If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
University of Sydney, School of Public Health, and the University of Technology, Centre for Health Economics Research and Evaluation, Sydney, Australia
University of Sydney, School of Public Health, and the University of Technology, Centre for Health Economics Research and Evaluation, Sydney, Australia
Long-active reversible contraceptives reduce unintended pregnancy and abortions, but uptake is low. Interventions to increase uptake in family medicine settings are untested.
Objective
The Australian Contraceptive ChOice pRoject, which was adapted from the successful US Contraceptive CHOICE study, aimed to evaluate whether a complex intervention in family medicine practices resulted in increased long-active reversible contraceptive uptake.
Study Design
This cluster randomized controlled trial was set in family practices in metropolitan Melbourne, Australia. From April 2016 to January 2017, we recruited 57 family physicians by mail invitation. Each family physician aimed to recruit at least 14 female patients. Eligible family physicians worked ≥3 sessions per week in computerized practices. Eligible women were English-speaking, sexually active, not pregnant, not planning a pregnancy in the next year, 16–45 years old, and interested in discussing contraception or in starting a new, reversible method. With the use of a randomization sequence with permuted blocks that were stratified by whether the family physician performed long-active reversible contraceptive insertion or not, family physicians were assigned randomly to a complex intervention that involved training to provide structured effectiveness-based contraceptive counselling and access to rapid referral to long-active reversible contraceptive insertion clinics. The 6-hour, online educational intervention was based on the US Contraceptive CHOICE Project and adapted for the Australian context. The control family physicians received neither the educational intervention nor access to the long-active reversible contraceptive rapid referral clinics and conducted their usual contraception counselling. We used the chi-square test, which was adjusted for clustering and stratification by whether the family physician inserted long-active reversible contraceptives, and binary regression models with generalized estimating equations and robust standard errors to compare, between the intervention and control groups, the proportions of women who had a long-active reversible contraceptive inserted. The primary outcome was the proportion of women with long-active reversible contraceptives that were inserted at 4 weeks. Secondary outcomes included women’s choice of contraceptive method, quality of life, and long-active reversible contraceptive use at 6 and 12 months. Analyses were performed according to intention-to-treat.
Results
A total of 25 intervention and 32 control family physicians recruited 307 and 433 women, respectively (N=740). Within 4 weeks, 19.3% of women in the intervention group and 12.9% of women in the control group had long-active reversible contraceptive inserted (relative risk, 2.0; 95% confidence interval, 1.1–3.9; P=.033). By 6 months, this number had risen to 44.4% and 29.3%, respectively (relative risk, 1.6; 95% confidence interval, 1.2–2.17; P=.001); by 12 months, it had risen to 46.6% and 32.8%, respectively (relative risk, 1.5; 95% confidence interval, 1.2–2.0; P=.0015). The levonorgestrel intrauterine system was the most commonly chosen long-active reversible contraceptive by women in the intervention group at all time points. Differences between intervention and control groups in mean quality-of-life scores across all domains at 6 and 12 months were small.
Conclusion
A complex intervention combination of family physician training on contraceptive effectiveness counselling and rapid access to long-active reversible contraceptive insertion clinics resulted in greater long-active reversible contraceptive uptake and has the potential to reduce unintended pregnancies.
International evidence shows that the increased use of long-acting reversible contraceptives (LARCs), defined as intrauterine devices (IUDs) and contraceptive implants, can reduce unintended pregnancy and abortion rates across all stages of a woman’s reproductive life.
The prevalence of unplanned pregnancy and associated factors in Britain: findings from the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3).
LARCs are the most effective reversible methods of contraception; with typical use, failure rates are 0.05–0.8% in the first-year of use compared with 9% with the oral contraceptive pill and 18% with male condoms.
Long-acting reversible contraceptive acceptability and unintended pregnancy among women presenting for short-acting methods: a randomized patient preference trial.
Despite this evidence, the prescription and use of LARCs remains low. In the United Kingdom, LARC prescription by family physicians (FPs) fell by 6% from 2014–2016.
