Turned Away and at Risk: Denial of Family Planning Services to Women in Malawi.

Family planning (FP) has been a development priority since the mid-1990s, yet barriers to access persist globally, including women being turned away from facilities without a method. This study aimed to assess the extent of, and reasons for, FP turnaway in three districts of Malawi. In 2019, data collectors screened women exiting 30 health facilities and surveyed those who had been denied a method. Follow-up surveys were conducted via telephone with turned away clients at six and 12 weeks postvisit. Of the 2,246 women who were screened, 562 were new or restarting users. Of these, 15% (83/562) reported having been turned away from the health facility without an FP method. Women cited 14 different reasons for turnaway; the top three were unavailability of method (34%), unavailability of a provider (17%), or a requirement to return on the scheduled FP day (15%). The multiple reasons cited for leaving the health facility without an FP method indicate that reducing turnaway will not be achieved easily. The top reasons for turnaway are related to health systems or management issues within health facilities. Facilities need additional support for staffing, training on long-acting and permanent methods, and a consistent supply of methods.

BACKGROUND

The Sustainable Development Goals (SDGs), adopted in 2015, recognized family planning (FP) as one component necessary to achieving several of the SDGs, including those for ending poverty, ending hunger, promoting good health, and achieving gender equality (United Nations 2015). In recognition of FP's role in reaching the SDGs, the goals highlight the challenge of 218 million women having an unmet need for FP (Sully et al. 2020). Nevertheless, throughout the world, effective contraception can be difficult to access given the many barriers to FP that exist (Starrs et al. 2018). The barriers to FP are varied and include financial barriers related to the cost of traveling to the facility, missing work, and/or the visit or method itself; geographic barriers such as long distances between a potential client and a health facility; and social barriers based on behavioral expectations for women or adolescents (Campbell, Sahin‐Hodoglugil, and Potts 2006; Bertrand et al. 1995). In addition, women may lack factual information about FP, which may cause them to delay or avoid seeking FP services (Campbell, Sahin‐Hodoglugil, and Potts 2006; Mathe, Kasonia, and Maliro 2011).

One important type of barrier is medical barriers. Bertrand and colleagues defined medical barriers to FP as “scientifically unjustifiable policies or practices, based at least in part on a medical rationale, that inappropriately prevent clients from receiving the contraceptive method of their choice or impose unnecessary process barriers to access FP services” (Bertrand et al. 1995). In 1992, Shelton et al. outlined seven types of medical barriers to FP: contraindications, eligibility, process hurdles, which providers dispense contraception (those not properly trained on contraindications, for instance), provider bias, inappropriate management of side effects, and regulation (Shelton, Jacobstein, and Angle 1992). The seven types of medical barriers included instances where providers require husbands to be present, deny methods to nulliparous women, or require unnecessary physical examinations or laboratory tests before dispensing a method.

A 2018 review highlighted many of these barriers, as well as stockouts of methods, resulting in limited method choice for women (Starrs et al. 2018). Many of these barriers can lead to women being denied, or “turned away,” from a facility without a method. The term turnaway was recently popularized by a study conducted in the United States comparing longitudinal data on women who received an abortion to those denied the service (Biggs et al. 2017). Although the term “turnaway” implies a direct action by a healthcare provider or another staff person, in some instances women are passively denied FP by long wait times, closed facilities, FP available only on designated “FP days,” or absence of adequately trained FP providers (Konje and Ladipo, 1999; Campbell, Sahin‐Hodoglugil, and Potts 2006). In addition, some turnaway is justified, because of medical ineligibility or clients deciding they do not wish to initiate a method after learning more about available options.

Occasionally, turnaway can be overcome relatively easily, if, for example, a woman can return to a clinic on a designated FP day. However, when the barrier is geographic (the health facility is far away or difficult to reach) or financial (transportation to get to the facility is costly) or when women's decision‐making is limited, return trips to a facility can be burdensome (Campbell, Sahin‐Hodoglugil, and Potts 2006; Pitorak, Lubaale, and Gurman 2014; Mathe, Kasonia, and Maliro 2011). Limited method choice at facilities has also been found to be problematic and challenging to overcome (Konje and Ladipo, 1999; Campbell, Sahin‐Hodoglugil, and Potts 2006).

