Contraception xx (2014) xxx – xxx
Original research article
The free perinatal/postpartum contraceptive services project for migrant women in Shanghai: effects on the incidence of unintended pregnancy☆,☆☆ Yongmei Huang a,⁎, Ruth Merkatz a , Haoping Zhu b , Kevin Roberts a , Régine Sitruk-Ware a , Linan Cheng b and The Perinatal/Postpartum Contraceptive Services Project for Migrant Women Study Group 1 a
b
Center for Biomedical Research Population Council, New York, NY 10021, USA Shanghai Institute of Planned Parenthood Research, 2140 Xie Tu Road, Shanghai 200032, P.R. China Received 6 October 2013; revised 6 February 2014; accepted 2 March 2014
Abstract Objective: In 2006, the incidence of unintended pregnancy among rural-to-urban migrant women (RUMW) in Shanghai was reported as 12.8 per 100 women-years during the first year postpartum. Among permanent residents of Shanghai, that same rate was 3.8 per 100 womenyears. An intervention study was designed to address the unmet need for family planning services among this underserved population of RUMW and reduce their high postpartum unintended pregnancy incidence. Study design: We enrolled 840 migrant women into an intervention study that provided free contraceptive counseling and a choice of methods. Subjects were recruited into the study during hospitalization for childbirth and offered a contraceptive method according to their choice prior to discharge. Counseling and further support were offered at 6 weeks and at 3, 6, 9 and 12 months postpartum via scheduled telephone calls and/or clinic visits. Results: Among all study participants, the median time to contraceptive initiation and sexual resumption was 2 months postpartum, respectively. The overall contraceptive prevalence at 12 months was 97.1%, and more than half of the women were using long-acting contraception. The incidence rate of unintended pregnancy during the first year postpartum was 2.2 per 100 women-years (95% confidence interval: 1.3–3.6). Conclusions: Integrating free family planning services into existing childbirth delivery services in a maternity setting in Shanghai was effective in addressing the unmet need for family planning and reduced the risk of unintended pregnancy during the first year postpartum. Implications: The maternity setting at the time of early labor and prior to postpartum hospital discharge is a practical venue and an optimal time to provide contraception counseling and for postpartum women to initiate use of contraceptive methods. Supporting services during the first year postpartum are also essential to encourage women to continue contraceptive use and reduce the incidence of postpartum unintended pregnancy. © 2014 Elsevier Inc. All rights reserved. Keywords: Free contraceptive services; Unintended pregnancy; Postpartum period; Migrants; China; Intervention study
☆
Conflicts of interest: There are no conflicts of interest among any of the investigators or authors. Sources of funding: This study was funded by Shanghai Municipal Population and Family Planning Commission (Grant Number: 2006JG03). ⁎ Corresponding author. Center for Biomedical Research Population Council, New York, NY 10065, USA. Tel.: +1-929-888-2015. E-mail address: [email protected] (Y. Huang). 1 The Perinatal/Postpartum Contraceptive Services Project for Migrant Women Study Group: Chanjuan Zhuang, Pu Jiang Community Health Center, Shanghai, P.R. China; Ziyin Xia, Minhang Central Hospital Affiliated to Shanghai JiaoTong University School of Medicine, P.R. China; Jianzhong Kang, International Peace Maternal and Child Health Hospital Affiliated to Shanghai JiaoTong University School of Medicine, P.R. China. ☆☆
http://dx.doi.org/10.1016/j.contraception.2014.03.001 0010-7824/© 2014 Elsevier Inc. All rights reserved.
2
Y. Huang et al. / Contraception xx (2014) xxx–xxx
1. Introduction
2. Materials and methods
During the past 20 years, increased government investment in health care in China has improved the overall health status of the Chinese population [1]. While registered residents in cities are entitled to a wide range of benefits, including health insurance from city governments, the household registration system in China (hukou) limits most rural-to-urban migrants from obtaining such benefits [2]. In 2011, Shanghai had approximately 9.3 million migrants who accounted for nearly 40% of its total population; about 80% of whom came from rural areas [3,4]. Lacking health insurance, rural-to-urban migrant women (RUMW) often do not obtain adequate medical care, and their reproductive health remains largely underserved [2,5–8]. As compared with permanent residents, migrant women have a higher unmet need for contraception and experience increased levels of unintended pregnancy [5,6]. They also have delayed and less frequent prenatal check-ups and are at a higher risk for maternal morbidity and mortality [7,8]. To encourage contraceptive use and reduce the unintended pregnancy rate among RUMW, free family planning services were established for this population in 2004 [6,9]. However, certain restrictions remained, i.e., only married couples who had temporary Shanghai resident permits and certificates of marriage and procreation (documentation of their reproductive history) were eligible [9]. There was no specific provision to ensure free contraceptive services among unmarried couples regardless of their household registration status [10]. Results of a survey conducted in 2006 to assess the results of this new program launched in 2004 showed that awareness and utilization of free family planning services among RUMW were low: 25% and 2%, respectively [9]. Another 2006 study described the unintended pregnancy incidence rate among a cohort of RUMW in Shanghai at 12.8 per 100 women-years during the first year postpartum [6], which was much higher than that observed among permanent residents (3.8 per 100 women-years) [11]. The high incidence of unintended pregnancy among the postpartum RUMW was attributed to nonuse of contraception (86%) or contraceptive failure (14%) [6]. These data indicated an urgent need to promote awareness of free family planning services among underserved RUMW in Shanghai and to make such services more accessible during the postpartum period, including to unmarried women. Accordingly, we designed the Perinatal and Postpartum Contraceptive Services Project for Migrant Women (PPCSP) to provide free contraceptive counseling and methods in the maternity setting prior to postpartum discharge, as well as additional support and services during the first postpartum year. Specifically, we sought to determine if free contraceptive services initiated in the maternity setting at the time of childbirth and supported over a 1-year period would affect the incidence of unintended pregnancy among RUMW.
