
Lost to follow-up (n=2) (analyzed as not placed)
3-month phone survey (n=22)
Control subjects analyzed (n=24)
47
Intervention (n=25)
2-week counseling call Appointment scheduling Appointment reminder call LARC placement 4-6 wks PP
3-month phone survey (n=25)
Intervention subjects analyzed (n=25)
Fig. 1. Flow of patients through the study.
48
K.B. Simmons et al. / Contraception 88 (2013) 45–51
LARC use were available for all 25 subjects through phone call or chart review. Within the control group, 22 of 24 (92%) subjects had 3-month follow-up data on LARC use available. 3.1. Comparison between intervention and control groups There was no significant difference in demographics between the groups (Table 1). However, there were slightly more partnered and married women in the intervention group and slightly more primiparous women in the intervention group. There was no significant difference between groups for type of LARC chosen, mode of delivery, availability of childcare, moving or having a phone disconnected within 3 months of delivery, owning a car, timing to initiate prenatal care, total of prenatal visits and number of no-show visits during prenatal care. The primary outcome of this study was placement of a LARC within 3 months postpartum. Within the control group, 16 out of 24 women (67%) had LARC placed as planned within 3 months. Eighteen of 25 women in the intervention group had LARC placed (72%). There was no significant difference in placement of LARC between the groups (p=.76) (Table 2). The percentages of LARC users in each group who selected each of the three LARC methods are described in Table 2.
Table 1 Baseline characteristics of study participants
Age (range, 18–38) 25 years or older Marital status Single/Divorced/Separated Partnered/Cohabitating/Married Education High school (HS) or less Some college or higher Race Caucasian Non-White Parity Primiparous Multiparous Mode of delivery Spontaneous vaginal delivery Cesarean section Vaginal birth after cesarean Childcare available (yes) Phone disconnected during study (yes) Own a car (yes) Moved in 3 months postpartum (yes) Initial prenatal visit (weeks) Total prenatal visits (range, 4–16) Prenatal no-shows
Control (n= 24)
Intervention (n=25)
25.5±4.6 16 (47.1)
25±4.9 7 (46.7)
9 (37.5) 15 (62.5)
4 (16.0) 21 (84.0)
12 (50.0) 12 (50.0)
12 (48.0) 13 (52.0)
16 (69.6) 7 (30.4)
18 (75.0) 6 (25.0)
9 (37.5) 15 (62.5)
13 (52.0) 12 (48.0)
18 (24.0) 5 (20.8) 1 (4.2) 18 (75.0) 7 (29.2) 17 (70.8) 8 (33.3) 13.1±6.4 10.8±3.1 0.9±1.3
12 (48.0) 11 (44.0) 2 (8.0) 21 (84.0) 9 (36.0) 17 (68.0) 6 (24.0) 12.6±5.9 10.2±3.1 1.7±2.3
Values are expressed as mean±standard deviation for continuous variables or number of subjects (column percentage).
Table 2 LARC placement by 3 months postpartum in control and intervention groups
LARC placed Type of LARC Levonorgestrel-IUS Copper T380 Etonogestrel implant LARC not placed Lost to follow up
Control (n= 24)
Intervention (n=25)
p
16 (66.7)
18 (72.0)
.76
9 (56.2) 5 (31.2) 2 (12.5) 6 (25.0) 2 (8.3)
12 (66.7) 1 (5.5) 5 (27.8) 7 (28.0) 0
Values are expressed as number of subjects (percentage). P values greater than .05 for all comparisons.
3.2. Factors associated with LARC uptake Women had multiple reasons for not following through with LARC placement at 3 months. Within the control group, two still desired LARC but had been unable to come to the clinic and reported using condoms or abstinence. Three control subjects changed their mind and were using other methods of contraception including condoms, depot medroxyprogesterone acetate and oral contraceptive pills. Two subjects were lost to follow up, and one refused to say why she was not using LARC. Within the intervention group, of the seven women who did not have LARC placed, five cited inability to get to the clinic as their reason and were using condoms instead, despite multiple attempts by the PA to help get them to the clinic. One subject changed her mind and was using abstinence instead, and one subject's partner had disapproved of the LARC plan and was using condoms instead. Demographic factors including age, owning a car, access to childcare, marital status, education level, ethnicity, having a phone disconnected during the postpartum period and prior LARC use were not related to LARC uptake when analyzed independent of study group (Table 3). However, parity was related to LARC uptake, with a greater portion of primiparous subjects (86.4%) following through with LARC placement than multiparous subjects (55.5%, p= .03). History of consistent prenatal care, measured by total number of prenatal visits and number of no-show visits during prenatal care, was also related to uptake. Total number of prenatal visits was positively associated with LARC placement (p for trend .007), while number of noshow visits during prenatal care was negatively associated with LARC placement (p for trend .02). Each prenatal visit increased the likelihood of a woman having LARC placed [odds ratio (OR), 95% confidence interval (CI) for each increased visit: 1.50, 1.15–1.96] (Fig. 2). Building on these findings of improved LARC uptake among women with better prenatal clinic attendance, we performed a secondary analysis looking at the effect of the intervention by clinic attendance history. The PA intervention did not significantly improve LARC placement in either subgroup. Placement of LARC increased slightly from 85%
K.B. Simmons et al. / Contraception 88 (2013) 45–51
Age, years Under 25 25 or older Private car Yes No Childcare available Yes No Marital status Single/Separated Partnered/Married Education HS or less Some college or more Race Caucasian Other Parity Primiparous Multiparous No-shows 2 or more Less than 2 Prenatal visits Each unit increase 10 or less More than 10 Prior LARC use Yes No
LARC placed, n (%)
Exact OR (95% CI)
p value
25.1±4.8 16 (69.6) 18 (69.2)
1.02 (0.25–4.12) 1
24 (70.6) 10 (66.7)
1.20 (0.25–5.17) 1
1.0
28 (71.8) 6 (60.0)
1.68 (0.29–8.80) 1
.72
9 (69.2) 25 (69.4)
1 1.01 (0.19–4.68)
1.0
18 (75.0) 16 (64.0)
1.67 (0.42–7.09) 1
.6
24 (70.6) 8 (61.5)
1.49 (0.30–6.81) 1
.79
19 (86.4) 15 (55.5)
4.90 (1.06–32.0) 1
.04
8 (50.0) 26 (78.8)
1 3.60 (0.85–16.2)
.09
10 (45.4) 24 (88.9)
1.5 (1.15–1.96) 1 9.10 (1.91–61.2)
.003 .003
8 (66.7) 26 (70.2)
1 1.18 (0.21–5.62)
1.0
.77 1.0
Values are expressed as mean±standard deviation for continuous variables or number of subjects (row percentage).