Although the issue of turnaway, or service denial, is not unique to FP, it has more often been analyzed through qualitative methods or provider self‐reporting on attitudes and practices, rather than estimates of the proportion of clients turned away (Sheffel et al. 2019; Wanyenze et al. 2017). The same is true of FP‐focused studies conducted throughout sub‐Saharan Africa, which have shown providers denying method provision based on age and parity, as well as conducting laboratory tests before providing a method or requiring direct observation of menses (proof of menstruation) for a woman to obtain a contraceptive method as a way of ensuring she is not already pregnant (Stanback and Twum‐Baah 2001; Sidze et al. 2014; Tumlinson, Okigbo, and Speizer 2015; Brunie et al. 2013; Speizer et al. 2000; Wado et al. 2019).

An exception is a study conducted by Tavrow et al., who conducted a quality assessment of provider attitudes toward FP in Malawi in the mid‐1990s. They examined client–provider interactions through the use of client exit interviews as well as simulated clients and found that 41% of simulated clients were turned away without a method, either before or after seeing a provider (Tavrow 1999; Tavrow, Namate, and Mpemba 1995).

In 2012, Baumgartner et. al. explored rates of denial of reinjection of injectable contraception to women returning after the reinjection deadline in South Africa and found that 22% of women coming one to 14 days late were denied a reinjection, as were 56% of women coming 15–30 days late and 74% of women coming 31–84 days late (Baumgartner et al. 2012). Some of these women, however, were given oral contraceptive pills (OCPs). More recently, in 2018, Hazel et al. engaged simulated clients in Malawi to study the quality of FP services in facilities and to report on a limited number of reasons for service denial (Hazel et al. 2021). In 12% of visits, clients were denied services for either not meeting provider requirements for HIV testing or tetanus vaccinations or because the facility was closed when it should have been open.

Given that the Tavrow et al. research was conducted over 20 years ago and Hazel et al. examined only a few reasons for turnaway, we saw a need to add to the evidence base. The overall goal of our study was to explore current barriers to FP leading to turnaway for women in government‐supported health facilities in Malawi. Specifically, the primary objective was to quantify the scope of turnaway—how many women did not receive an effective FP method on the day it was sought and why, and what characteristics were shared by women who were turned away. The secondary objective was to assess the short‐term FP outcomes of women who were turned away at the initiation visit. Did women return to the same facility, or a different facility? Were they able to initiate a method within a reasonable period, or did they continue to be at risk for unintended pregnancies?

METHODOLOGY

This quantitative study was the main component of a larger study that gathered data from FP units in 30 publicly supported health facilities in Malawi from October to December 2019. It included a client exit survey and two follow‐up surveys with turned‐away clients.

Study Design

A purposive sample of FP clinics run by the Malawian Ministry of Health and Population (MOHP) in the districts of Kasungu, Machinga, and Zomba was included in the study. In addition to being among the most highly populated districts, these districts were also selected based on the feasibility of data collection (e.g., travel budget and logistics for the data collectors, including travel time and available accommodations) and diversity in characteristics of the districts, such as religion and modern contraceptive prevalence rates (mCPR) (Table 1) (Malawi National Statistics Office 2019). The highest volume FP sites in each district were identified based on recent annual aggregate‐level District Health Information Software 2 (DHIS2) data on new FP clients and those restarting after a break of six months or more. Clients seeking an FP method from the 30 selected health facilities (10 per district), either as new users or after a break from previous use of at least six months, were asked if they were able to obtain an FP method and, if not, were invited to participate in an exit survey. Turned‐away clients who were willing to share their phone contact details were also asked to participate in six‐ and 12‐week follow‐up phone surveys. Data collector candidates were interviewed by members of the research team and selected based on prior data collection experience with priority on FP and health‐related experience, as well as experience collecting data with the use of electronic tablets. Upon recruitment, data collectors underwent a week‐long training prior to data collection. Surveys were translated into the predominant local language (Chichewa), and translations were reviewed with data collectors during data collector training. Data collectors were individually supervised by a member of the study team.

TABLE 1

Population, religious affiliation, and modern contraceptive prevalence rate (mCPR), in three districts in Malawi, 2018

				Religious affiliation (%)
	Total population	New FP users counseled	mCPR	Catholic	Muslim	Non‐Catholic Christian	Other religion
Kasungu	842,953	20,160	67.3	23	2	64	11
Machinga	735,438	30,638	45.9	7	67	24	2
Zomba	746,724	29,769	60.6	17	20	59	4

SOURCE: Population and religious affiliation data come from the 2018 Malawi Population and Housing Census, Main Report (Malawi National Statistics Office 2019). New FP users counseled were extracted from DHIS2 and provided by the Reproductive Health Directorate for the period from July 2018 to June 2019. mCPR data come from the 2015‐2016 DHS (Malawi National Statistics Office 2017).

