Measuring Unmet Need for Contraception Using a Person‐Centered Algorithm: An Application With a Community‐Based Sample of Married Rohingya Women in Bangladesh

Abstract

The standard measure of unmet need for contraception is not person‐centered and may not adequately represent women's contraceptive needs. To demonstrate the strength of a modified measure, we replicated the standard algorithm for unmet need, then created a person‐centered algorithm that considers (1) whether nonusers want to use contraception and (2) whether users want to use a different method. We applied the standard and person‐centered algorithms to a sample of 847 married Rohingya women aged 15–49 years living in camps in Cox's Bazar, Bangladesh, a population about whom little is known regarding contraceptive need. Forty‐six percent of respondents were currently using contraception. Among users, 14 percent wanted to use a different method and 36 percent of nonusers wanted to use a method. Using the standard algorithm, 39 percent had “unmet need,” 18 percent had “no need,” and 44 percent had “met need.” Using the person‐centered measure, 24 percent had “unmet need,” 38 percent had “no need,” and 38 percent had “met need.” The standard algorithm may overestimate unmet need among Rohingya nonusers, and the person‐centered measure provides evidence of method dissatisfaction among users. This measure also complements existing person‐centered measures of need and is an example of how incremental change can improve our understanding of women's contraceptive needs.

BACKGROUND

First developed in the 1980s, the measure “unmet need for contraception” has become a ubiquitous family planning indicator. The primary goal of the measure is to provide a population‐level estimate of the proportion of women who, based on their two‐year fertility intentions, are determined to have a “need” for contraception but are not using any contraceptive method. Today, unmet need is a key indicator that is heavily relied upon by governments and nongovernmental organizations (NGOs) to set and track population development goals and to design family planning programs (Speizer, Bremner, and Farid ²⁰²²; Senderowicz ²⁰²⁰).

Researchers in the field of family planning measurement have made efforts to address weaknesses in the unmet need measure. Bradley and Casterline (2014) extensively tested its robustness by estimating unmet need in various geographical contexts and under different assumptions about sexual activity, fertility intentions, postpartum amenorrhea duration, fertility intention reporting, infecundity, and other factors while acknowledging that the measure fails to include individual contraceptive preferences. Speizer, Bremner, and Farid (2022) highlighted how the language utilized in contraceptive need measurement contributes to misalignment between what measures—including unmet need—claim to estimate and what they actually capture, ultimately leading to misinterpretation of women's needs.

Moreover, the conceptualization, application, and interpretations of the measure have raised concerns. In their body of work on unmet need for contraception, studies of Senderowicz et al. highlight the limits of using unmet need as a proxy for contraceptive desires and the importance of incorporating women's self‐reported desires in unmet need and family planning indicators more broadly (Senderowicz 2020; Senderowicz et al. ²⁰²³; Senderowicz and Maloney ²⁰²²). Furthermore, they emphasize that one of the key issues with the measure is that it was designed to meet policymakers’ desires for a contraceptive indicator, rather than centering on individuals who may become pregnant (Senderowicz et al. 2023; Senderowicz and Maloney ²⁰²²). To address the measure's limitations concerning person‐centeredness, Rothschild, Brown, and Drake (2021) proposed incorporating women's contraceptive dissatisfaction into unmet need, while Moreau et al. (2019) and Senderowicz and Maloney (2022) advanced alternative frameworks centered around incorporating contraceptive demand into the indicator. Karra (2022) introduced a novel approach to estimating unmet need using population‐level data, while Bornstein et al. (2023) developed a construct focused on self‐reported contraceptive need. Additionally, Holt et al. (2023) and Senderowicz (2020) proposed alternative contraceptive indicators—preference‐aligned fertility management (PFM) and contraceptive autonomy, respectively—that redefine what constitutes success in family planning by shifting away from contraceptive use as the sole ideal outcome.

The provision of sexual and reproductive healthcare is a fundamental human right essential to limiting the burden of mortality and morbidity among women and girls in humanitarian settings. Currently, critical data gaps limit the existing evidence that informs service provision aiming to meet the needs of these populations (Warren et al. 2015; Singh et al. ²⁰¹⁸). Existing data on contraceptive use, nonuse, and need in humanitarian settings are sparse and typically rely on measuring “unmet need” for contraception. This measure, which aims to determine the proportion of women of reproductive age who do not want to become pregnant soon or at all, but are not using contraception, does not account for an individual's preferences, desires, and values around contraception making it inherently not person‐centered. More evidence using person‐centered approaches to measuring contraceptive need is essential to tailoring service provision for people in humanitarian settings.