(11 of 13) in control subjects to 93% (13 of 14) in intervention subjects who had N 10 prenatal visits (p=.6) (Fig. 3). Women with fewer prenatal visits showed no change in the incidence of LARC placement with the intervention (45% vs. 45%). Similarly, when subjects were divided by number of no-shows, placement increased slightly from 72% in control subjects with fewer than two no-shows during prenatal care to 86% in intervention subjects. Fifty percent of subjects with more than two no-shows received LARC in both control and intervention groups.
4. Discussion
The rate of IUD uptake in the control group (67%) is higher than has been previously reported in similar studies [11,12]. This may be related to baseline characteristics of our selected population. All subjects received their care through the OHSU resident clinic, where contraception is routinely discussed throughout pregnancy, and LARC methods are strongly emphasized. Therefore, our subjects may have had a higher baseline interest in LARC than the general postpartum population. At the time of enrollment, all but one subject rated their certainty of using a LARC 7 or higher on a 10point scale (1 = not at all certain, 10 = completely certain), with 33/50 subjects stating that they were 100% certain of their decision to use a LARC. In addition, the clinic has some built-in mechanisms for ensuring follow up. At the time of hospital discharge, all patients were given information on scheduling clinic appointments for postpartum follow up. In this study, eight of the intervention subjects already had appointments scheduled prior to the first phone call by the PA. All of the control subjects scheduled their follow up independently of the study. The benefit of a phone intervention was limited for some participants. This study involved a population with limited resources, some with unstable living and income situations. Reaching subjects by phone proved difficult in some cases. In the intervention group, 9 of 25 subjects had their phone disconnected within 3 months of delivery, and 6 of 25 subjects moved or became homeless. There were five women in the intervention group who still desired LARC but had been unable to get to the clinic despite the assistance of the PA phone call intervention. These subjects were all counseled on free medical transportation resources, confirmed their appointments the day prior and denied other specific barriers to coming to the clinic. Most of these women eventually stated that their lives were too busy to
1.0 0.9 Probability of LARC uptake
Table 3 Demographic characteristics related to uptake of LARC independent of study group
0.8 0.7 0.6 0.5 0.4 0.3 Control OR (95% CI): 1.47 (1.01, 2.16) Intervention OR (95% CI): 1.55 (1.06, 2.26) p (Study*Total-PNV)=.85
0.2 0.1
Our strategy of providing personalized assistance for postpartum contraception did not increase LARC placement by 3 months postpartum in this population. LARC placement rates in the control group (67%) were higher than anticipated, which may have masked the benefit of the intervention in this small sample size. Placement rate in the intervention group (72%) was close to the goal of 75%.
49
2
4
6
8 10 12 Total prenatal visits
14
16
18
Fig. 2. Association between number of prenatal visits and probability of LARC placement by study group. Total number of prenatal visits was positively associated with LARC placement in both groups (women with more prenatal visits were more likely to obtain LARC). No difference between control and intervention was observed across number of prenatal visits.
50
K.B. Simmons et al. / Contraception 88 (2013) 45–51
100%
100% 86% 80%
Control
80%
72%
60%
93% 85%
Intervention 50%
60% 45%
40%
40%
20%
20% 0%
0% No-shows <2
No-shows >=2
PNV >10
PNV <=10
Fig. 3. Effect of the study intervention when analyzed by reliability factors, including number of prenatal no-shows (left) and number of prenatal visits (right). Y axis shows percentage of subjects who had LARC placed. The PA intervention did not significantly improve LARC placement in either subgroup.
