Authors: Peter Kim Streatfield  ( Centre for Population, Urbanization, and Climate Change at the International Centre for Diarrhoeal Diseases Research, Bangladesh (ICDDR,B), Dhaka, Bangladesh. )
Nahid Kamal  ( MEASURE Evaluation, University of North Carolina at Chapel Hill, NC, USA. )

Introduction: The family planning programme (FPP) in Bangladesh and its role in reducing the fertility rate in the country has been at the centre of much scholarly debate. The current contraceptive prevalence rate (CPR) is 61%- double the rate observed in Pakistan, while the total fertility rate is 2.3 children per woman. This paper highlights both the supply-and demand-side factors explaining trends in contraceptive use over the last four decades. It identifies the challenges plaguing the programme today that range from funding and staffing deficiencies to bottlenecks in procurement processes.
Methodology: The paper uses secondary data sources such as Bangladesh Demographic and Health Survey (BDHS) 2004 and 2011 to show Family Planning trends in Bangladesh and uses the Spectrum software to forecast trends for the year 2015.
Results: The central message is that the current contraceptive method mix, which is heavily biased towards temporary methods, cannot support a sustained decline in fertility. The authors\\\' projected estimates reveal that in order to achieve replacement-level fertility by 2016, users of long-acting and permanent methods would need to be increased by 8-9 million. Drawing on global experiences, the paper outlines greater flexibility in the use of funds and a regionally-targeted approach, among other options, that could be adopted to ensure that national population targets are met.
Keywords: Family planning program, Bangladesh, Contraceptive method mix, Long-acting and permanent methods, Contraceptive procurement, Population, Fertility.

Although there is no dearth of literature on the demand-side aspects of family planning (FP) and fertility decline in Bangladesh, relatively little is known about the actual programme itself or the reasons behind the plateauing in the uptake of contraception during the 1990s. This paper starts off by providing the backdrop for the adoption of a national family planning programme (FPP) and explains the erratic nature of fertility decline since the1980s. The focus of the paper, however, is on the contraceptive methodmix in Bangladesh and supply-side aspects of the programme like staffing, procurement processes, and funding. We project (using SPECTRUM) the ideal method mix that is needed in order to achieve replacement-level fertility by 2016. The main section describing the national FPP is largely a desk review of published and unpublished literature spanning the programme\\\'s lifetime.
Background on Population and Family Planning
The latest 2011 census of Bangladesh estimates a population of 149 million for the country implying an increase of 19 million since the census of 2001. In 1901,the area that is now Bangladesh supported a population of roughly 25 million and population growth was negligible. By the late 20th century, due to a steep post-WWII mortality decline, population growth exploded, and by the beginning of the present century, at least a 100 million had been added.¹
Historically, Bangladeshi families had about seven children of whom less than 40%survived to adulthood to become parents themselves. This meant that only one son inherited the family land which, thus, was not divided. Following WWII, the crude death rate (CDR) fell by 50%in 15 years (40-20 deaths per 1,000 population), then again by 50% (20-10 deaths per 1,000 population) by the mid-1970s. The consequence was that child survival improved significantly such that five children per family reached adulthood, which had major implications for land inheritance and a rapid fragmentation of family landholdings.
The national FPP was initiated as a response to this rapid population growth, and fertility has been falling rapidly since the early 1980s.This fall has not been steady or even. The 1980s saw a steep decline in total fertility rate (TFR) from 6.5 to 3.3 by the early 1990s. This was followed by a decade-long plateau which was the consequence of a \\\'tempo effect.² The adoption of FP by Bangladeshi couples has always been after the first birth. The age at marriage did not change and there was no delay in age at first birth, and as such, no tempo effect was operating on first births. The 2004 Bangladesh Demographic and Health Survey (BDHS) showed the first nine percent reduction in fertility (TFR of 3.3 to 3.0) for a decade. The 2011 BDHS confirmed a further decline in TFR to 2.3 children per woman. Now, however, fertility levels are quite uneven - remarkably low in the west of the country (below replacement, on average) and worryingly high in the east (up to 1.5 children above replacement).
In order to attain any of the reasonable population estimates projected for mid-century (which range from 194 to 222 million)³ a substantial increase in the contraceptive prevalence rate (CPR) will be required in the next five years, and a CPR level in the order of 75% (mostly modern methods) by 2020. This target could theoretically be achieved if all current unmet need for FP (12% in 2011) were to be met.