Australia has similarly low rates, with national data from 2012–2013 reporting that only 11% of women were using a LARC (6.1% for IUDs and 4.9% for implants).
Long-acting reversible contraceptives are the most effective form of reversible contraception; however, the uptake of long-acting reversible contraceptives remains low. The Australian Contraceptive ChOice pRoject cluster randomized controlled trial investigated the impact of a complex family physician intervention on the uptake of long-acting reversible contraceptives.
Key findings
Training family physicians in effectiveness-based contraception counselling and providing rapid long-acting reversible contraceptive insertion clinics increased long-acting reversible contraceptive uptake in the intervention group compared with the control group.
What does this add to what is known?
Training family physicians in effectiveness-based contraceptive counselling and providing rapid-referral long-acting reversible contraceptive insertion clinics increases long-acting reversible contraceptives uptake and may reduce unplanned pregnancies. The Australian Contraceptive ChOice pRoject is the first trial to extend efficacy that was demonstrated by providing long-acting reversible contraceptive education to doctors in reproductive health/family planning clinics to family practice, where most contraceptives are prescribed.
In the US-based Contraceptive CHOICE Project (CHOICE), a prospective cohort study of 9526 women 14–45 years old,
the provision of evidence-based information about all reversible contraceptive options through structured counselling and free provision of implants and IUDs led to a significant increase in the uptake of LARC compared with national averages. This resulted in a 20-fold reduction in unplanned pregnancy rates at 3 years of follow up compared with contraceptive pill, patch, or ring users
A subsequent randomized controlled trial, also undertaken in reproductive health clinics in the United States, trained healthcare providers in LARC counselling and insertion but maintained normal costs to replicate real-life conditions. This study resulted in increased rates of counselling and LARC uptake in the intervention group and reduced pregnancy rates in women seeking family planning consultations.
These 2 studies, both undertaken in specialized clinic settings, demonstrated that improving healthcare provider knowledge and skills and addressing some of the financial and service access barriers
Current barriers and potential strategies to increase the use of long-acting reversible contraception (LARC) to reduce the rate of unintended pregnancies in Australia: an expert roundtable discussion.
can impact women’s uptake of LARC. However, in many countries, including Australia, specialized reproductive health services are not widely available, and women rely on their FP for contraceptive counselling and provision. Although the barriers to primary care provision of LARC have been well-documented,
Current barriers and potential strategies to increase the use of long-acting reversible contraception (LARC) to reduce the rate of unintended pregnancies in Australia: an expert roundtable discussion.
to our knowledge, no studies have tested interventions in this setting. Consequently, this study sought to compare a complex intervention on the uptake of LARC in the family medicine practice setting.
Materials and Methods
Trial design and oversight
The Australian Contraceptive ChOice pRoject (ACCORd) trial was set in metropolitan Melbourne, Australia, with the FP as the unit of randomization. Approved by the Monash University Human Research Ethics Committee (CF 14/3990-2014002066 and CF 16/188-2016000080) and conforming to CONSORT guidelines,
Increasing the uptake of long-acting reversible contraception in general practice: the Australian Contraceptive ChOice pRoject (ACCORd) cluster randomised controlled trial protocol.
The conduct of the trial was reviewed periodically by an independent data safety monitoring committee that comprised a statistician and 2 academic researchers (independent from the ACCORd study) who monitored recruitment, trial outcomes, and adverse events. The authors vouch for the accuracy and completeness of the data presented.
Trial Population and Recruitment Procedures
FPs were eligible if they worked ≥3 sessions (half days) per week, were based at a computerized practice, and had reception staff who could assist with recruiting. FP recruitment took place between May 2016 and January 2017, and all FPs who participated in the study gave written consent at enrolment. To avoid contamination because of cross-over effects, only 1 FP was included per practice. Participating FPs were accredited with Continuing Professional Development points necessary to maintain professional FP qualifications and received $500 (Australian dollars) as reimbursement for time spent on completion of the study.