come to the doctor's office for birth control, even with childcare and transportation provided. In addition, despite asking about transportation, many subjects did not disclose limited resources. Several subjects declined assistance with medical transportation, stating that they had a ride available, and then did not arrive to appointments. Unwillingness of subjects to disclose transportation limitations was a barrier to the success of the study intervention. A weakness of the study design is the inequality in contact with the health care team between study groups. It is not possible to say what specific services by the PA were most useful for subjects. Counseling, scheduling or just frequent telephone contact with a health care provider may have benefitted subjects, and combining all of these services in one group obscures the benefit of any one component of PA services. A control group using a noncontraceptive PA to provide increased contact with the health care system without specifically addressing contraception would have controlled for this inequality but would have made the control group less representative of routine care. The prenatal clinic attendance variables were better predictors of postpartum clinic follow up and LARC placement than any other variable we examined (Fig. 2). Women with a history of poor clinic attendance, as measured by fewer prenatal visits and more no-show visits during prenatal care, were less likely to keep postpartum clinic appointments for birth control, even if their stated desire to use a LARC was high at the time of delivery. This finding may be useful to identify women on the labor ward who are at higher risk for nonuptake of LARC. The phone-based intervention in this study did not affect LARC uptake in this subgroup of patients. These women may be ideal candidates for other strategies to increase LARC use, such as postplacental IUD or contraceptive implant placement. Baseline rates of IUD expulsion following interval insertion are between 2% and 10% during the first year of use [7,14]. Postplacental insertion appears to be associated with a higher rate of expulsion than interval placement, but reports are extremely variable, with 8–24% of postplacental devices expulsed by 6 months [12,15,16]. Even with this higher expulsion rate, ongoing IUD use at 6 months
postpartum appears to be equivalent for postplacental and routine 6-week placement. This is likely explained by a higher initial placement rate, replacement of some expelled devices in the postplacental group, and lower overall placement rates with 6-week placement [12]. Postplacental insertion of IUDs for all uninsured women is not costeffective from a hospital perspective but does reduce state costs for subsequent pregnancies [17]. However, postplacental placement may be appropriate and cost-effective in a select population of women who are at highest risk of not returning for delayed placement. Immediate postpartum placement of the contraceptive implant does not carry the risk of expulsion and may offer another mechanism for improving LARC use among highrisk postpartum women. Inpatient postpartum placement of the etonogestrel contraceptive implant is associated with excellent continuation at 6 and 12 months (96.9% and 86.3%, respectively, among a population of adolescent women) and low repeat pregnancy rates within 1 year (2.6%) [18]. We were not able to offer immediate postpartum contraceptive implant placement in this study due to the implant not being available on our hospital formulary. Changing hospital policies to support inpatient postpartum implant placement would give patients and providers another option for initiating reliable contraception before hospital discharge [18]. The preliminary finding in this study that prenatal care patterns may predict postpartum follow up for LARC placement should be confirmed in future studies as a possible mechanism to improve LARC use rates among high-risk populations. Strategies such as postplacental IUD placement and immediate implant insertion may improve access to reliable contraception among women who have difficulty returning for postpartum care.
Acknowledgments This project was funded by a grant from the Society of Family Planning. Special thanks to the OHSU Women's Health Research Unit for study coordination.
K.B. Simmons et al. / Contraception 88 (2013) 45–51
References [1] Jackson RA, Schwarz EB, Freedman L, Darney P. Advance supply of emergency contraception: effect on use and usual contraception – a randomized trial. Obstet Gynecol 2003;102:8–6. [2] Day T, Raker C, Boardman L. Factors associated with the provision of antenatal contraceptive counseling. Contraception 2008;78:294–9. [3] Finer LB, Zolna MR. Unintended pregnancy in the United States: incidence and disparities, 2006. Contraception 2011;84:478–85. [4] Delhlendorf C, Rodriguez MI, Levy K, et al. Disparities in family planning. AJOG 2010;202:214–8. [5] Smith G, Pell J, Dobbie R. Interpregnancy interval and risk of preterm birth and neonatal death: retrospective cohort study. BMJ 2003;327:313. [6] Cwiak C, Gellasch T, Zieman M. Peripartum contraceptive attitudes and practices. Contraception 2004;70:383–6. [7] Jensen JT, Nelson AL, Costales AC. Subject and clinician experience with the levonorgestrel-releasing intrauterine system. Contraception 2008;77:22–9. [8] Hohmann H, Creinin MD. The contraceptive implant. Clin Obstet Gynecol 2007;50:907–17. [9] Winner B, Piepert JF, Zhao Q, et al. Effectiveness of long-acting reversible contraception. N Engl J Med 2012;366:1998–2007. [10] Ogburn JA, Espey E, Stonehocker J. Barriers to intrauterine device insertion in postpartum women. Contraception 2005;72:426–9.
51
[11] Engin-Ustun Y, Ustin Y, Cetin F, et al. Effect of postpartum counseling on postpartum contraceptive use. Arch Gynecol Obstet 2007;275:429–32. [12] Chen BA, Reeves MF, Hayes JL, Hohmann HL, Perriera LK, Creinin MD. Postplacental or delayed insertion of the levonorgestrel intrauterine device after vaginal delivery: a randomized controlled trial. Obstet Gynecol 2010;116:1079–87. [13] Frost JJ, Singh S, Finer LB. Factors associated with contraceptive use and nonuse, United States, 2004. Perspect Sex Reprod Health 2007;39:90–9. [14] Allen RH, Goldberg AB, Grimes DA. Expanding access to intrauterine contraception. Am J Obstet Gynecol 2009;201:456.e1–5. [15] Hayes JL, Cwiak C, Goedken P, Zieman M. A pilot clinical trial of ultrasound-guided postplacental insertion of a levonorgestrel intrauterine device. Contraception 2007;76:292–6. [16] Grimes DA, Lopez LM, Schulz KF, Van Vliet H, Stanwood NL. Immediate post-partum insertion of intrauterine devices. Cochrane Database Syst Rev 2010:1–15, http://dx.doi.org/10.1002/14651858. CD003036.pub2 Issue 5. Art. No.: CD003036. [17] Rodriguez MI, Caughey AB, Edelman AE, et al. Cost-benefit analysis of state- and hospital-funded postpartum intrauterine contraception at a university hospital for recent immigrants to the United States. Contraception 2010;81:304–8. [18] Tocce KM, Sheeder JL, Teal SB. Rapid repeat pregnancy in adolescents: do immediate postpartum contraceptive implants make a difference? Am J Obstet Gynecol 2012;206:481.e1–7.
