Study design

The present analysis utilizes cross-sectional, mixed-methods data collected mid-pandemic from the Performance Monitoring for Action (PMA) Agile Youth Respondent-Driven Sampling Survey (YRDSS) cohort. The first round of data collection for this cohort was collected among adolescents and youth ages 15–24 (total n = 1,357; young men n = 690, young women n = 664) in Nairobi, Kenya recruited via respondent-driven sampling (RDS) between June and August 2019. Sampling and study procedures are outlined elsewhere [18, 21].

In 2020, a fully remote follow-up study was conducted with the study cohort to track contraceptive dynamics and assess the gendered impact of COVID-19. Quantitative surveys were conducted by phone in two sessions to limit participant burden: YRDSS Follow-up (young men n = 610, young women n = 613) and Gender/COVID-19 Survey (young men n = 605, young women n = 612). All data for the present analysis were taken from the 2020 follow-up surveys. All survey data was recorded via Open Data Kit (ODK) software. Surveys were conducted in English or Kiswahili per participant preference.

Qualitative data were collected concurrently from August to October 2020. Virtual qualitative methods included focus group discussions (FGDs) with unmarried youth ages 15–24 (n = 64, over eight gender-stratified groups), FGDs with youth-serving stakeholders (n = 32, over four groups), key informant interviews (KIIs) with higher-level family planning stakeholders (n = 12), and in-depth interviews (IDIs) with youth cohort members after completion of follow-up survey (n = 20).

Analytic sample

The quantitative analytic sample was restricted to those who completed Gender COVID and YRDS follow-up surveys (n = 605 young men, n = 612 young women). Due to skip patterns embedded within the survey, COVID-related contraceptive disruptions were only assessed for those who had ever had sex and were using contraception at follow-up survey (n = 337 young men, n = 347 young women). Contraceptive use was assessed via standard single item: Are you currently doing something or using any method to delay or avoid getting pregnant?” Once restricted to those with complete on COVID-related contraceptive difficulties (<10% missing data), the final analytic sample included 312 young men and 316 young women.

Measures

The primary dependent variable of interest, contraceptive disruption since COVID-19, was assessed via single multi-response item: “Have you experienced any of the following difficulties in accessing your usual contraceptive methods since the Coronavirus (COVID-19) restrictions began?” Specific contraceptive difficulties comprised: “healthcare facility or doctor’s office closed, appointment not possible”; “partner does not approve”; “no transportation to access healthcare services”; “unable to access services because of government restrictions on movement”; “unable to afford healthcare services”; “fear of being infected with COVID-19 at healthcare facilities”; and “other.” An affirmative response to any difficulty was classified as a contraceptive disruption since COVID-19.

Independent variables focused on postulated correlates for difficulty using contraception across the socioecological framework, including demographic characteristics (household, partner, individual), contraceptive characteristics, and COVID-19-related factors.

Standard items were used to assess demographic (self-reported household wealth ladder (response 0 (lowest)-10 (highest); analyzed as tertiles), marital status (dichotomous-unmarried vs. married), number of current sexual partners (dichotomous-no partner/one partner vs. multiple partners), age (dichotomous-16-20 vs. 21–26), level of education (categorical-less than secondary, secondary/A level, college/university), parity (dichotomous-nulliparous vs. parous)) and contraceptive characteristics (contraceptive method and place of procurement, both categorical) [22, 23]. Primary contraceptive method was also combined into a categorical variable to assess method effectiveness: long-acting reversible contraception (intrauterine device (IUD) and implant); short-acting (injectable and pills); barrier or hormonal coital-dependent (emergency contraception (EC), male condoms, and female condoms); and other (rhythm, withdrawal, and herbal pill method).

COVID-19-related items focused on changes to young adults’ personal circumstances since the onset of COVID-19 restrictions. Single items assessed change in amount of time spent with partner, control over healthcare, and control over how to spend money; responses of “more”, “less”, or “no change” categorized the relative frequency of each item in relation to the onset of restrictions. Additionally, a single four-response Likert scale assessed ability to meet basic needs since the onset of COVID-19 restrictions (responses: 1-very able to 4-not able at all); responses categories were collapsed into very able/somewhat able and not very able/not able at all to maximize statistical power.