The Rohingya are a predominantly Muslim ethnic minority who have lived in majority‐Buddhist Myanmar as a stateless group. They have been denied citizenship and access to fundamental human rights including education and healthcare. In 2017, nearly 750,000 Rohingya people were forcibly displaced from Myanmar to Bangladesh (Rohingya Emergency ²⁰²²). Today, about one million Rohingya, or forcibly displaced Myanmar Nationals, reside in 33 camps in the Cox's Bazar district of Bangladesh near the Myanmar border (Siddiqi 2021). In the camps, the Rohingya people rely on humanitarian aid from local and international NGOs and international bilateral agencies for education, food, clean water, and healthcare among other essential services (How to Help Rohingya Refugees, n.d.). Health facilities in the camps provide a range of free reproductive health services to Rohingya women, including contraceptive methods, menstrual regulation (MR),¹ and postabortion care (PAC). Methods offered at health facilities include intrauterine devices (IUDs), implants, oral contraceptive pills, emergency contraceptive pills, injectables, and condoms (Casey, Puls, and Jacobi ²⁰¹⁹). While contraceptive use in the camps has been documented (Khan et al. 2021; M. M. Islam, Khan, and Rahman ²⁰²¹; Abul Kalam Azad et al. ²⁰²²), less is known about the prevalence of unmet need for contraception among Rohingya women.

In this analysis, we measure contraceptive dynamics among Rohingya women residing in Cox's Bazar using both the standard measure of unmet need as well as a more person‐centered algorithm that accounts for contraceptive desires among both users and nonusers. This draws on the aforementioned conceptual and practical works by scholars critically engaging with unmet need. By examining the degree to which estimates of unmet need, met need, and no need among Rohingya women differ between the standard and person‐centered algorithms, we provide a critical examination of the standard measure and generate new evidence that highlights its limitations while proposing an approach for better capturing and understanding contraceptive dynamics, including in fragile contexts.

METHODS

Study Design, Setting, and Participants

This analysis was part of a larger study that was a collaborative effort between the BRAC James P. Grant School of Global Public Health, BRAC University, Bangladesh; the Association for Prevention of Septic Abortion, Bangladesh (BAPSA); and the Guttmacher Institute, USA. The study used a multistage systematic random sampling method to sample ever‐married Rohingya women of reproductive age (15–49 years) living in the camps in Cox's Bazar, Bangladesh. Ethical approval of the study protocol was obtained by both the Institutional Review Boards of the BRAC James P. Grant School of Public Health, BRAC University, and the Guttmacher Institute.

We fielded a community‐based survey (CBS) to gather information on a range of topics, including Rohingya women's reproductive history, family planning and fertility preferences, knowledge and management of unwanted pregnancy, and perceptions and knowledge of MR. The design of the CBS drew on previous CBS instruments from other countries and settings that were developed and fielded by the Guttmacher Institute and local researchers. For this study, the research team adjusted the CBS to incorporate context‐specific factors related to the Rohingya people, such as questions about their perceptions and experiences with sexual and reproductive healthcare in Myanmar and the camps, including MR, a practice that is legal in Bangladesh up to 12 weeks after a woman's last menstrual period but legally unavailable in Myanmar (Hossain et al. 2017; World Health Organization, Regional Office for Southeast Asia ²⁰²⁰). The design of the family planning and fertility intentions sections was also informed by ongoing work on the measurement of unmet need at the Guttmacher Institute and aimed to allow for the exploration of nuanced ways to measure unmet need. The CBS survey was developed in English, translated into Bangla, and then back‐translated into English to confirm the accuracy of the translation.

The survey was piloted in 60 households, revised, and then fielded September–November 2022 in four camps by 18 trained female Bangladeshi interviewers. We had no existing estimates of the study's primary outcomes (MR and PAC) within our study population prior to our study, and so we sought a sample of at least 275 eligible women in each of the four camps. This sample size enabled us to detect variations in estimates between camps while also ensuring that interviewers could complete all interviews within the period of time they were authorized to enter the camps. In each household, fieldwork supervisors identified the number of ever‐married women aged 15–49 years (ever‐married minors aged 15–17 were treated as emancipated) and used a Kish grid to select one eligible respondent per household. The response rate among the 1,260 eligible respondents identified was 93.1 percent. Data were analyzed using Stata 18.