Original research article
Personalized contraceptive assistance and uptake of long-acting, reversible contraceptives by postpartum women: a randomized, controlled trial☆ Katharine B. Simmons a,⁎, Alison B. Edelman a , Hong Li b , Keenan E. Yanit a , Jeffrey T. Jensen a b
a Department of Obstetrics & Gynecology, Oregon Health & Science University, Portland, OR 97239, USA Department of Biostatistics and Bioinformatics Resource, School of Public Health & Preventive Medicine, Oregon Health & Science University, Portland, OR 97239, USA Received 23 August 2012; revised 26 October 2012; accepted 31 October 2012
Abstract Background: Many women who intend to use long-acting, reversible contraceptives (LARCs) postpartum do not follow through with initiating use. The objectives of this study were to determine whether support from a contraceptive personal assistant could increase the uptake of LARCs by 3 months postpartum, and to identify risk factors for nonuptake of LARCs among women who planned LARC use. Study Design: This is a randomized, controlled trial of 50 low-income postpartum women who desired LARC. The intervention group received telephone contact from a personal assistant who provided contraception education, facilitation of insurance coverage, appointment scheduling and assistance with childcare and transportation. The control group received routine follow up. Women were surveyed immediately and 3 months postpartum regarding contraceptive use and anticipated barriers to LARC use. Results: A similar proportion of women in both groups received LARC [control 16/24 (67%), intervention 18/25 (72%), p= .76]. More primiparous (86.4%) than multiparous (55.5%) women obtained LARC (p=.04). In addition, women with more prenatal visits were more likely to have initiated LARC (odds ratio, 95% confidence interval for each increased visit: 1.50, 1.15–1.96). No other demographic factors were related to LARC uptake. Conclusions: Providing telephone assistance to help navigate barriers did not increase postpartum uptake of LARCs. A personal history of clinic visit no-shows and/or infrequent prenatal visits were related to poor uptake of LARCs postpartum. © 2013 Elsevier Inc. All rights reserved. Keywords: Postpartum contraception; Long-acting reversible contraception; Intrauterine device; Family planning
1. Introduction New parenthood involves shifts in lifestyle and priorities, and many women are motivated to avoid a repeat pregnancy during this time. The prenatal and postpartum periods are thus an important time to counsel women on contraception. Unintended pregnancies commonly occur after an interpregnancy interval of less than 12 months [1], with 20% of deliveries to multiparous women occurring within a year of a prior delivery [2]. Unintended pregnancies occur disproportionately in poor women [3], and young women who carry unintended pregnancies to term have decreased ☆ Financial support: The authors acknowledge grant support from the Society of Family Planning. ⁎ Corresponding author. Fax: +1 503 494 5680. E-mail address: [email protected] (K.B. Simmons).
0010-7824/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.contraception.2012.10.037
educational achievement and lower socioeconomic status than women who delay childbearing [4]. Likewise, interpregnancy interval under 1 year is an independent risk factor for preterm delivery and neonatal death [5]. These data provide important motivation for health care providers and society to ensure that women are given appropriate resources to avoid an unwanted or a sooner-than-wanted pregnancy. More effective use of contraception in the postpartum period could reduce unintended pregnancies and improve obstetric outcomes. To be effective in preventing pregnancy, contraception must be used correctly and reliably. An Emory University survey of 250 randomly selected postpartum women from around the United States found that they desired a contraceptive method that is easy to use, provides longterm protection, does not require frequent trips to a pharmacy or physician, and has no impact on breastfeeding [6]. Long-
46
K.B. Simmons et al. / Contraception 88 (2013) 45–51
acting reversible contraceptives (LARCs) including intrauterine devices (IUDs) and contraceptive implants meet these criteria and are ideal to reduce unintended pregnancies [7,8]. LARCs offer the most effective form of reversible contraception, with failure rates over the first year of 0.001%, 0.14% and 0.7%, respectively, for the etonogestrel contraceptive implant, levonorgestrel intrauterine system (IUS) and copper IUD [9]. Unfortunately, women who choose LARC methods for postpartum contraception often face barriers to initiation, including inadequate or incorrect counseling, financial hurdles, and difficulty getting to postpartum clinic visits for placement of the LARC device. Because of these and other unforeseen barriers, studies estimate that as many as 40–75% of women who are planning to have a postpartum IUD will not have their IUD placed [10–12]. Chen et al. [12] examined whether immediate postplacental IUD placement could lead to increased rates of IUD use by 6 months postpartum, compared to routine 6-week postpartum placement. Of the women assigned to receive a postplacental IUD, 98% had the device placed as planned, and 84% of these women were still using the device at 6 months postpartum. A similarly high number (90%) of women randomized to 6-week postpartum placement successfully had the device placed, and 76% of these women were using the device at 6 months postpartum [12]. Although usage rates at 6 months were not significantly different for these two groups, a third group emerged. Those women who were initially enrolled but who did not receive an IUD through the study (due to cesarean delivery, infection, hemorrhage or precipitous delivery) and were told to follow up with their provider for an IUD had a much lower rate of use (26.8%) at 6 months postpartum [12]. A possible explanation for this marked discrepancy is that participation in a clinical trial provides additional support that reduces the barriers to uptake present with usual care. If so, providing comprehensive contraception support services without a research component might assist patients to follow through on their expressed desire to use a LARC method during the postpartum period. The primary objective of the present project was to increase the use of LARCs among low-income postpartum women at an academic hospital. We hypothesized that logistical barriers prevent use of LARCs among women who specifically express a desire for these methods at hospital discharge, and that offering personalized assistance designed to anticipate and address these barriers, using currently available social and medical assistance, will increase the rates of LARC usage at 3 months postpartum. This project tested the effectiveness of a contraceptive “personal assistant” (PA), who provided personalized contraception education, navigated insurance coverage, provided referral to agencies with low-cost or sliding scale services, scheduled appointments for LARC placement, and helped arrange childcare and transportation. The secondary objective was to identify
risk factors for nonuptake of LARCs within the postpartum period.
