Gates Open Res Gates Open Research 2572-4754 F1000 Research Limited London, UK 10.12688/gatesopenres.14590.1 Research Article Articles Impact of the COVID-19 pandemic on women’s contraceptive use: a mixed-methods study in South Africa and Zambia

[version 1; peer review: 3 approved with reservations]

 Callahan Rebecca L. Conceptualization Funding Acquisition Methodology Supervision Writing – Original Draft Preparation Writing – Review & Editing https://orcid.org/0000-0001-5808-0320 1 Cartwright Alice F. Data Curation Formal Analysis Methodology Project Administration Writing – Original Draft Preparation Writing – Review & Editing https://orcid.org/0000-0002-1557-8062 a 1 Beksinska Mags Conceptualization Data Curation Investigation Methodology Project Administration Resources Software Supervision Writing – Review & Editing 2 Kasaro Margaret Conceptualization Data Curation Investigation Methodology Project Administration Resources Supervision Writing – Review & Editing https://orcid.org/0000-0001-7285-998X 3 Tang Jennifer H. Conceptualization Data Curation Investigation Methodology Project Administration Supervision Writing – Review & Editing 3 Milford Cecilia Conceptualization Data Curation Investigation Methodology Project Administration Supervision Writing – Review & Editing https://orcid.org/0000-0002-0026-7560 2 Wong Christina Data Curation Formal Analysis Writing – Review & Editing 1 Velarde Marissa Data Curation Formal Analysis Writing – Review & Editing 1 Maphumulo Virginia Data Curation Investigation Project Administration 2 Fawzy Maria Project Administration Supervision 1 Chinyama Manze Data Curation Investigation Project Administration 4 Chabu Esther Data Curation Investigation 4 Mudenda Mayaba Data Curation Investigation 4 Smit Jennifer Conceptualization Methodology Project Administration Resources Software Supervision Writing – Review & Editing 2 FHI 360, Durham, North Carolina, USA MatCH Research Unit (MRU), Department of Obstetrics and Gynecology, University of the Witwatersrand, Durban, South Africa Division of Global Women’s Health, Department of Obstetrics and Gynecology, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA UNC Global Projects, Lusaka, Zambia acartwright@fhi360.org

No competing interests were disclosed.

 27 4 2023 2023 7 61 19 4 2023 Copyright: © 2023 Callahan RL et al. 2023 This is an open access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background: The COVID-19 pandemic affected global access to health services, including contraception. We sought to explore effects of the pandemic on family planning (FP) service provision and use in South Africa and Zambia, including on implant and intrauterine device (IUD) users’ desire and ability to obtain removal.

Methods: Between August 2020 and April 2021, we conducted surveys with 537 women participating in an ongoing longitudinal contraceptive continuation study. We also carried out in-depth interviews with 39 of the survey participants and 36 key informants involved in FP provision. We conducted descriptive analysis of survey responses and thematic analysis of interviews.

Results: Contraceptive use changed minimally in this sample with the emergence of COVID-19. Fewer than half of women (n=220) reported attempting to access FP since the start of the pandemic, the vast majority of whom were using short-acting methods. Among those who sought services, 95% obtained their preferred method. The proportion of women not using a method before and after pandemic start did not change in Zambia (31%); in South Africa, the proportion increased from 8% to 10%. Less than 7% of implant or IUD users in either country reported wanting removal. Among those who sought removal (n=22), 91% (n=10) in Zambia and 55% (n=6) in South Africa successfully obtained removal. In qualitative interviews, women with challenges accessing FP services mentioned long queues, deprioritization of contraceptive services, lack of transportation, stock-outs, and fear of contracting COVID-19 at a facility. Key informants reported stock-outs, especially of injectables, and staff shortages as barriers.

Conclusions: We did not find a substantial impact of COVID-19 on contraceptive access among this sample; however, providers and others involved in service provision identified risks to continuity of care. As the COVID-19 pandemic wanes, it continues to be important to monitor people’s ability to access their preferred contraceptive methods.

 family planning contraception COVID-19 South Africa Zambia Long-acting reversible contraception (LARC) LARC removal mixed methods South African Medical Research Council Swedish International Development Cooperation Agency 2017/762965–0 United States Agency for International Development AID-OAA-A-15-00045 7200AA19CA00041 United Nations Population Fund This work was supported by the Gates Foundation (OPP1032115). This work was also supported by the United States Agency for International Development (USAID), provided to FHI 360 through Cooperative Agreement AID-OAA-A-15-00045 and Cooperative Agreement 7200AA19CA00041. This was also supported by the Swedish International Development Cooperation Agency (2017/762965–0), the South Africa Medical Research Council, and the United Nations Population Fund. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Introduction

The emergence of the novel coronavirus, COVID-19, as a worldwide pandemic affected provision of and access to basic health services starting in early 2020 1, 2 . At the outset of the pandemic, public health and medical professionals urged governments and non-governmental associations to ensure continued access to family planning (FP) services 3 . Strategies proposed in both developed and developing countries included expanding telemedicine for counseling and screening, sending prescriptions directly to pharmacies, increasing availability and supply of self-administered contraceptive methods (including through the outreach of community health workers (CHWs)), continuing to offer access to long-acting reversible contraception (LARC) as safety procedures allowed, and providing immediate postpartum and postabortion family planning care 3– 5 .

Non-governmental organizations, governments, and donors also made calls to ensure that financing and logistic support for contraceptive commodities remained a focus to avoid stock-outs and ensure a range of methods were available in the event of supply chain disruptions 5– 7 . In anticipation of possible limited access to removal services for long-acting reversible contraceptive (LARC) methods, such as contraceptive implants and intrauterine devices (IUDs), recommendations were also made that providers counsel users on method effectiveness beyond the labeled duration if they desired to continue their method 3, 6 . The impact of COVID-19 was anticipated to have a particularly deleterious impact on access to FP services for people living in low and middle-income countries.

More than 36 months into the global pandemic, evidence has begun to emerge on the impact COVID-19 has had on access to contraception globally, particularly in the early days of the pandemic when national lockdowns and curfews were common. Data collected specifically among women in union from Kenya and Burkina Faso as part of the Performance Monitoring for Action (PMA) initiative found that the majority (67% in Burkina Faso and 82% in Kenya) did not change their contraceptive status between late 2019 /early 2020 and May–July 2020, though a small percentage did mention COVID-19-related reasons for not using any contraception 8 . A population-level analysis of contraceptive need and use trends in Burkina Faso; Kenya; Lagos, Nigeria; and Kinshasa, Democratic Republic of the Congo using PMA data from before and after the start of the pandemic, determined that COVID-19 had not significantly impacted contraceptive need or use in these settings overall. However, some specific groups, such as nulliparous women, reported an increase in need, particularly in rural Burkina Faso and both rural and urban Kenya 9 . Relevant to the study reported here, a prospective study of outpatient clinic visit patterns in KwaZulu-Natal, South Africa from January 27 th 2020 to April 29 th 2020 found reductions in child health visits after the national lockdown in March 2020, but no difference in FP visits 10 . In contrast, another South African study in Gauteng Province documented a decline in the provision of injectable contraceptive methods and LARC, and an increase in oral contraceptive pills in the months leading up to and including the national lockdown compared to the year before 11 . Concerns remained that sexual and reproductive health services could be impacted with additional waves of infection, and the manufacture of contraceptive commodities affected if pharmaceutical plants were forced to suspend production 12 .

So far, few studies documenting providers’ experiences offering reproductive health services during the COVID-19 pandemic have been published. One exception was a survey of United States physicians, which found decreased LARC insertion and removal, but increased utilization of telehealth (the provision of health care services remotely using technology, such as a computer or mobile phone) for contraceptive counseling, renewal of prescriptions without an in-person visit, and allowing curbside pick-up or mailing of contraception 13 . No changes in support of self-administration of subcutaneous injectable contraception, counseling on extended use of LARC, or advanced provision of emergency contraception or additional months’ supply of oral contraceptive pills were reported in this sample of providers.