NOTE: Other reported religions not shown included traditional, other denominations, and no religion.

By looking at both new and restarting users, we estimated how many clients data collectors might expect to survey and how long they might need to stay at each facility to reach the desired sample size. For the number of new or restarting FP clients needed to estimate the proportion of clients turned away, we assumed that no more than 25% of the women seeking to initiate an FP method would be turned away without an effective method. This was based on the study conducted in Malawi in 2018 showing a 12% turnaway for a limited number of reasons for turnaway (Hazel et al. 2021). Because the focus of our research included any reason for turnaway, we assumed a higher turnaway proportion. Thus, we determined that a sample size of 289 women would be needed to estimate the proportion of women turned away, with a 95% confidence interval and 5% precision. However, we aimed to recruit more than this minimum to also be able to estimate with reasonable precision the proportion of those turned‐away clients who started an effective FP method by the time of the six‐week follow‐up. Based on monthly client volume data from the DHIS‐2, we estimated weekly attendance and determined that by spending four days at each facility we could intercept approximately 700 eligible clients, of whom approximately 175 might be turned away and eligible for a follow‐up interview. With this sample size, we would be able to estimate the proportion of those who start an effective FP method by the time of the six‐week follow‐up with 7% precision. This assumed that 67% of clients turned away initiate an effective method by the six‐week follow up (Egyptian Fertility Care Foundation 2004). Precision will be lower due to attrition, and if enrollment of the initial eligible sample is slower or the proportion of clients turnaway is smaller.

Data Collection

Before beginning data collection, a member of the study team met with district leaders to explain the purpose of the research and to gain their permission to visit facilities in their districts. All the facilities were publicly supported facilities; two had additional support from the Christian Health Association of Malawi (CHAM). District staff were requested not to provide detailed information about the purpose of the study to facility staff so as not to influence their treatment of clients or survey responses. For the same reason, we allowed several weeks to pass between informing them about the study and beginning data collection.

Upon arrival at the site on day one of data collection, data collectors introduced themselves to the person in charge of the facility, explained they had the support of the MOHP and district health office and asked where to find exiting FP clients. If pressed for details, data collectors explained they were there to collect information about FP client characteristics.

Data collectors attempted to screen all women exiting from FP services at each facility for a period of four days—generally Monday through Thursday. Data collectors arrived before the facility opened for the day and remained until the clinic officially closed. They approached all women leaving the FP services area of the facility to ask if they had a few minutes to answer some questions. If a woman agreed to answer the questions, a data collector then determined whether she was eligible to be included in the estimation of the turnaway proportion (those seeking FP for the first time or after a break of at least six months), as well as for the exit survey (those who had been turned away) and future follow‐up (turned‐away women who agreed to be contacted by phone) (Figure 1). Any client who did not get a method was considered to have been turned away.

FIGURE 1

Clients who sought FP services and were approached for inclusion in the study, and type of data collected from them, Malawi, October to December 2019

After the initial screening, only women who had consented and had been turned‐away took the exit survey, which covered aspects of services received that day; their experiences with group counseling and the provider; the reason for turnaway; cost, time, and effort to arrive at the facility; basic demographic information; and expected next steps in light of the turnaway (Figure 1). It was not possible to ask more than a few screening questions of all women exiting the facility without going through a full informed consent process, which would have risked our data collectors missing the opportunity to screen all women leaving the facility. Surveys were conducted on electronic tablets using Open Data Kit (ODK) (Hartung et al. 2010).

Clients agreeing to be followed up by phone were called six weeks after their initial turnaway. Those who reported having started a method or suspecting pregnancy were not followed up again at 12 weeks. Those who had not initiated a method or who did not suspect pregnancy at six weeks post turnaway were phoned again after another six weeks (12 weeks post turnaway). We phrased the pregnancy question on the survey as to whether they “worried they might be pregnant,” rather than whether they had actually tested positive, due to the difficulties women face obtaining pregnancy tests, along with cultural hesitations to declare pregnancy before the end of the first trimester.

