Immediate versus delayed postpartum use of levonorgestrel contraceptive implants: a randomized controlled trial in Uganda

Immediate versus delayed postpartum insertion of contraceptive implant and IUD for contraception

BACKGROUND

Long-acting reversible contraception (LARC), including intrauterine devices (IUDs) and contraceptive implants, are highly effective, reversible methods of contraception. Providing LARC methods during the postpartum period is important to support contraceptive choice, and to prevent unintended pregnancy and short interpregnancy intervals. Delaying offering contraception to postpartum people until the first comprehensive postpartum visit, traditionally at around six weeks postpartum, may put some postpartum people at risk of unintended pregnancy, either due to loss to follow-up or because of initiation of sexual intercourse prior to receiving contraception. Therefore, immediate provision of highly effective contraception, prior to discharge from hospital, has the potential to improve contraceptive use and prevent unintended pregnancies and short interpregnancy intervals.

OBJECTIVES

To compare the initiation rate, utilization rates (at six months and 12 months after delivery), effectiveness, and adverse effects of immediate versus delayed postpartum insertion of implants and IUDs for contraception.

SEARCH METHODS

We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase, and POPLINE for eligible studies up to December 2020. We examined review articles and contacted investigators. We checked registers of ongoing clinical trials, citation lists of included studies, key textbooks, grey literature, and previous systematic reviews for potentially relevant studies.

SELECTION CRITERIA

We sought randomized controlled trials (RCTs) that compared immediate postpartum versus delayed insertion of contraceptive implant and IUDs for contraception.

DATA COLLECTION AND ANALYSIS

Two review authors (JS, SK) independently screened titles and abstracts of the search results, and assessed the full-text articles of potentially relevant studies for inclusion. They extracted data from the included studies, assessed risk of bias, compared results, and resolved disagreements by consulting a third review author (PL, SA or PP). We contacted investigators for additional data, where possible. We computed the Mantel-Haenszel or inverse variance risk ratio (RR) with 95% confidence interval (CI) for binary outcomes and the mean difference (MD) with 95% CI for continuous variables.

MAIN RESULTS

In this updated review, 16 studies met the inclusion criteria; five were studies of contraceptive implants (715 participants) and 11 were studies of IUDs (1894 participants). We identified 12 ongoing studies. We applied GRADE judgements to our results; the overall certainty of the evidence for each outcome ranged from moderate to very low, with the main limitations being risk of bias, inconsistency, and imprecision. Contraceptive implants Immediate insertion probably improves the initiation rate for contraceptive implants compared with delayed insertion (RR 1.48, 95% CI 1.11 to 1.98; 5 studies, 715 participants; I2 = 95%; moderate-certainty evidence). We are uncertain if there was a difference between the two groups for the utilization rate of contraceptive implants at six months after delivery (RR 1.16, 95% CI 0.90 to 1.50; 3 studies, 330 participants; I2 = 89%; very low-certainty evidence) or at 12 months after insertion (RR 0.98, 95% CI 0.93 to 1.04; 2 studies, 164 participants; I2 = 0%; very low-certainty evidence). People who received an immediate postpartum contraceptive implant insertion may have had a higher mean number of days of prolonged vaginal bleeding within six weeks postpartum (mean difference (MD) 2.98 days, 95% CI -2.71 to 8.66; 2 studies, 420 participants; I2 = 91%; low-certainty evidence) and a higher rate of other adverse effects in the first six weeks after birth (RR 2.06, 95% CI 1.38 to 3.06; 1 study, 215 participants; low-certainty evidence) than those who received a delayed postpartum insertion. We are uncertain if there was a difference between the two groups for prolonged bleeding at six months after delivery (RR 1.19, 95% CI 0.29 to 4.94; 2 studies, 252 participants; I2 = 0%; very low-certainty evidence). There may be little or no difference between the two groups for rates of unintended pregnancy at six months (RR 0.20, 95% CI 0.01 to 4.08; one study, 205 participants; low-certainty evidence). We are uncertain whether there was a difference in rates of unintended pregnancy at 12 months postpartum (RR 1.82, 95% CI 0.38 to 8.71; 1 study, 64 participants; very low-certainty evidence). There may be little or no difference between the two groups for any breastfeeding rates at six months (RR 0.97, 95% CI 0.92 to 1.01; 2 studies, 225 participants; I2 = 48%; low-certainty evidence). IUDs Immediate insertion of IUDs probably improves the initiation rate compared with delayed insertion, regardless of type of IUD (RR 1.27, 95% CI 1.07 to 1.51; 10 studies, 1894 participants; I2 = 98%; moderate-certainty evidence). However, people who received an immediate postpartum IUD insertion may have had a higher expulsion rate at six months after delivery (RR 4.55, 95% CI 2.52 to 8.19; 8 studies, 1206 participants; I2 = 31%; low-certainty evidence) than those who received a delayed postpartum insertion. We are uncertain if there was a difference between the two groups in the utilization of IUDs at six months after insertion (RR 1.02, 95% CI 0.65 to 1.62; 6 studies, 971 participants; I2 = 96%; very low-certainty evidence) or at 12 months after insertion (RR 0.86, 95% CI 0.5 to 1.47; 3 studies, 796 participants; I2 = 92%; very low-certainty evidence). Immediate IUDs insertion may reduce unintended pregnancy at 12 months (RR 0.26, 95% CI 0.17 to 0.41; 1 study, 1000 participants; low-certainty evidence). We are uncertain whether there was difference in any breastfeeding rates at six months in people receiving progestin-releasing IUDs (RR 0.90, 95% CI 0.63 to 1.30; 5 studies, 435 participants; I2 = 54%; very low-certainty evidence).