The present analysis takes advantage of a unique sample of contraceptive users enrolled in a two-year longitudinal study of contraceptive use patterns to examine access to FP care during the COVID-19 pandemic in South Africa and Zambia. To explore more fully the factors surrounding supply, we also collected information from FP providers and others involved in service provision. The specific objectives of this study were to assess the effects of COVID-19 on family planning services and contraception access and use; to describe the effects of COVID-19 on LARC users’ desires and ability to remove their method; and to describe effects of COVID-19 on contraceptive supplies and service provision.

 Methods Ethics

All research participants gave written and verbal informed consent to participate in various aspects of the study. The main Contraceptive Use Beyond ECHO (CUBE) study, additional survey module, and in-depth interviews (IDIs) were approved by FHI 360’s Protection of Human Subjects Committee, the University of the Witwatersrand Human Research Ethics Committee (HREC), the University of North Carolina at Chapel Hill Institutional Review Board, and the University of Zambia Biomedical Research Ethics Committee (UNZABREC).

 Country contexts

In 2020 and 2021, COVID-19 was spreading globally, including across the African continent. South Africa had the highest number of documented COVID-19 cases in Africa at the time and roll-out of testing and vaccine availability, was slow the true impact may not yet be known in many countries. The first cases of COVID-19 were reported in South Africa on 5 th March 2020 14 and in Zambia on 18 th March 2020 15 . The government of Zambia enacted a series of public health measures on 13 th March 2020, and the government of South Africa declared a National State of Disaster on 15 th March 2020, implementing stringent lockdown measures designed to control the spread of COVID-19. As of 10 th March 2023, South Africa had reported over 4 million cases and 102,595 deaths, while Zambia had reported 343,135 cases and 4,057 deaths 16 .

 Study design

To better understand the experiences of accessing contraceptive methods within the context of COVID-19 lockdowns, we conducted a mixed methods study consisting of a survey module added to the CUBE study and IDIs with participants, FP providers, and other key informants involved with FP service provision. The methods for the CUBE have been described elsewhere 17 . In brief, the CUBE study enrolled a subset of South African and Zambian women who had previously participated in the Evidence for Contraceptive Options and HIV Outcomes (ECHO) Trial, had agreed to be re-contacted, were using one of three contraceptive methods at ECHO trial exit, and consented to participate 18 . These women were recruited to participate in a prospective longitudinal study (CUBE) after their exit from ECHO to explore contraceptive use dynamics and access to LARC removal services. The ECHO trial participants received either intramuscular depot medroxyprogesterone acetate (DMPA-IM), a levonorgestrel (LNG) implant, or a copper IUD for the 18-month ECHO trial period. Participants could switch or discontinue contraceptive methods both during the study period and at the time of study completion. For CUBE, participants were recruited from three of the 12 ECHO sites: two in the KwaZulu-Natal province in South Africa (an urban site situated in eThekwini District and a peri-urban site in uMgungundlovu District) and one in Lusaka, Zambia. Phone surveys were conducted with eligible women every six months for a total of four surveys over 24 months (at 6, 12, 18, and 24 months after ECHO exit). Since the COVID-19 survey module was initiated during the 24-month follow-up period, a portion of South African participants (less than 20%) were asked during their regularly scheduled 24-month call if they would be willing to participate in an additional module of questions related to COVID-19 during the current call or in a subsequent call. Most participants had already completed their 24-month survey and were re-contacted by phone and asked if they would be willing to complete the COVID-19 survey module. Surveys were conducted in South Africa between September 2020 and March 2021 and in Zambia from November 2020 to April 2021.

For the IDIs, interviews were conducted with a subset of CUBE participants along with FP providers and other key informants. Participant IDIs were conducted between September and November 2020 in South Africa and between April and June 2021 in Zambia. Key informant IDIs were conducted between August and October 2020 in South Africa and November 2020 and January 2021 in Zambia.

 Study population and sample

The main CUBE study enrolled 674 participants. For the COVID-19 survey module we sought to contact everyone who had completed the CUBE 18-month survey (n=626). Participants needed access to a phone to participate in both the CUBE study and the COVID-19 module. Written informed consent was obtained at initial CUBE enrollment and verbal consent was given to participate in the COVID-19 module. While the original eligibility criteria for the CUBE study was that participants had to be using one of the three ECHO contraceptive methods at enrollment (three-month injectable, LNG implant, or Copper IUD), there was no such restriction for participation in the COVID-19 survey module since participants could have switched from or discontinued the method they were using at CUBE enrollment. We were able to contact, consent, and interview 537 respondents for the COVID-19 survey.

A total of 39 CUBE participants (20 in South Africa and 19 in Zambia) and 36 key informants (16 in South Africa and 20 in Zambia) participated in the IDIs. IDI participants were purposefully sampled from the following four categories with an attempt to interview participants using all three contraceptive methods in each group (as applicable): 1) switched to a different method at ECHO exit; 2) discontinued their ECHO method while in CUBE; 3) continued their ECHO method into CUBE (and still using at 18 months); and 4) reported challenges getting LARC removal. For the key informant interviews, each country team created a list of FP providers, policy and program managers, and staff from FP commodity distribution centers in the public and NGO sectors, as well as other relevant stakeholders (e.g., NGOs, community advisory board (CAB) members, community advocates such as those representing sex workers, HIV-positive individuals, and youth groups). Key informant participants were purposively selected to represent local, district, provincial, and national levels.

 Data collection and analysis

Phone surveys. Research assistants (RAs) called eligible women participating in the CUBE study by phone either for their regularly scheduled 24-month follow-up survey or, for those who had already completed their 24-month survey, an additional survey call. RAs described the additional COVID-19 module and obtained verbal informed consent to participate. Due to social distancing restrictions, informed consent for the additional module of COVID-19 questions was conducted over the phone. Study staff confirmed with the participant that they were free to talk and in an area that provided privacy, which was particularly important as partners may have been more likely to be home due to restrictions of movement related to the pandemic. If requested by the participant, information on COVID-19 and/or related services was provided, in line with South African Department of Health and Zambian Ministry of Health guidelines. At the end of the interviews, study staff in South Africa gave women the phone number for the South African COVID-19 hotline and staff in Zambia offered to send women a text message with all relevant contact information to get COVID-19 related updates.

The phone surveys were conducted in isiZulu or English in South Africa and Bemba, Chinyanja or English in Zambia and lasted between 5 to 114 minutes, with a median length of 20 minutes. Respondents were reimbursed in mobile money or phone credit which was sent to the mobile device of participants’ choice to offset expenditures related to use of their phones. RAs in each country entered participant responses into an online, password-protected RedCap database pre-programmed with response options to limit errors in data entry 19 . Research staff in both countries and a research analyst at FHI 360, an international non-governmental organization, conducted weekly data quality checks. Any errors or inconsistencies were highlighted, and participants were recontacted as needed to correct responses.

In-depth interviews. Trained female interviewers conducted the IDIs face-to-face, over the phone, or via a web-based platform such as Zoom depending on participant’s preference, distance from the study site, and social distancing/quarantine guidelines at the time of the interview. IDIs were conducted in isiZulu or English in South Africa and Bemba, Chinyanja, or English in Zambia and lasted approximately 20 and 45 minutes for CUBE participants and between 10 and 55 minutes for key informants. CUBE participants and key informants were reimbursed for their participation, with the exception of Department of Health staff in South Africa, in line with the Department of Health policy.