Ethics Approval

The study protocol was reviewed and approved by FHI 360's Protection of Human Subjects Committee at Durham, NC, USA, as well as the National Health Sciences Research Committee (NHSRC) in Malawi. All study staff completed training on research ethics, the protocol, and informed consent administration. All clients provided their informed consent to participate, including a brief verbal consent at the time of screening, a full signed consent before beginning the survey, and a brief verbal reconsent on the phone at the time of follow‐up. Clients were not reimbursed for their participation in exit surveys or follow‐up calls because the NHSRC removed a requirement to do so before our data collection began.

Data Management and Data Analysis

Survey data were downloaded from ODK as .csv files for uploading into SPSS software, version 26.0 and Stata version 15 for data analysis (IBM Corp. 2019; StataCorp 2017). Both the exit survey and the follow‐up surveys included limited open‐ended questions. Open‐ended response coding was conducted in Excel by two project staff. Answers were compared and differences were discussed. When consensus could not be reached about whether to recode into either a new or existing response category, the response was left as “other.” These coded responses were merged into the final analysis datasets.

We used a 95% Clopper–Pearson binomial proportion confidence interval to estimate the proportion of clients who were not provided a method on the day they sought one (i.e., those who were turned away) and the proportion of turned‐away clients who initiated an FP method within six weeks of being turned away. Exit survey and follow‐up survey results were linked using unique identifiers. We also explored how turnaway might be associated with different characteristics of the women seeking FP. We used a t‐test, Fisher's exact, or the Freeman–Halton extension test, as appropriate, to explore the association between turnaway and age, method sought, district, and whether turnaway occurred on a designated FP day. Among women who completed six‐ or 12‐week follow‐up surveys, we performed similar analyses to explore the associations between having started a method by 12 weeks postturnaway and clients’ age, years of education, income, reasonableness of time to arrive at the facility, reasonableness of cost/effort to arrive at facility, number of living children, and income type. We also present descriptive information on services received as well as the cost, effort, mode of travel, and time to arrive at the facility, based on the exit survey. Significant results were assessed at the level of 0.05 for two‐sided comparisons.

RESULTS

In total, data collectors screened 2,246 women exiting from FP service areas. One person refused screening. Of those screened, 562 (25%) were new or restarting users (150 in Kasungu, 164 in Zomba, and 248 in Machinga) and were thus eligible for primary analysis of turnaway (Figure 2).

FIGURE 2

Turnaway Prisma for study participants who sought FP services in three districts of Malawi

Of the 562 new or restarting FP clients, 83 (15%) reported having been turned away from receiving an FP method on the day they sought it. All 83 of the turned‐away clients completed an exit interview the day they were turned away (Figure 2). The turnaway proportions varied by district (Table 2). Data collectors were unable to reach 55 (66%) women to determine their final FP/pregnancy outcome by 12 weeks (Figure 2).

TABLE 2

Proportion of clients receiving an FP method and turned away without a method by the district in Malawi, October to December 2019

	Total new/restarting clients	Received a method	Turned‐away clients
District	n (%)	n (%)	n (%), 95% CI
Kasungu	150 (26.7)	114 (76.0)	36 (24.0), 17.4–31.7
Zomba	164 (29.2)	136 (82.9)	28 (17.1), 11.7–23.7
Machinga	248 (44.1)	229 (92.3)	19 (7.7), 4.7–11.7
Overall	562 (100)	479 (85.2)	83 (14.8), 11.9–18.0

Table 3 compares women's characteristics by turnaway status and the results of the association analysis. The average age of the 562 eligible women was 24.9 years (range 16–48). We found statistically significant differences in turnaway based on age, method sought (short vs. long acting), district, and whether it was a designated FP day. Turned‐away women were older (average of 26.6 years versus 24.5), seeking a long‐acting or permanent method, coming from Kasungu or Zomba, or going to the facility on a non‐FP designated day. In Table 3 we also present the age data by category to show how the proportion turned away increases with age.

