Contraceptive decisions among individuals with medical conditions in Michigan, USA: A qualitative explanatory model informed by the Health Belief Model and the principle of respect for patient autonomy

Abstract

Objective:

We aimed to qualitatively describe the contraceptive experiences of individuals with medical conditions and develop an explanatory model of contraceptive decision-making in this context.

Study design:

We conducted 41 interviews with pre-menopausal patients with a range of medical conditions from Michigan, United States. We identified themes through open coding and comparative analysis until we reached theoretical saturation. We analyzed the themes in the context of the Health Belief Model constructs and patient autonomy. To develop the explanatory model, we qualitatively mapped out relationships between constructs and how they informed contraceptive-decision making.

Results:

The model posits that contraceptive decisions are affected by the perceived impact of chronic disease on pregnancy, parenting, and contraceptive options. These perceptions were strongly affected by others, particularly health care providers (HCPs). Most worried that pregnancy could threaten their health, fetal health, or ability to parent. Active symptoms of chronic disease (e.g., pain, fatigue) amplified this perceived threat. Patients appreciated when their HCPs acknowledged, rather than dismissed, their concerns about contraceptive side effects that affected their chronic disease symptoms and management (e.g., headaches, worsening mood). Patients with medical contraindications to hormonal contraception and preferred to use coital-dependent methods (e.g., condoms) did not feel supported to do so by their HCPs. Patients who were Black, young, living on low-incomes, or considered medically “high risk” felt discriminated against by HCPs who limited or coerced their contraceptive decisions.

Conclusions:

This explanatory model synthesizes factors that underlie, constrain, or promote reproductive and contraceptive decisions and autonomy for individuals with medical conditions.

Implications:

This new explanatory model can guide the development and evaluation of interventions that support the contraceptive decisions and autonomy of individuals with medical conditions.

1. Introduction

Nearly 40% of reproductive-aged people assigned female sex at birth in the United States have one or more medical conditions [1–4]. Individuals with medical conditions may have unique contraceptive considerations, such as hormonal drug interactions or an elevated risk of thrombosis with combined hormonal contraception (CHC) [5]. Contraceptive services for patients with medical conditions have increased over the last 10 years [6]. Despite this growing demand, there is little research to inform the delivery of contraceptive services for this heterogeneous population. A small yet growing body of literature has begun to identify contraceptive decision needs of those with medical conditions. In a study drawing from focus group discussions, participants with diabetes or hypertension reported they wanted to use contraception to avoid pregnancy complications but had limited knowledge about how their contraceptive choices were affected by their medical conditions [7].

Studies support the importance of using theory to design behavioral interventions prior to testing for effectiveness [8]. The Health Belief Model (HBM) has been successfully applied to understand health behavior and chronic disease across diverse conditions and populations [9]. The HBM has some conceptual limitations in contraceptive research. First, the HBM has been traditionally applied to sick-role contexts in which the goal is to prevent an undesired “illness.” However, pregnancy is not an illness and is not always undesired. Second, the overall focus of the HBM is on individual behavior, which may limit deeper exploration of external factors that limit one’s agency to carry out the behavior.

To address these gaps, we sought to qualitatively understand the contraceptive experiences of individuals with medical conditions and develop an explanatory model of their contraceptive decision-making. To theoretically ground this process and address the need to explore factors that affect patient agency to carry out contraceptive decisions, we applied an integration of constructs from the HBM and the principle of patient autonomy, including reproductive autonomy as described by Ross and colleagues [10].

2. Methods

Our conceptualization of the HBM constructs build upon Hall’s 2012 adaptation of HBM to contraceptive behavior [11] that we further refined [12] as follows: (1) perceived threat: perceptions that a medical condition may increase the risk of pregnancy-related harm to oneself, the fetus, or both; (2) perceived benefits and drawbacks (also referred to as “cost-benefit analysis” in the HBM): the pros and cons of using contraception; (3) cues to action: internal or external stimuli that trigger and shape one’s reproductive or contraceptive decisions; and (4) modifying factors : personal factors that impact the other constructs. We further extended Hall’s model by including self-efficacy, a more recent addition to the HBM model [13], and expanding it to 2 related dimensions: reproductive self-efficacy (the level of certainty or motivation to get pregnant or parent) and contraceptive self-efficacy (the level of certainty or motivation to use contraception). We conceptualized patient autonomy as reproductive autonomy (the right to decide if/when to parent) [10] or contraceptive autonomy (the right to decide if/when/which contraception to use).

The study protocol has been discussed elsewhere [14]. Participants were aged 18 to 50 years, pre-menopausal, non-pregnant, English-speaking, and had at least one medical condition. We purposefully sampled participants to include individuals with a range of medical conditions [15]. We recruited participants from primary care offices and subspecialty practices in southeast Michigan. The University of Michigan Institutional Review Board (HUM00128060) approved the protocol.