During the IDIs, CUBE participants were asked about issues associated with obtaining their contraceptive method of choice or any method; challenges in accessing FP services; method stock-out; gaps in use; method switching and discontinuation; and LARC removal. Key informant IDIs focused on contraceptive supply chain and procurement issues affecting method availability; effects of COVID-19 on provision of specific methods; and staffing and operational issues with FP service delivery including staff shortages, staff morale, safety concerns, and effects of social distancing. All IDIs were recorded and subsequently transcribed. If conducted in a language other than English, interviews were translated and transcribed simultaneously into English using a transcription protocol 20 .

Survey analysis. We linked each woman’s COVID-19 module to her full CUBE survey data (6, 12, 18 and 24-month survey responses), as well as additional demographic information reported in ECHO. We produced a descriptive summary of participant sociodemographic characteristics collected from ECHO, the 24-month CUBE survey, and the COVID-19 survey, as applicable. We compared sociodemographic characteristics between those who completed the COVID-19 module and those who were lost to follow-up. Using the dates of 24-month survey and COVID-19 module completion, we confirmed the contraceptive method that the participant was using before COVID-19 restrictions in their respective country. To assess the first two study objectives related to the effect of COVID-19 on contraception access and use and LARC users’ desires and ability to remove their method, we calculated descriptive statistics overall and by country. Summary statistics are provided, including percentages for categorical variables, and means/medians, standard deviations, and ranges for continuous variables.

IDI analysis. We used applied thematic analysis to analyze data from the IDIs 21 . A codebook was developed to structurally and thematically code the transcripts using NVivo 12 22 . Initially three analysts each coded the same transcript and then met to assess inter-coder reliability, discuss coding discrepancies, and to revise the codebook with emerging themes and to refine code definitions and inclusion/exclusion criteria. This process was repeated until inter-coder agreement was reached. Thereafter, analysts proceeded to code all the transcripts with frequent meetings held to ensure coding decisions were made jointly. After all transcripts were coded, coding reports were generated through NVivo 12, and four analysts worked on synthesizing data from each coding report by conducting inductive thematic analysis to identify major trends and thematic domains.

 Results

Characteristics of respondents are shown in Table 1. Mean age was slightly younger in South Africa (27.8 years) compared to Zambia (29.6 years). Women in South Africa had higher levels of education, were more likely to report being a student or employed, and more likely to report that they were not currently living with their partner compared with women in Zambia. Those lost-to-follow-up were significantly less likely to be students and more likely to be living with their partner compared to those who completed the COVID-19 module, though there were no significant differences in age, education, or parity. More women in South Africa (90.1%) were currently using contraception at the time of the COVID-19 survey compared with women in Zambia (68.7%). About a quarter of South African and just over two fifths of Zambian participants had ever had a COVID-19 test.

Respondent characteristics.
South Africa (N=342) n (%) Zambia (N=195) n (%) Total (N=537) n (%)
Age at COVID-19 survey (mean, SD) a 27.8 (3.9) 29.6 (4.7) 28.4 (4.2)
    19-24 88 (25.7) 48 (24.6) 136 (25.3)
    25-30 195 (57.0) 73 (37.4) 268 (49.9)
    31-38 59 (17.3) 74 (38.0) 133 (24.8)
Level of education b
    No schooling 0 (0.0) 13 (6.7) 13 (2.4)
    Primary school 0 (0.0) 73 (37.4) 73 (13.6)
    Secondary school, not complete 108 (31.6) 77 (39.5) 185 (34.5)
    Secondary school, complete 148 (43.3) 25 (12.8) 173 (32.2)
    Attended post-secondary school 86 (25.2) 7 (3.6) 93 (17.3)
Parity (mean, SD) b 1.2 (0.9) 2.6 (1.3) 1.7 (1.3)
Employment status c*
    Housewife/Unemployed/Other 183 (53.8) 134 (69.4) 317 (59.5)
    Student 61 (17.9) 4 (2.1) 65 (12.2)
    Part or full-time employment 96 (28.2) 55 (28.5) 151 (28.3)
Partner status c
    Living together (married/unmarried) 20 (5.9) 172 (88.7) 192 (36.0)
    Not living together (married/unmarried) 303 (89.1) 21 (10.8) 324 (60.7)
    No current partner † 17 (5.0) 1 (0.5) 18 (3.4)
Currently using contraception d 308 (90.1) 134 (68.7) 442 (82.3)
Received a COVID-19 test d 82 (24.0) 86 (44.1) 168 (31.3)
    Tested positive for COVID-19 d 7 (8.5) 1 (1.2) 8 (4.8)

aCalculated based on age at ECHO enrollment and date of COVID-19 survey; bCollected at ECHO enrollment; cCollected at CUBE 24 month survey; d Collected in COVID-19 survey; *2 missing from South Africa and 2 missing from Zambia; **2 missing from South Africa and 1 missing from Zambia; †Includes widowed/separated/divorced

Overall, 16% of respondents reported difficulty accessing sexual and reproductive health services (including FP and STI/HIV testing and treatment) since the beginning of the COVID-19 lockdowns and restrictions on movement 23 . The proportion citing difficulties was higher in South Africa (23%) than Zambia (2%) (data not shown). Figure 1 shows the distribution of reported contraceptive method use before and after enactment of COVID-19 restrictions in each country. The proportion of women in South Africa who reported using no method increased marginally from the pre- to post-lockdown period. The proportion in Zambia was unchanged, but much larger overall (almost one-third of Zambian respondents reported not using any method in either period). In terms of method switching, slight reductions in IUD and three-month injectable use and increased use of condoms were reported in South Africa; and small decreases in IUD, implant, and condom use were reported in Zambia, along with slight increases in oral contraceptive pills (OCP) and three-month injectable use.

Methods used pre- and post-COVID-19 lockdown, by country (n=537).

Half of COVID-19 module respondents across both countries had last obtained their method before COVID-19 restrictions went into place. However, the timing of method access was related to the type of method used. The vast majority of LARC users in both countries - 95% in South Africa and 93% in Zambia - obtained their method before COVID-19 restrictions, while all short-acting method users (with the exception of one condom user in Zambia) obtained their methods post-restrictions (not shown). For those who last obtained their method after COVID-19 restrictions came into place, 95% reported that they were able to get their preferred method, and only 4% (n=8) of those reported any impact of COVID-19 on their method use (not shown). For the 5% who were not able to obtain their preferred method post-COVID-19, it was almost exclusively because the method was out-of-stock.

Among all respondents, 14% reported switching or discontinuing their method after COVID-19 restrictions ( Table 2), and among those, only women in South Africa reported that COVID-19 restrictions affected why they switched or discontinued. All eight women in South Africa who reported that COVID-19 affected their contraceptive use were using three-month injectables prior to COVID-19. Seven of the eight reported the method was out of stock and one wanted to reduce her time in-person at the facility. Other reasons for changing methods included desire for a more reliable method (n=2) and concern that the facility would be closed due to COVID-19 (n=1).

Switched or discontinued method after COVID-19 restrictions, by country.
South Africa (n=342) n (%) Zambia (n=195) n (%) Total (n=537) n (%)
Switched or discontinued method
No 291 (85.1) 171 (87.7) 462 (86.0)
Yes 51 (14.9) 24 (12.3) 75 (14.0)
Did COVID-19 restrictions affect why switched method? (n=50)
No 28 (82.4) 16 (100) 44 (88.0)
Yes 6 (17.6) 0 (0.0) 6 (12.0)
Did COVID-19 restrictions affect why discontinued method? (n=25) *
No 15 (88.2) 7 (100) 22 (88.0)
Yes 2 (11.8) 0 (0.0) 2 (8.0)

*1 missing from Zambia

Respondents were also asked about pregnancy and pregnancy planning. Just under 4% of all respondents (n=20) reported that they were currently pregnant. Among those, three quarters reported that they wanted to get pregnant when they did, two wanted to wait until later, and three did not want to become pregnant at all. Of these mistimed pregnancies (n=5), three respondents (who were all using injectable contraception pre-COVID-19) reported that COVID-19 had impacted their ability to delay or avoid pregnancy because the method they wanted was not available or they had a fear of being infected with COVID-19 while obtaining services. Among those not currently pregnant (n=517), 4% reported that they wanted to further delay or avoid pregnancy as a result of COVID-19, and 3% were not sure if their plans had changed (not shown).