AUTHORS' CONCLUSIONS

Evidence from this updated review indicates that immediate postpartum insertion improves the initiation rate of both contraceptive implants and IUDs by the first postpartum visit compared to delayed insertion. However, it is not clear whether that there are differences in utilization rates at six and 12 months postpartum. We are uncertain whether there is any difference in the unintended pregnancy rate at 12 months. Provision of progestin-releasing implants and IUDs immediately postpartum may have little or no negative impact on breastfeeding. However, the expulsion rate of IUDs and prolonged vaginal bleeding associated with immediate implants appears to be higher.

Couples' decision-making on post-partum family planning and antenatal counselling in Uganda: A qualitative study

BACKGROUND

Although health workers have been trained to provide post-partum family planning (PPFP), uptake remains low in Uganda. An important reason is that women want the agreement of their partner, who is often absent at the time of delivery. In order to address this, we aimed to understand the views of couples and explore barriers and facilitators to implementation of antenatal couples' counselling on PPFP in Uganda.

METHODS

We conducted individual interviews with a purposive sample of 12 postpartum and 3 antenatal couples; and 34 focus groups with a total of 323 participants (68 adolescent women, 83 women aged 20-49, 79 men, 93 health workers) in four contrasting communities (urban and rural) in South-West and Central Uganda. These were recorded, transcribed, translated, and analysed thematically.

RESULTS

Although most participants felt that it is important for partners to discuss family planning, half of the couples were unaware of each other's views on contraception. Most had similar views on motivation to use family planning but not on preferred contraceptive methods. Most liked the idea of antenatal couples' counselling on PPFP. The main barrier was the reluctance of men to attend antenatal clinics (ANC) in health facilities. Respondents felt that Village Health Teams (VHTs) could provide initial counselling on PPFP in couples' homes (with telephone support from health workers, if needed) and encourage men to attend ANC. Suggested facilitators for men to attend ANC included health workers being more welcoming, holding ANC clinics at weekends and "outreach" clinics (in rural villages far from health facilities).

CONCLUSION

Antenatal couples' counselling has the potential to facilitate agreement PPFP, but some men are reluctant to attend antenatal clinics. Counselling at home by VHTs as well as simple changes to the organisation of antenatal clinics, could make it possible to deliver antenatal couples' counselling on PPFP.

Maternal health services utilization and maternal mortality in China: a longitudinal study from 2009 to 2016

Background

The Chinese government introduced the maternal health services as part of the Basic Public Health Service (BPHS) project in 2009. While China has made progress in reducing maternal mortality, the longitudinal association between the utilization rate of the maternal health services of the BPHS project and such reduction was rarely evaluated with robust methods.

Methods

We conducted a longitudinal study on the maternal health service utilization rates of the BPHS project and the maternal mortality ratios (MMR) in mainland China and its 31 provincial regions. The data were extracted from the National Health Statistic Yearbooks (2009–2016). Panel data models were used to evaluate the association between the utilization rate and the MMR after adjusting for available covariates on healthcare resources. Stepwise regression models were used to gauge the direction and magnitude of omitted variable bias.

Results

It was found that the service utilization rate increased from 79.24% in 2009 to 91.67% in 2016, while MMR decreased from 30.90 to 17.88 per 100,000 livebirths at the national level. The results of the fixed effects panel data model revealed that the utilization rate was associated with MMR deduction continuously. With every 1% increase in utilization rate, the maternal death would decrease by 0.35 per 100,000 livebirths after having the health resources variables adjusted.

Conclusions

The utilization of maternal health services increased continuously after the introduction of BPHS project in 2009 and MMR went downward from then on. The utilization of maternal health services did have significant association with MMR and could reduce MMR effectively. Such improvement could be attributed to the fact that this program was designed to serve the targeted population throughout the continuum of maternal care and the government’s rigorous efforts in stressing equality and standard care in program implementation.