A researcher with expertise in qualitative research and contraception (JPW) led the study team. Trained research assistants who are both white women (ST, EJ) conducted semistructured interviews with participants from April 2018 to January 2020. We asked open-ended questions to elicit participants’ understanding of and concerns about the impact of their medical conditions on pregnancy, childbirth, parenting, and their contraceptive options. We also assessed the influence of others on their reproductive and contraceptive decisions, especially health care professionals (HCPs). We audio recorded interviews in private rooms at clinic sites or other settings (e.g., library). We used MAXQDA (version 12.3.6) to analyze professionally transcribed interview data [16] and grounded theory techniques for analysis [17]. After independently open-coding the first few transcripts, we resolved coding differences through discussion [18]. We repeated this process until we developed a consensus coding template. Through sorting and connecting codes, we advanced from descriptive codes to analytic codes and themes. We continued sampling and analysis until we reached theoretical understanding of contraceptive decision-making [18]. To generate the explanatory model, we triangulated these themes in context of the HBM constructs and reproductive/contraceptive autonomy.

3. Results

Sixty-one percent of participants (N = 41) had 3 or more medical conditions; the most common conditions reported were psychiatric (73%), endocrine (54%), or cardiovascular (39%) (Table 1). Just over half identified as white (56%) and 27% identified as Black. Sixty-one percent had a college education and private insurance; 31% had Medicaid.

Table 1.

Characteristics of individuals with medical conditions from Michigan, USA who participated in qualitative interviews 2018-2020

Age group	n(%)
≤30	15 (37)
31-40	14 (34)
≥40	12 (29)
Race/Ethnicity*
Non-Hispanic White	23 (56)
Non-Hispanic Black	11 (27)
Hispanic White	4 (10)
Asian	2 (7)
Other	3 (7)
Education
High school/GED	4 (10)
Some college	12 (29)
College degree or higher	25 (61)
Insurance*
Private	25 (61)
Medicaid	13 (32)
Medicare	2 (5)
Other	2 (5)
Medical condition categories*
Psychiatric (depression, anxiety, bipolar, attention deficit ADHD)	30 (73)
Endocrine (diabetes, thyroid, polycystic ovarian syndrome)	22 (54)
Cardiovascular (hypertension, heart disease)	16 (39)
Neurologic (migraines, seizures, multiple sclerosis)	12 (29)
Respiratory (asthma)	12 (29)
Musculoskeletal/Pain (fibromyalgia, chronic back pain)	10 (24)
Hematologic (anemia, sickle cell trait)	3 (7)
Immunologic (rheumatoid disease, systemic lupus erythematosus)	2 (5)
Gynecologic/Urologic (dysmenorrhea, endometriosis, fibroids)	2 (5)
Number of medical conditions
One medical condition	8 (20)
Two medical conditions	8 (20)
Three or more medical conditions	25 (61)
Current birth control *
Long-acting reversible contraception (IUDs, implant)	16 (33)
Condoms (penile)	13 (27)
Combined estrogen/progestin methods (ring, patch, pills)	7 (15)
Withdrawal	6 (13)
Tubal Sterilization	4 (8)
Progestin shot (Depo-Provera®)	2 (4)

Column percents may not add up to 100% because of rounding or *categories are not mutually exclusive.

IUD, intrauterine device.

3.1. Explanatory model of contraceptive decision-making

The final explanatory model posits that four HBM constructs are interrelated psycho-cognitive antecedents to contraceptive decision making (highlighted in the dotted box): (1) the p erceived threat of pregnancy on one’s health or fetal health, which in turn, drives reproductive self-efficacy and the motivation to seek or avoid pregnancy; (2) the perceived pros and cons of contraception, which then drive contraceptive self-efficacy and the motivation to use or avoid contraception. Cues to action are triggers that affect individual perceptions, self-efficacy, motivations, and decision-making. These cues can be internal (physical or psychological symptoms of chronic disease) or external (provider counseling, advice from others). Modifying factors refer to personal factors that affect one’s experience of the other constructs (e.g., a prior obstetrics complication) or patient autonomy. Patient autonomy, the extent to which individuals exert agency over their reproductive and contraceptive decisions, can be differentially affected by others (health care providers) based upon personal characteristics (age, class, race/ethnicity, and health status). Below, we describe the themes most salient to the HBM constructs and explanatory model (Fig. 1).

Fig. 1.

