Factors influencing the likelihood of acceptance of postpartum intrauterine devices across four countries: India, Nepal, Sri Lanka, and Tanzania

To examine the factors that positively influenced the likelihood of accepting provision of postpartum intrauterine devices (PPIUDs) across four countries: Sri Lanka, Nepal, Tanzania, and India.


| INTRODUCTION
In recent years, there has been increasing international recognition of the importance of offering effective contraceptive services to women immediately after childbirth. The benefits for women and their children are clear: contraceptive use has the potential to improve perinatal outcomes and child survival by widening the birth interval, 1 and reduce maternal morbidity and mortality associated with high parity and unintended pregnancies. 2 However, the ability of services to meet women's contraceptive needs postpartum has historically been problematic, with estimates that around two-thirds of women have an unmet need for contraception in the first year after birth. 3 The frequent tradition of deferring family planning services to the 6-week postpartum visit fails many women who may desire to postpone or limit childbearing but face barriers to accessing postnatal care 4 or prioritize services for their newborn over their own care. 5 Services are responding to these challenges at both a local and global level, with the development of initiatives offering contraceptive services immediately postpartum. These initiatives take advantage of prenatal care and delivery as opportunities to address the postpartum contraceptive needs of women. 6 Effective contraceptive counselling lies at the core of these initiatives, and is essential to enabling uptake of contraception among postpartum women prior to discharge.
Despite this increased focus on offering effective methods to women immediately postpartum, there has been little corresponding research into how best to counsel women to make choices. 7 Counselling factors such as timing and provider status have the potential to significantly impact on uptake of postpartum contraception 8 ; however, there is a dearth of evidence exploring these associations. We were able to identify two studies that aimed to analyze the impact of timing on contraception uptake. One study, based in the USA, identified significantly increased rates of postpartum use of a more effective method of contraception among patients counselled in both the prenatal and postpartum period as opposed to one time period only; the authors did not however comment on the comparative impact of prenatal versus postnatal counselling. 9 Another study, based in Egypt, identified no significant difference in the effect of prepartum versus postpartum counselling on verbal acceptance of a receiving a postpartum intrauterine device (PPIUD). 10 Even less is known about the impact of the status of the counselling provider. Those providing counselling come from a wide variety of backgrounds, 7 and may operate in health systems where obstetrics and family planning services have traditionally been separate. 6 As part of their study, Mohamed et al. 10 commented that provider status had no impact on subsequent acceptance of PPIUD, but provided no data to support this claim, and no further detail on the backgrounds of the counselling providers included in their study. More research on the impact of the timing of counselling and provider status is needed to better understand how we can best assist women in making contraceptive choices postpartum.
The aim of the present study was to examine the factors that influenced the outcome of counselling for PPIUD, as part of a FIGO initiative, across four countries: India, Sri Lanka, Nepal, and Tanzania.
Although they are all low-or lower-middle-income countries, the profiles of the four countries selected in this study are different when contraceptive use is taken into account. This is likely to impact on the outcomes. Recent figures compiled by the UN on world contraceptive use, published in 2018, 11 are outlined in Table 1. IUDs relate mainly to interval IUDs as PPIUD was not significantly in practice in 2015-2016 when the surveys were collated. These figures show that IUDs are not popular, particularly in India, Nepal, and Tanzania, with contraceptive prevalence ranging from 0.9-1.5. If IUDs are not popular or common in the country, counselling will be a key element of the implementation process, hence the need for a deeper understanding of its effects.

| MATERIALS AND METHODS
The present study forms part of a larger multicountry PPIUD initia- Women delivering in those facilities were asked for their consent to take part in a short 15-minute face-to-face structured interview.
In those cases where consent was obtained, in-country data collection officers (DCOs) conducted the interview prior to discharge. All data were collected by the DCOs on tablets and stored using T A B L E 1 Contraceptive prevalence and unmet need for family planning in India, Nepal, Sri Lanka, and Tanzania. Tanzania. These dates were selected because more than 80% of all women delivering in the chosen facilities in each country had been interviewed by DCOs in these periods and data could then be considered representative. Although six countries participated in the initiative, Kenya and Bangladesh did not achieve high enough interview rates among delivered women to be included in this analysis.
This was also the reason for not including the other 12 facilities in Sri Lanka, as interview rates in these units did not reach the required threshold.   Information was not available on parity, survival of recent birth, and counsellors for Nepal because this information was not collected on their interview form.

| RESULTS
b Information was not available on Counsellors for Sri Lanka and Tanzania as they did not employ this strategy for counselling women.
times more likely to say yes to PPIUD, whereas in Nepal and Tanzania those counselled by doctors were more likely to respond positively (OR 1.48 and 1.39, respectively). In Nepal and Sri Lanka, women who were seen by nurses/midwives were less likely to consent to PPIUD as a contraceptive option.