This study was approved by the Ethics Committee of the International Peace Maternal and Child Health Hospital and was conducted at the Pu Jiang Community Health Center. Study participants were enrolled between July and October 2006. After admission to the maternity ward during early labor, RUMW were informed about the PPCSP and provided a one-page leaflet that briefly described the study. Migrant women who were interested in learning about or participating in the study were asked to provide written informed consent to complete the Willingness to Participate (WTP) Questionnaire. The questionnaires were used to (1) collect demographic and reproductive history data and (2) document prenatal health status information. Following data collection, the investigator discussed with each respondent the inclusion criteria for the PPCSP, including (1) both women and their partners, regardless of marital status, lived in rural areas before migrating to the city and did not have a household registration in Shanghai; (2) did not plan to have another child within 2 years after the index childbirth; (3) were healthy and did not have any pregnancy complications or severe systemic disease; (4) planned to live in Shanghai with their partners during the first year postpartum and (5) were able to provide a telephone number for follow-up. Women who met these criteria were tracked through delivery, at which point they were excluded if they experienced severe delivery complications, had a stillbirth or gave birth to an infant with birth defects or any other serious health problem. Women who continued to be eligible provided additional informed consent to participate in the follow-up study. After enrollment, each participant and her partner received contraceptive counseling, which was guided by a multi-page study pamphlet that contained information about the (1) PPCSP study; (2) efficacy, side effects, risks and benefits of long-acting and short-acting contraceptive methods; (3) timing of return-to-fertility and resumption of sexual activity after delivery; (4) risks associated with unintended pregnancies; (5) accessibility of free family planning services and (6) telephone numbers of investigators if questions about maternal and child health and contraceptive use arose. During counseling sessions, each woman had the option of selecting the contraceptive method of her choice. The following long-acting contraceptive methods were provided prior to hospital discharge: (1) tubal ligation, performed during the time of cesarean delivery; (2) copper intrauterine device (Cu-IUD), inserted immediately after cesarean or vaginal delivery or (3) depot-medroxyprogesterone acetate (DMPA) injections at the time of discharge. Women who wished to initiate one of these long-acting contraceptives were screened for contraindications and, if eligible, provided written informed consent. A second counseling session was scheduled after delivery if time did not allow for completion of the entire counseling prior to childbirth. Women who did not initiate long-acting
Y. Huang et al. / Contraception xx (2014) xxx–xxx
contraceptives prior to discharge were provided with male condoms and instructions for correct usage. After hospital discharge, telephone interviews were conducted at 6 weeks and at 3, 6, 9 and 12 months for all participants. During each follow-up contact, participants were asked about their health status and that of their infants, including feeding patterns, resumption of sexual intercourse, contraceptive practices and return of menses. Contraceptive counseling was provided. The possibility/suspicion of being pregnant was discussed and clinic visits were scheduled for women who suspected they were pregnant or wanted further contraceptive services, including tubal ligation, IUD insertion, DMPA injection or oral contraceptive pills, emergency contraception pills, male condoms and spermicides. If pregnancy was confirmed, an ultrasound examination was used to estimate gestational weeks and date of conception. An additional follow-up was arranged at 15 months if pregnancy was suspected at the 12-month interview. Our sample size determination was based on achieving an unintended pregnancy rate of 5 per 100 women-years during the first year postpartum. For a relative precision equal to 10% and a confidence level of 95%, a sample of at least 385 women was required [12]. SAS statistical software version 9.1 (Cary, NC, USA) was employed for data analysis. Women's characteristics and reproductive history were compared by t test for continuous variables or chi-square test for categorical variables. The percentage of contraceptive initiation, sexual activity resumption and return of menses at various time points postpartum was summarized based on the total number of available participants at each follow-up. The incidence rate and 95% confidence intervals (CIs) of unintended pregnancy during the first year postpartum were computed by survival analysis. The log-rank test was used to compare the median time to sexual activity resumption, contraceptive initiation and return of menses. Logistic regression models were used to identify the predictors for long-acting contraceptive use by the end of first year postpartum. Data analysis was two tailed, and pb.05 was considered statistically significant.
3
Additional analyses were completed to compare outcomes from the current study with those from two earlier non-intervention studies, which included migrant women who gave birth at the same health center but prior to the initiation of PPCSP. The demographics and reproductive history of women in two non-intervention cohorts were similar to that of women in the current study. Cohort 2006 enrolled 720 migrants who delivered between January and May in 2006 (unpublished data), and Cohort 2005 included 588 migrants who gave birth in the calendar year of 2005 [6]. Data about these two cohorts were collected through a review of medical records and telephone interviews. Statistical comparisons using results from the other two non-intervention cohorts were not planned and therefore pvalues were not computed.
3. Results From 1 July through 31 October 2006, a total of 1513 women completed the WTP questionnaires. Of those, 858 (56.7%) met the eligibility criteria following childbirth, 18 of whom were excluded due to delivery complications or neonate health problems. Accordingly, 840 women were enrolled for the intervention study and 601 (71.5%) completed one full year of follow-up (Fig. 1). The demographics, reproductive history and delivery information of study participants (n= 840) and those who declined entry or did not meet eligibility criteria (i.e., study non-participants n= 673) are presented in Table 1. There were no significant differences between study participants and non-participants (each pN.05) as well as between women who completed one full year of follow-up (n= 601) and those who dropped out of the study (n= 239). Prior to hospital discharge, 54 (6.4%) women adopted contraceptive methods: 17 women underwent tubal ligation during cesarean delivery, 32 had an IUD inserted during delivery (cesarean delivery, 25; vaginal delivery, 7) and 5 chose DMPA injection. By 6 weeks postpartum, however,
Fig. 1. Flow chart of participants.
4
Y. Huang et al. / Contraception xx (2014) xxx–xxx
Table 1 Demographics and reproductive histories of study participants and nonparticipants Characteristic
Study participants n=840
Demographic Age at delivery 26.4 (4.9) (years) [mean (SD)] Range 18–44 Marital status at delivery n (%) Married 625 (74.4) Unmarried 215 (25.6) Educational level n (%) Primary school or 156 (18.6) no schooling Middle school 601 (71.6) High school+ 83 (9.9) Annual family income (RMB) a n (%) b12,000 72 (8.6) 12,000–17,999 439 (52.3) 18,000–23,999 138 (16.4) ≥24,000 191 (22.7)
Study non-participants n= 673 27.1 (5.0) 15–43 471 (70.0) 202 (30.0) 145 (21.5) 477 (70.9) 51 (7.6) 62 (9.2) 341 (50.7) 120 (17.8) 150 (22.3)
Reproductive history Number of previous live birth n (%) ≥1 359 (42.7) 304 (45.2) 0 481 (57.3) 369 (54.8) Induced abortion history prior to index delivery n (%) Yes 281 (33.5) 216 (32.1) No 559 (66.5) 457 (67.9) Method of delivery for index childbirth n (%) Vaginal delivery 610 (72.6) 503 (74.7) Cesarean delivery 230 (27.4) 170 (25.3) Sex of newborn for index child n (%) Male 450 (53.6) 378 (56.1) Female 390 (46.4) 295 (43.9) a
100 RMB=US$16.
approximately 30% of women had initiated contraception and resumed sexual activity (Fig. 2). By 3 months, the percentage of contraceptive users more than doubled, to 71.1%, and by 12 months, 97% were using contraception. The Kaplan-Meier curve (Fig. 3) shows the median time to contraceptive initiation and resumption of sexual activity,
Fig. 2. Contraceptive initiation, sexual resumption and menses return, by time, among study participants.