Quantitative analysis

Descriptive analyses examined characteristics of the analytic sample and classified contraceptive disruption overall and by gender; significant differences by gender were assessed via design-based F-statistics. Proportions of youth experiencing contraceptive disruption were examined by contraceptive method effectiveness, stratified by gender. Unadjusted and adjusted logistic regression models, using GLM link log and family binomial, examined sociodemographic, contraceptive, and COVID-19-related correlates of contraceptive disruption overall and by gender; adjusted models included correlates with p<0.1 from bivariate models (overall or gender stratified), after assessing for multicollinearity.

All analyses were conducted in STATA 16 (College Station, TX) with statistical significance set a priori at p<0.05. Sampling weights accommodate the RDS study design using RDS-II (Volz-Heckathorn) weights, post-estimation adjustment based on 2014 Kenya DHS (age, sex, education levels), and modest loss-to-follow-up. All analyses are weighted, with statistical testing accounting for robust standard error clustering by node (recruiter for RDS design at baseline).

Qualitative data collection and analysis

Youth participants and youth-serving stakeholders were selected to participate in FGDs, IDIs, and KIIs in August to October 2020. Youth participants were identified by community-partnered recruitment through the assistance of local youth organizations; participation in the baseline survey was not a requirement for youth to be selected for the qualitative study. Eligibility criteria were as follows: 1) youth FGDs—15–24 years old, unmarried, living in Nairobi county for at least one year; 2) youth-serving stakeholder FGDs—currently or recently working at a youth-serving organization in a role in direct contact with youth, in Nairobi county, or otherwise knowledgeable about youth health issues in Nairobi; 3) KIIs—health and social service officials and other government officers managing or providing services in the area of adolescent/youth health, social supports, gender equity, and microfinance/economic empowerment; 4) youth IDIs—youth participant in YRDSS cohort with completion of 2020 quantitative survey.

Interviews for youth were conducted via Zoom videoconferencing by trained research assistants. Extensive privacy and confidentiality protections were in place, including password requirements, waiting rooms, no video or chat with other participants, and changing participant display name to an ID. Prior to beginning the interview, the research assistant asked a series of questions to confirm privacy for participants; at this time, a code word was also designated for use, should privacy be compromised during the interview. These virtual sessions were audio recorded and transcribed from Kiswahili to English. FGDs, IDIs and KIIs followed a semi-structured guide focused on norms and opinions related to youth family planning, relationships, and impacts of COVID-19.

A team of five qualitative researchers coded the 44 transcripts generated by FGDs, KIIs, and IDIs using a deductive approach. An initial codebook was created based on baseline data and existing literature on youth contraceptive behaviors and relationship dynamics. To maximize interrater reliability, all transcripts were dual coded by two researchers. After a set of 2–4 transcripts were coded, the research team met to discuss emergent themes, interpretations of youth experiences, and potential codes to add to the codebook. Retro-coding was used to ensure the quality of analysis and triangulation between transcripts. Matrices were created to synthesize emergent challenges to obtaining contraception during the pandemic (by gender), including COVID-19 concerns/behaviors, government response to the pandemic, and COVID restrictions impacting contraceptive access. For the purpose of this analysis, quotes from youth voices were prioritized to best understand issues youth were directly facing during COVID-19 restrictions.

Ethical considerations

The Institutional Review Board (IRB) at the Johns Hopkins Bloomberg School of Public Health and the Ethics Review Committee (ERC) at Kenyatta National Hospital/University of Nairobi approved all study procedures. Due to the remote nature of data collection, oral consent activities occurred for both quantitative and qualitative phases, with oral consent responses recorded in ODK (quantitative) or a participant call log (qualitative). The IRB/ERC waived parental consent for this study; therefore, only participant consent was required. Participants’ compensation (500 KES/5 USD), decided in consultation with community partners and resident enumerators, was transferred electronically via MPesa. All procedures aligned with ethical best practices for sensitive topics [24, 25], including specialized training, extensive privacy and confidentiality protections (including password protections, auditory privacy screener, no video, and code word in case of interruptions), and provision of resource referrals, specifically for COVID-19 and violence support.