The influence of basic public health service project on maternal health services: an interrupted time series study

BACKGROUND

Reducing maternal mortality remains a global priority. In 2000, the United Nations Member States pledged to work towards a series of Millennium Development Goals (MDGs), in which the fifth target was to reduce maternal mortality ratio by 75% from 1990 to 2015. The Chinese government introduced Basic Public Health Service project in 2009 to the further improvement of maternal health services and reduction in maternal mortality. China had achieved the goal of MDG5 1 year ahead of the schedule in 2014, but the effects of the project on reducing maternal mortality were rarely evaluated with robust methods.

METHODS

We conducted a longitudinal study on maternal mortality ratio by extracting mortality data from the National Maternal Mortality Surveillance System (1991-2016) and maternal health services measures from the China health statistic yearbook (2001-2016). We utilized the segmented linear regression model to assess changes and trends of maternal mortality ratio and maternal health services before and after the introduction of Basic Public Health Service project. Pearson correlation analysis was conducted to measure the strength of association between the maternal mortality ratio and maternal health services.

RESULTS

The yearly trend change of national maternal mortality ratio was - 1.76 (p < 0.01) after the introduction of Basic Public Health Service project in 2009, while the yearly trend change of maternal health record establish rate, prenatal examination rate, postpartum visit rate was 0.77 (p < 0.01), 0.61 (p < 0.01) and 0.83 (p < 0.01) separately. The negative correlations were also found between national maternal mortality ratio and prenatal examination rate (r = - 0.95, p < 0.01), maternal health record establish rate (r = - 0.93, p < 0.01) and postpartum visit rate (r = - 0.92, p < 0.01).

CONCLUSIONS

The Basic Public Health Service project was found to be associated with the improvements in the maternal health services and reduction in maternal mortality. The design and implementation of the project may serve as a positive example for other developing countries. Continued monitoring and assessment of project effects should be stressed.

The effect of immediate postpartum levonorgestrel contraceptive implant use on breastfeeding and infant growth: a randomized controlled trial

OBJECTIVE

This study assessed whether immediate postpartum insertion of levonorgestrel contraceptive implants is associated with a difference in infant growth from birth to 6 months, onset of lactogenesis, or breastfeeding continuation at 3 and 6 months postpartum compared to delayed insertion at 6 to 8 weeks postpartum.

STUDY DESIGN

We conducted a randomized trial of women in Uganda who desired contraceptive implants postpartum. We randomly assigned participants to receive either immediate (within 5 days of delivery) or delayed (6 to 8 weeks postpartum) insertion of a two-rod levonorgestrel contraceptive implant system. This is a prespecified secondary analysis evaluating breastfeeding outcomes. The primary outcome of this secondary analysis was change in infant weight; infants were weighed and measured at birth and 6 months. We used a validated questionnaire to assess onset of lactogenesis daily in person while participants were in the hospital, and then daily by phone after they left the hospital, until lactogenesis was documented. We used interviewer-administered questionnaires to assess breastfeeding continuation and concerns at 3 months and 6 months postpartum.

RESULTS

Among the 96 women randomized to the immediate group and the 87 women to the delayed group, the mean change in infant weight from birth to 6 months was similar between groups: 4632 g in the immediate group and 4407 g in the delayed group (p=.26). Among the 97 women who had not experienced lactogenesis prior to randomization, the median time to onset of lactogenesis did not differ significantly between the immediate and delayed groups (65 h versus 63 h; p=.84). Similar proportions of women in the immediate and delayed groups reported exclusive breastfeeding at 3 months (74% versus 71%; p=.74) and 6 months (48% versus 52%; p=.58).

CONCLUSION

We found no association between the timing of postpartum initiation of levonorgestrel contraceptive implants and change in infant growth from birth to 6 months, onset of lactogenesis, or breastfeeding continuation at 3 or 6 months postpartum.

IMPLICATIONS

This study provides evidence that immediate postpartum initiation of contraception implants does not have a deleterious effect on infant growth or initiation or continuation of breastfeeding.

Two-year contraceptive continuation rates among immediate postpartum implant users at a district hospital in Malawi: a prospective cohort study

OBJECTIVE

To compare 2-year continuation rates in Malawian women undergoing immediate postpartum insertion of the levonorgestrel implant or etonorgestrel implant.

STUDY DESIGN

We followed 159 women who underwent immediate postpartum levonorgestrel implant or etonorgestrel implant insertion at Kasungu District Hospital for up to 2 years.

RESULTS

We analyzed continuation data on 145 (92.4%) implant users. The 2-year continuation rates were 93.4 (95% CI 86.5-96.8) for levonorgestrel implant and 96.3 (95% CI: 76.5-99.5) for etonorgestrel implant (p=.268).

CONCLUSIONS

Immediate postpartum implant insertion of both the levonorgestrel and etonorgestrel implant had high continuation rates at 2 years in Malawian women.

IMPLICATIONS

Immediate postpartum implant insertion of both the levonorgestrel and etonorgestrel implant had continuation rates of over 90% at 2 years among our population of Malawian women. Both implants should be offered routinely to eligible and interested women prior to hospital discharge after delivery.