which was both 2 months postpartum [interquartile ranges (IQRs) 2–3 and 1.5–2.5, respectively], significantly earlier than the median time to return of menses, i.e., 4 months (IQR 2–6.5) (pb0.05). The male condom was the most common contraceptive method used during the study period with a proportion ranging from 22.5% to 48.5%. IUD use increased significantly from 6 months postpartum and became the most popular method by the end of first year. The overall contraceptive prevalence at 12 months was 97%, including IUD (39.8%), male condom (39.3%), tubal ligation (8.2%), DMPA (6.8%) and other methods (3%) (Fig. 4). By the end of the first year postpartum, 54.7% of women had a tubal ligation or were using an IUD or DMPA. Multivariate logistic regression modeling indicated that longacting contraceptive use was associated with age greater than 30 years, having one or more children prior to the index child and having a male newborn as the index child (Table 2). During the first year postpartum, the incidence rate of unintended pregnancy among all study participants was 2.2 per 100 women-years (95% CI: 1.3–3.6). Fifteen individual women experienced unintended pregnancies. Seven of them were due to nonuse of contraception and 8 were related to
Fig. 3. Kaplan-Meier curves for time to contraceptive initiation, sexual resumption and menses return for study participants.
Y. Huang et al. / Contraception xx (2014) xxx–xxx
5
Fig. 4. Methods of postpartum contraception used among study participants over time after childbirth.
contraceptive failure, i.e., condom failure (7) and IUD expulsion (1). One woman carried her pregnancy to term while the remaining 14 terminated early. When compared with the two non-intervention cohorts (2006 and 2005), the median time to sexual resumption after delivery (2 months) was the same in all three cohorts; however, the intervention cohort initiated contraceptive use earlier (2 months versus 8.5 months and 7.5 months, respectively). The intervention cohort also had a higher prevalence of contraceptive use at the end of first year postpartum (97% versus 73.6% and 62.9%, respectively). As a result, the incidence of unintended pregnancy in the intervention cohort was much lower than that of the two nonintervention cohorts (2.2 per 100 women-years versus 12.8 and 9.6, respectively).
4. Discussion Our study results indicated that migrant women enrolled in the PPCSP initiated postpartum contraception earlier than women in two other non-intervention cohorts, and they had a much higher percentage of contraceptive use throughout the first year postpartum. Consequently, they had fewer unintended pregnancies. The incidence rate of unintended pregnancy reported in the current study is particularly relevant since it is comparable to that described for permanent residents in Shanghai (3.8 per 100 womenyears) who have access to free family planning services without restrictions [11]. Our study demonstrated that promoting awareness of free family planning services and ensuring access to these services, including among unmarried women, is vital for reducing the high level of unintended pregnancy in postpartum RUMW in Shanghai. Integrating free family planning services into existing childbirth services is a practical approach for this underserved population.
Since ovulation can occur by 4 weeks postpartum among non-lactating women, utilization of effective contraceptive methods should be initiated in a timely fashion [13–19]. In our study, the practice of contraception was high and initiation of contraception paralleled the resumption of sexual activity throughout the study period. The median time for both was 2 months. Results from our previous studies revealed that, without provision of contraceptive services, postnatal RUMW started contraceptive use 5.5 months later than when they resumed sexual intercourse [6]. The results of the current study suggest that providing contraceptive counseling in the maternity setting prior to childbirth discharge, along with offering free contraceptives, significantly improves access to family planning services among migrant women and promotes their early use of contraception post-delivery. Results also underscore the importance of sustaining such services during the first year postpartum and encouraging women to continue contraceptive use. By the end of the first year postpartum, more than half of study participants were using long-acting contraceptives. Such use was often observed in women who were older than 30 years, had children prior to the index infant and had given birth to a male newborn. These results reflect China's current family planning policy and cultural custom of preferences for male offspring, particularly in rural areas [20,21]. After achieving the desired family size and the preference for sex of their child, married women tend to use long-acting contraceptives [22]. Accordingly, demographics and reproductive histories as well as awareness of local cultural customs can be important predictors of women's choices of postpartum contraceptive methods. In addition, we identified that the male condom was a popular contraceptive method used by couples during the postpartum period. Providing a free supply of condoms and clear instructions for correct use appears to be beneficial and is recommended, especially among younger and unmarried couples who are less likely to choose long-acting contraception.
6
Y. Huang et al. / Contraception xx (2014) xxx–xxx
Table 2 Factors associated with long-acting contraceptive methods among study participants by the end of first year postpartum Variables
Long-acting methods (n= 329) n (%)
Age at delivery (years) 18–24 117 (45.4) 25–29 101 (59.4) 30–44 111 (64.2) Education status Primary school or no schooling 77 (61.1) Middle school 228 (54.4) High school+ 23 (41.1) Marital status at delivery Married 262 (58.6) Unmarried 68 (43.9) Number of living children before index delivery 0 168 (51.6) ≥1 173 (62.9) Sex of newborn for index child Male 200 (61.2) Female 129 (47.1)
No long-acting methods (n=272) n (%)
Unadjusted odds ratios (95% CI)
Adjusted odds ratios (95% CI)
141 (54.6) 69 (40.6) 62 (35.8)
1.0 1.8 (1. 2, 2.6) 2.2 (1.5, 3.2)
1.0 1.3 (0.9, 2.4) 2.1 (1.2, 3.7)
49 (38.9) 191 (45.6) 33 (58.9)
2.3 (1.2, 4.3) 1.7 (0.9, 3.0) 1.0
185 (41.4) 87 (56.1)
1.8 (1.2, 2.6) 1.0
158 (48.4) 102 (37.1)
1.0 1.6 (1.2, 2.2)
1.0 1.2 (1.0, 2.2)
127 (38.8) 145 (52.9)
1.8 (1.3, 2.5) 1.0
1.5 (1.0, 2.3) 1.0
Long-acting methods: tubal ligation/IUD/DMPA use.
