The Discordance Between Planned Use and Actual Receipt of Immediate Postpartum Depot Medroxyprogesterone Among Low-Income Women

Abstract

Background:

This study describes the patterns of planned use and actual receipt of immediate postpartum depot medroxyprogesterone (DMPA) prior to hospital discharge among low-income breastfeeding initiators.

Materials and

Methods: Bivariate analyses among DMPA recipients by prenatal planned/unplanned use and the sensitivity of DMPA self-report relative to pharmacologic record were calculated.

Results:

Among immediate postpartum DMPA recipients (n=58), 72.4% (n=42) did not plan to use DMPA. The sensitivity of self-reported DMPA use was 89.7% (95% confidence interval, 85.2, 94.2).

Conclusions:

Clinically, it is unclear if the immediate postpartum period is the appropriate time to obtain consent and administer a long-acting contraceptive method. In our sample, women accurately recalled receiving DMPA in the immediate postpartum period. However, the majority did not plan to use this contraceptive method. Further high-quality qualitative and quantitative research regarding women's contraceptive plans and perception of the postpartum DMPA consent process and the healthcare provider's attitudes regarding consent and prescription of immediate postpartum DMPA are warranted.

Introduction

Unintended pregnancies remain a public health concern, representing approximately 50% of all pregnancies; unintended pregnancies are most prevalent among low-income women (more than five times that of women of higher income).¹ Short interpregnancy intervals (<6 months) can contribute to poor perinatal outcomes (increased risk of preterm birth, small for gestational age, and low birth weight).² Unintended pregnancies are also associated with adverse economic and social consequences at both the individual and the population level.^1,3

Strategies to increase access to contraception, especially in the low-income population, have historically been included in the Healthy People Objectives.⁴ However, women at high risk of short-interval, repeat pregnancy have limited contraceptive options in the early postpartum period.⁵ In addition to clinical reasons, these limited options are often due to hospital formulary restrictions and reimbursement requirements within state-funded insurance programs. This population may be less likely to use nonhormonal contraceptive methods because of the inability to consistently negotiate sexual activity or usage of either behavioral (e.g., abstinence, withdrawal) or barrier (e.g., condoms) methods. As a result, not all forms of available contraception are utilized by low-income women.

Depot medroxyprogesterone (DMPA) is an injectable, progestin-only method that provides effective, long-acting (3 months) contraception.⁶ When systemic methods are needed because of noncompliance with barrier methods or postpartum follow-up, DMPA is often the provider's contraceptive method of choice.⁷ The DMPA package insert indicates exclusively breastfeeding women should not receive the injection until 6 weeks postpartum.⁶ However, the Centers for Disease Control and Prevention endorsed general use of DMPA in the immediate and early postpartum period (<1 month) regardless of lactation status.⁸ The American Congress of Obstetricians and Gynecologists supports these recommendations.⁹ Of note is that these recommendations conflict with those contained in the World Health Organization's Medical Eligibility Criteria for Contraceptive Use.¹⁰ The few available studies evaluating breastfeeding outcomes among various low-income populations report incidence rates of immediate postpartum DMPA use (prior to hospital discharge) in the range of 26.0–45.3%.^11–15

Data describing maternal and healthcare provider decision-making processes regarding contraception in the immediate postpartum period are limited. Miller et al.¹¹ interviewed 299 Medicaid-eligible women prenatally and postpartum to determine the concordance between planned and actual postpartum contraceptive methods. In their sample, 12.1% of women planned to use DMPA prior to the 6-week postpartum obstetric follow-up appointment, and 32.4% actually received DMPA. Among women with no planned contraceptive method, 60.8% (n=25) received postpartum DMPA.

It is unclear if the immediate postpartum period is the appropriate time to obtain consent and administer a long-acting contraceptive method. This study describes the funnel effect of planned use and receipt of DMPA in the immediate postpartum period among low-income women who initiated breastfeeding. Additionally, we report the sensitivity of DMPA self-report relative to pharmacologic records.