Reception staff from ACCORd FPs invited women to complete an online eligibility survey that included contact details, with the use of an iPad (Apple, Cupertino, CA) in the waiting room. Women were eligible to participate if they were 16–45 years old, had been sexually active with a male partner in the previous 6 months or anticipated sexual activity in the subsequent 6 months, had not undergone tubal ligation or hysterectomy, had sexual partners who had not undergone a vasectomy, were neither pregnant nor anticipating a pregnancy in the next 12 months, spoke proficient English, and were interested in discussing contraception or in starting a new reversible contraceptive method.
All eligible women were contacted by telephone by an ACCORd researcher to obtain consent and complete baseline questionnaires. After enrolment, women were asked to return to their ACCORd FP within 1 week for a contraceptive counselling appointment. Any additional charges for this visit were covered by ACCORd to ensure that the women did not bear out-of-pocket costs for this additional visit. ACCORd did not provide coverage for the cost of individual contraceptive products.
Randomization and masking
The trial statistician generated a randomization sequence with permuted blocks (block sizes of 4, 6, and 8), stratified by whether the FP performed LARC insertion (IUDs/implants).
This sequence was then held by a research assistant who was not involved in the ACCORD trial. When a FP was recruited, ACCORd staff contacted the research assistant to assign the FP to the next allocation in the sequence.
Interventions
FPs in the intervention group were trained to deliver structured contraceptive counselling and given access to rapid referral to LARC insertion clinics through an online booking system. Materials from the “LARC first” (contraceptive effectiveness) online training site of the Contraceptive CHOICE project
were adapted to the Australian context with input from an advisory group comprising the project investigators, FPs, and consumers. Training was delivered online through a 6-hour training package with additional practice visits, email messages, and telephone support, where required. Structured contraceptive counselling
that consisted of nonbiased, scripted descriptions of all available contraceptive methods, with particular reference to the safety and efficacy of each method, was then delivered to the participating women by the intervention-trained FPs. FPs also collected clinical information from the women to identify any contraindications or conditions that may influence the choice of contraception. Women were able to choose their contraception method, provided that it was not medically contraindicated. The FP was then advised to screen the woman for pregnancy (history and urine pregnancy test) and chlamydia (according to clinical practice guidelines published by the Royal Australian College of General Practitioners).
The online training recommended ruling out pregnancy before (1) providing a prescription for the method of choice, (2) offering “same day” insertion of the LARC method or at a subsequent time at the FP clinic, or (3) providing an appointment for insertion of the LARC method at 1 of the insertion clinics. Emergency contraception was advised for women who had recent unprotected intercourse, although “quick start” contraception (ie, commencing contraception at any time rather than at the start of the next menstrual cycle) was recommended for women in cases in which pregnancy could be ruled out (as per the Faculty of Sexual and Reproductive Healthcare guidelines).
In both of these cases, a return appointment in 3–4 weeks for a LARC insertion (and a repeat pregnancy test) was also recommended.
A rapid referral pathway to a LARC insertion clinic with 2 local private gynecologists was implemented through an online booking system for intervention FPs who did not or chose not to perform insertions in their own rooms. Gynecologists who provided these LARC insertion clinics received payment of $300 (Australian dollars) per 3.5-hour clinic use and were free to charge patients their usual fees at these clinics.
FPs in the control group provided usual contraceptive care to women who were recruited to this group and did not have access to the rapid referral LARC insertion clinics. At the conclusion of the trial, the control group of FPs were invited to undertake the online contraceptive effectiveness training.
Fidelity checking
To ensure fidelity of the counselling, a researcher (blinded to the allocation of the FP to intervention group or control group) visited FPs in both groups. During this visit, the researcher observed a single consultation and completed a checklist regarding the content of the contraceptive counselling that was provided to ascertain whether the counselling was structured with an emphasis on effectiveness.