TABLE 3

Proportion of clients receiving a method and turned away by selected client characteristics in three districts in Malawi, October to December 2019

Characteristics	Turnaway
	Received a method	Turned‐away	Total	p‐value
	n = 479	n = 83	n = 562
Age
Mean (SD)	24.5 (5.8)	26.6 (7.0)	24.9 (6.0)	0.013
Range	16–48	17–42	16–48
Age group				0.024
16–20	150 (88.8)	19 (11.2)	169
21–30	253 (86.1)	41 (13.9)	294
31 and older	76 (76.8)	23 (23.2)	99
FP method, n (%) a				0.015
Short‐acting	330 (88.2)	44 (11.8)	374
Injectable	295 (87.3)	43 (12.7)	338
Oral contraceptive pills (OCPs)	34 (97.1)	1 (2.9)	35
Condoms	1 (100.0)	0 (0.0)	1
Long‐acting or permanent	149 (80.1)	37 (19.9)	186
Implant	142 (84.5)	26 (15.5)	168
IUD	5 (50.0)	5 (50.0)	10
Tubal ligation	2 (25.0)	6 (75.0)	8
District, n (%)				<0.001
Machinga	229 (92.3)	19 (7.7)	248
Kasungu	114 (76.0)	36 (24.0)	150
Zomba	136 (82.9)	28 (17.1)	164
Designated FP day at the clinic, n (%)				<0.001
Yes/NA	450 (88.4)	59 (11.6)	509
No	29 (54.7)	24 (45.3)	53

Two women who were turned away did not have a method in mind when they were turned away. The p‐value for this variable compares short versus long‐acting methods.

Demographic information for the 83 turned‐away women is presented in Table 4. One of the 83 turned‐away women reported having no living children, while two had more than eight. The mean number of children with those fewer than eight was 2.6. A majority (57%) had given birth in the past six months. Over 80% of the women were married (82%) with some variation in marital status by district—Zomba had the greatest percentage of single women (18%) and Machinga had no single women in the sample. None of the women were widowed. No other demographic data were collected at the time of turnaway.

TABLE 4

Demographic results

Demographic information for turned‐away women	n = 83
Mean number of living children a	2.6 (0–7)
Marital status – n (%)
Married	68 (82)
Single	9 (11)
Divorced	6 (7)
Given birth in the last six months – n (%)	47 (57)

n = 81, two women had more than eight children (exact number of children not reported).

Reasons for Turnaway

The most cited reason for turnaway was “preferred method not available” (34%), followed by “provider not available” (17%), and “told to come back on an FP day” (15%) (Table 5). Specifically, of the 28 women who were told that their preferred method was not available, 68% were seeking injectable contraception, and 21% were seeking an implant. All of those told to come back on an FP day were seeking either injectables (n = 7) or implants (n = 5).

TABLE 5

Exit survey results: Primary reason for not starting FP and travel information at three clinics in Malawi, October to December 2019

Reason for not starting FP the day of the exit survey (n = 83)	n (%)
The method I wanted was not available	28 (34)
Provider not available	14 (17)
Told to come back on an FP day	12 (15)
The wait was too long	6 (7)
Did not have a health book	5 (6)
Pregnancy test required but not available	3 (4)
Not currently menstruating	3 (4)
Told to return on a different day without explanation	3 (4)
Told there was a medical reason I couldn't start (headaches, high blood pressure, other medication I'm taking)	2 (2)
I am breastfeeding/told I needed to wait longer after giving birth	3 (4)
I decided I didn't want to start FP or couldn't decide on a method	3 (4)
Told I was too old	1 (1)

Six women were turned away because of an inability to show they were not pregnant—either a pregnancy test was not available or they were not currently menstruating (Table 5). A 37‐year‐old woman seeking an injectable was told she was too old for her preferred method. Two women, both seeking an implant, were told they needed to wait longer after delivery before receiving an implant. Of the 83 women who were turned away, about half (51%) were seeking injectable contraception followed by an implant, either the three‐ or five‐year options (30%). Online Appendix 1 provides a breakdown of reasons for turnaway by method sought.

No women were turned away for refusing other procedures such as HIV testing, physical exams, or a pregnancy test, although in some cases these procedures were suggested and accepted. For example, six clients were asked to submit to a pelvic exam. In all six cases, the clients agreed to the exam; two women were seeking a tubal ligation, two an implant, one an IUD, and one an injectable.

Eight out of nine clients agreed to take a pregnancy test. These women were seeking OCPs, injectables, implants, and tubal ligation. All four women who were asked to show proof of menstruation for two different methods (injectables and implants) agreed to do so. One woman agreed to a vaccination (tetanus toxoid), two to an HIV test, and one to a malaria test.