2. Materials and methods A randomized, controlled trial was conducted at Oregon Health & Science University (OHSU) in Portland, OR, from May 2011 to February 2012. The institutional review board at OHSU approved the study protocol. Fifty postpartum women with public insurance (Oregon Medicaid) were recruited on the OHSU postpartum units and signed an informed, written consent. All subjects meeting inclusion criteria during the enrollment period were identified by daily review of the postpartum unit census. We focused on the Medicaid population because women on Medicaid are more likely than women with private insurance to have gaps in contraceptive use [13]. Subjects were 18–38 years old, English or Spanish speaking, received prenatal care through the OHSU resident clinic and had a stated intention to use a LARC (etonogestrel implant, levonorgestrel IUS and copper IUD). All women received routine contraceptive counseling during prenatal care and had a contraceptive plan in place prior to delivery. Women were excluded if they did not meet inclusion criteria or were incarcerated at the time of delivery due to their inability to return for follow-up visits. LARC devices were not provided as part of the study, in order to ensure that the observed effect was related to the intervention only and not the removal of financial barriers. All participants were asked a baseline questionnaire of 27 multiple-choice questions regarding basic demographics, future pregnancy plans, contraceptive history, work and childcare plans and access to resources like regular medical care and transportation. Medical records were reviewed to obtain information on prenatal care visits. Subjects were randomized into two groups: control (standard care, n= 25) and intervention (standard care plus PA, n= 25). Randomization was generated by a computer-generated randomization scheme in a 1:1 ratio. Women in the control group had routine postpartum follow up including scheduling their own 6-week postpartum appointment. This meant that women desiring LARC had to request it at the time of appointment scheduling in order for the LARC to be available for placement at the 6-week postpartum appointment. This control group was believed to most accurately represent the postpartum care received by women at our center. Women in the intervention group had additional assistance from one of two contraceptive PAs. Both of these assistants had experience with contraception counseling; one was a resident in obstetrics/gynecology (KBS), and the second was a fourth year medical student planning to pursue ob/gyn (KEY). Two weeks after hospital discharge, the PA telephoned each subject to confirm her interest in LARC and to answer questions and provide counseling as needed.
K.B. Simmons et al. / Contraception 88 (2013) 45–51
During this call, insurance coverage and cost were explained. For patients without insurance coverage, the PA referred subjects to assistance programs that provide full coverage for a LARC method. Further assistance was provided by scheduling the postpartum visit for LARC placement. Three days prior to the scheduled appointment, the PA made a second call to the subject to confirm the appointment time and availability of childcare and transportation for the medical appointment. Assistance with transportation to the appointment was arranged as needed using medical or public transportation. Subjects who did not keep their appointments received additional phone calls for rescheduling, up to three times, and additional targeted contraceptive counseling, as needed. Placement of the LARC for all subjects occurred at their routine postpartum clinic visit between 4 and 6 weeks postpartum. This was not a research visit. Providers were not aware of subject involvement in the study and were free to provide additional contraceptive counseling as needed. At 3 months postpartum, all subjects were contacted by phone to answer a follow-up questionnaire regarding current contraceptive method, other methods used since delivery, and ease of obtaining LARC. Subjects who had indicated that they would like to use a LARC but who were not using a LARC at 3 months postpartum were asked additional questions to specifically evaluate why they changed their plan. Subjects were compensated US$50 each at the completion of their follow-up phone call. Based on prior reports, we estimated that approximately 25% of subjects receiving standard care (control group) would have LARC placed by 3 months postpartum [11,12]. Power calculations for group size were based on a goal of
75% LARC use by 3 months in the intervention group as has been previously reported [12]. A sample size of 25 in each group would provide 90% power to detect this 50% difference at an alpha level of .05. All analyses were performed using Statistical Analysis System (SAS) software (SAS version 9.2, SAS Institute). Differences between groups were compared using t test and chi-square/Fisher's Exact test according to an intention-to-treat basis. The associations between LARC placement and prenatal visit history were determined using Cochran–Armitage trend test. Two-sided p values were presented, and p values less than .05 were considered as statistically significant.
3. Results Project enrollment started on 5/15/2011 and was completed on 11/1/2011. Flow of study subjects is documented in Fig. 1. During this time period, there were 56 eligible patients at OHSU out of 302 resident patient deliveries. A total of 51 subjects signed informed consent. One subject was excluded after enrollment but before randomization upon learning that she was incarcerated (exclusion criteria). Fifty subjects were randomized. One subject in the control group was excluded after randomization because she was not planning a LARC device (incorrectly stated in her medical chart). Within the intervention group, 23 of 25 (92%) subjects had at least one of the intervention phone calls completed. All were analyzed as intention-to-treat within the intervention group. Three-month follow-up data on
302 resident patient deliveries 56 eligible patients 1 excluded (incarcerated, exclusion criteria) 2 declined 3 not approached (missed) 50 women randomized Control (n=25)
Routine care LARC placement 4-6 wks PP Excluded for not planning LARC (n=1)
Lost to follow-up (n=2) (analyzed as not placed)
3-month phone survey (n=22)
Control subjects analyzed (n=24)
47
Intervention (n=25)
2-week counseling call Appointment scheduling Appointment reminder call LARC placement 4-6 wks PP
3-month phone survey (n=25)
Intervention subjects analyzed (n=25)
Fig. 1. Flow of patients through the study.
