Abstract

Contraception behavior is a form of behavior comprising the actions people take to prevent pregnancy, from choosing a method to using it consistently over time. Cognitive psychology treats it not as a single decision but as a chain from beliefs and intentions through to repeated action, and that chain is where most failures occur. This article explains the main models predicting contraceptive intention and use, distinguishes a method’s perfect-use efficacy from its lower typical-use effectiveness, and shows why intentions so often fail to become consistent use. It then examines counseling, the tension between promoting the most effective methods and respecting reproductive autonomy, and how unmet need is measured. Three interactive demonstrations let the reader combine attitudes, norms, and perceived control into an intention; compare perfect-use and typical-use failure across methods; and see how unmet need is computed from childbearing intentions.

Keywords: contraceptive behavior, theory of planned behavior, perfect versus typical use, contraceptive counseling, unmet need

Introduction

Preventing an unintended pregnancy is rarely a single act. It is a sequence: forming an intention to avoid pregnancy, selecting a method, obtaining it, using it correctly on every relevant occasion, and sustaining that use across the fertile years or switching methods without a gap. Each link can hold or break, and the behavior that demographers and clinicians actually measure is the whole chain, not the initial choice (Finer & Zolna, 2016). Cognitive psychology is interested in contraception precisely because the gap between what people intend and what they do is wide, systematic, and consequential.

The word behavior is doing real work in the MeSH heading. Contraceptive knowledge, a favorable attitude toward a method, and even a firm intention to avoid pregnancy are not themselves contraception behavior; they are antecedents of it. The behavior is the observable pattern of acquiring and using a method over time. This distinction matters because interventions that raise knowledge or shift attitudes often fail to change the behavior, and the models below exist largely to explain why the translation is so lossy (Fisher & Fisher, 1992).

Scale fixes the stakes. In the United States, roughly 45% of pregnancies were unintended at the start of the last decade, a figure that then fell as use of highly effective methods rose (Finer & Zolna, 2016; Kavanaugh & Jerman, 2018). Most unintended pregnancies occur not among people using no method but among people using a method inconsistently or incorrectly, which is the empirical signature of a behavior problem rather than an access problem alone (Trussell, 2011).

Key Takeaways

  • Contraception behavior is the whole chain from intention through method choice to consistent, sustained use, not a single decision.
  • The health belief model, the theory of planned behavior, and the information-motivation-behavioral-skills model each predict contraceptive intention and use from distinct cognitive antecedents.
  • A method’s perfect-use efficacy and its typical-use effectiveness can differ enormously, and the gap is a behavioral quantity that depends on how forgiving the method is of human error.
  • Intentions fail to become consistent use because of forgetting, ambivalence, partner and social factors, and the cognitive load of user-dependent methods.
  • Effective counseling and respect for reproductive autonomy can pull against each other, and measuring unmet need requires inferring people’s actual childbearing intentions rather than assuming them.

Theories of Contraceptive Behavior

Three cognitive models dominate the prediction of contraceptive behavior, and each identifies a different set of antecedents. The oldest is the health belief model, which holds that a health-protective behavior becomes likely when a person believes they are susceptible to a bad outcome, judges that outcome severe, expects the behavior to reduce the risk, and sees few barriers to performing it (Rosenstock, 1974). Applied to contraception, it predicts use when a person feels genuinely at risk of an unwanted pregnancy they judge serious, trusts the method, and faces low cost, inconvenience, and side-effect burden.

The most widely used model is the theory of planned behavior, which grew from the earlier theory of reasoned action (Fishbein & Ajzen, 1975; Ajzen, 1991). It holds that behavior follows from an intention, and that intention is predicted by three constructs: the person’s attitude toward the behavior, the subjective norm (perceived social pressure), and perceived behavioral control (how easy or hard the person judges the behavior to be). Perceived control was the addition that distinguished planned behavior from reasoned action, and it matters enormously for contraception, where a favorable attitude and supportive norms are useless if the person cannot reliably obtain a method or negotiate its use (Ajzen & Madden, 1986). Crucially, perceived control predicts behavior both through intention and directly, because a behavior one cannot actually execute will not happen however strong the intention.

The information-motivation-behavioral-skills model was developed specifically for health behaviors such as contraception and HIV prevention, and it makes the lossy translation explicit (Fisher & Fisher, 1992). It holds that accurate information and the motivation to act are necessary but not sufficient; between them and the behavior lies a layer of behavioral skills, such as negotiating condom use with a partner or remembering a daily pill, that must be present and felt to be within reach. The model’s empirical signature is that information and motivation affect behavior largely through their effect on skills, which is why education campaigns that raise knowledge without building skills change so little.