Variables and Measurement

The survey question we used to determine contraceptive use was: “Are you currently doing anything or using any method to delay or avoid getting pregnant?” For determining infecundity, we used the self‐reported responses of being in menopause or having had a hysterectomy.²

There are two key distinctions between the standard algorithm (Figure 1) (Bradley et al. 2012) and the proposed person‐centered algorithm (Figure 2). First, the person‐centered algorithm considers two factors that the standard algorithm does not: (1) whether users wish they were using a different method by asking, “Do you wish you were using a different method?” with the aim of capturing method dissatisfaction, similar to Rothschild, Brown, and Drake (2021), and (2) whether nonusers wish they were using a method, by asking, “Do you want to be using a method of family planning?” with the aim of capturing self‐reported contraceptive need, similar to Bornstein et al. (2023) and Holt et al. (2023).

FIGURE 1.

The standard unmet need algorithm, 2012 revision (Bradley et al., 2012)

FIGURE 2.

Person‐centered algorithm of unmet need used to estimate unmet need among the Rohingya

Second, the standard measure considers fertility intentions when assessing whether nonusers are experiencing unmet need through the questions, “Would you like to have (a/another) child, or would you prefer not to have any (more) children?” and “How long would you like to wait from now before the birth of (a/another) child?” whereas the person‐centered algorithm decouples contraceptive use from fertility intentions and does not ask about fertility preferences (Bradley et al. 2012). This separation challenges the assumption that contraceptive use is always linked to fertility desires, supporting evidence that women may use contraception for reasons other than avoiding or delaying a pregnancy (Frederiksen, Diep, and Salganicoff ²⁰²⁴) and that there are legitimate reasons for nonuse even when not actively trying to conceive.

This analysis first replicates the 2012 standard unmet need algorithm (Bradley et al. 2012), with the exception of two components: (1) whether a respondent is postpartum amenorrheic as this information was not available in our dataset; and (2) the inclusion of currently pregnant women. Our survey did not include retrospective or prospective questions about contraceptive desire for currently pregnant women, preventing us from including them in the person‐centered algorithm calculation. As we therefore could not compare the estimates of each algorithm for currently pregnant women, they were excluded from the analysis altogether. Additionally, the standard unmet need algorithm uses a calendar method to determine infecundity whereas our study relies on self‐reports of menopause or having had a hysterectomy to determine who is infecund.

In the standard algorithm, respondents are considered to have “no need” for contraception if they were (1) infecund or (2) not using a method and wanted a child within two years (Table 1). The person‐centered indicator classifies respondents as having “no need” if they were infecund or did not want to use or were unsure about using contraception. For “met need,” the standard algorithm classification includes all current users; however, the person‐centered algorithm only classifies current users who did not wish to use or were unsure about using a different method as having “met need.” In our sample, we treated the responses “Don't know/Unsure” (n = 3) to the contraceptive desire questions as indicative of ambivalence to use and classified those respondents as having either a “met need” or “no need” rather than an “unmet need.” While “don't know” responses may reflect a need for contraception, we cannot assume that uncertainty necessarily indicates a lack of knowledge about contraception or a need for it. Classifying all women who responded “don't know” as having a need could lead to overestimating a population's true level of need.

TABLE 1.

Classification of respondents into the unmet need, met need, and no need categories using the standard and person‐centered algorithms

Standard algorithm			Person‐centered algorithm
Unmet need	Met need	No need	Unmet need	Met need	No need
Nonusers who do not want another child at all	Current users	Nonusers who want a child within the next two years	Users who wish to use a different method	Current users who do not want to use a different method or are unsure about wanting to use a different method	Nonusers who do not want to use a method or are unsure about wanting to use a method
Nonusers who do not want a child within the next two years		Respondents classified as infecund	Nonusers who wish to use a method		Respondents classified as infecund

In the standard algorithm, nonusers who do not want a/another child ever or within the next two years and are not using a method are ascribed “unmet need.” The person‐centered algorithm assigns “unmet need” to users who wished they were using a different method and nonusers who wished they were using a contraceptive method, regardless of if or when they wanted a/another child.