Most respondents who were using a method and last obtained it after COVID-19 restrictions went into place, obtained their method from the public sector in both South Africa (67%) and Zambia (81%) ( Figure 2). These proportions were even higher (85% and 96%, respectively) when limited to injectable contraceptives (the most common method obtained after restrictions went into place). For those in South Africa who obtained condoms post-restrictions (n=39), 23% obtained them from the public sector, 37% from private health facilities, and 40% from another source (most commonly from a shop or market). Of the 220 women who last obtained their method post-COVID-19, 93% reported that COVID-19 did not play a role in where they obtained their method; rather people reported that they went to the location where they usually obtained FP and that this place was close to their home or work. For the 7% who reported COVID-19 related reasons (all in South Africa), the most common were hindrance of movement due to government restrictions or their usual source was not offering FP services (not shown).

Location where methods were obtained post-COVID-19, by country (n=220).

For those respondents not using a method at the time of the survey (n=95), 14 said that they had wanted to obtain a method since COVID-19 and all but one attempted to obtain a method, mostly through the public sector. Most reported non-COVID-19-related reasons for not obtaining a method, including that they changed their mind (n=5), the method was not available (n=3), or they became pregnant (n=2). Only two respondents reported that they could not get a method due to FP services not being offered or a provider not available. Respondents who did not want to use a method were most commonly pregnant at the time of the survey or trying to get pregnant or not having sex, though a small number mentioned that they feared COVID-19 infection (n=2) or that they could not currently see their partner due to COVID-19 restrictions on movement (n=2). Finally, among all respondents, only 3% (n=14) reported wanting to change from their contraceptive method due to COVID-19, most commonly because they wanted a more reliable method (n=5) or wanted to reduce their time at the health facility (n=4). Smaller numbers were concerned about stockouts or facility closures due to COVID-19.

The overwhelming majority of LARC users at the time of the COVID-19 survey (n=233) had not wanted to get their method removed since restrictions went into place (>93% across both methods and country settings) ( Figure 3). Of the 11 LARC users who wanted to get their method removed post-COVID-19, six had attempted removal. Of these, three reported that they were unable to get removal because the provider told them to keep their method, one said a trained provider was not available, one reported that removal was too expensive, and one was told to go back to where the method was inserted for removal. Two of the six also said that COVID-19 had affected their ability to obtain removal. For those who wanted removal but did not try (n=5), reasons included the facility being closed or too full/busy, not being able to go to the provider due to COVID-19 restrictions, or changed their mind or partner wanted them to keep the method.

LARC removal desires and attempts, by country.

The proportion of successful LARC removals (respondents who wanted to have their method removed and obtained removal) was higher in Zambia (91%) than South Africa (55%) after COVID-19 restrictions went into place ( Table 3).

Obtained LARC removal (among those who wanted and tried to get their method removed), by country.
South Africa Zambia Total
Implant (n=8) IUD (n=3) Total (n=11) Implant (n=8) IUD (n=3) Total (n=11) Implant (n=16) IUD (n=6) Total (n=22)
Did not obtain removal 3 (37.5) 2 (66.7) 5 (45.5) 1 (12.5) 0 (0) 1 (9.1) 4 (25.0) 2 (33.3) 6 (27.3)
Obtained removal 5 (62.5) 1 (33.3) 6 (54.5) 7 (87.5) 3 (100) 10 (90.9) 12 (75.0) 4 (66.7) 16 (72.7)
IDI results

Impact of COVID-19 on access to reproductive health services. Of the 39 CUBE participants interviewed, 22 said they accessed reproductive healthcare services, including contraception and HIV/STI testing and treatment, after COVID-19 restrictions came into place. Of these, five South African participants said that they encountered challenges in accessing services, while no participant in Zambia reported challenges. As this South African participant explains, challenges included long queues at clinics and contraceptive services being deprioritized compared with other conditions:

Other challenges mentioned by South African participants (including those who did and did not attempt to access services) included limited capacity on public transport, the need to carry appointment cards or referral letters to travel during lockdown, and stock-outs of some contraceptives.

Fear of contracting COVID-19 was also mentioned by participants and led at least three women in South Africa to avoid visiting a facility for services. In general, more South African participants reported fear of infection at a clinic compared to Zambian participants; however, the concern was raised in both countries. Fear of contracting COVID-19 led this South African woman to discontinue contraceptive use:

Providers and other key informants in South Africa mentioned the same challenges that women described in accessing reproductive health services. In addition, providers and key informants also reported other challenges, including clients not able to get referral letters to other clinics because their usual clinic was closed, women desiring pregnancy termination but unable to access services, and method users finding it difficult to leave the house to get contraceptives during the lockdown because their partners did not know they were using contraceptives.

In both South Africa and Zambia, providers and key informants talked about the rise in unplanned pregnancies during lockdown. A key informant in Zambia shared how it has been particularly challenging for adolescents:

Impact of COVID-19 on family planning use. When COVID-19 restrictions were implemented in March 2020, 29 of the 39 CUBE participants who participated in IDIs were using a method, most commonly injectables, followed by implants, IUDs, condoms, and oral contraceptive pills. Those who were not using any contraceptive said that they were pregnant or that their partner was not staying with them during the lockdown. Six of the 29 method users (four in South Africa and two in Zambia) said they had switched or discontinued use after COVID-19 restrictions were implemented. Of the four South African participants, three said that they were not able to get their method because facilities were either closed or had no stock of their preferred method, or that they were afraid of going to the clinic and contracting COVID-19. The fourth switched methods due to menstrual side effects. The two participants from Zambia switched from their implants to other methods for personal reasons not influenced by the pandemic.

Of the 23 CUBE participants who had not switched methods since the start of the pandemic, about half (n=12) were using LARC. These users, especially in South Africa, said that they were glad they were on a long-acting method, so they did not need to visit a clinic. Several FP providers in South Africa also reported that women asked to switch to a LARC method during the pandemic, and others reported that women were asking for tubal ligations so that they would not have to go to a clinic, as this provider explains:

Overall, FP providers and other key informants described more negative impacts on FP use in South Africa compared to Zambia. Most Zambian key informants reported that services had largely not been interrupted and that women were not discontinuing use during the pandemic, some surmising that women did not want to become pregnant during the pandemic.