Figure 1

The theory of planned behavior applied to contraceptive use

A path diagram from three antecedents to intention to contraceptive use Three labelled nodes, attitude, subjective norm, and perceived behavioral control, feed into an intention node, which in turn feeds a contraceptive use node; perceived behavioral control also connects directly to contraceptive use. Attitude Subjective norm Perceived behavioral control Intention Contraceptive use
Note. Attitude, subjective norm, and perceived behavioral control jointly predict the intention to use contraception, which predicts use; perceived behavioral control (dashed path) also predicts use directly, because a method one cannot actually execute will not be used however strong the intention (Ajzen, 1991). Original schematic.

Interactive: the theory-of-planned-behavior pathway

In Ajzen’s model, three beliefs — your attitude toward using contraception, the subjective norm (what you think important others expect), and your perceived behavioral control — combine into an intention. But intention alone does not determine behavior: perceived control also feeds the behavior directly. Lower control and watch a strong intention fail to become protection.

Attitude toward using70
Subjective norm45
Perceived behavioral control60
direct control pathAttitude70Subjective norm45Control60Intention60Contraceptive use60
Intention stands at 60 and the predicted behavior at 60 — use is likely but inconsistent — the intention-behavior gap bites. Raise attitude and norm to their maximum but hold control low, and intention climbs while behavior lags: that gap between wanting to and doing is exactly what the direct control path encodes, and why access and skill matter as much as motivation.

Perfect Use Versus Typical Use

The single most important behavioral fact about contraception is that a method’s stated efficacy and its real-world effectiveness are different numbers, and the difference is a measure of behavior. Perfect-use efficacy is the probability of avoiding pregnancy when the method is used correctly and consistently on every occasion; typical-use effectiveness is the probability observed across real users, who forget, run out, use late, or use incorrectly (Trussell, 2011). The gap between them is not a property of the chemistry; it is the behavioral cost of a method’s demands on the user.

The gap is tiny for methods that ask almost nothing of the user and large for methods that ask a lot. A contraceptive implant or intrauterine device, once placed, requires no further action, so its perfect and typical use rates are nearly identical and both excellent. The oral pill, by contrast, has excellent perfect-use efficacy but a typical-use first-year failure near 9%, because a daily action is easy to miss (Trussell, 2011). The long-acting reversible contraception methods owe their effectiveness almost entirely to removing the user from the loop: the Contraceptive CHOICE Project found that users of pills, patches, and rings had failure rates roughly twenty times higher than users of implants and IUDs, with the disparity largest among the youngest users (Winner et al., 2012; Secura et al., 2014).

Table 1 Approximate first-year percentage of users experiencing an unintended pregnancy, perfect versus typical use
Method Perfect use (%) Typical use (%) User demand
Contraceptive implant 0.05 0.05 None after placement
Hormonal IUD 0.2 0.2 None after placement
Injectable 0.2 6 Injection every 3 months
Oral pill 0.3 9 A pill every day
Male condom 2 18 Correct use every act
Withdrawal 4 22 Control every act
No method 85 85 —

Interactive: the effectiveness gap, compounded over time

A method’s perfect-use and typical-use failure rates differ by how forgiving it is of human error. Slide adherence between the two, and watch the chance of at least one unintended pregnancy compound across years: a small annual rate becomes a large cumulative one. First-year rates are Trussell’s (2011).

Method
Adherence (typical → perfect)50%
Years of use5
0%25%50%75%100%0246810years of use
Over 5 years of oral pill use, the cumulative chance of at least one unintended pregnancy is 1.5% with perfect use, 21% at your adherence, and 38% with typical use. When perfect and typical rates are equal (the implant and IUD), adherence does nothing — that forgiveness of error is precisely what makes long-acting methods so effective in practice.

Why Intentions Fail

If most people who experience an unintended pregnancy were using or meant to use a method, the central cognitive question is why intention fails to produce consistent behavior. The models in the first section name the usual suspects. Perceived behavioral control is often overestimated before the fact: a person confident they will take a daily pill discovers that routines break, supplies run out, and refills lapse, and the method’s unforgiving daily demand converts a small lapse into a real risk (Ajzen, 1991; Trussell, 2011). Behavioral skills that seemed adequate in the abstract prove thinner in the moment, particularly skills that depend on a partner (Fisher & Fisher, 1992).