Explanatory model of contraceptive decision-making in context of medical conditions: An integrated application of the Health Belief Model and the principle of patient autonomy

3.2. Perceived threat of pregnancy

Most participants worried about pregnancy-related harm to oneself or the fetus because of their chronic disease and related medication use. A participant with kidney disease explained that “with the state of…my kidneys, if I had done it [pregnancy] at that time it would have been a very, very high risk pregnancy…I could die.” Other perceived threats of pregnancy included “controlling” their medical conditions during pregnancy (“my diabetes would probably go out of control”); neonatal risks (“the baby will come out bigger”); and fetal harm from medications (“Paxil, that’s known to cause heart defects”). Mental health deterioration was a common concern. A participant described; “I’m scared of not having that [anti-depressant] available because… my depression can be really, really bad if I’m not on medication.” One exception was a participant with multiple sclerosis (MS) who recalled feeling “great” during pregnancy, because MS “turns off some of your body’s immune response so that you don’t eat the baby, if you will.”

3.3. Reproductive self-efficacy

Reproductive decisions are complicated by the potential impact of chronic disease on pregnancy health, parenting, or both. Some participants felt they could not actively seek pregnancy, despite wanting biological children, because pregnancy would be too risky or parenting would be too difficult. A participant with hypertension, diabetes, and anemia admitted, “Oh my gosh, I got baby fever, but I don’t want to cause a high risk when I know better, and I know that my body is not… ready for that [pregnancy].” Another person felt fibromyalgia would affect her “capacity to be a mother” because “I already struggle cooking and doing everything for my-self.” For others, the strength of their desire for a biological child outweighed any health risks, even when cautioned by their PCPs to avoid pregnancy. A participant with hypertension and a history of severe preeclampsia recalled: “I had informed [my doctor] that I was gonna try to have a baby. And he sat me and my husband down, and told me that, it’s a 50/50 chance [of death] and we just did anyway.” HCPs also played an important role in reassuring patients that they can safely get pregnant. A participants with MS was grateful to her HCP for affirming that she could become a parent just like others with MS: “They have children, they have lives, it’s okay. Like she kind of like talked me off the ledge in that moment… it’s indelible in my memory that she did that for me.”

3.4. Perceived contraceptive benefits or drawbacks

Participants want their HCPs to address their concerns about contraceptive side effects, as well as respect their decisions to use coital-dependent methods (withdrawal, condom, and fertility awareness methods). Participants reported concerns about contraceptive side effects that worsened their chronic disease symptoms, such as headaches, weight gain, fatigue, and worsening mood symptoms. They wanted their HCPs to acknowledge rather than dismiss these concerns. A patient with migraines and depression recalls the frustration of having to convince to her doctor that she was having pill side effects: “Okay, you see that one sucked too. And so I’m going through these cycles all these different pills, if I have to prove to you. And finally I said look; I gonna go to a different doctor if you’re not willing to help me.” In contrast, another participant with a mood disorder was appreciated her doctor for supporting her decision to stop oral contraceptive pills: “every single pill I tried turned me into a raging bitch. Like total lunatic. Thankfully I had a really good OBGYN at that point in time.”

Participants reported that their PCPs did not routinely discuss or made disrespectful comments about the use of coital-dependent methods. A participant with diabetes felt “scoffed” and “shut down” by her OBGYN who “discouraged the use of condoms-like a lot” despite her reporting consistent use (“we are responsible adults”). Another noted, “I have heard some horror stories from other women of the pushback they get [for using coital-dependent methods]. You know when I say, we’re not using hormonal, we’re using natural family planning.”

3.5. Contraceptive self-efficacy

Fear of being unable to remove longer-acting contraceptive methods or stop their effects dissuade some from choosing these methods. Some were hesitant about longer acting reversible methods because they worried about “being stuck” with side effects. A participant with depression stopped oral contraceptive pills because of mood side effects. She was “wary to get an injection because I don’t want all of that in—like injected into my body at once and having like a mental reaction or something like that. …Like, it’s in your body and you don’t know what’s gonna happen.” She was also apprehensive about long-acting reversible contraception (LARC) based upon the misassumption that LARC require invasive surgery for removal: “And… something goes wrong, like you’re gonna have to get at surgically removed probably.”

3.6. Cues to action (internal triggers)

“Active” symptoms of medical conditions trigger interrelated motivations to avoid pregnancy and use contraception. For some, the disruptions of daily life related to chronic disease strengthened their motivations to avoid pregnancy and use contraception. A participant who suffered daily pain and mood swings described taking multiple medications “ just so that I can function” and was concerned about stopping her medications during pregnancy. She chose to be abstinent, citing worries that “most of my health conditions are genetic…it’s one of the reasons why I have chosen not to have children.” A patient with diabetes described, “ I measure my blood sugar all the time. My fingers hurt. I can’t just eat what I want. I feel like if I got pregnant…oh my God that really freaks me out my diabetes probably go out of control; that, that really stresses me out to think about that.”