Since the person‐centered algorithm does not incorporate fertility intentions and classifies respondents differently than the standard algorithm, it does not further distinguish “met need” into “using to limit/space” or “unmet need” into “unmet need for limiting/spacing” as the standard algorithm does.

Study Size

Of the 1,173 respondents interviewed, 42 widowed women were excluded from the analytical sample, as they were not asked questions about their fertility intentions; 46 separated/divorced women were excluded to ensure that our findings were comparable to existing estimates of unmet need among currently married women only. Respondents who were currently pregnant at the time of the survey (n = 173) were excluded from the analysis because they did not answer contraceptive preference questions.

An additional 39 respondents who stated they were using injectable contraception but who received their last injection over three months before the survey were excluded. The survey routing of these respondents was revised during fieldwork, resulting in some being routed to the contraceptive desire question applicable to users, and others being routed to the contraceptive desire question applicable to nonusers. Due to the inconsistent assignment of user or nonuser status of these respondents, they were excluded from the analysis. One respondent was excluded due to a data entry error.

The number of respondents for whom we had incomplete data for variables used in either the standard or person‐centered algorithms was negligible (n = 25); these cases were excluded from the study sample. The final analytical sample included 847 women with complete information on all the variables of interest.

RESULTS

Participants

Of the 847 respondents included in the analysis, 95 percent were living with their spouses and almost two‐thirds (65 percent) were aged 20–34. The same proportion of respondents had no formal schooling, while a quarter had less than five years of schooling. Just 7 percent of respondents had never given birth, while over half of respondents (54 percent) had given birth four or more times; 5 percent self‐identified as infecund (Table 2).

TABLE 2.

Demographic Characteristics and Reproductive History, Community‐Based Survey, 2022

	N = 847	Frequency	Percentage
Age	15–19	55	6.5
	20–24	189	22.3
	25–29	200	23.6
	30–34	163	19.2
	35–39	110	13.0
	40–44	82	9.7
	45–49	48	5.7
Highest education level completed a	None	554	65.4
	Some (<5 years)	214	25.3
	Primary or more	56	6.6
Lives with husband b	Yes	808	95.4
	No	38	4.5
Sexual frequency in the past month	0	93	11.0
	1	33	3.9
	2 or more	685	80.9
	Refused to answer	18	2.1
	Don't know/unsure	18	2.1
Parity c	0 births	60	7.1
	1–3 births	324	38.3
	4–7 births	344	40.6
	8–15 births	117	13.8
Fecundity	Fecund	805	95.0
	Infecund	42	5.0

^aTwenty‐three did not know or the data was invalid;

^bOne observation where the data was invalid;

^cTwo missing observations.

Descriptive Statistics

Among self‐identified fecund respondents, 46 percent reported using a contraceptive method at the time of the survey. The two most common methods used were injectables and birth control pills (46 percent and 45 percent of users, respectively), followed by implants (8 percent). Among users, 14 percent expressed wanting to use a different method, while 36 percent of nonusers reported wanting to use a contraceptive method (Table 3). Among both users and nonusers, the method most reported as the one they would rather use or like to use was injectables (55 percent and 46 percent, respectively). No respondents reported wanting to use a non‐biomedical method, such as withdrawal, lactational amenorrhea, or periodic abstinence/rhythm method (data not shown).

TABLE 3.

Contraceptive Use and Desires, and Fertility Intentions, Community‐Based Survey, 2022 ^a

		Frequency	Percentage
Current contraceptive use (n = 805)	Yes	369	45.8
	No	436	54.2
Contraceptive methods used (n = 369) b	Injectables	171	46.3
	Birth control pills	165	44.7
	Implants	31	8.4
	IUD	6	1.6
	Other c	4	1.1
Desire to use a different method (users only, n = 369)	Yes	50	13.5
	No	318	86.2
	Don't know/unsure	1	0.3
Desire to be using a method (nonusers only, n = 436)	Yes	156	35.8
	No	278	63.8
	Don't know/unsure	2	0.5
Desire to have another child (n = 799) d	Yes	453	56.7
	No	287	35.9
	Up to Allah	59	7.4
How long would you like to wait before having a/another child? (n = 512)	Less than a year	27	5.3
	1–2 years	19	3.7
	At least 2 years	177	34.6
	More than 2 years	197	38.5
	Up to Allah	92	18.0

^aExcludes respondents determined to be infecund (n = 42).