Ambivalence is a second source. Many people are neither firmly trying to conceive nor firmly trying to avoid it, and a weak or conflicted intention predicts weak behavior. Depression and other mental-health symptoms are associated with less consistent contraceptive use and higher risk of unintended pregnancy, plausibly because they erode the motivation and executive follow-through that consistent use requires (Hall et al., 2014). The behavior is embedded in a life, and competing goals, relationship dynamics, and fluctuating desire to avoid pregnancy all modulate it.

Subjective norms and the social context supply a third. Partner preferences, perceived or real family and community expectations, and misinformation about side effects all shape whether an intention survives contact with daily life. Because the models treat these as measurable antecedents rather than noise, they point to specific intervention targets: build real skills rather than only knowledge, choose methods whose demands match the person’s actual capacity for the behavior, and address the ambivalence directly rather than assuming a firm avoid-pregnancy goal.

Counseling, Autonomy, and the LARC Debate

The perfect-versus-typical gap has an obvious clinical implication: steering people toward the methods that ask least of them would prevent the most unintended pregnancies. Patient-centered contraceptive counseling tries to do this while keeping the choice genuinely the patient’s, eliciting preferences and values rather than issuing recommendations, and its quality predicts both satisfaction and continuation (Dehlendorf et al., 2014). Good counseling is itself a behavioral intervention, because a method chosen against a person’s values is a method they will abandon.

But the same effectiveness logic can curdle into coercion. The enthusiasm for long-acting methods has drawn a sharp critique: a LARC-first approach that promotes implants and IUDs because they work best risks prioritizing demographic outcomes over the person in front of the clinician, and it carries the historical shadow of coercive sterilization and contraceptive programs aimed at marginalized groups (Gomez et al., 2014). A method that is hard for the user to discontinue without a clinician is effective precisely because it removes user control, and that same property makes it a vehicle for coercion if a provider is reluctant to remove it. The autonomy critique insists that the goal is not the most effective method but the method the informed person actually wants, including the freedom to stop.

This reframes the field’s success metric. Counting averted pregnancies rewards steering; a reproductive-autonomy framework instead asks whether people are able to have the number of children they want, when they want, by means they have freely chosen (Senderowicz, 2020). The two metrics usually agree, because freely chosen methods are used more consistently, but where they diverge the autonomy framework treats a freely chosen less-effective method as a success and a coerced highly effective one as a failure.

Unmet Need and Measuring Intentions

To know whether contraception behavior is meeting people’s goals, one must measure the goals, and this is harder than it looks. The workhorse demographic indicator is unmet need for contraception: the share of fecund, sexually active people who want to avoid or delay pregnancy but are not using any method (Bradley & Casterline, 2014). The indicator has a long and contested history precisely because every term in it, “want to avoid,” “fecund,” and “using a method,” requires inferring a mental state or classifying an ambiguous behavior, and small changes in how intentions are elicited move the estimate substantially.

The deeper problem is that pregnancy intentions are not the stable binary the indicator assumes. People are often genuinely ambivalent, and retrospective questions about whether a pregnancy was “intended” conflate a prior plan with a post-hoc feeling. Prospective measures try to capture the attitude before the fact: the Desire to Avoid Pregnancy scale measures how much a person wants to avoid a pregnancy in the near future as a continuous disposition, and it predicts subsequent contraceptive behavior and pregnancy better than a single intention question (Rocca et al., 2019). Treating the desire to avoid pregnancy as a measurable, graded quantity rather than a yes-or-no fact is the methodological counterpart of treating contraceptive use as a behavior rather than a one-time choice.

Interactive: what counts as unmet need

“Unmet need” is a precise demographic measure, not a headcount of non-users: it is the women who are fecund and sexually active, want to avoid or delay a pregnancy, yet are using no method. Build the cohort of 1,000 down to that group and see how each filter narrows it.

Fecund & sexually active72%
…who want to avoid/space80%
…currently using a method78%
1,000 women of reproductive ageusing a method: 449unmet need: 127
Of 1,000 women, 720 are fecund and sexually active, 576 want to avoid or delay pregnancy, and 449 are using a method — leaving 127 with unmet need (17.6% of the fecund, active group). Raising the share who want to avoid pregnancy can increase measured unmet need even as contraceptive use rises: the denominator is desire, not merely non-use, which is why the measure tracks autonomy rather than simple coverage.