There are a number of factors which influence future population growth. Bangladesh has considerable built-in population momentum because of high fertility in the past, and even with reduced fertility, many young women will pass through reproductive ages over the coming decades. For example, during the first decade of the 20th century, the number of women of reproductive age increased from around 32 million to 41 million as the children born in the higher fertility 1970s and early 1980s entered their childbearing years, according to UN estimates. This trend will continue for several decades.
There needs to be a demand for fertility limitation in order to reduce fertility in a non-coercive environment. The perceived value of children has long been recognized as being a determinant of desired family size.⁴ Historical demographic experience suggests that as recent investments in female primary and secondary education in Bangladesh manifest themselves in improved opportunities for formal sector employment for young women, parents will tend to favour smaller families, investing more per child in education-quality versus quantity. This trend will also be influenced by the saturation of the rural labour force and the fragmentation of agricultural land holdings such that there will be deceasing employment opportunities for unskilled workers.
Current status of the family planning programme
Bangladesh is unique in terms of having a large non-government organization (NGO) sector presence. While various donor partners have supported the national FPP since its inception, the country has also benefitted from the research and technical support of local NGOs. The International Centre for Diarrhoeal Diseases Research, Bangladesh (ICDDR,B) works closely with the government and has been instrumental in shaping the health programme over the years. A number of interventions, including FP integration with mother and child health (MCH), were tried and tested by ICDDR, before being incorporated in the national programme. Other NGOs, notably BRAC, Bangladesh, have provided extensive reproductive health services in urban Bangladesh where public health infrastructure is limited. One of the largest non-government FP-MCH service delivery projects is the United States Agency for International Development (USAID)-funded NGO Service Delivery Programme which serves a catchment population of 20 million. These NGOs work in partnership with the government in the provision of FP-MCH services.
Under the Directorate of Family Planning, the success of the Bangladesh FPP has been based since the 1970s primarily on a nationwide rural network of female fieldworkers called family welfare assistants (FWAs) who have been providing outreach services to couples, particularly married women with limited mobility outside the home or compound (bari). These FWAs were usually all married women, only moderately educated, and had a work schedule where they visited each of the houses in their coverage units every two months.
The outreach approach largely explains the contraceptive method mix; oral pills have always been the predominant method, currently accounting for almost half of all users. Condom use remains low, though serving a useful purpose, sometimes complementing other methods. The provision of clinical services like intrauterine contraceptive devices (IUCDs), injectables, or permanent methods require referral to higher-level staff, such as family welfare visitors (FWVs), and for some methods, clinical staff of the Directorate of Health.
The CPR has increased eightfold over the last four decades to 61.2%, in 2011.⁵ This is equal to about 25 million of 41 million eligible couples. The trend line for CPR suggests a plateau in 2004 when the level of injectables users fell by three percent in the BDHS 2007 due to a nationwide stock-out. The downturn in CPR recovered in 2007 when supplies became available again in 2008.⁶
There is wide recognition that to achieve replacement fertility or below, a much greater proportion of eligible couples will need to be using long-term and permanent methods. With the average age at marriage for women (apparently7) still well below the legal minimum age of 18 years, many women have completed their childbearing by their mid- to late 20s. This leaves them with 20 years or so of reproductive life to protect themselves from unwanted pregnancies.


Family planning method mix
Oral pills: Oral pills tend not to account for more than 30% of couples across the developing world with the minor exceptions of Algeria and Morocco. Bangladesh is close to that apparent ceiling. This is not to say that further testing would not be useful, with the provision of a second oral pill with different hormonal content to counteract present side-effects, or progestin-only pills for breastfeeding postpartum mothers, and of course, emergency contraception for responding to unintended pregnancies without recourse to menstrual regulation. Clinical methods have been declining both in absolute and relative terms.
Injectables: These account for one in ten eligible couples, a level similar to Nepal and Sri Lanka, but there is much greater potential than that. In the ICDDR,B Matlab field site, they account for 25% of couples, as they do in Indonesia, a Muslim country sharing similar population challenges and characteristics.
Intrauterine contraceptive devices: These are virtually non-existent, having declined to less than one percent of couples. This method has achieved great popularity in other developing countries with similar levels of clinical services, but the reluctance to adopt IUCDs in Bangladesh remains widespread. Historically, there has been a perception that Muslim women will not be comfortable with any method that increases bleeding and extends periods of impurity. This is why IUCDs were only promoted in Hindu Bali within the Indonesian FP programme in its early years, and not in the predominantly Muslim provinces.