Service Information

Just over a quarter (26%) of turned‐away clients participated in group counseling before being turned away (n = 82, one client skipped this question). Of those who did not participate in group counseling, 72% cited as the reason for arriving at the clinic late. A fifth (20%) stated that group counseling was not offered. The occurrence of those not attending group counseling was evenly distributed between districts. None of the women cited missing group counseling as the primary reason for turnaway, and of 17 women who reported having missed group counseling, two failed to see a provider because of arriving late. Nearly all (90%) clients waited less than an hour for group counseling to begin after their arrival, and 94% were seen by the provider within an additional 10–30 minutes.

When asked “Since you were not able to start a method today, what will you do?”, most (73%) said they would return another day, and 9% said they were not sure when they would return.

Cost and Effort to Arrive at the Facility

Most women who were turned away traveled to the clinic by walking, using a bicycle taxi, or using her own bicycle (Table 5). The cost of transport was similar across districts. Nearly 90% of clients reported that the facility they attended the day of the interview was the facility nearest to their home.

Follow‐Up Calls at Six and 12 Weeks

When asked whether they could be contacted for follow‐up, 82 of 83 women agreed and 38 of those were reached for a follow‐up interview approximately six weeks after the day of turnaway (Figure 2). Of those reached at six and 12 weeks postturnaway, 24 (52.6%, 95% CI: 35.8% – 69.0%) had started a method by the time we last spoke with them, and four (10.5%) suspected they might be pregnant (Figure 2). Eleven of 20 women who had started a method after having been turned away did so within one week of the turnaway; two did so within two weeks, and five waited a month or longer before initiating. One person did not answer this question, and one could not remember when she had initiated.

For the women for which a final FP/pregnancy outcome is known (n = 32), we did not find statistically significant differences in whether women started a method or not based on age, years of education, income, reasonableness of time to arrive at the facility, reasonableness of cost/effort to arrive at facility, number of living children, or income type.

When asked why they had not yet started a method, either at six‐ or 12‐week visits, 36% of those who had not yet initiated a method (including those who suspected they might be pregnant at six weeks) reported it was because the preferred method was still not in stock. One woman said that she did not want a medical procedure requested by the provider—a pregnancy test she was told she would have to pay for herself. At the time of the 12‐week follow‐up, she cited the same reason, saying she had not yet been able to purchase the pregnancy test. She was seeking an implant.

DISCUSSION

The results presented here are the main findings of a larger study exploring turnaway for FP services from contraceptive clinics in three districts in Malawi. We found 15% of women seeking FP services were turned away without a method from October through December 2019. Although clients cited 14 unique reasons for being turned away, over half of the turnaway reasons we found were related to the structure or management of health facilities, including stockouts, unavailability of providers, or clients being asked to come back on a designated FP day (Table 5).

Although our methodology differed from the work of Tavrow et. al. conducted in the 1990s, if both studies measured the same turnaway phenomenon with reasonable accuracy, our results suggest that turnaway is becoming rarer. Both studies revealed similar primary reasons for turnaway. Tavrow et. al. found that the largest group of turned‐away women were turned away at the clinic entrance before seeing a provider, for reasons such as missing group counseling; coming on the “wrong” day, even after having been told services were available that day; provider unavailability; and stockouts of commodities (Tavrow, Namate, and Mpemba 1995). In our research, no women cited missing group counseling as the primary reason they were turned away, although it was cited by two women as a reason for not seeing a provider. It was also a reason for noninitiation at the time of our follow‐up calls.

In our research, stockouts were least common in Machinga, which has greater support from nongovernmental organizations than the other two districts—a likely reason the turnaway proportion was lower there. The method most frequently sought—injectables—was also the most frequently out of stock. Almost half of turnaway clients who were seeking injectables said the primary reason for turnaway was that the method was not available. Injectable contraception is the most commonly used modern method in Malawi and is highly valued for its ability to be concealed (Malawi National Statistics Office 2017; Bisika 2008). There was, however, a worldwide shortage of injectable contraception in 2019 due to the shutdown of a manufacturer sterilizing facility, which likely affected supply during the time of our research (U.S. Food and Drug Administration 2019). Even so, clients were more likely to be turned away when seeking long‐acting or permanent methods. These methods require more technical training to administer compared to OCPs or injectables, so there may be an insufficient number of providers trained specifically on these procedures.