Worked Example

The perfect-versus-typical gap compounds over time, and the arithmetic makes the behavioral stakes concrete. First-year failure rates are annual probabilities, so the probability of still avoiding pregnancy after several years is the single-year success probability raised to the number of years, and the cumulative failure probability is one minus that.

Take the oral pill. Its perfect-use first-year failure is about 0.3%, so the single-year success probability is 0.997; its typical-use failure is about 9%, so the single-year success probability is 0.91. Over three years of continuous reliance, perfect use gives a cumulative failure of 1 − 0.9973 = 1 − 0.991 = 0.009, about 0.9%. Typical use gives 1 − 0.913 = 1 − 0.754 = 0.246, about 25%. The same pill, the same chemistry, produces a one-in-four three-year failure under real behavior against a one-in-a-hundred failure under flawless behavior.

Now contrast the implant, whose perfect and typical rates are both about 0.05%. Its three-year cumulative failure is 1 − 0.99953 = 1 − 0.9985 = 0.0015, about 0.15%, and it is the same figure whether or not the user is diligent, because the method asks nothing of them. The comparison isolates the behavioral quantity exactly: the pill’s 25% versus 0.9% is the cost of its daily demand, while the implant has no such cost to pay. A person choosing between them is, whether they realize it or not, choosing how much their protection will depend on their own future consistency. The interactive demonstration above lets the reader set a method and an adherence level and watch the cumulative failure curve bend as the years accumulate.

Discussion

The cognitive account of contraception is organized around one asymmetry: deciding is easy and doing is hard. The theoretical models all locate the action in the translation from favorable beliefs to a firm intention, and from a firm intention to repeated behavior, and they agree that this translation is where campaigns and good intentions leak away. The perfect-versus-typical gap quantifies the leak method by method, and it shows that the most reliable way to improve the behavior is often to choose a method that demands less behavior.

That conclusion is in productive tension with the autonomy critique, and the tension is the discipline’s healthiest feature. Effectiveness logic alone would steer everyone to the methods that remove the user from the loop; autonomy logic insists that removing the user from the loop is exactly what makes a method a potential instrument of coercion, and that a freely chosen method used well is the real goal. The resolution is not to pick a side but to keep both metrics in view: build the behavioral skills and match the method to the person, so that the method someone genuinely wants is also one they can use consistently.

For the practitioner and the educated reader, the operative variables are again the ones that can be moved. Perceived control can be calibrated by honest discussion of what a method actually demands; behavioral skills can be taught and rehearsed; ambivalence can be surfaced and addressed rather than assumed away; and the method can be matched to the person’s real capacity for the behavior rather than to a population-level effectiveness ranking. None of this requires pretending that intention and behavior are the same thing, which is the error the whole field exists to correct.

Current Directions

The most active front is the measurement of intentions themselves. Prospective, graded instruments such as the Desire to Avoid Pregnancy scale are displacing the retrospective yes-or-no intention question, on the evidence that a continuous disposition predicts behavior and pregnancy better than a binary recalled after the fact (Rocca et al., 2019). The shift is consequential because the field’s central indicators, including unmet need, are only as good as the intention data feeding them (Bradley & Casterline, 2014).

The second front is the reframing of success. A reproductive-autonomy framework is increasingly proposed as the field’s organizing goal in place of averted pregnancies, with new indicators intended to capture whether people can enact their own childbearing preferences freely rather than whether they adopt the most effective methods (Senderowicz, 2020; Gomez et al., 2014). Alongside these conceptual shifts, continued monitoring of method-use trends documents the real-world uptake of long-acting methods and the decline in unintended pregnancy that accompanied it, keeping the behavioral and the demographic pictures in contact (Kavanaugh & Jerman, 2018; Finer & Zolna, 2016).

Common Misconceptions

A method’s effectiveness is a fixed property of the method.
A method has two numbers: its perfect-use efficacy and its typically much lower typical-use effectiveness, and the gap between them is a behavioral quantity reflecting how much the method depends on the user (Trussell, 2011).
Unintended pregnancies happen mainly to people using nothing.
A large share of unintended pregnancies occurs among people using a method inconsistently or incorrectly, which is why the problem is behavioral and not only one of access (Trussell, 2011; Finer & Zolna, 2016).
Knowing the facts is enough to produce consistent use.
Information and motivation affect behavior largely through behavioral skills; education that raises knowledge without building skills changes little (Fisher & Fisher, 1992).
The best counseling always steers people to the most effective method.
Patient-centered counseling elicits values rather than issuing recommendations, and a reproductive-autonomy framework treats a freely chosen method as the goal even when it is less effective (Dehlendorf et al., 2014; Senderowicz, 2020).