When CUBE participants were asked if COVID-19 had influenced whether they wanted to get pregnant in the next one to two years, seven out of 38 participants (two in South Africa and five in Zambia) said that COVID-19 had changed their plans about becoming pregnant. All seven said that the pandemic period was not the time to have a child because of the risk of the child getting infected with COVID-19, and that they would prefer to wait to become pregnant, as this Zambian participant explained:

Several providers/key informants in South Africa and a few in Zambia reported that they feared an increase in unintended pregnancies, having already seen a rise in antenatal care since the start of the pandemic, as this Zambian provider shared:

Impact of COVID-19 on contraceptive supplies and provision of family planning services. Providers and other key informants in both countries reported contraceptive stock-outs and other procurement issues as major challenges during the pandemic. Implants and injectables (both two-month and three-month) were the most commonly mentioned methods with procurement issues, but some also reported stock-outs for IUDs and pills. Some key informants said that procurement challenges were attributed to reduced transportation during lockdown, reduced patient flow because of restrictions, funding limitations, and competing priorities posed by the pandemic. In response to not having desired methods available, key informants noted that patients were often given another method, while less commonly noting that clients may be referred elsewhere or made to wait until their method was back in stock. A community advocate in South Africa explained how the pandemic affected method availability:

Many South African key informants also reported staff shortages, while only a few reported the same in Zambia. South African interviews described shortages due to infections among staff, time out for frequent testing, and the need to take on shifts in the COVID-19 wards. A few facilities addressed social distancing requirements by rotating staff on alternate days. Staff shortages led to the remaining staff feeling overwhelmed and increased client wait times. LARC services were sometimes not offered if trained staff were not present. Because of staff shortages in South Africa, one respondent said that they had to supplement services by bringing in Cuban doctors and referring patients to other facilities (usually in the private sector).

Providers and other key informants also reported that staff morale dropped because of the pandemic. In both countries many said staff were afraid of getting infected, with the fear causing further demoralization in their work. They also felt anxious when clients did not comply with safety guidelines, such as wearing masks:

Others felt that morale was low because providers were tired from a high influx of patients, staff shortages, having to self-isolate when infected, and practicing safe distancing at work which had been difficult as they could not shake hands, pray together, or give hugs to comfort patients/co-workers when they received bad news (e.g., the death of a family member).

Impact of COVID-19 on LARC users’ desires and ability to remove their method. A few providers reported that demand for LARC removal decreased during the pandemic, with a few surmising that women were putting up with side effects longer (rather than seeking removal) because of the pandemic-related restrictions as this provider reported:

Others, however, said they had not seen a change in requests for either LARC removal or insertion. One provider in South Africa described greater challenges to LARC service provision in the public sector compared with the private sector due to pandemic lockdown restrictions.

 Discussion

In this sample of contraceptive users who had been followed prospectively over more than two years, we generally did not find a substantial impact of the COVID-19 pandemic on respondents’ ability to access their preferred contraceptive methods or LARC removal. Our findings are in line with findings to date on the minimal impact of COVID-19 on contraceptive use in South Africa and other sub-Saharan country settings 9, 10 . However, interviews with providers and other key informants involved with FP service delivery revealed negative effects of the pandemic on provision and uptake of FP. Overall, we documented higher contraceptive use among our sample in South Africa compared to Zambia both before and after COVID-19 restrictions were implemented. The proportion not using a method during the pandemic rose slightly in South Africa compared to pre-pandemic levels but stayed flat in Zambia.

Encouragingly, we found that most survey respondents who wanted to use a method were currently using one. Among the small number (n=14) who were not using a method but wanted to, COVID-19 was not the primary reason reported for non-use. Rather, users cited changing their minds, becoming pregnant, or the method they wanted was not available. While providers and other key informants did describe negative effects of the pandemic on contraceptive supply chains, stock outs were identified as a problem pre-pandemic 17 . In addition, we did not find extensive evidence of switching or discontinuation due to COVID-19 restrictions, again in line with what has been documented in other sub-Saharan African countries 8 . To our knowledge, ours is the first study to inquire specifically about desires and attempts to obtain LARC removal in the COVID-19 period and we found that the overwhelming majority of LARC users in our sample did not want to get their method removed and most who sought removal were able to access services.

While several FP providers and other stakeholders mentioned in the IDIs that demand for LARC methods had increased, we did not see an increase in LARC adoption in the survey data among women who were not using a LARC pre-pandemic. This could be related to the de-prioritizing of more intensive family planning services at health facilities during the early months of the pandemic, which was mentioned by providers and women alike. Almost all respondents who obtained a method after COVID-19 restrictions had gotten a short-acting method since they generally have to be obtained every 1-3 months. The majority seeking short-acting methods were still able to get them in the public sector. It may also be that due to their involvement in both the ECHO and the CUBE studies over the previous three and a half years, these women had had multiple opportunities to switch to a LARC method and were satisfied with their short-acting methods. However, it is possible that use patterns and method access continued to evolve as COVID-19 infections persisted. In the face of resurgence of the pandemic in the future, more people may want their LARC removed or face challenges in accessing their preferred short-acting methods if health facilities again become crowded or focused on managing COVID-19 patients.

Limitations

One of the limitations of this study is that it constitutes a unique sample. Due to their participation in the ECHO and the CUBE studies, participants likely had more knowledge about contraceptive methods and where to obtain services. In addition, their use of LARC methods was much higher (40–45% of the sample across both countries) than the general population of women of reproductive age in either South Africa or Zambia. However, such a large number of LARC users allowed us to explore the impact of COVID-19 on method switching and discontinuation, specifically the ability to access LARC removal. Since only women were enrolled as participants in the original ECHO and CUBE study samples, we are unable to comment on how COVID-19 may have impacted men’s contraceptive use patterns. In addition, the participants were from ECHO study sites in urban and peri-urban areas. It is possible that the impact of the COVID-19 pandemic may have been more severe in rural areas.

The surveys and interviews described here were completed at slightly different points during the pandemic: between August 2020-March 2021 in South Africa and November 2020-June 2021 in Zambia. These time periods were after the initial wave of cases in Africa (particularly South Africa) and during the summer in the Southern hemisphere 16 , when cases were likely lower overall. The experiences shared by respondents in this study may be subject to some recall bias and they may not reflect the situation as the pandemic progressed. Finally, the impact of the restrictions on family planning services in both countries was likely dynamic and possibly changed over time as the pandemic persisted.

 Conclusion

Among this group of contraceptive users in South Africa and Zambia, including a large number of LARC users, COVID-19 did not have a detrimental impact on contraceptive access, at least in the early days of the pandemic. However, interviews with family planning providers and other key stakeholders paint a slightly different, and more precarious, picture of the reproductive health care situation for the population at large. It will be important to continue to monitor the effects of the COVID-19 and other pandemics on health systems and access to primary services including family planning. Studies such as this one should be repeated to ensure that potential problems are identified proactively and access to contraception is protected.

 Data availability Underlying data

Full transcripts are not available for ethical reasons, to ensure anonymity of participants.

However, relevant excerpts are available from the corresponding author. ECHO data is available for researchers who provide a methodologically sound proposal, which will be reviewed by the ECHO Management Committee. Proposals should be directed to icrc@uw.edu and data requestors will need to sign a data access agreement.

Harvard Dataverse: CUBE Study COVID-19 Survey

https://doi.org/10.7910/DVN/TJKRU4 24

This project contains the following underlying data:

CUBE COVID_Participant Survey.tab

 Extended data

Harvard Dataverse: CUBE Study COVID-19 Survey

https://doi.org/10.7910/DVN/TJKRU4 24

This project contains the following extended data:

CUBE COVID_Participant Survey_codebook.pdf

CUBE COVID IDI Guide_Participants SA.pdf

CUBE COVID IDI Guide_Participants ZM.pdf

CUBE COVID IDI Guide_Key Informants SA.pdf

CUBE COVID IDID Guide_Key Informants ZM.pdf

Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).

 Software availability

RedCap software was used for programming and recording participant responses to the COVID-19 survey. A similar program available at no cost is Open Data Kit. NVivo software was used for coding and thematic analyses of qualitative data. A similar program available at no cost is Taguette.

 Acknowledgements

We are grateful for support from the Department of Health in South Africa in facilitating the in-depth interviews. We thank Pai-Lien Chen for his feedback on the analysis plan and Sophie Gao and Mu-Tien Lee for their help in verifying study results. We also thank Elizabeth Knippler for support with qualitative analysis. We are grateful to Laneta Dorflinger and Mario Chen for their constructive feedback on the manuscript. We thank Timothy Mastro for his help in developing the study concept and securing support from the ECHO Management Committee. Finally, we appreciate the time given by all of the research participants.