Although the sampled facilities reported providing FP services every day, we found that, in practice, many facilities have specific days of the week earmarked as FP days. Unavailability of a provider and being asked to come back on the “official” FP day were our second and third most cited reasons for turnaway (Table 5), despite the MOHP expectation that all publicly supported facilities in Malawi provide FP five days per week. Grouping services on certain days of the week can cause problems for clients beyond just being turned away. For example, if it is widely known in the community that a certain day is an FP day, it may be difficult for women to conceal the reason for their visit that day (Government of Malawi 2015). When facilities do not or cannot offer the full method mix every day, women desiring confidentiality may have to choose between confidentiality and accessing a method. Better integration of FP into other services such as postnatal care or childhood vaccinations may help to alleviate this situation. More research is needed to understand the reasons why facilities designate specific FP days.

While not unique to Malawi, long wait times were a common complaint during the time of the Tavrow et al. research, which reported the average wait time for FP services at three hours (Tavrow, Namate, and Mpemba 1995; Assaf, Wang, and Mallick 2017; Chavane et al. 2017; Zaky, Khattab, and Galal 2007). We found that most clients waited 90 minutes or less in total for the group counseling and the provider. Although the numbers were small, the long wait was still the fourth most common reason women left a facility without a method (Table 5), indicating there is room for more improvement in this area.

Medical barriers were a less common reason for turnaway in our study and did not reflect the level of provider bias reported by Tavrow et al. (Tavrow, Namate, and Mpemba 1995). Few women in our study were asked to submit to a pelvic exam, but of those who were, all agreed. Although pelvic exams are not part of a standard protocol for the provision of OCPs, injectables, or implants, they may be needed if the client has other symptoms, such as unexplained bleeding (Ministry of Health and IntraHealth International 2010). Similarly, all four women in our study asked to show proof of menstruation agreed to do so. Providers, however, have other options available rather than subjecting women to the embarrassment of showing someone else their menstrual pads, such as using “the reasonably sure not pregnant” checklist included in their preservice training or simply asking a woman whether she is currently menstruating (Ministry of Health and IntraHealth International 2010). Because all women agreed to both pelvic exams and providing proof of menstruation, we do not know if they would have been denied a method without it. We would need to use a simulated client approach to test this more thoroughly.

WHO guidance does not limit injectables by age, nor implants in the postpartum period nonetheless, one woman was told she was too old for injectables and two were told they needed to wait longer after delivery before receiving an implant (World Health Organization 2015). The infrequent occurrence of these clinically questionable decisions indicates the problem is not widespread, but certainly frustrating for the turned‐away women. In contrast to provider attitudes reported by Tavrow et al., no clients reported being turned away and asked to return with a husband, partner, or parent, nor were any told they were too young to be using FP or that they should be having more children (Tavrow, Namate, and Mpemba 1995). Additionally, in contrast with Hazel et al., no clients were turned away for refusing HIV testing or a vaccine, unofficial policies enacted in some districts but not in others (Hazel et al. 2021).

Although our numbers were small, we found that just under half of the turned‐away women we reached at the time of the six‐week follow‐up had still not initiated a method and some not even by 12 weeks (Figure 2). Still others suspected pregnancy. These women represent missed opportunities for initiation of FP and in some cases, possibly resulted in unwanted or mistimed pregnancies.

The most common reason for not initiating a method by the time of the follow‐up surveys was that the preferred method was still out of stock, highlighting the importance of a reliable supply chain. Related literature has shown that women were resistant to switching methods when their first choice was not available. For example, research in Uganda found that women were hesitant to change methods in the case of a stockout, and rather than switching to a different hormonal method, most commonly resorted to using condoms, withdrawal, or abstaining from sex until their preferred method could be obtained (Grindlay et al. 2016). A discrete choice experiment in Malawi showed that young women (aged 15–24) placed the highest value on having friendly providers and knowing all commodities were in stock when selecting a service provider (Michaels‐Igbokwe et al. 2015). Implementing a supply chain management program for FP commodity distribution might be an option, such as the informed push model, which showed great success in Senegal (Daff et al. 2014; World Health Organization 2017). Under this model, FP commodities are resupplied to facilities monthly rather than waiting for the facilities to order them.

While there are other FP service provision options available to women, such as private facilities, community health workers, and pharmacies, women generally returned to the same facility where they had been initially turned away to make another attempt at initiating a method, citing the convenience of the facility. Given that several noted returning only to find their method was still out of stock, the affordability of methods at public facilities, where they are normally free, is likely an important secondary factor to convenience.