Glossary

Attitude.
In the theory of planned behavior, a person’s overall positive or negative evaluation of performing the behavior, here of using contraception.
Behavioral skills.
In the information-motivation-behavioral-skills model, the concrete abilities, such as negotiating condom use or remembering a pill, that stand between motivation and behavior.
Contraception behavior.
The whole observable pattern of acquiring and using a method to prevent pregnancy over time, as distinct from the knowledge, attitudes, or intentions that precede it.
Desire to Avoid Pregnancy scale.
A prospective instrument measuring, as a continuous disposition, how much a person wants to avoid a pregnancy in the near future.
Health belief model.
A model predicting health behavior from perceived susceptibility, perceived severity, perceived benefits, and perceived barriers.
Information-motivation-behavioral-skills model.
A health-behavior model in which information and motivation act on behavior largely through a mediating layer of behavioral skills.
Intention.
The proximal cognitive determinant of behavior in reasoned-action models, representing a person’s readiness to perform the behavior.
Long-acting reversible contraception.
Implants and intrauterine devices that, once placed, require no ongoing user action, so their typical-use effectiveness nearly equals their perfect-use efficacy.
Perceived behavioral control.
A person’s judgment of how easy or hard a behavior is to perform; it predicts behavior both through intention and directly.
Perfect use.
The rate of avoiding pregnancy when a method is used correctly and consistently on every occasion.
Reproductive autonomy.
The capacity to decide freely whether and when to have children and by what means, proposed as the field’s organizing goal in place of averted pregnancies.
Subjective norm.
In the theory of planned behavior, the perceived social pressure to perform or not perform the behavior.
Typical use.
The rate of avoiding pregnancy observed across real users, who sometimes use a method inconsistently or incorrectly.
Unmet need for contraception.
The share of fecund, sexually active people who want to avoid or delay pregnancy but are using no method.

Key Researchers

Icek Ajzen

(b. 1942). University of Massachusetts Amherst; originated the theory of planned behavior, the most widely used model of the intention-to-behavior link in contraceptive research. Google Scholar - Wikipedia

Christine Dehlendorf

University of California, San Francisco; leads the research program on patient-centered contraceptive counseling and its effect on method satisfaction and continuation. Google Scholar

Kelli Stidham Hall

Tulane University; studies the links between mental health and contraceptive behavior in young women.

Jeffrey F. Peipert

University of Vermont; principal investigator of the Contraceptive CHOICE Project, which quantified the effectiveness advantage of long-acting methods. ORCID - Google Scholar

Corinne H. Rocca

University of California, San Francisco (ANSIRH); developed the Desire to Avoid Pregnancy scale, a prospective measure of pregnancy preferences.

James Trussell

(1949-2018). Princeton University (Office of Population Research); authority on contraceptive failure rates and the perfect-versus-typical-use distinction.

Frequently Asked Questions

What is contraception behavior?

It is the whole observable pattern of acquiring and using a method to prevent pregnancy over time, from choosing a method through using it consistently. It is distinct from the knowledge, attitudes, and intentions that precede it, which is why changing those antecedents does not automatically change the behavior (Fisher & Fisher, 1992).

What is the difference between perfect use and typical use?

Perfect use is the rate of avoiding pregnancy when a method is used correctly on every occasion, while typical use is the rate observed across real users who sometimes forget or use the method incorrectly. The gap between them measures how much the method depends on user behavior (Trussell, 2011).

Why do people with a method still get pregnant?

Because consistent, correct use over months and years is hard, especially for methods that demand a daily or per-act action. A favorable intention often overestimates the person’s actual control, and small lapses in an unforgiving method translate into real risk (Ajzen, 1991; Trussell, 2011).

Which theory best predicts contraceptive use?

No single theory dominates, but the theory of planned behavior is the most widely applied, predicting use from attitude, subjective norm, and perceived behavioral control via intention (Ajzen, 1991). The information-motivation-behavioral-skills model adds the insight that skills mediate the effect of information and motivation (Fisher & Fisher, 1992).

Why are long-acting methods so much more effective in practice?

Implants and intrauterine devices require no ongoing user action once placed, so they remove the main source of failure. The Contraceptive CHOICE Project found their failure rates roughly twenty times lower than those of pills, patches, and rings, with the largest advantage among younger users (Winner et al., 2012; Secura et al., 2014).

What is the LARC-first critique?