 Luciani S Caixeta R Chavez C : What is the NCD service capacity and disruptions due to COVID-19? Results from the WHO non-communicable disease country capacity survey in the Americas region. BMJ Open. 2023;13(3):e070085. 36863746 10.1136/bmjopen-2022-070085 9990165 Lal A Lim C Almeida G : Minimizing COVID-19 disruption: Ensuring the supply of essential health products for health emergencies and routine health services. Lancet Reg Health Am. 2022;6:100129. 34909753 10.1016/j.lana.2021.100129 8645973 Nanda K Lebetkin E Steiner MJ : Contraception in the Era of COVID-19. Glob Health Sci Pract. 2020;8(2):166–168. 32312738 10.9745/GHSP-D-20-00119 7326510 Pfitzer A Lathrop E Bodenheimer A : Opportunities and Challenges of Delivering Postabortion Care and Postpartum Family Planning During the COVID-19 Pandemic. Glob Health Sci Pract. 2020;8(3):335–343. 33008851 10.9745/GHSP-D-20-00263 7541126 Mickler AK Carrasco MA Raney L : Applications of the High Impact Practices in Family Planning during COVID-19. Sex Reprod Health Matters. 2021;29(1):1881210. 33599196 10.1080/26410397.2021.1881210 8009015 Weinberger M Hayes B White J : Doing Things Differently: What It Would Take to Ensure Continued Access to Contraception During COVID-19. Glob Health Sci Pract. 2020;8(2):169–175. 32561528 10.9745/GHSP-D-20-00171 7326519 Dasgupta A Kantorová V Ueffing P : The impact of the COVID-19 crisis on meeting needs for family planning: a global scenario by contraceptive methods used [version 2; peer review: 1 approved, 2 approved with reservations]. Gates Open Res. 2020;4:102. 33330836 10.12688/gatesopenres.13148.2 7720714 Karp C Wood SN Guiella G : Contraceptive dynamics during COVID-19 in sub-Saharan Africa: longitudinal evidence from Burkina Faso and Kenya. BMJ Sex Reprod Health. 2021;47(4):252–260. 33579717 10.1136/bmjsrh-2020-200944 7886665 Wood SN Karp C OlaOlorun F : Need for and use of contraception by women before and during COVID-19 in four sub-Saharan African geographies: results from population-based national or regional cohort surveys. Lancet Glob Health. 2021;9(6):e793–e801. 34019835 10.1016/S2214-109X(21)00105-4 8149322 Siedner MJ Kraemer JD Meyer MJ : Access to primary healthcare during lockdown measures for COVID-19 in rural South Africa: a longitudinal cohort study. medRxiv. 2020; [cited 14 Jun 2021]. 32511504 10.1101/2020.05.15.20103226 7273272 Adelekan T Mihretu B Mapanga W : Early Effects of the COVID-19 Pandemic on Family Planning Utilisation and Termination of Pregnancy Services in Gauteng, South Africa: March–April 2020. Wits J Clin Med. 2020;2(2):145–152. 10.18772/26180197.2020.v2n2a7 Govender D Naidoo S Taylor M : Don’t Let Sexual and Reproductive Health become Collateral Damage in the Face of the COVID-19 Pandemic: A Public Health Perspective. Afr J Reprod Health. 2020;24(s1):56–63. 34077054 Zapata LB Curtis KM Steiner RJ : COVID-19 and family planning service delivery: Findings from a survey of U.S. physicians. Prev Med. 2021;150:106664. 34081938 10.1016/j.ypmed.2021.106664 8316381 South Africa National Institute for Communicable Diseases: First Case of COVID-19 Coronovirus Reported in SA.2020; [cited 11 Aug 2021]. Reference Source Chipimo PJ Barradas DT Kayeyi N : First 100 Persons with COVID-19 — Zambia, March 18–April 28, 2020. MMWR Morb Mortal Wkly Rep. 2020;69(42):1547–1548. 33090982 10.15585/mmwr.mm6942a5 7583505 Center for Systems Science and Engineering (CSSE), Johns Hopkins University.COVID-19 Dashboard. 2023; [cited 3 Apr 2023]. Reference Source Cartwright AF Callahan RL Beksinska M : Contraceptive Continuation and Experiences Obtaining Implant and IUD Removal Among Women Randomized to Use Injectable Contraception, Levonorgestrel Implant, and Copper IUD in South Africa and Zambia. Stud Fam Plann. 2023. 36727169 10.1111/sifp.12222 Evidence for Contraceptive Options and HIV Outcomes (ECHO) Trial Consortium: HIV incidence among women using intramuscular depot medroxyprogesterone acetate, a copper intrauterine device, or a levonorgestrel implant for contraception: a randomised, multicentre, open-label trial. Lancet. 2019;394(10195):303–313. 31204114 10.1016/S0140-6736(19)31288-7 6675739 Harris PA Taylor R Minor BL : The REDCap consortium: Building an international community of software platform partners. J Biomed Inform. 2019;95:103208. 31078660 10.1016/j.jbi.2019.103208 7254481 McLellan E MacQueen KM Neidig JL : Beyond the Qualitative Interview: Data Preparation and Transcription. Field Methods. 2003;15(1):63–84. 10.1177/1525822X02239573 Guest G MacQueen KM Namey EE : Applied Thematic Analysis.Thousand Oaks, CA: SAGE Publications, Inc.;2012. 10.4135/9781483384436 QSR International Pty Ltd: NVivo.QSR International Pty Ltd.,2020. Reference Source South African Government: COVID-19/Novel Coronavirus: About alert system.[cited 17 Apr 2023]. Reference Source Cartwright A Callahan R Beksinska M : CUBE Study COVID-19 Survey.Harvard Dataverse, V1, UNF: 6: 9/dgmLBA2uDTjSj3vRtTDw== [fileUNF],2023. http://www.doi.org/10.7910/DVN/TJKRU4 10.21956/gatesopenres.15906.r35217 Reviewer response for version 1 Fuseini Kamil 1 Referee Population Council, Accra, Ghana

Competing interests: No competing interests were disclosed.

 14 11 2023 Copyright: © 2023 Fuseini K 2023 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

This paper titled “Impact of the COVID-19 pandemic on women’s contraceptive use: a mixed-methods study in South Africa and Zambia” sought to explore effects of the pandemic on family planning (FP) service provision and use in South Africa and Zambia, including on implant and intrauterine device (IUD) users’ desire and ability to obtain removal.

The study is a mixed-method study conducted with a subset of participants from the ECHO study who were using either injectable Depo-Provera, LNG implant, or copper IUD at the end of the ECHO trial and who agreed to participate in a longitudinal follow up study (CUBE). The CUBE participants responded to an additional COVID-19 module, the dataset for this study. Key informant interviews were also conducted with selected key personnel in the FP space.

The paper is clearly written, very informative and will make a good contribution to the literature; however, there are parts of the paper that can be strengthened.

The title of the paper is a little confusing as it suggests that the paper is looking at “… women’s contraceptive use …” yet looking at the objectives of the paper it seems to suggest that the paper is looking at various aspects of contraception including access. I suggest that the authors consider revising the title to include access.

The abstract needs to be edited. Parts of it do not read well. The conclusion of the abstract is also a little problematic as the title of the paper talks about use of contraceptives and the conclusion is talking about access. Which relates to my concern in comment 1.

The study design described in the paper is excellent, however, I do not think the statistical analytical methods are sufficient to assess “impact”. The analyses are descriptives and not sufficient to draw conclusions on the impact of an event. I would suggest that the paper be re-positioned and oriented. For example, change the title to “Use and access to contraceptive methods before - and during COVID-19 … ; Family Planning before and during".