The strength of our study was its prospective study design among a hard-to-reach migrant population likely to experience frequent residency changes within the city [23]. More than 70% of the participants completed the 1-year follow-up, which exceeded the expected sample size. Our results demonstrated that study participants who initiated contraception during the early postpartum period had a high contraceptive prevalence and a low level of unintended pregnancy during the first year postpartum. Our study findings were limited, however, by lack of a randomized control group for valid statistical comparisons. Without the appropriate comparison group, we were also limited in terms of evaluating the impact of our intervention on the incidence of unintended pregnancy among women who were lost to follow-up. A future well-designed randomized controlled trial is recommended to address these concerns and to assess whether the findings from this study can be replicated among other groups of postpartum women who have an unmet need for family planning. In conclusion, contraceptive counseling and access to free contraceptive methods during postpartum hospitalization with support of such services over a 1-year postpartum period appears to be an effective approach for promoting early use of contraception, decreasing the unmet need for contraception and the incidence of postpartum unintended pregnancies among RUMW in Shanghai. This approach can be introduced for the benefit of postpartum women and their infants in other impoverished settings where access to health care and resources are limited and the benefits of optimal child spacing are especially important. Acknowledgment We acknowledge all RUMW and their families who participated in the study and shared their experiences with
us. We thank all the doctors and nurses working at the Department of Obstetrics and Gynecology at Pu Jiang Community Health Center for their support during study period. We gratefully acknowledge Drs. Lisa Cowen, Gary Bologh and Harvey Kushner for their thoughtful reviewing and insightful comments. The first author received the support of the Fred H. Bixby fellowship from the Population Council while completing the manuscript.
References [1] You F, Huo K, Wang R, Xu D, Deng J, Wei Y, et al. Maternal mortality in Henan Province, China: changes between 1996 and 2009. PLoS ONE 2012;7(10):e47153, http://dx.doi.org/10.1371/journal. pone.0047153. [2] Hu X, Cook S, Salazar MA. Internal migration and health in China. Lancet 2008;15:1717–9. [3] Shanghai statistical yearbook 2012. Available at http://www.stats-sh. gov.cn/tjnj/nj12.htm?d1=2012tjnj/C0201.htm. [4] Shanghai Bureau of Statistics. The status and characteristics of crossprovince migrant. Available at http://www.stats-sh.gov.cn/fxbg/201109/ 232741.html Accessed September 23, 2011. [5] Zhou JP. Decreasing the rate of induced abortion in Shanghai. Shanghai Med 2002;25:449–50. [6] Huang YM, Merkatz R, Kang JZ, Roberts K, Hu XY, Di Donato F, et al. Postpartum unintended pregnancy and contraception practice among rural-to-urban migrant women in Shanghai. Contraception 2012;86(6):731–8. [7] Zhao Q, Huang ZJ, Yang S, Pan J, Smith B, Xu B. The utilization of antenatal care among rural-to-urban migrant women in Shanghai: a hospital-based cross-sectional study. BMC Public Health 2012;12:1012. [8] Zhu L, Qin M, Du L, Jia W, Yang Q, Walker MC, et al. Comparison of maternal mortality between migrating population and permanent residents in Shanghai, China, 1996–2005. BJOG 2009;116:401–7. [9] Huang YM, Zhu HP, Xia ZY, Hu XY, Kang JZ, Chen LN. Awareness of free contraceptive service among migrant pregnant women and related factors. Chin J Women Child Health 2010;1:24–8.
Y. Huang et al. / Contraception xx (2014) xxx–xxx [10] Xu JSH. Case study of adolescent reproductive health policy process and feasibility of implementation in China. FuDan University. Doctorate dissertation,2012:4. [11] Huang YM, Kang JZ, Hu XY, Chen RZ, Cheng LN. Study on unintended pregnancy and risk factors among postnatal women from inner city district in Shanghai. Reprod Contracept 2008;28:286–91. [12] Lwanga SK, Lemeshow S. Sample size determination in health studies: a practical manual. Geneva: World Health Organization; 1991, pp. 27, Available at http://www.tbrieder.org/publications/books_english/ lemeshow_samplesize.pdf. [13] Lopez LM, Hiller JE, Grimes DA, Chen M. Education for contraceptive use by women after childbirth. Cochrane Database Syst Rev 2012;15(8):CD001863, http://dx.doi.org/10.1002/14651858. CD001863.pub3. [14] Speroff L, Mishell Jr DR. The postpartum visit: it's time for a change in order to optimally initiate contraception. Contraception 2008;78:90–8. [15] Jackson E, Glasier A. Return of ovulation and menses in postpartum nonlactating women: a systematic review. Obstet Gynecol 2011;117: 657–62.
7
[16] Ezebialu IU, Eke AC. Resumption of vaginal intercourse in the early postpartum period: determinants and considerations for child spacing in a Nigerian population. J Obstet Gynaecol 2012;32:353–6. [17] Glazer AB, Wolf A, Gorby N. Postpartum contraception: needs vs. reality. Contraception 2011;83:238–41. [18] Yee L, Simon M. Urban minority women's perceptions of and preferences for postpartum contraceptive counseling. J Midwifery Womens Health 2011;56(1):54–60. [19] Quinlivan JA, Box H, Evans SF. Postnatal home visits in teenage mothers: a randomized controlled trial. Lancet 2003;361:893–900. [20] Wang C. History of the Chinese family planning program: 1970–2010. Contraception 2012;86:563–9. [21] Bongaarts J. The implementation of preference for male offspring. Population and Development Review 2013;39(2):85–208. [22] Wang C. Trends in contraceptive use and determinants of choice in China: 1980–2010. Contraception 2012;85:570–9. [23] Qian X, Smith H, Huang W, Zhang J, Huang Y, Garner P. Promoting contraceptive use among unmarried female migrants in one factory in Shanghai: a pilot workplace intervention. BMC Health Serv Res 2007; 7:77.