It is the argument that promoting long-acting methods because they are most effective can prioritize demographic outcomes over the individual and can enable coercion, since these methods are hard to discontinue without a clinician. The autonomy response treats a freely chosen method as the goal, including the freedom to stop (Gomez et al., 2014; Senderowicz, 2020).

What does unmet need for contraception mean?

It is the share of fecund, sexually active people who want to avoid or delay pregnancy but are using no method. Every term in the definition requires inferring a mental state or classifying an ambiguous behavior, so the estimate is sensitive to how pregnancy intentions are elicited (Bradley & Casterline, 2014).

How are pregnancy intentions measured?

Traditionally by a retrospective yes-or-no question, which conflates a prior plan with a post-hoc feeling. Prospective instruments such as the Desire to Avoid Pregnancy scale measure the desire to avoid pregnancy as a continuous disposition and predict subsequent behavior better (Rocca et al., 2019).

References

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211. https://doi.org/10.1016/0749-5978(91)90020-T

Ajzen, I., & Madden, T. J. (1986). Prediction of goal-directed behavior: Attitudes, intentions, and perceived behavioral control. Journal of Experimental Social Psychology, 22(5), 453–474. https://doi.org/10.1016/0022-1031(86)90045-4

Bradley, S. E. K., & Casterline, J. B. (2014). Understanding unmet need: History, theory, and measurement. Studies in Family Planning, 45(2), 123–150. https://doi.org/10.1111/j.1728-4465.2014.00381.x

Dehlendorf, C., Krajewski, C., & Borrero, S. (2014). Contraceptive counseling: Best practices to ensure quality communication and enable effective contraceptive use. Clinical Obstetrics and Gynecology, 57(4), 659–673. https://doi.org/10.1097/GRF.0000000000000059

Finer, L. B., & Zolna, M. R. (2016). Declines in unintended pregnancy in the United States, 2008–2011. New England Journal of Medicine, 374(9), 843–852. https://doi.org/10.1056/NEJMsa1506575

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Addison-Wesley.

Fisher, J. D., & Fisher, W. A. (1992). Changing AIDS-risk behavior. Psychological Bulletin, 111(3), 455–474. https://doi.org/10.1037/0033-2909.111.3.455

Gomez, A. M., Fuentes, L., & Allina, A. (2014). Women or LARC first? Reproductive autonomy and the promotion of long-acting reversible contraceptive methods. Perspectives on Sexual and Reproductive Health, 46(3), 171–175. https://doi.org/10.1363/46e1614

Hall, K. S., Kusunoki, Y., Gatny, H., & Barber, J. (2014). The risk of unintended pregnancy among young women with mental health symptoms. Social Science & Medicine, 100, 62–71. https://doi.org/10.1016/j.socscimed.2013.10.037

Kavanaugh, M. L., & Jerman, J. (2018). Contraceptive method use in the United States: Trends and characteristics between 2008, 2012 and 2014. Contraception, 97(1), 14–21. https://doi.org/10.1016/j.contraception.2017.10.003

Rocca, C. H., Ralph, L. J., Wilson, M., Gould, H., & Foster, D. G. (2019). Psychometric evaluation of an instrument to measure prospective pregnancy preferences: The Desire to Avoid Pregnancy scale. Medical Care, 57(2), 152–158. https://doi.org/10.1097/MLR.0000000000001048

Rosenstock, I. M. (1974). Historical origins of the health belief model. Health Education Monographs, 2(4), 328–335. https://doi.org/10.1177/109019817400200403

Secura, G. M., Madden, T., McNicholas, C., Mullersman, J., Buckel, C. M., Zhao, Q., & Peipert, J. F. (2014). Provision of no-cost, long-acting contraception and teenage pregnancy. New England Journal of Medicine, 371(14), 1316–1323. https://doi.org/10.1056/NEJMoa1400506

Senderowicz, L. (2020). Contraceptive autonomy: Conceptions and measurement of a novel family planning indicator. Studies in Family Planning, 51(2), 161–176. https://doi.org/10.1111/sifp.12114

Trussell, J. (2011). Contraceptive failure in the United States. Contraception, 83(5), 397–404. https://doi.org/10.1016/j.contraception.2011.01.021

Winner, B., Peipert, J. F., Zhao, Q., Buckel, C., Madden, T., Allsworth, J. E., & Secura, G. M. (2012). Effectiveness of long-acting reversible contraception. New England Journal of Medicine, 366(21), 1998–2007. https://doi.org/10.1056/NEJMoa1110855