^bMultiple response questions, eight respondents reported using two methods, so the percentage sums to >100 percent.

^cincludes female sterilization (n = 2), lactational amenorrhea (n = 1), menstrual regulation (n = 1).

^dSix missing answers.

At the time of the survey, 57 percent of respondents reported wanting a/another child, while 7 percent reported that whether they have a/another child is “up to Allah.” Of these women, almost three‐quarters (73 percent) reported wanting to wait for two years or more, while 18 percent said that the timing is “up to Allah.” Among users, 53 percent wanted to have a/another child, and among these respondents, a large majority (87 percent) wanted to wait two years or more. Among nonusers, 59 percent wanted to have a/another child; 63 percent wanted to wait two years or more, while a quarter (24 percent) reported that the timing is “up to Allah” (data not shown).

Unmet Need Using the Standard and Person‐Centered Algorithms

Using the standard measure, the proportion of women in our sample experiencing “unmet need” was 39 percent. Using the person‐centered measure, this proportion decreased to 24 percent (Figure 3). The proportion of women assigned “no need” was 18 percent using the standard measure and increased to 38 percent using the person‐centered measure. These differences are driven by the nonusers who were assigned “unmet need” with the standard algorithm because they did not want a/another child within the next two years or at all, and who were instead assigned “no need” using the person‐centered measure because they did not want to use a contraceptive method.

FIGURE 3.

Estimates of unmet need, met need, and no need using the standard and person‐centered algorithms

The proportion of women ascribed “met need” decreased from 44 percent using the standard measure to 38 percent using the person‐centered measure. This decrease is driven by the current contraceptive users who wished they were using a different method and were assigned “met need” using the standard measure and instead assigned “unmet need” using the person‐centered measure.

DISCUSSION

Our findings support and further emphasize the limitations of using fertility intentions, combined with contraceptive use, as an indicator of contraceptive need, and the importance of developing and applying more person‐centered definitions and measurements. When using the standard algorithm, we found that estimates of unmet need and met need for contraception among our respondents were higher compared to the same estimates using the person‐centered algorithm. In contrast, the estimate of no need was higher using the person‐centered algorithm. The largest observed difference is in the estimate of unmet need, which decreased by 15 percentage points when the person‐centered algorithm was applied. This suggests that the standard measure may overestimate the unmet need for contraception among Rohingya nonusers in Cox's Bazar.

When using the person‐centered algorithm, we also found that unmet need exists among current users in the form of method dissatisfaction. By contrast, the standard measure fails to capture this phenomenon (Rothschild, Brown, and Drake ²⁰²¹; Rominski and Stephenson ²⁰¹⁹; Steinberg et al. ²⁰²¹). Current contraceptive users are rarely, if ever, included in measures of contraceptive unmet need despite evidence that they experience method dissatisfaction, and this is a significant gap in family planning research (Rothschild, Brown, and Drake ²⁰²¹; Burke, Potter, and White ²⁰²⁰). Currently, the most used definition of contraceptive need is often too narrow because it is framed around contraceptive uptake with the goal of delaying or stopping childbearing. Future research would benefit if the definition of what constitutes contraceptive need was widened to include the need to switch or discontinue a current method.

Standardized family planning survey instruments and population‐based surveys rarely ask respondents directly about their contraceptive preferences and desires (especially among current users), despite these types of questions being essential to the development of person‐centered family planning measurements and programs (Senderowicz et al. 2023). In this study, we aim to address this issue by basing the measurement of unmet need for contraception on the expressed desires of respondents as opposed to inferring unmet need based on an externally perceived misalignment between reported contraceptive behaviors and fertility intentions. This enables us to recenter the “need” of unmet need on what an individual actually reports wanting. Our approach also helps address a key limitation of the standard measure described by Senderowicz, which is that unmet need for contraception is too often used as a proxy for contraceptive desire, despite not containing any data on an individual's desires (Senderowicz et al. 2023).