Original research article
The free perinatal/postpartum contraceptive services project for migrant women in Shanghai: effects on the incidence of unintended pregnancy☆,☆☆ Yongmei Huang a,⁎, Ruth Merkatz a , Haoping Zhu b , Kevin Roberts a , Régine Sitruk-Ware a , Linan Cheng b and The Perinatal/Postpartum Contraceptive Services Project for Migrant Women Study Group 1 a
b
Center for Biomedical Research Population Council, New York, NY 10021, USA Shanghai Institute of Planned Parenthood Research, 2140 Xie Tu Road, Shanghai 200032, P.R. China Received 6 October 2013; revised 6 February 2014; accepted 2 March 2014
Abstract Objective: In 2006, the incidence of unintended pregnancy among rural-to-urban migrant women (RUMW) in Shanghai was reported as 12.8 per 100 women-years during the first year postpartum. Among permanent residents of Shanghai, that same rate was 3.8 per 100 womenyears. An intervention study was designed to address the unmet need for family planning services among this underserved population of RUMW and reduce their high postpartum unintended pregnancy incidence. Study design: We enrolled 840 migrant women into an intervention study that provided free contraceptive counseling and a choice of methods. Subjects were recruited into the study during hospitalization for childbirth and offered a contraceptive method according to their choice prior to discharge. Counseling and further support were offered at 6 weeks and at 3, 6, 9 and 12 months postpartum via scheduled telephone calls and/or clinic visits. Results: Among all study participants, the median time to contraceptive initiation and sexual resumption was 2 months postpartum, respectively. The overall contraceptive prevalence at 12 months was 97.1%, and more than half of the women were using long-acting contraception. The incidence rate of unintended pregnancy during the first year postpartum was 2.2 per 100 women-years (95% confidence interval: 1.3–3.6). Conclusions: Integrating free family planning services into existing childbirth delivery services in a maternity setting in Shanghai was effective in addressing the unmet need for family planning and reduced the risk of unintended pregnancy during the first year postpartum. Implications: The maternity setting at the time of early labor and prior to postpartum hospital discharge is a practical venue and an optimal time to provide contraception counseling and for postpartum women to initiate use of contraceptive methods. Supporting services during the first year postpartum are also essential to encourage women to continue contraceptive use and reduce the incidence of postpartum unintended pregnancy. © 2014 Elsevier Inc. All rights reserved. Keywords: Free contraceptive services; Unintended pregnancy; Postpartum period; Migrants; China; Intervention study
☆
Conflicts of interest: There are no conflicts of interest among any of the investigators or authors. Sources of funding: This study was funded by Shanghai Municipal Population and Family Planning Commission (Grant Number: 2006JG03). ⁎ Corresponding author. Center for Biomedical Research Population Council, New York, NY 10065, USA. Tel.: +1-929-888-2015. E-mail address: [email protected] (Y. Huang). 1 The Perinatal/Postpartum Contraceptive Services Project for Migrant Women Study Group: Chanjuan Zhuang, Pu Jiang Community Health Center, Shanghai, P.R. China; Ziyin Xia, Minhang Central Hospital Affiliated to Shanghai JiaoTong University School of Medicine, P.R. China; Jianzhong Kang, International Peace Maternal and Child Health Hospital Affiliated to Shanghai JiaoTong University School of Medicine, P.R. China. ☆☆
http://dx.doi.org/10.1016/j.contraception.2014.03.001 0010-7824/© 2014 Elsevier Inc. All rights reserved.
2
Y. Huang et al. / Contraception xx (2014) xxx–xxx
1. Introduction
2. Materials and methods
During the past 20 years, increased government investment in health care in China has improved the overall health status of the Chinese population [1]. While registered residents in cities are entitled to a wide range of benefits, including health insurance from city governments, the household registration system in China (hukou) limits most rural-to-urban migrants from obtaining such benefits [2]. In 2011, Shanghai had approximately 9.3 million migrants who accounted for nearly 40% of its total population; about 80% of whom came from rural areas [3,4]. Lacking health insurance, rural-to-urban migrant women (RUMW) often do not obtain adequate medical care, and their reproductive health remains largely underserved [2,5–8]. As compared with permanent residents, migrant women have a higher unmet need for contraception and experience increased levels of unintended pregnancy [5,6]. They also have delayed and less frequent prenatal check-ups and are at a higher risk for maternal morbidity and mortality [7,8]. To encourage contraceptive use and reduce the unintended pregnancy rate among RUMW, free family planning services were established for this population in 2004 [6,9]. However, certain restrictions remained, i.e., only married couples who had temporary Shanghai resident permits and certificates of marriage and procreation (documentation of their reproductive history) were eligible [9]. There was no specific provision to ensure free contraceptive services among unmarried couples regardless of their household registration status [10]. Results of a survey conducted in 2006 to assess the results of this new program launched in 2004 showed that awareness and utilization of free family planning services among RUMW were low: 25% and 2%, respectively [9]. Another 2006 study described the unintended pregnancy incidence rate among a cohort of RUMW in Shanghai at 12.8 per 100 women-years during the first year postpartum [6], which was much higher than that observed among permanent residents (3.8 per 100 women-years) [11]. The high incidence of unintended pregnancy among the postpartum RUMW was attributed to nonuse of contraception (86%) or contraceptive failure (14%) [6]. These data indicated an urgent need to promote awareness of free family planning services among underserved RUMW in Shanghai and to make such services more accessible during the postpartum period, including to unmarried women. Accordingly, we designed the Perinatal and Postpartum Contraceptive Services Project for Migrant Women (PPCSP) to provide free contraceptive counseling and methods in the maternity setting prior to postpartum discharge, as well as additional support and services during the first postpartum year. Specifically, we sought to determine if free contraceptive services initiated in the maternity setting at the time of childbirth and supported over a 1-year period would affect the incidence of unintended pregnancy among RUMW.