Trial measures
At baseline, eligible women undertook an initial telephone-based questionnaire that was drawn from the CHOICE Project
Further surveys were conducted online at 6 months (including the Medical Outcomes Survey) and at 12 months (including the Health Literacy Questionnaire and Medical Outcomes Survey). After completing each survey, women were given an entry into a monthly prize draw for a $150 gift voucher.
Participating FPs and gynecologists working in the LARC insertion clinics were asked to complete a standardized data collection form at every consultation that involved an ACCORd participant.
Primary and secondary outcomes
The primary outcome was the proportion of women who had a LARC inserted within 4 weeks of the initial contraceptive consultation with their FP. Secondary outcomes included women’s choice of contraceptive method, quality of life, and LARC use at 6 and 12 months. These outcomes were measured with the use of data that were sourced from the standardized data collection forms and from the 6- and 12-month surveys.
Statistical analysis
Current LARC use increased from 2.3–11% of all contraceptives used in Australia over a 13-year time frame.
A British study estimated that, if 5% of British women who used oral contraceptives used LARC instead, the decrease in contraceptive failure would result in 7500 annual unplanned pregnancies.
LARC Guideline Development Group The cost-effectiveness of long-acting reversible contraceptive methods in the UK: analysis based on a decision-analytic model developed for a National Institute for Health and Clinical Excellence (NICE) clinical practice guideline.
Therefore, we chose an effect size of 10%. We estimated that we would require 24 FPs and 24 women per FP in each of the 2 study groups (intervention and control) to detect a 10% increase in the LARC insertion rate, with 80% power and a significance level of 5% that allowed for stratification according to whether FPs inserted LARCs and a clustering effect (intracluster correlation) of 0.05. This corresponds to the maximum intracluster correlation for variables that are associated with FP–patient encounters in a recent cluster randomized control trial
Screening and counselling in the primary care setting for women who have experienced intimate partner violence (WEAVE): a cluster randomised controlled trial.
We aimed to recruit 27 FPs and 27 women per FP in each of the 2 study groups to allow for up to a 10% drop-out rate among FPs and a 10% drop-out among participants.
We calculated counts and proportions for descriptive characteristics of FPs and women at baseline. We used the χ2 test, adjusted for clustering and stratification by whether the FP inserted LARCs, and binary regression models with generalized estimating equations and robust standard errors to compare the proportions of women who had a LARC inserted (the primary outcome) between the intervention and control groups for women who had outcome data available. The outcomes for women were analyzed according to their randomized group (intention-to-treat analysis). This method was also applied to the secondary outcomes of LARC use at 6 and 12 months. Linear regression models, which were also adjusted for study design, were used to compare mean quality-of-life scores between groups. We conducted sensitivity analyses by adjusting for the following variables: FP sex, FP age group, women’s age group, parity, and use of LARC at baseline. Additional sensitivity analyses were carried out assuming that women with missing outcome data were not missing at random. For these analyses, we used multiple imputation under plausible missing data scenarios; women with missing outcome data had (1) the same probability of the outcome as those from the same group, (2) the same probability of the outcome as those from the control group, (3) the same probability of the outcome as those from the intervention group, and (4) no LARC inserted. Twenty imputation datasets were created in each analysis, and the results were combined with the use of Rubin’s rules. In the binary regression models and the use of interaction terms, we investigated whether the effect of the intervention varied across subgroups that were defined by age, parity, use of LARC at baseline, marital status, socioeconomic status, education, previous unintended pregnancy, and previous abortion. All analyses were carried out with SAS software (version 9.4; SAS Institute Inc, Cary, NC).
Stakeholder involvement
Before commencement of recruitment and before final ethics submission, the study tools (FP surveys) were piloted among 5 FPs who provided suggestions for amendment. FPs were also asked to assess the burden of intervention and the time required to participate in the study.