Materials and Methods

These analyses used data from a retrospective study designed to evaluate the effect of immediate postpartum DMPA (prior to hospital discharge) on the cessation of any breastfeeding within 6 weeks postpartum.¹⁵ In brief, at low income-serving obstetric and pediatric clinics in Rochester, NY, from January 2010 through December 2012, eligible mothers (≥18 years of age who spoke English, with singleton, healthy, term infants ≤1 year old, who gave birth at one of three local hospitals and initiated breastfeeding) were recruited within 1 year postpartum to participate in a retrospective cohort study.¹⁵ The 183 participants consented to release perinatal maternal/infant medical records and hospital-provided birth certificate data; these data were merged with a self-administered survey to assess breastfeeding experience and outcomes.¹⁵ The current analyses were restricted to participants with complete data who received DMPA in the immediate postpartum period (n=58).

Bivariate chi-squared analyses were conducted among women receiving DMPA by prenatal planned use (relative to unplanned use). Additionally, the sensitivity of DMPA self-report relative to medical records (medical record DMPA receipt used as the “gold standard”) was calculated.

This exploratory, descriptive study used a convenience sample from a preexisting cohort; no power analysis was conducted. Statistical significance was set at alpha <0.05, and all p values were two-sided. Analyses were conducted in Statistical Analysis Systems version 9.3 software (SAS Institute, Cary, NC). This study was approved by the University of Rochester Institutional Review Board.

Results

Table 1 gives a description of the DMPA recipients who initiated breastfeeding. Over half of the recipients presented with unintended pregnancies (56.3%; n=27), almost all participants reported social support regarding their individual decision to breastfeed (88.2%; n=45), the majority received at least adequate prenatal care (70.0%; n=35), and over three-quarters of the study population participated in the Special Supplemental Nutrition Program for Women, Infants, and Children while pregnant (77.6%; n=45).

Table 1.

Key Variables Among Breastfeeding Initiators and Immediate Postpartum Depot Medroxyprogesterone Recipients

	Total (n=58)
Age (years) [mean (SD)]^a,b	23.9 (5.1)
Race^b
White	19 (33.9)
Black	27 (48.2)
Other	10 (17.9)
Education^b
≤High school	44 (78.6)
>High school	12 (21.4)
Adequacy of prenatal care^b,c
Adequate/adequate plus^b	35 (70.0)
Intermediate	6 (12.0)
Inadequate	9 (18.0)
Parental cohabitation^b
Yes	22 (39.3)
No	34 (60.7)
Parity/previous breastfeeding experience^b
Primparous	28 (52.8)
Multiparous with breastfeeding experience	20 (37.7)
Multiparous without breastfeeding experience	5 (9.5)
Breastfeeding plan^b
Breastfeeding goal/milestone	27 (55.1)
None	22 (44.9)
Healthy prepregnancy body mass index^b,d
Yes	21 (38.9)
No	33 (61.1)
Intendedness of pregnancy^b
Intended	21 (43.7)
Unintended	27 (56.3)
Baby Friendly birth hospital^b
Yes	17 (30.4)
No	39 (69.6)
Academic medical center birth hospital^b
Yes	32 (57.1)
No	24 (42.9)
Maternal transience within 2 years^b
0 moves	18 (36.7)
1 move	17 (34.7)
>1 move (2–10)	14 (28.6)
Maternal tobacco use during pregnancy^b
Yes	18 (32.1)
No	38 (67.9)
Social support of breastfeeding^b,e
Yes	45 (88.2)
No	6 (11.8)
Breastfeeding confidence at hospital discharge^b
Very	26 (49.0)
Fairly/somewhat	18 (34.0)
Not at all	9 (17.0)
WIC participation during pregnancy
Yes	45 (77.6)
No	13 (22.4)

Data are n (%) unless indicated otherwise.