The person‐centered measure also addresses other limitations of the standard measure, including its assumption that all married women are sexually active and therefore at risk of pregnancy, and its reliance on self‐reported sexual activity among unmarried women where in many contexts, it may be underreported (Bradley and Casterline 2014; Cleland, Harbison, and Shah ²⁰¹⁴). While the standard measure incorporates recent sexual activity in its assessment of contraceptive need, the person‐centered measure presented here relies solely on individuals’ expressed contraceptive desires, regardless of reported sexual activity. As such, factors like sexual frequency or abstinence do not directly influence the classification of need within this framework, rather these factors may be implicitly included in an individual's family planning desires and preferences (i.e., an individual may not want to use contraception because they are abstinent). Additionally, this approach acknowledges that there are valid reasons for choosing to use—or not to use—contraception, regardless of sexual activity.

Exploring the contraceptive desires of Rohingya women is crucial for better understanding the dynamics of contraceptive use within this population and for generating evidence to guide the effective allocation of family planning resources. While Rohingya women's unmet need for contraception may be overestimated by the standard measure, this does not imply that the overall need for family planning services is also overestimated or that fewer resources should be devoted to them. Rather, a more precise assessment of family planning needs can inform how resources are allocated, potentially shifting resources to family planning services that are not as emphasized as contraceptive methods, such as contraceptive education, counseling, and sexually transmitted infection screening and treatment.

In recent years, international aid for the Rohingya people has been decreasing (World Food Programme ²⁰²⁴). In 2023, the UN Refugee Agency (UNHCR) called for an additional almost $900 million in funding to ensure the provision of shelter, food, education, and healthcare, among other essential needs, to the Rohingya people, but with many competing humanitarian demands, UNHCR is stretched thin and not always able to meet all need (Strangio 2023). In addition, recent suspensions of US foreign aid—which amounted to $301 million in 2024, accounting for more than half of all assistance to the Rohingya—further jeopardize this population's access to essential services (Shamim 2025). Given these constraints, it is essential to accurately identify which family planning services are most needed to ensure that available resources are optimized and that reproductive health services in the camps remain available. By aligning service provision more closely with individuals’ self‐identified needs, family planning programs can become more impactful.

Person‐centered measures of contraceptive need that do not rely on reports of fertility intentions are particularly important in populations such as the Rohingya, where women express types of fertility intentions that the standard measure of unmet need struggles to capture (e.g., a substantial proportion of our respondents reported that whether and when they have a/another child is “up to Allah”). The standard measure's lack of nuance and inability to capture dynamic intentions and belief systems when determining fertility intentions therefore limits its analytical uses in measures of need among the Rohingya and other populations. Family planning programmers should draw from the person‐centered algorithm findings that some nonusers who do not want to become pregnant may not be open to using contraceptive methods, and that current users experience contraceptive needs. This could lead to efforts that better meet the needs of Rohingya women. For example, the needs of current users could be addressed if they were given opportunities to change methods. Addressing women's reasons for being dissatisfied with their contraceptive method might also lead to more and better contraceptive counseling for both users and nonusers.

Future Research

In our study, though we asked current users about their desire to use a different method, we did not ask them about their desire to use contraception at all, meaning we did not capture the proportion of women who may want to stop using contraception altogether or who were coerced into using contraception. In addition, we specifically asked respondents if they were using contraception to avoid or delay getting pregnant. We therefore did not capture women who may have been using contraception for reasons other than delaying or avoiding a pregnancy, including for health reasons or sexually transmitted infection prevention (Claringbold, Sanci, and Temple‐Smith ^2019; What Are the Benefits & Advantages of Depo‐Provera?, n.d.; What Are the Benefits & Advantages of Birth Control Pills?, n.d.). Future research should ask questions about using contraception for other reasons than delaying or preventing a pregnancy.