The first paragraph of the study design does not also seem to suggest that the study is assessing impact and rightfully so, I do not think the analytical techniques are sufficient to do so. It would be clearer to revise the title to fit the content of the paper and make it clear that the paper is looking at use and access.

This section in the methods section “ … COVID-19 cases in Africa at the time and roll-out of testing and vaccine availability, was slow the true impact may not yet be known in many countries.” needs to be revised.

 In general, the paper is a very good paper and if revised to reflect a descriptive paper, it would be great.

Is the work clearly and accurately presented and does it cite the current literature?

Yes

If applicable, is the statistical analysis and its interpretation appropriate?

Yes

Are all the source data underlying the results available to ensure full reproducibility?

Yes

Is the study design appropriate and is the work technically sound?

Partly

Are the conclusions drawn adequately supported by the results?

Partly

Are sufficient details of methods and analysis provided to allow replication by others?

Yes

Reviewer Expertise:

Population Health, Sexual and Reproductive Health (including family planning)

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Cartwright Alice

Competing interests: No competing interests were disclosed.

 8 7 2024

1. This is a good observation and suggestion.  We have changed the title of the paper to:  Contraceptive access and use before and during the COVID-19 pandemic: A mixed-methods study in South Africa and Zambia

2. We are not sure which parts of the abstract the reviewer refers to as not reading well, but we have made some edits, which hopefully are improvements.  We have adjusted the abstract conclusion and paper title to match.

3. Thanks for this suggestion. We have changed the title of the paper as described above.

4. We have addressed this through the title change.

5. We have deleted this sentence.

 10.21956/gatesopenres.15906.r35216 Reviewer response for version 1 Geary Rebecca 1 Referee https://orcid.org/0000-0003-1417-1057 University of Liverpool, Liverpool, UK

Competing interests: No competing interests were disclosed.

 3 11 2023 Copyright: © 2023 Geary R 2023 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

Thank you for the opportunity to review this paper on the "Impact of the COVID-19 pandemic on women’s contraceptive use: a mixed-methods study in South Africa and Zambia". The paper reports minimal impact of the COVID-19 pandemic on both contraceptive access, and LARC removal. 

Among the survey sample of 537 women, only a small number reported difficulties accessing contraception, or wanting removal of their LARC method, limiting the sample and therefore inferences possible on the relative importance of different reasons for accessing contraception or removal. I would have welcomed the inclusion of confidence intervals to aid the interpretation of differences between countries, or pre- and post-COVID described in the text. I would also  recommend reviewing the presentation of small numbers in table cells to avoid presenting potentially disclosive results. These latter two points I consider important changes.

Participants had previously participated in two other studies (CUBE and ECHO) and their survey responses for the current study were linked to those from CUBE and ECHO. It would have been helpful to have a figure illustrating which data used were from which study, and when and how those data were collected. 

I was interested in the findings that varied between Zambia and South Africa and would have welcomed the author's reflections on potential reasons for these differences in the Discussion for those less familiar with these settings. I would also have welcomed more contextualisation of the results in the discussion in terms of the international literature.

Is the work clearly and accurately presented and does it cite the current literature?

Yes

If applicable, is the statistical analysis and its interpretation appropriate?

Partly

Are all the source data underlying the results available to ensure full reproducibility?

Yes

Is the study design appropriate and is the work technically sound?

Yes

Are the conclusions drawn adequately supported by the results?

Yes

Are sufficient details of methods and analysis provided to allow replication by others?

Yes

Reviewer Expertise:

Public health, contraceptive access, sexual health, health inequalities

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Cartwright Alice

Competing interests: No competing interests were disclosed.

 8 7 2024

1. In response to the reviewer’s comment, we have included confidence intervals around the proportions in Tables 1, 2, and 3. However, we would still like to remind the reviewer that this was a descriptive study, and while we present results from South Africa and Zambia to provide some degree of comparison, the study was not designed to draw conclusions about statistically significant differences.

In response to the reviewer’s second comment, we have removed specific n’s from Tables 1, 2, and 3. In addition, for Figure 3 and Table 3, where the n’s and proportions are very small, we have combined those seeking implant and IUD removal into overall LARC removal to avoid any risks of deductive disclosure.

2. For the sake of simplicity, we have not included a high level of detail about rates of participation in the ECHO and CUBE studies in this manuscript. We have acknowledged in the Notes to Table 1 the sources for the different pieces of information included. For timing of participation and other details on the methodology, we would refer the reviewer to these citations (included in the paper):

ECHO study: Evidence for Contraceptive Options and HIV Outcomes (ECHO) Trial Consortium. HIV incidence among women using intramuscular depot medroxyprogesterone acetate, a copper intrauterine device, or a levonorgestrel implant for contraception: a randomised, multicentre, open-label trial. Lancet Lond Engl. 2019;394: 303–313. doi:10.1016/S0140-6736(19)31288-7

CUBE study: Cartwright AF, Callahan RL, Beksinska M, Kasaro MP, Tang JH, Milford C, et al. Contraceptive Continuation and Experiences Obtaining Implant and IUD Removal Among Women Randomized to Use Injectable Contraception, Levonorgestrel Implant, and Copper IUD in South Africa and Zambia. Stud Fam Plann. 2023. doi:10.1111/sifp.12222

 3. We have added mention in the Discussion section that the more stringent lockdown in South Africa seemed to have a more severe impact on accessing FP services including access to LARC removal: " While overall minimal, we did find a slightly greater effect of COVID restrictions on FP access and use, including access to LARC removal, in South Africa compared to Zambia.  This is likely because the lockdown measures in South Africa were more stringent than in Zambia, affecting access to and provision of methods."

 10.21956/gatesopenres.15906.r34939 Reviewer response for version 1 A Cohen Megan 1 Referee https://orcid.org/0000-0002-9902-7806 Division of Complex Family Planning, Department of Gynecology and Obstetrics, Emory University, Atlanta, Georgia, USA

Competing interests: No competing interests were disclosed.

 28 9 2023 Copyright: © 2023 A Cohen M 2023 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

This is a mixed-methods study of the effect of the COVID-19 pandemic on women's access to, uptake, preferences for, and continuation of contraceptive use in South Africa and Zambia. It is very well written, timely, and of interest to an international audience. The study is performed on a subset of participants from the ECHO study who were using either injectable Depo-Provera, LNG implant, or copper IUD at the end of the ECHO trial and who agreed to participate in a longitudinal follow up study (CUBE). The CUBE participants were then also consented for the additional COVID-19 module which served as the dataset for this study. A subset of participants as well as key informants in the Family Planning sector were then also selected for key informant interviews. 

The methods are adequately described and the study design appears to be able to answer the study objectives. The qualitative component and use of key informant interviews adds context to the study, particularly regarding the larger population who were not participants in the ECHO study, which the authors correctly identify as a special population with likely increased contraceptive knowledge compared to the population at large. 

Overall the results are well-presented, but there were several areas of confusion or where I felt the presentation of the results and discussion could be strengthened. 

Do the 75 (14%) of individuals who stated they stopped or discontinued their contraceptive methods include the 20 pregnant patients (I assume so)? What is the breakdown by country of this, especially for the mistimed/unplanned pregnancies? Or are they separated? If they are not included in the individuals who stopped using a method, why not? They must have stopped given that they were pregnant?

Similarly, in the results on page 8 it states that 95 individuals were not using a method at the time of the survey, implying that more than 75 individuals either discontinued or switched methods. I think I may be missing that a certain proportion of individuals had already stopped using a method after ECHO but before the COVID-19 restrictions were in place? If this is the case, it would be very helpful to clarify this earlier in the results as the numbers do not seem to be adding up.  