But three other Muslim countries have high levels of IUCD use — Egypt at 36% (2008 DHS), Jordan at 22% (2007 DHS), and Indonesia at five percent across all provinces (DHS 2007). They are also widely used in China and Vietnam. This would be a suitable time to promote new hormone-impregnated IUCDs, like Mirena, which reduce bleeding and do not exacerbate anaemia.


Implants: These have received some exposure in Bangladesh but never reached a significant level. Now that the multi-rod Norplant has ceased production and newer single or dual rod models such as Implanon, Jadelle, and Sino-Implant, are available, it may be time to promote this relatively low-cost, low-maintenance method, which could be a useful long-term temporary method.
Sterilization: Sterilization declined to one in ten users from its peak of one in three users (9.1% of eligible couples in 1991) in the 1980s when the incentive paid to fieldworkers and agents to recruit tubectomy clients, was stopped.
Male sterilization: This has never been significant (peak of 1.5% of couples in 1985) and continues to decline. While it is common in a number of western countries, it is almost non-existent in developing countries, with the exception of Bhutan — a very small country. Nevertheless, male involvement in FP is a subject of considerable discussion and research hand it may be hoped that levels higher than the present one percent can be achieved with a better understanding of the barriers and misunderstandings. These include men suffering weakness and not being able to work productively after vasectomies, and the common confusion between sterilization and castration.
Female Sterilization
There is reason to expect that female sterilization could be made more popular and achieve higher than the current five percent of eligible women. India has a long and controversial history of reliance on tubectomy. Currently, 37% of Indian women are sterilized. Other countries have achieved high levels too, although not necessarily without coercion. These include China, Iran, Thailand, Nepal, and Sri Lanka. But to achieve this in Bangladesh, high-quality services where couples do not fear surgical procedures will be required.
It is noteworthy that the decline in female sterilization in Bangladesh as a percentage of couples conceals a small increase among women aged under 30 years, which has been more than counterbalanced by a decline among older women who are now \\\'ageing out\\\' of their reproductive ages. So there are some encouraging signs of an upturn in sterilizations (from a low base), due partly to activities in northern Bangladesh.
It needs to be remembered that with this population size, an increase of more than 300,000 couples is required to register a one percentage point increase in national CPR.
Finally, geographic differences are persistently very large as might be expected from the geographic fertility differentials (Figure-3).

Although there are substantial geographic differentials in age at marriage, it is really the FP differentials across divisions that determine fertility levels. The CPR is much higher in the western districts than in the eastern districts of the country. This geographic pattern of wide fertility differentials calls for a targeted approach focusing mainly on the current high fertility divisions in the east of the country. Then again, when division-level populations are considered, a greater national impact may be achieved by targeting the much larger central divisions like Dhaka, where fertility has reached replacement level of 2.2 children. More scenarios need to be developed along these lines.
Table shows the method mix — as proportions, not percentages - for modern and any traditional methods, in 2010.⁸ It also makes a rough projection to 2015 for a more desirable method mix, rather than the current one which is overly dependent on oral pills and under-represents long-acting and permanent methods (LAPM).
If the replacement level fertility target is to be achieved by 2016, Bangladesh will need a CPR somewhere around 74% for all methods, and 69%for modern methods.

Table-1 is an approximation showing the additional requirements if such a method mix is to be achieved in a situation where the number of women of reproductive age increases by 8.3% (41.06 million to 44.47 million) and the number of currently married couples increases by 9.6%, assuming constant age-specific marriage rates. This latter assumption may lead to a small overestimate of the number of married couples, e.g., if marriage starts to be delayed.
The desired pattern results in a decline in the number of oral pill users (-1.5 million), condom users (-225,000), and traditional methods (-1.2 million), with increases in all other methods. Intrauterine contraceptive device users would need to increase by 1.4 million, injectable users by 2.7 million, implant users by 1.5 million, female sterilization users by 1.7 million, and male sterilization acceptors by 1.5 million. These are very substantial increases in the numbers of modern method users and have implications for staffing, training, logistics, supplies, and financing.
Family Planning Market Share of Public, Private, and NGO Providers
According to the 2011 DHS, over 52% of all users of modern contraception obtained their supplies from the public sector, followed by the private sector (around 43%),and NGOs\\\' (five percent). Bangladesh\\\'s public sector is the primary provider of clinical methods whereas the private sector, comprising primarily Social Marketing Company (SMC) pharmacies, is increasingly becoming an important source of temporary methods. The SMC is a not-for-profit social marketing organization selling its own brand of pills, condoms, and injectables at subsidized prices through private sector outlets.