48
K.B. Simmons et al. / Contraception 88 (2013) 45–51
LARC use were available for all 25 subjects through phone call or chart review. Within the control group, 22 of 24 (92%) subjects had 3-month follow-up data on LARC use available. 3.1. Comparison between intervention and control groups There was no significant difference in demographics between the groups (Table 1). However, there were slightly more partnered and married women in the intervention group and slightly more primiparous women in the intervention group. There was no significant difference between groups for type of LARC chosen, mode of delivery, availability of childcare, moving or having a phone disconnected within 3 months of delivery, owning a car, timing to initiate prenatal care, total of prenatal visits and number of no-show visits during prenatal care. The primary outcome of this study was placement of a LARC within 3 months postpartum. Within the control group, 16 out of 24 women (67%) had LARC placed as planned within 3 months. Eighteen of 25 women in the intervention group had LARC placed (72%). There was no significant difference in placement of LARC between the groups (p=.76) (Table 2). The percentages of LARC users in each group who selected each of the three LARC methods are described in Table 2.
Table 1 Baseline characteristics of study participants
Age (range, 18–38) 25 years or older Marital status Single/Divorced/Separated Partnered/Cohabitating/Married Education High school (HS) or less Some college or higher Race Caucasian Non-White Parity Primiparous Multiparous Mode of delivery Spontaneous vaginal delivery Cesarean section Vaginal birth after cesarean Childcare available (yes) Phone disconnected during study (yes) Own a car (yes) Moved in 3 months postpartum (yes) Initial prenatal visit (weeks) Total prenatal visits (range, 4–16) Prenatal no-shows
Control (n= 24)
Intervention (n=25)
25.5±4.6 16 (47.1)
25±4.9 7 (46.7)
9 (37.5) 15 (62.5)
4 (16.0) 21 (84.0)
12 (50.0) 12 (50.0)
12 (48.0) 13 (52.0)
16 (69.6) 7 (30.4)
18 (75.0) 6 (25.0)
9 (37.5) 15 (62.5)
13 (52.0) 12 (48.0)
18 (24.0) 5 (20.8) 1 (4.2) 18 (75.0) 7 (29.2) 17 (70.8) 8 (33.3) 13.1±6.4 10.8±3.1 0.9±1.3
12 (48.0) 11 (44.0) 2 (8.0) 21 (84.0) 9 (36.0) 17 (68.0) 6 (24.0) 12.6±5.9 10.2±3.1 1.7±2.3
Values are expressed as mean±standard deviation for continuous variables or number of subjects (column percentage).
Table 2 LARC placement by 3 months postpartum in control and intervention groups
LARC placed Type of LARC Levonorgestrel-IUS Copper T380 Etonogestrel implant LARC not placed Lost to follow up
Control (n= 24)
Intervention (n=25)
p
16 (66.7)
18 (72.0)
.76
9 (56.2) 5 (31.2) 2 (12.5) 6 (25.0) 2 (8.3)
12 (66.7) 1 (5.5) 5 (27.8) 7 (28.0) 0
Values are expressed as number of subjects (percentage). P values greater than .05 for all comparisons.
3.2. Factors associated with LARC uptake Women had multiple reasons for not following through with LARC placement at 3 months. Within the control group, two still desired LARC but had been unable to come to the clinic and reported using condoms or abstinence. Three control subjects changed their mind and were using other methods of contraception including condoms, depot medroxyprogesterone acetate and oral contraceptive pills. Two subjects were lost to follow up, and one refused to say why she was not using LARC. Within the intervention group, of the seven women who did not have LARC placed, five cited inability to get to the clinic as their reason and were using condoms instead, despite multiple attempts by the PA to help get them to the clinic. One subject changed her mind and was using abstinence instead, and one subject's partner had disapproved of the LARC plan and was using condoms instead. Demographic factors including age, owning a car, access to childcare, marital status, education level, ethnicity, having a phone disconnected during the postpartum period and prior LARC use were not related to LARC uptake when analyzed independent of study group (Table 3). However, parity was related to LARC uptake, with a greater portion of primiparous subjects (86.4%) following through with LARC placement than multiparous subjects (55.5%, p= .03). History of consistent prenatal care, measured by total number of prenatal visits and number of no-show visits during prenatal care, was also related to uptake. Total number of prenatal visits was positively associated with LARC placement (p for trend .007), while number of noshow visits during prenatal care was negatively associated with LARC placement (p for trend .02). Each prenatal visit increased the likelihood of a woman having LARC placed [odds ratio (OR), 95% confidence interval (CI) for each increased visit: 1.50, 1.15–1.96] (Fig. 2). Building on these findings of improved LARC uptake among women with better prenatal clinic attendance, we performed a secondary analysis looking at the effect of the intervention by clinic attendance history. The PA intervention did not significantly improve LARC placement in either subgroup. Placement of LARC increased slightly from 85%
K.B. Simmons et al. / Contraception 88 (2013) 45–51
Age, years Under 25 25 or older Private car Yes No Childcare available Yes No Marital status Single/Separated Partnered/Married Education HS or less Some college or more Race Caucasian Other Parity Primiparous Multiparous No-shows 2 or more Less than 2 Prenatal visits Each unit increase 10 or less More than 10 Prior LARC use Yes No
LARC placed, n (%)
Exact OR (95% CI)
p value
25.1±4.8 16 (69.6) 18 (69.2)
1.02 (0.25–4.12) 1
24 (70.6) 10 (66.7)
1.20 (0.25–5.17) 1
1.0
28 (71.8) 6 (60.0)
1.68 (0.29–8.80) 1
.72
9 (69.2) 25 (69.4)
1 1.01 (0.19–4.68)
1.0
18 (75.0) 16 (64.0)
1.67 (0.42–7.09) 1
.6
24 (70.6) 8 (61.5)
1.49 (0.30–6.81) 1
.79
19 (86.4) 15 (55.5)
4.90 (1.06–32.0) 1
.04
8 (50.0) 26 (78.8)
1 3.60 (0.85–16.2)
.09
10 (45.4) 24 (88.9)
1.5 (1.15–1.96) 1 9.10 (1.91–61.2)
.003 .003
8 (66.7) 26 (70.2)
1 1.18 (0.21–5.62)
1.0
.77 1.0
Values are expressed as mean±standard deviation for continuous variables or number of subjects (row percentage).