We ruled out non‐support of FP use by husbands/partners as a limiting factor to a subsequent visit to a facility because nearly all the followed‐up clients reported being supported in their decision to use FP. This may not be the case in the larger population, however, because our participants were self‐selected based upon willingness to receive a future phone call for follow‐up surveys, and we were not able to reach all these participants. Research in Malawi has shown women may be motivated to conceal their use of FP from their male partners to avoid any possible perception of reduced sexual pleasure associated with its use (John, Babalola, and Chipeta 2015).

Our results highlight the need to gather more data from providers about their knowledge, attitudes, and practices of provision of FP. In addition, this research did not examine how willing women were to accept a second‐choice method in the case of their first choice was unavailable. An important area for future research is the number and proportion of women accepting an alternative method, and the most accepted and rejected second‐choice methods. Lastly, it is also important to compare the results from Malawi to other countries to better understand if the improvements are specific to the country, or part of a broader movement toward reduced barriers in accessing FP.

Strengths and Limitations

This study is one of few to examine and quantify the issue of client turnaway from a facility in sub‐Saharan Africa. While much recent research has rightfully focused on the impact of community‐provided services or self‐care options such as self‐injection, the extent to which women continue to face barriers at facilities has not been as thoroughly examined (Burke et al. 2018, 2014; Kalanda 2010; Masiano et al. 2019). The impact of being turned away has also not been examined. We struggled with the sample size to examine the impact due to a lower turnaway proportion than expected, as well as difficulties reaching women for follow‐up without threatening their confidentiality. While all except one of our 83 turned‐away clients agreed to receive a follow‐up call, many clients did not provide a phone number for the follow‐up. Instead, they took the data collector's number and indicated they would call the data collector later with a number for follow‐up, in most cases telling the data collector they did not know the number of a phone they were sharing. Even when numbers were offered, some participants could not be reached. As a result, clients reached at follow‐up were likely skewed toward those who were not attempting to conceal their use of FP, as well as toward those who owned a phone rather than sharing one with someone else—a common practice in the region (James 2014; Shava and Chinyamurindi 2018; Gunnlaugsson et al. 2020; Marron et al. 2020). Still, we were able to capture descriptive data on the consequences of turnaway.

This study provides important insights on the barriers faced by women seeking FP; however, the purposively selected sample may not represent the client populations of other facilities in the respective districts or other districts in Malawi.

Although providers at health facilities were not aware of our exact intentions on the days data collectors were present, our results may have been affected by the Hawthorne effect (Eckmanns et al. 2006). In fact, in a few cases data collectors gave specific examples where either they or the clients felt services had been provided because of the presence of the data collectors. We were also limited by not being able to use simulated clients, both to reduce the chances of the Hawthorne effect, as well as to test specific scenarios such as what happens when a woman refuses a suggested procedure. In our study design, it would not be possible to debrief providers on the day of data collection, or when a simulated client exited a facility because it would expose further research at that or nearby facilities. For this reason, we were prevented from using simulated clients by our US‐based institutional review board.

CONCLUSIONS

This study provides evidence suggesting that the occurrence of facility‐based providers turning away clients seeking to initiate a method of FP has declined since the 1990s. We have documented, however, that turnaway remains too high. The main barriers to accessing preferred contraceptive methods were the availability of providers or services—including initiation of more complex methods—and stockout of the methods themselves. These findings reflect the previous research and demonstrate the need for the MOHP and district authorities to improve the family planning infrastructure at health facilities in Malawi. For example, the MOHP may consider an examination of publicly supported health facilities’ needs for increased staffing, appropriate training on more complicated FP methods, and how to ensure a reliable supply of methods. These measures will help to ensure that women seeking a method of contraception do not leave facilities empty‐handed and at risk for unintended pregnancies. Recent research on turnaway rates in other countries is missing from the literature and could help to put our results in context, as well as to improve FP services and access in those countries. Additional research on reasons for designated FP days, the number and proportion of women accepting an alternative method when their first choice is not available, and more robust turnaway outcome data would also advance efforts to understand and reduce turnaway. Ensuring pregnancies are wanted and spaced according to women's needs will help Malawi to achieve related targets put forth in the SDGs.

CONFLICTS OF INTEREST

The authors report no conflicts of interest.

APPENDIX 1 Turnaway Reason by Method, n (column %)

Click here for additional data file.