Programmatic Challenges
Two wings of the Ministry of Health and Family Welfare
As a consequence of the bifurcated structure within the Ministry of Health and Family Welfare (MOHFW), lower-level workers in the Directorate of FP have to refer couples to the clinical staff of the Directorate of Health Services (HS) for FP methods such as implants and male or female sterilization. This involves the couples also having to go to the Upazila Health Centre or District Hospital under the Directorate General of Health Services (DGHS). It was to circumvent this cross-directorate barrier that the Directorate General of Family Planning (DGFP) upgraded 60 of the 90 maternal and child welfare centres (MCWCs) during the 1990s. Even though these female-staffed centres are all virtually located in the district headquarters, they have been very popular with women seeking high-quality clinical contraception.
This issue of the desirability of merging the two wings of the MOHFW has been discussed for many years. Indeed, the start of the first sector-wide programme, the Health and Population Sector Programme (HPSP), was delayed by one year (1998 instead of 1997, as planned) over a disagreement between the MOHFW and its development partners on this issue. One of the aspects of this perennial issue is that if the two directorates were merged, many of the merged positions would be taken by the staff of the DGHS.
Staff-related challenges
Many DGFP staff are now approaching retirement age. The first batch of FWAs was hired in the 1970s, followed by a second batch of 10,000 or so, in the late 1980s. Many of these 23,000 staff members are now in their 50s. Decisions have to be made about replacing them, especially as it is projected that some 18,000 community clinics (CCs) will require an FWA on the staff.
Many of the FWVs, who are responsible for clinical contraception (inserting IUCDs, giving injectables) and referring for permanent methods, are approaching retirementage, as well. There are plans now to recruit new FWVs as part of the discussion about strengthening the Union Health and Family Welfare Centres (UHFWCs). This issue comes up from time to time when considering the possibility of providing round-the-clock service for safe delivery by FWVs at UHFWCs. Although many UHFWCs have been designed to provide accommodation for one FWV, they are rarely used for that purpose.
Another issue that continues to trouble the FPP is that of vacancies in remote areas. The low-performing division of Sylhet in the east of the country is plagued by long-term vacancies, partly because it is economically advantaged compared to other rural divisions. Sylheti women, both for economic and religious reasons, are reluctant to take up fieldworker (FWA) posts. Thus, those who do take this level of employment tend to be from the minority religious sectors who have limited access to the majority Sylheti households.
There are proposals to try providing services in Sylhet division with some combination of public and private sector workers. Outreach workers may well need to be kept in the two eastern divisions, Sylhet and Chittagong, where FP use is far lower than national rates.
It should not be ignored that the training materials for the incoming FWAs and FWVs may need updating. It is believed that many of the materials have not been updated in two or three decades and do not incorporate best practices or World Health Organization (WHO) and other international standards and guidelines.
When the CCs were designed under the HPSP (1998-2003), too many tasks were added to the responsibilities of the FWAs and health assistants (HAs) who were expected to staff the CC. It should be mentioned that the catchment populations of the field staff have greatly increased since the system was first designed and an average FWA was given 800 households to cover. Population growth has resulted in up to double that number, now.
At present, FWAs are only authorized to deliver oral pills and condoms. Some FWAs have been trained to deliver injectables in conjunction with Blue Staroutlets — Blue Star is an SMC social franchise programme that recruits and trains private sector practitioners, mainly non-medically trained pharmacists, to deliver injectables in addition to providing other FP/health services. However, the current pool of FWAs is not adequate to administer injectables. If the prevalence of injectable users is to be increased to 5.3 million-each of whom needs four injections, annually — there would need to be sufficient staff to give over 21 million injections, annually. The existing 5,000 FWVs, many of whom will soon retire, cannot possibly serve such numbers. Similar issues relate to nurses and their authority and training to insert IUCDs.
Issues with logistics and procurement
The SMC procures its FP supplies independently through USAID support and distributes them through a network of over 200,000 pharmacies across the country. The government of Bangladesh procures almost all FP commodities distributed by the public and NGO sectors.⁹ The FP logistics system within the country is satisfactory, but obstacles to timely procure FP commodities internationally are creating stock-outs at a time when unmet demand for FP needs to be met for the CPR to increase.