(11 of 13) in control subjects to 93% (13 of 14) in intervention subjects who had N 10 prenatal visits (p=.6) (Fig. 3). Women with fewer prenatal visits showed no change in the incidence of LARC placement with the intervention (45% vs. 45%). Similarly, when subjects were divided by number of no-shows, placement increased slightly from 72% in control subjects with fewer than two no-shows during prenatal care to 86% in intervention subjects. Fifty percent of subjects with more than two no-shows received LARC in both control and intervention groups.
4. Discussion
The rate of IUD uptake in the control group (67%) is higher than has been previously reported in similar studies [11,12]. This may be related to baseline characteristics of our selected population. All subjects received their care through the OHSU resident clinic, where contraception is routinely discussed throughout pregnancy, and LARC methods are strongly emphasized. Therefore, our subjects may have had a higher baseline interest in LARC than the general postpartum population. At the time of enrollment, all but one subject rated their certainty of using a LARC 7 or higher on a 10point scale (1 = not at all certain, 10 = completely certain), with 33/50 subjects stating that they were 100% certain of their decision to use a LARC. In addition, the clinic has some built-in mechanisms for ensuring follow up. At the time of hospital discharge, all patients were given information on scheduling clinic appointments for postpartum follow up. In this study, eight of the intervention subjects already had appointments scheduled prior to the first phone call by the PA. All of the control subjects scheduled their follow up independently of the study. The benefit of a phone intervention was limited for some participants. This study involved a population with limited resources, some with unstable living and income situations. Reaching subjects by phone proved difficult in some cases. In the intervention group, 9 of 25 subjects had their phone disconnected within 3 months of delivery, and 6 of 25 subjects moved or became homeless. There were five women in the intervention group who still desired LARC but had been unable to get to the clinic despite the assistance of the PA phone call intervention. These subjects were all counseled on free medical transportation resources, confirmed their appointments the day prior and denied other specific barriers to coming to the clinic. Most of these women eventually stated that their lives were too busy to
1.0 0.9 Probability of LARC uptake
Table 3 Demographic characteristics related to uptake of LARC independent of study group
0.8 0.7 0.6 0.5 0.4 0.3 Control OR (95% CI): 1.47 (1.01, 2.16) Intervention OR (95% CI): 1.55 (1.06, 2.26) p (Study*Total-PNV)=.85
0.2 0.1
Our strategy of providing personalized assistance for postpartum contraception did not increase LARC placement by 3 months postpartum in this population. LARC placement rates in the control group (67%) were higher than anticipated, which may have masked the benefit of the intervention in this small sample size. Placement rate in the intervention group (72%) was close to the goal of 75%.
49
2
4
6
8 10 12 Total prenatal visits
14
16
18
Fig. 2. Association between number of prenatal visits and probability of LARC placement by study group. Total number of prenatal visits was positively associated with LARC placement in both groups (women with more prenatal visits were more likely to obtain LARC). No difference between control and intervention was observed across number of prenatal visits.
50
K.B. Simmons et al. / Contraception 88 (2013) 45–51
100%
100% 86% 80%
Control
80%
72%
60%
93% 85%
Intervention 50%
60% 45%
40%
40%
20%
20% 0%
0% No-shows <2
No-shows >=2
PNV >10
PNV <=10
Fig. 3. Effect of the study intervention when analyzed by reliability factors, including number of prenatal no-shows (left) and number of prenatal visits (right). Y axis shows percentage of subjects who had LARC placed. The PA intervention did not significantly improve LARC placement in either subgroup.