Age range 17–36 years; this represents age at the time of delivery, not age at the time of consent and data collection.

Observations <58.

Adequacy of prenatal care as calculated by the Kotelchuck Index (accounting for month of initial prenatal care visit, number of visits, and gestational age).

Healthy prepregnancy BMI defined as 18.5–24.9 kg/m².

Social support of breastfeeding defined as a positive response regarding the father, other family members, or a friend supporting a woman's decision to breastfeed.

WIC, Special Supplemental Nutrition Program for Women, Infants and Children.

Among immediate postpartum DMPA recipients (n=58), 51.7% (n=30) had not previously used DMPA, and 72.4% (n=42) did not plan to use DMPA in the immediate postpartum period. Of participants reporting unplanned DMPA use (n=42), half (50%; n=21) had used this contraceptive method previously. Differences between planned and unplanned DMPA use by previous DMPA use were not statistically significant (Table 2). The sensitivity of DMPA use (self-report relative to pharmacologic record) was 89.7% (95% confidence interval, 85.2, 94.2).

Table 2.

Planned Depot Medroxyprogesterone (DMPA) Use Among Women Who Initiated Breastfeeding and Received Immediate Postpartum DMPA

	Planned DMPA use ^b
Previous DMPA use ^a	Yes (n=16)	No (n=42)
Yes	7 (43.7)	21 (50.0)
No	9 (56.3)	21 (50.0)

Data are n (%).

Self-report of ever previously using DMPA (could be in the immediate postpartum for previous pregnancy or any other DMPA use).

Comparing those with and without planned DMPA use, p>0.05 by chi-squared test.

Discussion

This study evaluated the concordance between planned use and actual immediate postpartum DMPA receipt. Among women who initiated breastfeeding and received DMPA in the immediate postpartum period, the majority (72.4%) did not plan to use this contraceptive method.

Our results are specific to the immediate postpartum period, and we did not collect data regarding what other contraceptive method a woman may have planned to use (or none). As a result, it is not possible to directly compare our results with those of Miller et al.¹¹ However, these are the only estimates that exist regarding planned and actual postpartum DMPA use, and the implications are similar; among two low-income urban study populations, the majority of women not planning to use DMPA received DMPA.

The sensitivity of self-reported immediate postpartum DMPA use was 89.7%. In the larger cohort, among the non-DMPA users, the specificity was 95.4% (95% confidence interval, 92.3%, 98.5%). This suggests women recalled whether or not they received DMPA in the immediate postpartum period with reasonable accuracy. The sensitivity and specificity of DMPA use have previously been evaluated only once in a cohort of 122 women conducted in rural Kenya. Maggwa et al.¹⁶ interviewed women and compared their self-report contraceptive method with clinical records. Within their study population, 80 women (65%) used DMPA previously; the sensitivity and specificity of maternal self-report relative to medical records were 94% and 93%, respectively. Our findings are similar to those of Maggwa et al.,¹⁶ but our estimates are specific to DMPA use in the immediate postpartum period. Combined, these results suggest that women recall DMPA use with relatively high accuracy regardless of timing of receipt (e.g., immediate postpartum versus the entire reproductive life course).

The fact that the majority of women in this study did not intend to use DMPA as their contraceptive method of choice but nonetheless received DMPA generates concerns regarding issues such as health literacy, informed consent, and ultimately the trust patients have in their healthcare provider. Low health literacy is more prevalent among those with limited education and increased poverty, and it is associated with poorer overall health status and higher mortality.¹⁷ Trust, effective communication, and informed consent are critical components in the delivery of quality medical care. Patients with a low level of trust in their healthcare provider have been found to have lower use of recommended preventive services and are less likely to comply with physician recommendations relative to patients with greater patient/healthcare provider trust.¹⁸