This study was approved by the Ethics Committee of the International Peace Maternal and Child Health Hospital and was conducted at the Pu Jiang Community Health Center. Study participants were enrolled between July and October 2006. After admission to the maternity ward during early labor, RUMW were informed about the PPCSP and provided a one-page leaflet that briefly described the study. Migrant women who were interested in learning about or participating in the study were asked to provide written informed consent to complete the Willingness to Participate (WTP) Questionnaire. The questionnaires were used to (1) collect demographic and reproductive history data and (2) document prenatal health status information. Following data collection, the investigator discussed with each respondent the inclusion criteria for the PPCSP, including (1) both women and their partners, regardless of marital status, lived in rural areas before migrating to the city and did not have a household registration in Shanghai; (2) did not plan to have another child within 2 years after the index childbirth; (3) were healthy and did not have any pregnancy complications or severe systemic disease; (4) planned to live in Shanghai with their partners during the first year postpartum and (5) were able to provide a telephone number for follow-up. Women who met these criteria were tracked through delivery, at which point they were excluded if they experienced severe delivery complications, had a stillbirth or gave birth to an infant with birth defects or any other serious health problem. Women who continued to be eligible provided additional informed consent to participate in the follow-up study. After enrollment, each participant and her partner received contraceptive counseling, which was guided by a multi-page study pamphlet that contained information about the (1) PPCSP study; (2) efficacy, side effects, risks and benefits of long-acting and short-acting contraceptive methods; (3) timing of return-to-fertility and resumption of sexual activity after delivery; (4) risks associated with unintended pregnancies; (5) accessibility of free family planning services and (6) telephone numbers of investigators if questions about maternal and child health and contraceptive use arose. During counseling sessions, each woman had the option of selecting the contraceptive method of her choice. The following long-acting contraceptive methods were provided prior to hospital discharge: (1) tubal ligation, performed during the time of cesarean delivery; (2) copper intrauterine device (Cu-IUD), inserted immediately after cesarean or vaginal delivery or (3) depot-medroxyprogesterone acetate (DMPA) injections at the time of discharge. Women who wished to initiate one of these long-acting contraceptives were screened for contraindications and, if eligible, provided written informed consent. A second counseling session was scheduled after delivery if time did not allow for completion of the entire counseling prior to childbirth. Women who did not initiate long-acting
Y. Huang et al. / Contraception xx (2014) xxx–xxx
contraceptives prior to discharge were provided with male condoms and instructions for correct usage. After hospital discharge, telephone interviews were conducted at 6 weeks and at 3, 6, 9 and 12 months for all participants. During each follow-up contact, participants were asked about their health status and that of their infants, including feeding patterns, resumption of sexual intercourse, contraceptive practices and return of menses. Contraceptive counseling was provided. The possibility/suspicion of being pregnant was discussed and clinic visits were scheduled for women who suspected they were pregnant or wanted further contraceptive services, including tubal ligation, IUD insertion, DMPA injection or oral contraceptive pills, emergency contraception pills, male condoms and spermicides. If pregnancy was confirmed, an ultrasound examination was used to estimate gestational weeks and date of conception. An additional follow-up was arranged at 15 months if pregnancy was suspected at the 12-month interview. Our sample size determination was based on achieving an unintended pregnancy rate of 5 per 100 women-years during the first year postpartum. For a relative precision equal to 10% and a confidence level of 95%, a sample of at least 385 women was required [12]. SAS statistical software version 9.1 (Cary, NC, USA) was employed for data analysis. Women's characteristics and reproductive history were compared by t test for continuous variables or chi-square test for categorical variables. The percentage of contraceptive initiation, sexual activity resumption and return of menses at various time points postpartum was summarized based on the total number of available participants at each follow-up. The incidence rate and 95% confidence intervals (CIs) of unintended pregnancy during the first year postpartum were computed by survival analysis. The log-rank test was used to compare the median time to sexual activity resumption, contraceptive initiation and return of menses. Logistic regression models were used to identify the predictors for long-acting contraceptive use by the end of first year postpartum. Data analysis was two tailed, and pb.05 was considered statistically significant.
3
Additional analyses were completed to compare outcomes from the current study with those from two earlier non-intervention studies, which included migrant women who gave birth at the same health center but prior to the initiation of PPCSP. The demographics and reproductive history of women in two non-intervention cohorts were similar to that of women in the current study. Cohort 2006 enrolled 720 migrants who delivered between January and May in 2006 (unpublished data), and Cohort 2005 included 588 migrants who gave birth in the calendar year of 2005 [6]. Data about these two cohorts were collected through a review of medical records and telephone interviews. Statistical comparisons using results from the other two non-intervention cohorts were not planned and therefore pvalues were not computed.
3. Results From 1 July through 31 October 2006, a total of 1513 women completed the WTP questionnaires. Of those, 858 (56.7%) met the eligibility criteria following childbirth, 18 of whom were excluded due to delivery complications or neonate health problems. Accordingly, 840 women were enrolled for the intervention study and 601 (71.5%) completed one full year of follow-up (Fig. 1). The demographics, reproductive history and delivery information of study participants (n= 840) and those who declined entry or did not meet eligibility criteria (i.e., study non-participants n= 673) are presented in Table 1. There were no significant differences between study participants and non-participants (each pN.05) as well as between women who completed one full year of follow-up (n= 601) and those who dropped out of the study (n= 239). Prior to hospital discharge, 54 (6.4%) women adopted contraceptive methods: 17 women underwent tubal ligation during cesarean delivery, 32 had an IUD inserted during delivery (cesarean delivery, 25; vaginal delivery, 7) and 5 chose DMPA injection. By 6 weeks postpartum, however,
Fig. 1. Flow chart of participants.
4
Y. Huang et al. / Contraception xx (2014) xxx–xxx
Table 1 Demographics and reproductive histories of study participants and nonparticipants Characteristic
Study participants n=840
Demographic Age at delivery 26.4 (4.9) (years) [mean (SD)] Range 18–44 Marital status at delivery n (%) Married 625 (74.4) Unmarried 215 (25.6) Educational level n (%) Primary school or 156 (18.6) no schooling Middle school 601 (71.6) High school+ 83 (9.9) Annual family income (RMB) a n (%) b12,000 72 (8.6) 12,000–17,999 439 (52.3) 18,000–23,999 138 (16.4) ≥24,000 191 (22.7)
Study non-participants n= 673 27.1 (5.0) 15–43 471 (70.0) 202 (30.0) 145 (21.5) 477 (70.9) 51 (7.6) 62 (9.2) 341 (50.7) 120 (17.8) 150 (22.3)
Reproductive history Number of previous live birth n (%) ≥1 359 (42.7) 304 (45.2) 0 481 (57.3) 369 (54.8) Induced abortion history prior to index delivery n (%) Yes 281 (33.5) 216 (32.1) No 559 (66.5) 457 (67.9) Method of delivery for index childbirth n (%) Vaginal delivery 610 (72.6) 503 (74.7) Cesarean delivery 230 (27.4) 170 (25.3) Sex of newborn for index child n (%) Male 450 (53.6) 378 (56.1) Female 390 (46.4) 295 (43.9) a
100 RMB=US$16.
approximately 30% of women had initiated contraception and resumed sexual activity (Fig. 2). By 3 months, the percentage of contraceptive users more than doubled, to 71.1%, and by 12 months, 97% were using contraception. The Kaplan-Meier curve (Fig. 3) shows the median time to contraceptive initiation and resumption of sexual activity,
Fig. 2. Contraceptive initiation, sexual resumption and menses return, by time, among study participants.