come to the doctor's office for birth control, even with childcare and transportation provided. In addition, despite asking about transportation, many subjects did not disclose limited resources. Several subjects declined assistance with medical transportation, stating that they had a ride available, and then did not arrive to appointments. Unwillingness of subjects to disclose transportation limitations was a barrier to the success of the study intervention. A weakness of the study design is the inequality in contact with the health care team between study groups. It is not possible to say what specific services by the PA were most useful for subjects. Counseling, scheduling or just frequent telephone contact with a health care provider may have benefitted subjects, and combining all of these services in one group obscures the benefit of any one component of PA services. A control group using a noncontraceptive PA to provide increased contact with the health care system without specifically addressing contraception would have controlled for this inequality but would have made the control group less representative of routine care. The prenatal clinic attendance variables were better predictors of postpartum clinic follow up and LARC placement than any other variable we examined (Fig. 2). Women with a history of poor clinic attendance, as measured by fewer prenatal visits and more no-show visits during prenatal care, were less likely to keep postpartum clinic appointments for birth control, even if their stated desire to use a LARC was high at the time of delivery. This finding may be useful to identify women on the labor ward who are at higher risk for nonuptake of LARC. The phone-based intervention in this study did not affect LARC uptake in this subgroup of patients. These women may be ideal candidates for other strategies to increase LARC use, such as postplacental IUD or contraceptive implant placement. Baseline rates of IUD expulsion following interval insertion are between 2% and 10% during the first year of use [7,14]. Postplacental insertion appears to be associated with a higher rate of expulsion than interval placement, but reports are extremely variable, with 8–24% of postplacental devices expulsed by 6 months [12,15,16]. Even with this higher expulsion rate, ongoing IUD use at 6 months
postpartum appears to be equivalent for postplacental and routine 6-week placement. This is likely explained by a higher initial placement rate, replacement of some expelled devices in the postplacental group, and lower overall placement rates with 6-week placement [12]. Postplacental insertion of IUDs for all uninsured women is not costeffective from a hospital perspective but does reduce state costs for subsequent pregnancies [17]. However, postplacental placement may be appropriate and cost-effective in a select population of women who are at highest risk of not returning for delayed placement. Immediate postpartum placement of the contraceptive implant does not carry the risk of expulsion and may offer another mechanism for improving LARC use among highrisk postpartum women. Inpatient postpartum placement of the etonogestrel contraceptive implant is associated with excellent continuation at 6 and 12 months (96.9% and 86.3%, respectively, among a population of adolescent women) and low repeat pregnancy rates within 1 year (2.6%) [18]. We were not able to offer immediate postpartum contraceptive implant placement in this study due to the implant not being available on our hospital formulary. Changing hospital policies to support inpatient postpartum implant placement would give patients and providers another option for initiating reliable contraception before hospital discharge [18]. The preliminary finding in this study that prenatal care patterns may predict postpartum follow up for LARC placement should be confirmed in future studies as a possible mechanism to improve LARC use rates among high-risk populations. Strategies such as postplacental IUD placement and immediate implant insertion may improve access to reliable contraception among women who have difficulty returning for postpartum care.
Acknowledgments This project was funded by a grant from the Society of Family Planning. Special thanks to the OHSU Women's Health Research Unit for study coordination.
K.B. Simmons et al. / Contraception 88 (2013) 45–51
References [1] Jackson RA, Schwarz EB, Freedman L, Darney P. Advance supply of emergency contraception: effect on use and usual contraception – a randomized trial. Obstet Gynecol 2003;102:8–6. [2] Day T, Raker C, Boardman L. Factors associated with the provision of antenatal contraceptive counseling. Contraception 2008;78:294–9. [3] Finer LB, Zolna MR. Unintended pregnancy in the United States: incidence and disparities, 2006. Contraception 2011;84:478–85. [4] Delhlendorf C, Rodriguez MI, Levy K, et al. Disparities in family planning. AJOG 2010;202:214–8. [5] Smith G, Pell J, Dobbie R. Interpregnancy interval and risk of preterm birth and neonatal death: retrospective cohort study. BMJ 2003;327:313. [6] Cwiak C, Gellasch T, Zieman M. Peripartum contraceptive attitudes and practices. Contraception 2004;70:383–6. [7] Jensen JT, Nelson AL, Costales AC. Subject and clinician experience with the levonorgestrel-releasing intrauterine system. Contraception 2008;77:22–9. [8] Hohmann H, Creinin MD. The contraceptive implant. Clin Obstet Gynecol 2007;50:907–17. [9] Winner B, Piepert JF, Zhao Q, et al. Effectiveness of long-acting reversible contraception. N Engl J Med 2012;366:1998–2007. [10] Ogburn JA, Espey E, Stonehocker J. Barriers to intrauterine device insertion in postpartum women. Contraception 2005;72:426–9.
51
[11] Engin-Ustun Y, Ustin Y, Cetin F, et al. Effect of postpartum counseling on postpartum contraceptive use. Arch Gynecol Obstet 2007;275:429–32. [12] Chen BA, Reeves MF, Hayes JL, Hohmann HL, Perriera LK, Creinin MD. Postplacental or delayed insertion of the levonorgestrel intrauterine device after vaginal delivery: a randomized controlled trial. Obstet Gynecol 2010;116:1079–87. [13] Frost JJ, Singh S, Finer LB. Factors associated with contraceptive use and nonuse, United States, 2004. Perspect Sex Reprod Health 2007;39:90–9. [14] Allen RH, Goldberg AB, Grimes DA. Expanding access to intrauterine contraception. Am J Obstet Gynecol 2009;201:456.e1–5. [15] Hayes JL, Cwiak C, Goedken P, Zieman M. A pilot clinical trial of ultrasound-guided postplacental insertion of a levonorgestrel intrauterine device. Contraception 2007;76:292–6. [16] Grimes DA, Lopez LM, Schulz KF, Van Vliet H, Stanwood NL. Immediate post-partum insertion of intrauterine devices. Cochrane Database Syst Rev 2010:1–15, http://dx.doi.org/10.1002/14651858. CD003036.pub2 Issue 5. Art. No.: CD003036. [17] Rodriguez MI, Caughey AB, Edelman AE, et al. Cost-benefit analysis of state- and hospital-funded postpartum intrauterine contraception at a university hospital for recent immigrants to the United States. Contraception 2010;81:304–8. [18] Tocce KM, Sheeder JL, Teal SB. Rapid repeat pregnancy in adolescents: do immediate postpartum contraceptive implants make a difference? Am J Obstet Gynecol 2012;206:481.e1–7.