3.7. Cues to action (external triggers)

HCPs advise against hormonal contraception for patients who are medically “high risk” but inconsistently educate them about differences in “hormones” or support the use of alternative methods. Participants at risk for thrombotic complications vividly recalled their HCPs telling them to avoid “all hormonal” contraception. Few reported HCP education about the distinctions between CHC (estrogen and progestin) versus progestin-only methods with respect to risks of blood clots. Others did not recall any counseling about alternative methods. Referring to “complex patients” like herself, one participant with musculoskeletal conditions said “we don’t fit in a box…the doctors can’t easily say ‘well this is what we normally advise’.” She felt abandoned to find contraception by herself: “they go ‘I don’t have anything’ and you can’t use this and you can’t use this…you’re on your own.”

3.8. Modifying factors

Some reported that their HCPs rushed or coerced them into contraceptive decisions based upon their health, age, race, income, or a combination of these factors. A participant whose pregnancy was complicated by diabetes and a heart attack regrets making a “fast decision” to get a tubal ligation even though she was only in her 20s. She felt her doctor “scared her” into believing “this is the best thing to do… and then come down, years down the line I realized it wasn’t.” A Black participant recalled her doctors “dismissing me”, “not listening to me” and telling her to use the shot with-out offering other methods. She felt this treatment was related to her “being a person of color, and being a person with no income, and being a young person…there’s a lot of discrimination when it comes to many aspects. And that includes birth control.” Another Black participant described difficulties getting her IUD removed: “I saw 3 different providers to get my IUD out… I don’t make every issue a race issue…but I felt disrespected. And I felt like she looked at me as if like, you know, I’m just gonna be out here havin’ babies. Because I’ve seen friends of, that are not the same color as me and easily taken their birth control out.”

4. Discussion

In this qualitative study, we developed an explanatory model to synthesize factors that affect contraceptive decision-making for individuals with a range of medical conditions. This work advances the contraceptive and behavioral health research in several ways. First, we expanded upon Hall’s conceptual model by adding reproductive self-efficacy and contraceptive self-efficacy as constructs. Second, in contrast to the traditional HBM model that treats the condition of interest as a harmful outcome to be avoided, we demonstrated the uniqueness of pregnancy as a potentially desired outcome and not a disease. Third, we integrated HBC constructs with patient autonomy to enhance our theoretical sensitivity to factors that affected participants’ contraceptive decisions and agency to act upon these decisions – including the right to use their most preferred method and the right to not use contraception.

As described by Holt and colleagues, person-centered contraception care prioritizes positive patient experiences and respect for patient decision-making over “the narrower focus of preventing unintended pregnancy at the potential expense of people’s autonomy”[19]. The study findings highlight potential counseling approaches to promote person-centered contraception care for patients with medical conditions, which may also be broadly applied for all patients. First, patients should be given realistic estimates of pregnancy-related risks in the context of their chronic disease and medication use. Patient desires to get pregnant and parent should be acknowledged and explored, even if in conflict with medical advice. This process may require consultation with specialists or colleagues to support fully informed pregnancy decisions. HCPs should acknowledge and respond to, rather than minimize, patient concerns about contraceptive side effects and the impact on their chronic disease symptoms. Patients need to understand that not all “hormones” are the same, and if they are medically eligible for CHC, progestin-only methods, or both. Those who can-not safely take any hormonal contraception may prefer hormone-free options, including coital-dependent methods. To help patients feel supported rather than “shut down”, HCPs can routinely discuss coital-dependent methods as options that are no less acceptable than other methods. Finally, HCPs should universally counsel all patients that LARC can be removed “early” and fulfill these requests in a timely manner.

Our findings that Black individuals experienced reproductive coercion by health care providers through racially-based discriminatory counseling and practices have been reported by others in different clinical contexts [20–22]. As described by Boyd and colleagues, researchers have a responsibility to name racism and its role in perpetuating health inequities [23]. To that end, we draw attention to a limitation in our ability to identify more robust themes at the intersections of chronic disease, reproductive autonomy, and racism. While under sampling of Black participants may have contributed, our interviewing methods may have also played a role. For example, we did not specifically acknowledge or ask about racism in reproductive health care. Our study interviewers were white individuals. There is a need for future research that centers on the experiences of Black participants with medical conditions and creates a safe space for disclosure of experiences of racism in reproductive health care [24, 25].

There were other study weaknesses. Nearly 60% of the study sample were white insured participants with some college education or a college degree from Michigan, United States, and may not reflect the experiences of those with different sociode-mographic traits. We aimed to understand contraceptive decisions across a range of medical conditions, which limited our ability to probe differences between conditions. We focused on contraceptive decision-making as an outcome, which may not predict actual contraceptive use over time.

There were study strengths. We used rigorous qualitative methods in study sampling, data collection and analysis, and model development [8]. The theoretical basis for this study increases the likelihood its findings can be clinically translated to effective contraceptive interventions [8]. Future research is needed to validate this explanatory model and explore adaptations for its use across different medical conditions and clinical settings.