Results
Trial sites and participants
From April 2016 to May 2017, 43 FPs were allocated randomly to the intervention group (with 25 subsequent withdrawals), and 44 FPs were allocated to the control group (with 23 subsequent withdrawals). A total of 25 intervention FPs recruited at least 1 participant, as did 32 control FPs (Figure). The characteristics of the FPs were well-balanced between the intervention and control groups (Table 1). Most of the FPs, who were women who were 35–54 years old, inserted implants but not IUDs. Most FPs (81%) had ≥10 years of experience. Recognized training in contraception had been undertaken by 25% of FPs; 40% of intervention FPs and 34% of control FPs also had specific training in IUD insertion (Table 1).
Between June 2016 and July 2017, intervention FPs recruited 410 women (103 women initially expressed an interest in the study but did not consent) and control FPs recruited 622 women (189 women initially expressed an interest in the study but did not consent) that resulted in 307 and 433 women in the intervention and control groups, respectively (N=740). The characteristics of the women were also well-balanced between the 2 groups (Table 1). This balance was retained among women with available data from the Standardized Data Collection Forms and from the 6- and 12-month survey. Most women were <35 years old, had no children, and were not currently using a LARC. The rate of cohort retention was 71% in both groups.
Primary and secondary outcomes
Within 4 weeks of the contraceptive counselling consultation, 8% more women in the intervention group than in the control group had had a LARC inserted (95% confidence interval, 1.5–15.4; P=.018; Table 2), with intracluster correlation of 0.13.
Table 2Outcomes at 4 weeks, 6 months, and 12 months
LARC uptake continued to rise with time at 6 and 12 months, with a greater proportion of women in the intervention group (44% and 47%, respectively) currently using a LARC compared with the control group (29% and 33%, respectively; Table 2).
The levonorgestrel intrauterine system was the most commonly chosen LARC in the intervention group; the etonogestrel implant was most commonly chosen in the control group at the 4-week, 6-month, and 12-month time points. (Table 3). None of the interaction tests indicated a differential effect of the intervention across subgroups defined by age, parity, use of LARC at baseline, marital status, socioeconomic status, education, previous unintended pregnancy or previous abortion (Supplemental Table 1).
Table 3Choice of contraceptive method
Variable
Hormone intrauterine system, n (%)
Copper intrauterine device, n (%)
Implant, n (%)
Injection, n (%)
Oral contraceptive pill (combined or progestogen only), n (%)
Ring, n (%)
Condoms n (%)
Withdrawal, n (%)
Nothing, n (%)
Other, n (%)
Not answered, n (%)
Contraception recorded at baseline for women with data available from standardized data collection forms
Of the women, 78% had the baseline survey completed after the initial family physician visit (For these women, baseline contraception information was derived from the data collected at this initial visit. Only 1 form of contraception was recorded at these visits; however, the baseline questionnaire allowed for multiple forms. To reconcile the 2 data sources, women were been assigned the most effective method if they recorded use of multiple methods. The baseline questionnaire also did not differentiate between hormonal and copper intrauterine devices)
Intervention (n=248)
16 (6.5)
13 (5.2)
3 (1.2)
114 (46.0)
4 (1.6)
61 (24.6)
14 (5.6)
34 (13.7)
9 (3.6)
Control (n=378)
16 (4.2)
29 (7.7)
5 (1.3)
173 (45.8)
1 (0.3)
87 (23.0)
9 (2.4)
65 (17.2)
7 (1.9)
Contraception method recorded within 4 wks of initial contraceptive counseling consultation
Of the women, 78% had the baseline survey completed after the initial family physician visit (For these women, baseline contraception information was derived from the data collected at this initial visit. Only 1 form of contraception was recorded at these visits; however, the baseline questionnaire allowed for multiple forms. To reconcile the 2 data sources, women were been assigned the most effective method if they recorded use of multiple methods. The baseline questionnaire also did not differentiate between hormonal and copper intrauterine devices)
Women were not asked whether they were currently using withdrawal.