Overall I would recommend moving down the numbers of participants into results, where that is traditionally reported. It may be helpful to have a Figure 1 or supplementary figure be how many out of the 626 who completed the CUBE 18-month survey were contacted, how many were not able to be reached, how many declined to consent to the COVID-19 module, how many were excluded if any for any reason, and then how many were ultimately enrolled. It could be helpful to also include at time of enrollment into the module if they were or were not using contraception and if so which type. It could be helpful to then also include how many were using contraception at the beginning of the enrollment for this module and how many were not. 

I'm confused about the coding of "method not available" or stock-outs as COVID-19 related in some places (i.e. in results on page 6 for South African women), but then not COVID-19 related later in the results (in the paragraph on the 95 patients not on contraception) and in the discussion. Is this specifically referring to a stock-out that occurred prior to COVID-19? It seems that if the method continued to be stocked out such that a patient could not reinitiate that would still be COVID-19 related, especially since it also came out as a theme in the Key Informant interviews. 

I agree that luckily most patients were able to continue their methods if desired and few sought to remove their LARC devices, but it does seem that there were some difficulties in South Africa that warrant a bit more discussion. Almost 50% of the individuals who desired LARC removal in South Africa were unable to obtain removal, which is a fairly big deal even though they were a small proportion of the study population. Additionally overall 23% of South African participants compared to 2% of participants in Zambia stating difficulties in accessing SRH services. It might be helpful to focus on this a bit more and perhaps place it into the context of what seemed to be more stringent lockdown in South Africa? Similarly it is a pretty big deal to have 3 out of 5 mistimed pregnancies as a result of COVID-related concerns or inability to obtain a method. So I might suggest a more nuanced discussion of the COVID-19 impacts in the discussion section, especially in regards to South Africa. 

Other thoughts:

The introduction is quite long. Could consider moving some of the results from other countries down into the discussion section to provide more context in the discussion and also to streamline the introduction. 

I would recommend that Figure 3 and Table 3 be organized in the same fashion. For example, consider presenting SA implant and SA IUD in the figure followed by Zambia implant and then Zambia IUD so it follows the same organization as the Table. 

Is the work clearly and accurately presented and does it cite the current literature?

Yes

If applicable, is the statistical analysis and its interpretation appropriate?

Yes

Are all the source data underlying the results available to ensure full reproducibility?

No source data required

Is the study design appropriate and is the work technically sound?

Yes

Are the conclusions drawn adequately supported by the results?

Partly

Are sufficient details of methods and analysis provided to allow replication by others?

Partly

Reviewer Expertise:

family planning, global contraception, obstetrics and gynecology

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Cartwright Alice

Competing interests: No competing interests were disclosed.

 8 7 2024

1. To clarify – the 75 individuals in Table 2 changed or discontinued their method after COVID-19 restrictions went into place (comparing the method they were using before COVID-19 restrictions and afterward). This was true for 10 of the 20 currently pregnant respondents. We take this to mean that the other 10 pregnant respondents discontinued their methods prior to COVID-19 restrictions but became pregnant after the restrictions went into place.

Among the 10 pregnant respondents who stopped using their method after COVID-19 restrictions, only 1 South African respondent reported COVID-19 impacted their method use.

However, when all of the 20 pregnant respondents were asked separately about the timing of their pregnancies post-COVID-19 restrictions, 15 of the 20 pregnant respondents (75%) reported that they wanted the pregnancy at that time, 2 said they wanted another pregnancy but wanted to wait until later, and 3 said they did not want to get pregnant at all. Among the 5 with mistimed or unwanted pregnancies (4 of whom were in South Africa), only 3 (all in South Africa) reported that they felt that COVID-19 restrictions impacted their ability to avoid or delay pregnancy (all 3 reported that the method they wanted was not available and 1 reported fear of being infected with COVID-19). We have described this group in the 6th paragraph of the Results section, but we have added some additional identifiers so that it is clear that the majority of mistimed or unwanted pregnancies were in South Africa.

We acknowledge that it is challenging to align these slightly conflicting responses to different ways of asking the questions about pregnancy intention and contraceptive use, but ultimately 3 pregnant respondents, all in South Africa, reported that COVID-19 contributed to their mistimed or unwanted pregnancy. We have added a sentence addressing this to the Limitations section:

“ Relatedly, with many factors affecting the health systems and contraceptive supplies, it is difficult to know if respondents accurately interpreted their situations as resulting from the COVID-19 pandemic. For example, in South Africa, stock-outs of contraceptive methods are not uncommon, and therefore, respondents may not necessarily see their inability to access their preferred contraceptive method as being related to COVID-19.”

2. Table 1 captures the method that respondents are using at the time of the COVID-19 survey (where 95 respondents report they are not using a method). Before COVID-19 restrictions went into place, 70 respondents were already not using a method (and continued not using a method after restrictions went into place). We have added a sentence to clarify this in the second paragraph of the Results section: “ Most of the participants not using contraception had discontinued method use before COVID-19 restrictions went into place (n=70)”.

Then from the time before to after COVID-19 restrictions went into place, 25 additional respondents discontinued their method (as reflected in Table 2). The 75 individuals in Table 2 either switched methods (n=50) or discontinued their method (n=25).

3. We have moved the number of participants from the Methods to the Results section.

We appreciate the suggestion to add an additional Figure; however, we are not sure how much additional information an additional Figure would provide to readers that could not be provided in the text. Therefore, we have added the following detail to the Methods in the "Study population and sample" section:

“The main CUBE study enrolled 674 participants. For the COVID-19 survey module we sought to contact everyone who had completed the CUBE 18-month survey (n=626; n=390 in South Africa and n=236 in Zambia).”

We have also added the following text to the beginning of the first paragraph of the Results section:

“For the COVID-19 survey module, we were able to contact, consent, and survey 537 CUBE respondents. In South Africa, we were able to reach 87.9% of the 18-month CUBE participants, though one survey was only partially completed and excluded, resulting in a final sample of 342 respondents. In Zambia, we re-contacted 82.6% of the 18-month participants, resulting in a final sample of 195.”

All participants were enrolled into the COVID-19 module *after* country-level restrictions had gone into place. As noted in the paper, the COVID-19 survey module was conducted between September 2020-April 2021. All information about their contraceptive use at the beginning of the COVID-19 module is provided in Figure 1.

4. These points refer to slightly different groups. The former is those that switched or discontinued their methods *after* the date COVID-19 restrictions went into place and asks if respondents perceived that COVID-19 restrictions themselves impacted why they switched their method or discontinued. (5th paragraph in Results section and Table 2).

The group of 95 people discussed in the 8th paragraph of the Results section are all of those who reported not using contraception at the time of the COVID-19 survey module. This includes the 25 who discontinued a method, but also includes those who were not using a method before the restrictions and continued not using a method. The results described in this paragraph are specifically focused on those who were not using a method but wanted to use one (n=14) to determine whether COVID-19 impacted their ability to re-start using a method. We found that for the majority of those people, COVID-19 was not the main determining factor in access.

5. We have added mention in the Discussion that the more stringent lockdown in South Africa seemed to have a more severe impact on accessing FP services including access to LARC removal: " While overall minimal, we did find a slightly greater effect of COVID restrictions on FP access and use, including access to LARC removal, in South Africa compared to Zambia.  This is likely because the lockdown measures in South Africa were more stringent than in Zambia, affecting access to and provision of methods."

Other thoughts:

1. We have moved references to other studies from the Introduction to the Discussion.

2. We have combined the implant and IUD removal interest into LARC overall to align with Reviewer 2's concern about small n's and risk of disclosure. However, we have ordered Figure 3 and Table 3 such that South Africa and Zambia are presented consistently.