| DISCUSSION
Counselling had a much greater impact in some countries than in others and it is interesting to elucidate why that was the case. Across the six facilities in India, 86% of women were counselled and the proportion of women consenting to PPIUD was the highest at 34%.
In 2016, the documented background interval IUD prevalence was reported as 1.5% at national level. 11 Which factors may have influenced this success?
Odds ratios for women's age and parity in India suggest that women who had fewer children and were younger were more likely to agree to PPIUD. In the six Indian facilities included in the initiative, the The value of multiple encounters was clear, particularly in countries where the IUD is not popular. Multiple encounters also allowed for relatives to become involved in the counselling sessions, which is key in communities where family and relatives play a significant role in decision making around family planning. In India in particular, women were encouraged to come back with their husbands and mother-in-laws at subsequent appointments so that they could also hear about the benefits of family planning, but particularly the value of PPIUD as an effective nonhormonal long-acting reversible contraceptive.
The value of the counsellor was also evident in India where the ORs of accepting PPIUD was 1.17 times greater among women who were counselled by a counsellor compared with those who were not.
Reproductive Child Health (RCH) counsellors are part of the existing government healthcare system in Indian hospitals. They counsel not only on family planning but also other reproductive and child health topics. The value that counsellors add to clinical consultations is clearly evident when it comes to contraception-a topic requiring indepth, private conversations with every woman.
The initiative in Nepal saw smaller proportions of women receiving counselling in family planning and PPIUD (36% and 32% of those interviewed, respectively), and therefore the PPIUD consent rate was also lower (10%). The method mix was different in Nepal, where hormonal implants are on offer but only after 6 weeks. Fewer practitioners were trained in counselling and insertion and although counsellors were employed, this analysis could not demonstrate their impact on acceptance of PPIUD because the interview forms did not collect the relevant information.
Family planning counsellors in Nepal were employed directly by the initiative and so their sustainability remains in question. They were also a new cadre of health staff in the health system, and during monitoring visits it was evident that in some facilities they were not well accepted and sometimes viewed with suspicion. Nepali women tend to have more trust in medical staff, in particular doctors, and it was mentioned by providers that without the doctor's ratification women would be less likely to accept a new concept such as PPIUD. Further qualitative research into this aspect is currently underway.
The six Nepali facilities involved are large institutions with a relatively low staff to patient ratio. This meant that doctors simply did not have the time to spend counselling women at length. Interestingly, the odds ratio of consenting for PPIUD was 0.67 when women were counselled by nurses. On monitoring visits it was noticed that many nurse/midwives were not motivated to counsel or insert PPIUD. Often it was seen as an extra burden that had previously been undertaken by the family planning department and not maternity staff. In an already stretched and busy environment, these nurses were often not willing to take on an extra task. In many instances they were the gateway to knowledge and information for the patient and without their support, it is not surprising that counselling and consent rates were low.
However, it should be reiterated that the national IUD prevalence in Nepal was 1.4% in 2016, 11 and therefore that a consent rate of 10% is still quite an achievement. Initiating new practices in an environment where the system is already under strain is not an easy task and is likely to take considerable time.
Sri Lanka is a country with much lower unmet need for contra- and counselling and insertion were primarily conducted by midwives.
The results suggest that despite the overall low modern contraceptive prevalence, there is demand for long-acting methods, such as PPIUD, in Tanzania. This study suggests that nationalization of PPIUD services predominantly through midwives has the potential to be a highly successful intervention in Tanzania.
The present study adopted a quantitative approach when looking at different factors influencing consent for provision of PPIUD. Clearly there are multiple variables: family structures and who is the decision maker, organization of health care in the country, alternative methods and adverse effects, as well as a woman's perceptions of her need for family planning. It is likely that further studies using a qualitative methodology in each country would help to better understand how these factors interact and hence help inform policy makers.

| CONCLUSION
It is not possible to generalize the results across all four countries with regard to factors that influence women to consent to provision of PPIUD. These appear to be very much context specific. The only consistent factor across all countries that resulted in a greater likelihood of acceptance of PPIUD was having multiple counselling episodes.
This should be taken into account by policy makers when designing implementation programs.

AUTHOR CONTRIBUTIONS
AM wrote the manuscript with help from NT. Data cleaning, analysis, and tables were prepared by MS and KM. SA planned the initiative and directed implementation together with AM. PM, AB, KT, and GP coordinated activities in their respective countries. All authors reviewed the final version of the paper. MS worked on the project while employed by FIGO.

ACKNOWLEDGMENTS
We wish to acknowledge the past and present members of the PPIUD team at FIGO headquarters and each country team that has worked hard to make this initiative possible. We would like to thank the We would like to thank the ministries of health of the governments of India, Nepal, Tanzania, and Sri Lanka for their participation in the initiative. Finally, we would also like to thank our anonymous donors for their generous grant and continued support.

CONFLICTS OF INTEREST
The authors have no conflicts of interest to declare.