Intervention (n=219)
63 (28.8)
6 (2.7)
26 (11.9)
4 (1.8)
68 (31.1)
0 (0)
67 (30.6)
—
4 (1.8)
4 (1.8)
Control (n=308)
39 (12.7)
11 (3.6)
49 (15.9)
2 (0.7)
106 (34.4)
2 (0.7)
98 (31.8)
—
15 (4.9)
3 (1.0)
Mazza et al. Increasing LARC uptake: the ACCORd cluster RCT. Am J Obstet Gynecol 2020.
a Of the women, 78% had the baseline survey completed after the initial family physician visit (For these women, baseline contraception information was derived from the data collected at this initial visit. Only 1 form of contraception was recorded at these visits; however, the baseline questionnaire allowed for multiple forms. To reconcile the 2 data sources, women were been assigned the most effective method if they recorded use of multiple methods. The baseline questionnaire also did not differentiate between hormonal and copper intrauterine devices)
b Note only 1 form of contraception recorded at family physician visits
c Women could record multiple methods
d Women were not asked whether they were currently using withdrawal.
The results of the primary outcome analysis were similar, although the effects were smaller, when covariates were adjusted for or when missing data were imputed under various assumptions. The probability values for the comparison of binary outcomes were similar when calculated with the χ2 test, with adjustment for clustering and stratification, or with the use of binary regression with generalized estimating equation for all outcomes, except for insertion at 4 weeks where the probability values were .20 and .03, respectively (Supplemental Table 2).
The differences between intervention and control groups in mean quality-of-life scores across all domains at 6 and 12 months were small and unlikely to be of practical importance or clinical significance, despite 2 of the comparisons being statistically significant. The statistically significant differences did not persist at 12 months (Table 4).
Table 4Participant quality-of-life scales at baseline and 6 and 12 months
Adjusted for clustering by family physician, stratification (whether family physician inserts long-acting reversible contraceptives and baseline values.
Adjusted for clustering by family physician, stratification (whether family physician inserts long-acting reversible contraceptives and baseline values.
P value
Physical functioning
Intervention group
93 (11.7)
94 (10.7)
2.4 (0.04–4.7)
.05
93 (12)
1.3 (–1.4–4.1)
.34
Control group
93 (14.9)
91 (16.9)
91 (17.6)
Role limitations because of physical health
Intervention group
73 (38.9)
87 (27.7)
5.4 (–0.2–1.1)
.06
87 (29.5)
2.2 (–2.7–7.2)
.37
Control group
76 (35.3)
83 (31.6)
84 (32.4)
Role limitations because of emotional problems
Intervention group
73 (36.6)
74 (37.8)
1.3 (–5.2–7.8)
.70
75 (36)
0.6 (–4.7–5.9)
.83
Control group
75 (36.4)
73 (39.0)
74 (38.5)
Energy/fatigue
Intervention group
55 (19.3)
51 (19.9)
0.4 (–2.6–3.3)
.81
51 (21.1)
–0.5 (–4.1–3.2)
.80
Control group
52 (20.8)
50 (19.8)
50 (20.6)
Emotional well-being
Intervention group
76 (15.1)
71 (17.2)
2.3 (–0.2–4.8)
.07
72 (16.7)
0.8 (–1.9–3.5)
.56
Control group
75 (16.6)
69 (19.1)
70 (18.3)
Social functioning
Intervention group
82 (18.7)
84 (18.1)
2.3 (–1.6–6.1)
.24
82 (19.9)
–0.1 (–3.0–2.8)
.94
Control group
82 (19.6)
82 (20.3)
82 (20.2)
Pain
Intervention group
74 (21.5)
81 (18.4)
2.2 (–0.6–5.0)
.13
78 (21.9)
–0.3 (–3.1–2.4)
.81
Control group
76 (21.7)
79 (20.7)
79 (21.0)
General health
Intervention group
71 (19.1)
68 (18.4)
2.2 (1.2–3.2)
<.0001
67 (19.4)
0.7 (–2.9–3.3)
.62
Control group
70 (19.8)
66 (19.6)
66 (19.5)
Note: Question 23 of SF-36, which contributes to the energy/fatigue scale, was not included in the survey. Results were similar when missing data were imputed, with the assumption that women with missing outcome data have similar outcomes as (1) those from same group or (2) those in the control group.