To develop a more comprehensive understanding of the dynamics of contraceptive use and need that can more accurately convey programmatic needs, other constructs that shape contraceptive desires and use such as access (Senderowicz et al. 2023), autonomy (Senderowicz 2020; Loll et al. ²⁰¹⁹; Wollum et al. ²⁰²³), intention to use (Boydell et al. 2024), relational perspectives (e.g., couple dynamics) (Upadhyay, Raifman, and Raine‐Bennett ²⁰¹⁶; Kusunoki and Barber ²⁰²⁰), and agency must be considered in evaluations of need. While some of these constructs, such as agency, are still in the early stages of conceptualization (Holt et al. 2024), a person‐centered measure of contraceptive need that does not consider them provides a limited understanding of the drivers of contraceptive need and may prevent programs from allocating resources where they are most needed. In addition, measures of contraceptive need should account for other factors that may shape desires and use, such as abstinence, contraceptive coercion, fear, beliefs, and experiences around side effects and health impacts (Jonas et al. 2022; Dingeta et al. ²⁰²¹; M. Islam and Habib ²⁰²⁴; Polis, Hussain, and Berry ²⁰¹⁸), and fatalistic attitudes toward pregnancy (Jones 2018).

Limitations

Our measurement of infecundity relied on self‐reports; in the absence of biomedical information on fecundity, this may have led to less accurate estimations of infecund respondents. However, because an individual's own perception of their fecundity informs their contraceptive decisions (Gemmill 2018), we consider self‐reported infecundity more relevant to contraceptive desire than biomedical information.

The interviews were conducted in Rohingya, a spoken dialect that was not the native language of our interviewers. The interviewers, therefore, translated the Bengali version of the survey into Rohingya while interviewing the respondents. While keywords were standardized and interviewers were thoroughly trained in how to ask each question in Rohingya, this language barrier left room for potential variability in how questions were asked and how responses were understood.

Never‐married Rohingya women were not interviewed as the community would have seen it as inappropriate to include them because of the social prohibitions on their sexual activity, and their participation could have threatened their standing in their community. Excluding divorced/separated and never‐married women limits our ability to more accurately measure unmet need since these populations also have contraceptive needs (Sedgh, Ashford, and Hussain ²⁰¹⁶).

During some of the interviews, people other than the respondent and the interviewer were circulating in and out of the household, jeopardizing efforts to interview respondents privately. This lack of privacy may have affected some respondents’ willingness to answer sensitive questions truthfully, or at all.

CONCLUSION

Over half of the Rohingya people living in the camps in Cox's Bazar are women and girls, yet there is a gap in knowledge about contraceptive dynamics among Rohingya women as a whole as well as from person‐centered perspectives (Warren et al. 2015; Singh et al. ²⁰¹⁸; Rohingya Refugee Crisis Explained ²⁰²³). The findings of this study provide some insights into the nuances of contraceptive use among Rohingya women, using a person‐centered approach that highlights the limitations of the standard measure of unmet need, which operates under assumptions about contraceptive behaviors and fertility intentions that are externally determined and do not cohere with women's stated preferences.

Unmet need for contraception continues to be a heavily used indicator that influences family planning investment and policy decisions worldwide. While the person‐centered algorithm presented in this paper cannot address all the underlying limitations of the standard measure of unmet need or capture the full scope of outcomes relevant to advocacy and programming, it contributes to the ongoing development of a new measure of contraceptive need. It complements measures such as PFM and preference‐adjusted PFM (Holt et al. 2023; Rothschild et al. ²⁰²⁴) and is an example of how incremental change—through the addition of a reasonably small set of questions to quantitative research instruments—can improve our understanding of women's contraceptive needs.

AUTHOR CONTRIBUTIONS

Ann M. Moore (AMM) and Kaosar Afsana (KA) conceptualized the research project. All authors contributed to the development of the survey. Pragna Paramita Mondal (PMM) contributed to the training of the fieldworkers, supervised and supported by KA, AMM, and Altaf Hossain (AH). KA, PMM, Octavia Mulhern (OM), and Mira Tignor (MT) supervised fieldwork. OM and MT processed and cleaned the data. OM conceptualized this analysis and wrote the original draft of the paper, supported by Rubina Hussain (RH). RH, Joe Strong (JS), and AMM provided theoretical input. All authors contributed equally to the further drafting and approval of the manuscript.

CONFLICTS OF INTEREST STATEMENT

The authors declare that they have no known competing financial interests or personal relationships that could have influenced the work presented in this paper.

ETHICS APPROVAL STATEMENT

This study was reviewed and approved by the Institutional Review Boards of the BRAC James P. Grant School of Public Health, BRAC University, and the Guttmacher Institute.

DATA AVAILABILITY STATEMENT

The data that support these findings are publicly available on OSF (OSF | MR/Abortion/PAC Among the Rohingya in Cox's Bazar, Bangladesh).