Glasier et al.¹⁹ suggested postpartum contraception options should not be discussed in the hospital; instead, they should be discussed in the antenatal period. These recommendations are also supported by the American Congress of Obstetricians and Gynecologists.²⁰ Similarly, a recent Cochrane review concluded there is insufficient evidence to suggest early postpartum educational interventions targeting contraceptive use are effective (regarding outcomes such as unplanned pregnancies, attendance at family planning clinics, use of contraception at 40 days, 9 weeks, or 3 months, and breastfeeding cessation at 3 months).²¹

Our results include findings that are both consistent with the existing literature and are novel. First, these results regarding unplanned postpartum DMPA use are similar to the estimates of Miller et al.,¹¹ but to our knowledge, this is the first time that planned DMPA use specific to the immediate postpartum period has been reported. Second, we used an objective measure of DMPA use, compared the “gold standard” with the maternal self-report, and measured the sensitivity of postpartum DMPA receipt. Third, because our observed high sensitivity indicates that low-income women in this sample are aware of postpartum DMPA use, further research studies could use DMPA self-report as an exposure measure if collected within 1 year postpartum. Last, as a result of the demonstrated high sensitivity and specificity, we believe the data collection items used to assess immediate postpartum DMPA receipt are likely accurate. Additionally, they may well be valid indicators of DMPA use, but these items should be validated in additional and diverse populations.

These findings have several limitations. Because this is a secondary analysis, the findings are hypothesis-generating and only generalizable to women satisfying the original study's inclusion criteria (low-income breastfeeding initiators). However, because the initial sampling frame was restricted to low-income women and this population represents the highest prevalence of immediate postpartum DMPA use, poor breastfeeding outcomes, and short-interval, repeat pregnancy, we believe the results are generalizable to that target population. Second, this study was exploratory and descriptive in nature, and because of the relatively small number of observations (n=58), we did not conduct any multivariable analyses. The observed results may be confounded by maternal demographics or birthing hospital. Third, data regarding planned and previous DMPA use are self-reported. However, because all observations included in this analysis received DMPA, we would not expect systematic differences in recall between those who did and did not plan to use DMPA in the immediate postpartum period; these nondifferential misclassification biases would only dilute our observed results. Fourth, because these data are self-reported, planned DMPA use may be a proxy for contraception knowledge (e.g., if a woman was not aware of DMPA as a contraceptive option, she would likely indicate she did not plan to use that method). One would expect this proxy of contraception knowledge to only be relevant to first-time DMPA recipients, which was only about one-third of our sample (n=21; 36.2%). Fifth, the primary study was designed to evaluate the effect of immediate postpartum DMPA on breastfeeding duration. We did not collect data on planned use of other contraceptive methods or clinic/birthing hospital policy regarding immediate postpartum DMPA use.

Conclusions

In our sample, women accurately recalled receiving DMPA in the immediate postpartum period. However, the majority did not plan to use this contraceptive method. Although it is not possible to draw causal inferences from our data, these results suggest a potential disconnect among prenatal education, healthcare literacy, and DMPA consent in the immediate postpartum period. Furthermore, it remains unclear how individuals are consented for immediate postpartum DMPA receipt, and this discussion should occur in the antenatal care setting. Because of the often highly stressful and emotional state typical of the immediate postpartum period, it may not be appropriate to seek consent for a long-acting contraceptive method during the immediate postpartum period. A study of informed consent for postpartum family planning method administration is needed. Further high-quality qualitative and quantitative research regarding women's contraceptive plans and perception of the postpartum DMPA consent process and the healthcare provider's attitudes regarding consent and prescription of immediate postpartum DMPA is also warranted.

Footnotes

Acknowledgments

This research study was conducted using data collected in partial fulfillment of the requirements for a degree of Doctor of Philosophy, Division of Epidemiology, Department of Public Health Sciences, University of Rochester School of Medicine and Dentistry (E.B.).

Disclosure Statement

No competing financial interests exist.