which was both 2 months postpartum [interquartile ranges (IQRs) 2–3 and 1.5–2.5, respectively], significantly earlier than the median time to return of menses, i.e., 4 months (IQR 2–6.5) (pb0.05). The male condom was the most common contraceptive method used during the study period with a proportion ranging from 22.5% to 48.5%. IUD use increased significantly from 6 months postpartum and became the most popular method by the end of first year. The overall contraceptive prevalence at 12 months was 97%, including IUD (39.8%), male condom (39.3%), tubal ligation (8.2%), DMPA (6.8%) and other methods (3%) (Fig. 4). By the end of the first year postpartum, 54.7% of women had a tubal ligation or were using an IUD or DMPA. Multivariate logistic regression modeling indicated that longacting contraceptive use was associated with age greater than 30 years, having one or more children prior to the index child and having a male newborn as the index child (Table 2). During the first year postpartum, the incidence rate of unintended pregnancy among all study participants was 2.2 per 100 women-years (95% CI: 1.3–3.6). Fifteen individual women experienced unintended pregnancies. Seven of them were due to nonuse of contraception and 8 were related to
Fig. 3. Kaplan-Meier curves for time to contraceptive initiation, sexual resumption and menses return for study participants.
Y. Huang et al. / Contraception xx (2014) xxx–xxx
5
Fig. 4. Methods of postpartum contraception used among study participants over time after childbirth.
contraceptive failure, i.e., condom failure (7) and IUD expulsion (1). One woman carried her pregnancy to term while the remaining 14 terminated early. When compared with the two non-intervention cohorts (2006 and 2005), the median time to sexual resumption after delivery (2 months) was the same in all three cohorts; however, the intervention cohort initiated contraceptive use earlier (2 months versus 8.5 months and 7.5 months, respectively). The intervention cohort also had a higher prevalence of contraceptive use at the end of first year postpartum (97% versus 73.6% and 62.9%, respectively). As a result, the incidence of unintended pregnancy in the intervention cohort was much lower than that of the two nonintervention cohorts (2.2 per 100 women-years versus 12.8 and 9.6, respectively).
4. Discussion Our study results indicated that migrant women enrolled in the PPCSP initiated postpartum contraception earlier than women in two other non-intervention cohorts, and they had a much higher percentage of contraceptive use throughout the first year postpartum. Consequently, they had fewer unintended pregnancies. The incidence rate of unintended pregnancy reported in the current study is particularly relevant since it is comparable to that described for permanent residents in Shanghai (3.8 per 100 womenyears) who have access to free family planning services without restrictions [11]. Our study demonstrated that promoting awareness of free family planning services and ensuring access to these services, including among unmarried women, is vital for reducing the high level of unintended pregnancy in postpartum RUMW in Shanghai. Integrating free family planning services into existing childbirth services is a practical approach for this underserved population.
Since ovulation can occur by 4 weeks postpartum among non-lactating women, utilization of effective contraceptive methods should be initiated in a timely fashion [13–19]. In our study, the practice of contraception was high and initiation of contraception paralleled the resumption of sexual activity throughout the study period. The median time for both was 2 months. Results from our previous studies revealed that, without provision of contraceptive services, postnatal RUMW started contraceptive use 5.5 months later than when they resumed sexual intercourse [6]. The results of the current study suggest that providing contraceptive counseling in the maternity setting prior to childbirth discharge, along with offering free contraceptives, significantly improves access to family planning services among migrant women and promotes their early use of contraception post-delivery. Results also underscore the importance of sustaining such services during the first year postpartum and encouraging women to continue contraceptive use. By the end of the first year postpartum, more than half of study participants were using long-acting contraceptives. Such use was often observed in women who were older than 30 years, had children prior to the index infant and had given birth to a male newborn. These results reflect China's current family planning policy and cultural custom of preferences for male offspring, particularly in rural areas [20,21]. After achieving the desired family size and the preference for sex of their child, married women tend to use long-acting contraceptives [22]. Accordingly, demographics and reproductive histories as well as awareness of local cultural customs can be important predictors of women's choices of postpartum contraceptive methods. In addition, we identified that the male condom was a popular contraceptive method used by couples during the postpartum period. Providing a free supply of condoms and clear instructions for correct use appears to be beneficial and is recommended, especially among younger and unmarried couples who are less likely to choose long-acting contraception.
6
Y. Huang et al. / Contraception xx (2014) xxx–xxx
Table 2 Factors associated with long-acting contraceptive methods among study participants by the end of first year postpartum Variables
Long-acting methods (n= 329) n (%)
Age at delivery (years) 18–24 117 (45.4) 25–29 101 (59.4) 30–44 111 (64.2) Education status Primary school or no schooling 77 (61.1) Middle school 228 (54.4) High school+ 23 (41.1) Marital status at delivery Married 262 (58.6) Unmarried 68 (43.9) Number of living children before index delivery 0 168 (51.6) ≥1 173 (62.9) Sex of newborn for index child Male 200 (61.2) Female 129 (47.1)
No long-acting methods (n=272) n (%)
Unadjusted odds ratios (95% CI)
Adjusted odds ratios (95% CI)
141 (54.6) 69 (40.6) 62 (35.8)
1.0 1.8 (1. 2, 2.6) 2.2 (1.5, 3.2)
1.0 1.3 (0.9, 2.4) 2.1 (1.2, 3.7)
49 (38.9) 191 (45.6) 33 (58.9)
2.3 (1.2, 4.3) 1.7 (0.9, 3.0) 1.0
185 (41.4) 87 (56.1)
1.8 (1.2, 2.6) 1.0
158 (48.4) 102 (37.1)
1.0 1.6 (1.2, 2.2)
1.0 1.2 (1.0, 2.2)
127 (38.8) 145 (52.9)
1.8 (1.3, 2.5) 1.0
1.5 (1.0, 2.3) 1.0
Long-acting methods: tubal ligation/IUD/DMPA use.