A two-year supply of FP commodities can take more than 18 months to procure due to a cumbersome chain of approvals; some 65 signatures within the MOHFW and three serial (not simultaneous) approvals from the World Bank/Washington. This lengthy process applies to routine procurements, not to new items.
Some countries facing similar delays are switching from two-year procurement cycles to five-year approval cycles with simple annual purchase orders within the approval period. New ways for procurement are needed, but some of the barriers must first be recognized.
Issues with contraceptive procurement mechanisms
A few countries have relied on external agencies like the United Nations Population Fund (UNFPA) to procure all their FP needs, but there have been concerns about procurement charges (five percent in general), tax exemption (UN agencies normally expect to be accorded this status), and bureaucratic delays common in large agencies. On the other hand, such approaches can take advantage of economies of scale in negotiating lower prices for multi-country orders. A good example is Peru; between 1999 and 2007, the country successfully shifted its contraceptive procurement from using public funds to purchase all commodities through an agreement with UNFPA, to an arrangement whereby a local NGO, PRISMA would purchase commodities from local suppliers at competitive prices.¹⁰ This was done chiefly through encouraging private and public sector providers to pool their contraceptive purchases in order to negotiate a lower per-unit price.
Forecasting contraceptive requirements
The main set of forecasts that is used by the DGFP is from 2001.¹¹ This has now been replaced by a new project, \\\'strengthening pharmaceutical systems.\\\' But it should be said that the MOHFW does not rely solely on these forecasts when deciding on procurement. The size of orders is much more dependent on funding availability than on estimated requirements.
The completion of an order takes 18 months provided there are no obstacles. It can take up to three years if there are problems with specifications or approvals. For example, an order of 100,000 IUCDs which arrived in Dhaka recently was rejected because the good manufacturing practice (GMP) certificate did not meet requirements. This will require retendering. An order of 4.2 million injectables was also rejected recently due to manufacturing problems.
The SMC has fewer bureaucratic obstacles and can complete an order in 8-9 months. The MOHFW system was also faster in the fourth programme (1992-1997) than under the sector-wide approach (SWAps) — 1998 to present. The government\\\'s annual budgeting system creates some difficulties for procurement as the funds for contraceptives are not guaranteed in the budget. There was a shortage of such funds in 2010-11. The government estimated a USD 45 million need, but received only USD 6 million.
Local production of contraceptives
Procuring commodities locally could reduce some of the delays. But there are some barriers to local manufacturers bidding as the requirements for using World Bank International Development Association (IDA) funds are that the company exports to at least two European G8 countries. This is proving difficult for the German-Bangladesh Condom factory, even though it is exporting to certain countries, for example, in the Middle East. The government\\\'s own Essential Drug Company Ltd. (EDCL) is making condoms and supplying about 50%of the total 9.5 million monthly condom requirements. Apparently, IDA funds cannot be used to purchase them.
As pills are the major national method, the government is the main supplier at nine million, monthly, followed by the SMC at 3.5 million, monthly. One company, Renata has started production, but the government is not buying them yetas it cannot use IDA funds for this. It is planned that Renata will eventually supply 30 million cycles, annually to the SMC. The older and more expensive brands, Marvelon and Ovostat, are also manufactured, locally. Popular is manufacturing a brand of emergency contraception and Incepta has plans to commence production.
IUCDs are a very minor method used by just 0.9%of couples, so requirements are only 22,000, monthly. It is most unfortunate that this cost-effective method is so underutilized when the unit cost is only 43 cents, thus making the annual requirement only USD 110,000. It would not be worthwhile for Bangladesh to attempt local production of an item in such small numbers. In addition, the new hormone-impregnated IUCDs will require more sophisticated production facilities and are probably not worth considering at this time.
About one million injectables (Depo-Provera) are required, monthly, including 100,000 coming from the SMC. One local Bangladeshi company, Technohaven, is producing injectables, but again, the government cannot use IDA funds to procure them. Implants account for a small proportion of users at 0.7%, but the cessation of Norplant production has created some problems with delays in selecting a replacement.
Family Planning Financing
The funding for the national FPP comes from the Health, Nutrition, and Population Sector Programme (HNPSP) with additional commodities supplied from USAID and other funding agencies like UNFPA and Deutsche Gesellschaftfür Internationale Zusammenarbeit (GIZ). The proportion of HNPSP allocated to the FPP is only three percent — much less than in other comparable developing countries.