References

Finer

, Zolna

. Unintended pregnancy in the United States: Incidence and disparities, 2006. Contraception, 2011; 84:478–485.

Conde-Agudelo

, Rosas-Bermudez

, Kafury-Goeta

. Effects of birth spacing on maternal health: A systematic review. Am J Obstet Gynecol, 2007; 196:297–308.

Sonfield

, Kost

, Gold

, et al. The public costs of births resulting from unintended pregnancies: National and state-level estimates. Perspect Sex Reprod Health, 2011; 43:94–102.

U.S. Department of Health and Human Services. Topics & Objectives Index—Healthy People. 2011. www.healthypeople.gov/2020/topicsobjectives2020/objectiveslist.aspx?topicId=13 (accessed May 10, 2014).

Queenan

. Contraception and breastfeeding. Clin Obstet Gynecol, 2004; 47:734–739.

Pfizer. DEPO-PROVERA^® Contraceptive Injection Medroxyprogesterone Acetate Injectable Suspension, USP, Physician Information [package insert]. New York: Pfizer, 2006.

ACOG committee opinion No. 361: Breastfeeding: Maternal and infant aspects. Obstet Gynecol, 2007; 109:479–480.

Farr

, Folger

, Paulen

, et al. U S. Medical Eligibility Criteria for Contraceptive Use, 2010: Adapted from the World Health Organization Medical Eligibility Criteria for Contraceptive Use, 4th edition. MMWR Recomm Rep, 2010; 59(RR-4):1–86.

Committee opinion no. 505: Understanding and using the U.S. Medical Eligibility Criteria for Contraceptive Use, 2010. Obstet Gynecol, 2011; 118:754–760.

10.

World Health Organization. Medical Eligibility Criteria for Contraceptive Use. Geneva: World Health Organization, 2009. http://whqlibdoc.who.int/publications/2010/9789241563888_eng.pdf?ua=1 (accessed May 10, 2014).

11.

Miller

, Laken

, Ager

, et al. Contraceptive decision making among Medicaid-eligible women. J Community Health, 2000; 25:473–480.

12.

Guiloff

, Ibarra-Polo

, Zanartu

, et al. Effect of contraception on lactation. Am J Obstet Gynecol, 1974; 118:42–45.

13.

Halderman

, Nelson

. Impact of early postpartum administration of progestin-only hormonal contraceptives compared with nonhormonal contraceptives on short-term breast-feeding patterns. Am J Obstet Gynecol, 2002; 186:1250–1256.

14.

Hannon

, Duggan

, Serwint

, et al. The influence of medroxyprogesterone on the duration of breast-feeding in mothers in an urban community. Arch Pediatr Adolesc Med, 1997; 151:490–496.

15.

Brownell

, Fernandez

, Fisher

, et al. The effect of immediate postpartum depot medroxyprogesterone on early breastfeeding cessation. Contraception, 2013; 87:836–843.

16.

Maggwa

, Mati

, Mbugua

, et al. Validity of contraceptive histories in a rural community in Kenya. Int J Epidemiol, 1993; 22:692–697.

17.

Berkman

, Sheridan

, Donahue

, et al. Low health literacy and health outcomes: An updated systematic review. Ann Intern Med, 2011; 155:97–107.

18.

O'Malley

, Gonzalez

, Sheppard

, et al. Primary care cancer control interventions including Latinos: A review. Am J Prev Med, 2003:25:264–271.

19.

Glasier

, Logan

, McGlew

. Who gives advice about postpartum contraception?. Contraception, 1996; 53:217–220.

20.

Committee on Health Care for Underserved Women, Committee on Obstetric Practice. Breastfeeding: Maternal and infant aspects. Int J Gynaecol Obstet, 2001; 74:217–232.

21.

Hiller

, Griffith

, Jenner

. Education for contraceptive use by women after childbirth. Cochrane Database Syst Rev, 2002(3):CD001863.