Quantitative results

Demographic, contraceptive, and COVID-related factors are presented in Table 1, both overall and by gender. At COVID-19 survey, the majority of the cohort was between ages 21–26 (72.1%) and had completed secondary or A level education (61.3%). Approximately one in ten had married in the past year (11.0%), however, recent marriage was slightly higher for young women (young men = 7.7%; young women = 13.5%; p = 0.09). No significant differences by gender were observed in other demographic characteristics.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of young women and young men in Nairobi using contraception at 2020 follow-up survey (n = 628), weighted.

https://doi.org/10.1371/journal.pgph.0001005.t001

Primary contraceptive method and method effectiveness ranged substantially by gender (p<0.001), with 90.8% of young men relying on barrier or hormonal coital-dependent methods, primarily the male condom (86.9%). Barrier or hormonal coital-dependent methods were similarly most widely used for young women (46.1%), however, large proportions also utilized short-acting (30.4%) and long-acting reversible contraceptive methods (21.7%). Most prominent contraceptive methods reported for young women were male condom (38.5%), injectable (25.1%), and implant (18.8%). Place of contraceptive procurement similarly differed by gender (p<0.001), with young men predominantly seeking their method from a pharmacy (56.3%), whereas women were more divided between pharmacies (41.1%) and health centers (29.1%).

COVID-19 had mixed impact on youth, including time with partner, change in control over healthcare, change in control over how to spend money, and ability to meet basic needs since COVID-19 restrictions; however, only change in control over how to spend money differed significantly by gender (more control: 40.4% of young men vs. 27.4% of young women; p = 0.04).

Over one in three (37.3%) contraceptive users faced contraceptive difficulty during COVID-19 (40.4% young men vs. 34.6% young women). Fear of infection at healthcare facilities (19.6%) and closure of facilities (11.5%) were the most prominent difficulties reported for both genders, however, young men additionally faced disproportionate difficulty accessing services due to government restrictions on movement (15.5% young men vs. 6.1% young women; p = 0.003).

Contraceptive difficulty was most prominent for those using barrier or hormonal coital-dependent methods (42.6% young men; 51.5% young women; Fig 1). For young women, contraceptive difficulty decreased with longer method effectiveness, however, for young men, larger proportions of long-acting users faced contraceptive difficulties than short-acting users.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Difficulty accessing contraception by method effectiveness, by gender.

Method effectiveness: Long-acting reversible contraception = intrauterine device (IUD), implant; short-acting = injectable, pills; hormonal/barrier coital-dependent = emergency contraception.

https://doi.org/10.1371/journal.pgph.0001005.g001

Table 2 displays bivariate and multivariable logistic regression results, overall and by gender. Within combined gender models, married youth, compared to those unmarried, displayed trend towards decreased contraceptive disruptions (aOR = 0.58; 95% CI = 0.33–1.01; p = 0.06), whereas multiple partnerships trended towards increased odds of disruption (aOR = 1.34; 95% CI = 0.98–1.83; p = 0.06). When examining COVID-related-factors, youth with more control over healthcare since COVID-19 restrictions, compared to those with unchanged control, showed protection against contraception disruption (aOR = 0.59; 95% CI = 0.42–0.83; p = 0.003).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Logistic regression examining any difficulty accessing contraception during COVID-19, among contraceptive users at follow-up, by gender, weighted (n = 628).

https://doi.org/10.1371/journal.pgph.0001005.t002

Young men’s contraceptive disruptions centered around COVID-19-related factors. Specifically, young men with higher decision-making control surrounding healthcare reported protection towards contraceptive procurement difficulty (aOR = 0.49; 95% CI = 0.31–0.79; p = 0.004), whereas those with inability to meet basic needs since COVID-19 restrictions faced heighted difficulties (aOR = 1.50; 95% CI = 1.00–2.14; p = 0.05).

Conversely, for young women, procurement difficulties concentrated at the partner-level, where heightened difficulty was seen for those with more than one current partner (aOR = 1.91; 95% CI = 1.20–3.02; p = 0.01), whereas recently marriage trended towards protective effect (aOR = 0.54, 95% CI = 0.27–1.08; p = 0.08). In examining contraceptive factors, those who sought their method from a hospital (aOR = 1.71; 95% CI = 1.02–2.88; p = 0.04) or a clinic/other health provider (aOR = 2.14; 95% CI = 1.09–4.18; p = 0.03) reported increased odds of procurement difficulty, compared to young women who procured their primary method of contraception from a pharmacy.