The strength of our study was its prospective study design among a hard-to-reach migrant population likely to experience frequent residency changes within the city [23]. More than 70% of the participants completed the 1-year follow-up, which exceeded the expected sample size. Our results demonstrated that study participants who initiated contraception during the early postpartum period had a high contraceptive prevalence and a low level of unintended pregnancy during the first year postpartum. Our study findings were limited, however, by lack of a randomized control group for valid statistical comparisons. Without the appropriate comparison group, we were also limited in terms of evaluating the impact of our intervention on the incidence of unintended pregnancy among women who were lost to follow-up. A future well-designed randomized controlled trial is recommended to address these concerns and to assess whether the findings from this study can be replicated among other groups of postpartum women who have an unmet need for family planning. In conclusion, contraceptive counseling and access to free contraceptive methods during postpartum hospitalization with support of such services over a 1-year postpartum period appears to be an effective approach for promoting early use of contraception, decreasing the unmet need for contraception and the incidence of postpartum unintended pregnancies among RUMW in Shanghai. This approach can be introduced for the benefit of postpartum women and their infants in other impoverished settings where access to health care and resources are limited and the benefits of optimal child spacing are especially important. Acknowledgment We acknowledge all RUMW and their families who participated in the study and shared their experiences with
us. We thank all the doctors and nurses working at the Department of Obstetrics and Gynecology at Pu Jiang Community Health Center for their support during study period. We gratefully acknowledge Drs. Lisa Cowen, Gary Bologh and Harvey Kushner for their thoughtful reviewing and insightful comments. The first author received the support of the Fred H. Bixby fellowship from the Population Council while completing the manuscript.
References [1] You F, Huo K, Wang R, Xu D, Deng J, Wei Y, et al. Maternal mortality in Henan Province, China: changes between 1996 and 2009. PLoS ONE 2012;7(10):e47153, http://dx.doi.org/10.1371/journal. pone.0047153. [2] Hu X, Cook S, Salazar MA. Internal migration and health in China. Lancet 2008;15:1717–9. [3] Shanghai statistical yearbook 2012. Available at http://www.stats-sh. gov.cn/tjnj/nj12.htm?d1=2012tjnj/C0201.htm. [4] Shanghai Bureau of Statistics. The status and characteristics of crossprovince migrant. Available at http://www.stats-sh.gov.cn/fxbg/201109/ 232741.html Accessed September 23, 2011. [5] Zhou JP. Decreasing the rate of induced abortion in Shanghai. Shanghai Med 2002;25:449–50. [6] Huang YM, Merkatz R, Kang JZ, Roberts K, Hu XY, Di Donato F, et al. Postpartum unintended pregnancy and contraception practice among rural-to-urban migrant women in Shanghai. Contraception 2012;86(6):731–8. [7] Zhao Q, Huang ZJ, Yang S, Pan J, Smith B, Xu B. The utilization of antenatal care among rural-to-urban migrant women in Shanghai: a hospital-based cross-sectional study. BMC Public Health 2012;12:1012. [8] Zhu L, Qin M, Du L, Jia W, Yang Q, Walker MC, et al. Comparison of maternal mortality between migrating population and permanent residents in Shanghai, China, 1996–2005. BJOG 2009;116:401–7. [9] Huang YM, Zhu HP, Xia ZY, Hu XY, Kang JZ, Chen LN. Awareness of free contraceptive service among migrant pregnant women and related factors. Chin J Women Child Health 2010;1:24–8.
Y. Huang et al. / Contraception xx (2014) xxx–xxx [10] Xu JSH. Case study of adolescent reproductive health policy process and feasibility of implementation in China. FuDan University. Doctorate dissertation,2012:4. [11] Huang YM, Kang JZ, Hu XY, Chen RZ, Cheng LN. Study on unintended pregnancy and risk factors among postnatal women from inner city district in Shanghai. Reprod Contracept 2008;28:286–91. [12] Lwanga SK, Lemeshow S. Sample size determination in health studies: a practical manual. Geneva: World Health Organization; 1991, pp. 27, Available at http://www.tbrieder.org/publications/books_english/ lemeshow_samplesize.pdf. [13] Lopez LM, Hiller JE, Grimes DA, Chen M. Education for contraceptive use by women after childbirth. Cochrane Database Syst Rev 2012;15(8):CD001863, http://dx.doi.org/10.1002/14651858. CD001863.pub3. [14] Speroff L, Mishell Jr DR. The postpartum visit: it's time for a change in order to optimally initiate contraception. Contraception 2008;78:90–8. [15] Jackson E, Glasier A. Return of ovulation and menses in postpartum nonlactating women: a systematic review. Obstet Gynecol 2011;117: 657–62.
7
[16] Ezebialu IU, Eke AC. Resumption of vaginal intercourse in the early postpartum period: determinants and considerations for child spacing in a Nigerian population. J Obstet Gynaecol 2012;32:353–6. [17] Glazer AB, Wolf A, Gorby N. Postpartum contraception: needs vs. reality. Contraception 2011;83:238–41. [18] Yee L, Simon M. Urban minority women's perceptions of and preferences for postpartum contraceptive counseling. J Midwifery Womens Health 2011;56(1):54–60. [19] Quinlivan JA, Box H, Evans SF. Postnatal home visits in teenage mothers: a randomized controlled trial. Lancet 2003;361:893–900. [20] Wang C. History of the Chinese family planning program: 1970–2010. Contraception 2012;86:563–9. [21] Bongaarts J. The implementation of preference for male offspring. Population and Development Review 2013;39(2):85–208. [22] Wang C. Trends in contraceptive use and determinants of choice in China: 1980–2010. Contraception 2012;85:570–9. [23] Qian X, Smith H, Huang W, Zhang J, Huang Y, Garner P. Promoting contraceptive use among unmarried female migrants in one factory in Shanghai: a pilot workplace intervention. BMC Health Serv Res 2007; 7:77.