It is preferable for an essential programme like the FP programme to be funded by a permanent stable source like the government rather than by development partners. There are also a variety of ways in which required funding can be reduced. For example, by segmenting the market into sectors where well-off clients buy FP products at the full commercial cost (as is the case in most developed countries), middle-income clients obtain subsidized supplies through organizations like the SMC, and the poor receive free FP commodities.
Funding for the social marketing company
The USAID/Washington provides SMC with USD 7-8 million in commodities every year. They sell a variety of pills, but Femicon accounts for 80% of their pill sales. The injectable, Depo-Provera is sold under the name Somaject through their 3,600 Blue Star outlets.
Incentives and disincentives for adopting family planning
Another approach used in many national FPPs has been incentives - the use of payments to clients and/or service providers, and in some cases, disincentives — the use of penalties for not adopting FP, or alternatively, for having more than a certain prescribed number of children. China is an obvious example of the latter, and other examples include Indonesia (at one time, public servants would not receive their salaries if they exceeded two or three children) and Singapore (at one time, a third child would not be permitted to attend school).
There was a major push during the 1980s to increase the utilization of permanent and long-acting methods. Incentives were offered directly to sterilization acceptors and these proved effective in increasing uptake - from 189,782 in 1979-80 to 552,167 in 1983-84. However, the incentives were offered to recruiting agents or FP field staff, and it became apparent that pressure was being applied in some cases. The incentive payment for the sterilization client scheme was closed down in the late 1980s.

The FP programme in Bangladesh is much talked about; the contraceptive-use rate increased eightfold over four decades with the current CPR standing at 61%, roughly double the rate observed in a similar Muslim country like Pakistan. After the programme\\\'s initial success, the rate of CPR increase slowed down, and at one point, even declined. This phenomenon is attributable to both supply-and demand-side issues. There is no doubt that an innate demand for smaller families has to be created. Further research is needed to try and understand why contraceptive uptake is low in the east of the country before a regionally-targeted approach can be adopted. One of the central messages in this paper is that with an over reliance on temporary methods, the current method mix cannot support a sustained fertility decline. The immediate need to promote long-acting and permanent methods cannot be overemphasized if the country is to achieve replacement-level fertility by 2016. This would require behaviour change communication (BCC) to remove socio-cultural barriers, and addressing inherent problems within the MOHFW. We identify the challenges and loopholes in the current programme that range from staffing and funding deficiencies to complicated FP procurement procedures. A more efficient mechanism of procurement and logistics will minimize stock-outs and ensure smoother programme implementation at the grassroots level. Greater flexibility in funds usage and a more decentralized approach are among the possible actions that could be adopted by policymakers to ensure that national population targets are met. Many of the issues raised in this paper have wider implications and may provide context for countries like Pakistan.

1. Kamal N. Population trajectories of Bangladesh and West Bengal during the twentieth century: A comparative study. London School of Economics and Political Science; 2008.
2. Bongaarts J. and Feeney G. On the Quantum and Tempo of Fertility. Popul Dev Rev 1998; 24(2): 271-91.
3. Streatfield PK and Ahsan AZ. Population challenges for Bangladesh in the coming decades. J Health Popul Nutr 2008; 26(3): 261-72.
4. Cain M. The economic activities of children in a Bangladesh village. Popul Dev Rev 1977; 3(3): 201-27.
5. National Institute of Population Research and Training. Mitra and Associates, and Macro International. Bangladesh Demographic and Health Survey 2011. Dhaka and Calverton.
6. Utilization of essential service delivery (UESD) survey. Dhaka: National Institute of Population Research and Training,Associates for Community and Population Research.; 2010.
7. Streatfield PK, Kamal N, Ahsan KZ and Nahar Q. Early marriage in Bangladesh: Not as early as it appears. 2013.
8. Bangladesh Maternal Health Services and Maternal Mortality Survey 2011; Government of Bangladesh, USAID, UNFPA.
9. Bates J. Bangladesh: Contraceptive procurement in transition — How context can affect outcomes. United States Agency for International Development (USAID); 2008.
10. Lopez-Linares R. Contraceptive procurement in Peru: Diversifying suppliers. Arlington (VA): United States Agency for International Development (USAID); 2008.
11. Islam MA and Chakraborty N. Projected contraceptive commodity requirements: 2000-2015. Dhaka: University of Dhaka, Department of Statistics, for the US Agency for International Development; 2000.