Qualitative results

Most young people did not directly experience difficulty in obtaining contraception since the onset of COVID-19; however, those who did faced significant barriers. While economic disparities in Nairobi predated the COVID-19 pandemic, all groups of young people participating in FGDs and IDIs mentioned economic barriers as a predominant and ongoing difficulty in accessing contraception.

Specifically, young people who experienced job loss due to COVID-19 often discussed difficulty accessing contraception. With less available income, an individual facing loss of employment must decide between affording food, rent, and contraception, among other basic needs, which often meant that contraception was deprioritized.

I think when corona came, the outbreak increased poverty in our country. Like all these youths who lost their jobs job due to the virus, they cannot…if they were getting like 500 a day, now if they do not have the money, what they hustle for is for food and not buying those contraceptives, there are those who do not use them.–Young man, 20–24 years

Economic barriers to obtaining contraception were reported by young people patronizing both the public and private sectors. Among young people interviewed, pharmacies tended to be a popular place to access contraception. However, those who generally rely on pharmacies commented on diminished purchasing power at private health facilities due to higher costs relative to the public sector, which were exacerbated by changes in employment and/or income generation.

Coronavirus has affected apart from affecting the economic status of people… Let’s say, for example, people who don’t usually like going to public hospitals to pick these contraceptives they do go to these chemists, chemists in private hospitals to pick them and this is money. So, when it has come because it has ruined the financial status of most people, some have been laid off from their jobs, others even their business had to be closed. Something like that so people don’t have money for… somebody choses between contraceptives and whatever, food, which will he/she buy? So, it is better to buy food, but contraceptives… She/he says eeh!–Young woman, 20–24 years

Covid-19, has affected the youth in terms of getting the contraceptives… mostly is, financially because the contraceptives are available even in chemistries (chemists)… So the main problem is finances because the Covid-19… has many people staying at home.–Young man, 15–19 years

Most young people surveyed were dependent on condoms as their primary contraceptive method. Prior to COVID-19, many young people, especially young men, relied on programs supplying free condoms in public places, including schools and hospitals, to obtain contraception. Yet, due to COVID-19 restrictions and generalized fear of COVID-19 exposure, young people who relied on these programs no longer had access to free services.

Now for example with young men, before Covid started in the hospital there was this as in place of keeping this condom. Right now, they don’t keep it. So that they can try to avoid that overcrowding or those people who keep coming to hospital.–Female, 20–24 years

I can say that Covid-19 pandemic has affected the access of contraceptives especially to the men. So, for instance you find in, men used to get contraceptives in their washrooms. But currently because the schools are closed. The access to contraception has been cut down.–Young man, 20–24 years

Economic losses and COVID-19 restrictions have also affected NGOs operating programs supplying contraception to young people, specifically those using health centers to distribute contraceptives, therefore causing disruptions to youth dependent on these services.

Now like my area we used to have this as in people of health were coming and giving us services in this health centres, they give us this contraceptives and stuff. But since Corona started those health facilities are not coming.–Young woman, 20–24 years

It has affected because you find you can get most of them were depending on the government to give those NGOs, but right now, there are no gatherings. No one to give them, so you can find it has affected so much, so those who were using are affected, so it can damage them so much.–Young woman, 15–19 years

Beyond economic barriers, young people living with their parents may encounter unique challenges to obtaining contraceptives. Though not specific to COVID-19 restrictions, young women described facing stigma from family and community members for using contraception. Further, as a result of stay-at-home orders, youth worried that their parents would find their contraceptive methods and know that they are sexually active.

You find that most of the time you spend time with parents they are not going to work because of Covid at the moment, so might go access let’s say those contraceptives and carry things like condoms. And come and put them in the house. So, that time that the parent might pass by your room. And comes and find them. So, you find there is fear they will come and find them.–Young man, 15–19 years

Notably, several young women shared in in-depth interviews that they did not experience disruptions to their contraceptive use during COVID-19 because they were using a long-acting method: “I haven’t had concern because already I’m in, I’m using it the five, the long term one” (Young woman, 20–24 years).