Mazza et al. Increasing LARC uptake: the ACCORd cluster RCT. Am J Obstet Gynecol 2020.
a Adjusted for clustering by family physician, stratification (whether family physician inserts long-acting reversible contraceptives and baseline values.
Fidelity checks were completed for 9 intervention FPs and 12 control FPs. Initiation of structured efficacy-based contraceptive counselling was observed for 44% of the intervention FPs (n=4) compared with 8% of the control FPs (n=1). Also, the data monitoring committee met every 3 months during the recruitment and data collection phases of the study. No unexpected complications or adverse effects were noted in either group.
Comment
Principal findings
The ACCORd trial data demonstrated that a family medicine practice–based intervention that consists of online training in structured effectiveness-based contraceptive counselling and the provision of a rapid referral pathway to LARC insertion clinics results in an increased LARC uptake. Participants of FPs who had received these interventions were significantly more likely to have had a LARC inserted 4 weeks from the receipt of contraceptive counselling by their FP. This number increased by 6 months and increased further at 12 months.
Results (in context of what is known)
Although ACCORd was modelled on the successful CHOICE study in the United States,
our intervention differed from CHOICE in that it did not focus on reducing the cost of contraceptive methods, which suggests that, in contexts such as Australia, where LARC uptake is poor despite universal health coverage and subsidized contraception, the cost of contraception for an individual woman may not impact on contraceptive decision-making as much as receiving structured effectiveness-based contraceptive counselling and the availability of a timely pathway to LARC insertion. Indeed, the effect of the intervention did not differ by socioeconomic status.
Lack of FP training in LARCs and LARC insertion has been identified as a barrier to increasing LARC uptake.
Current barriers and potential strategies to increase the use of long-acting reversible contraception (LARC) to reduce the rate of unintended pregnancies in Australia: an expert roundtable discussion.
Even with training, FPs often face difficulties sustaining practice in LARC insertion; 1 study found that only approximately 30% of those FPs who were trained in LARC insertions continued to insert ≥12 devices per year, which is the minimum suggested by experts to maintain skill levels.
The ACCORd intervention did not train FPs to insert LARCs. Despite this, it still achieved increased rates of LARC uptake, which may be because the ACCORD intervention addressed other barriers that have been well-described in the literature, such as tackling the myths and misconceptions concerning LARCs held by both FPs (through the training) and women (through structured effectiveness focused counselling) ,and by making LARC insertion more accessible through rapid referral pathways to insertion clinics.
Clinical implications
Our findings are important because ACCORd is the first trial to extend the efficacy demonstrated by providing LARC education to doctors in reproductive health and family planning clinics
to a new and important site: family practice. Extending LARC education to primary care can assist the large number of women who access general practice for their health care. In many countries internationally, there is a paucity of specialized contraceptive clinics, and general practice is the main provider of women’s sexual and reproductive health services, particularly contraception.
Research implications
Although the trial demonstrated that a complex intervention that involved training FPs to deliver structured effectiveness-based contraceptive counselling and making available timely access to LARC insertion clinics is effective at increasing LARC uptake, we cannot identify which aspect of the intervention mattered the most. Although LARC uptake increased in both the intervention and control groups, the intervention group had higher uptake of the hormonal intrauterine system, which may indicate the importance of timely access to insertion clinics, especially because only 44% of intervention fidelity checks witnessed the delivery of structured efficacy based contraceptive counselling.
Strengths and limitations
The strengths of this study include the evaluation of intervention in routine general practices and the examination of the sustainability of the effects after the availability of the intervention has ceased. We undertook randomization of doctors rather than women in our cluster randomized controlled trial, which reduced contamination that would have occurred if women had been randomly assigned individually, because individual women in the same practice may have been in different groups of the study.
The intervention effect and the high cohort retention rate are also strengths that provided us with the opportunity to demonstrate the longevity of the effect of the ACCORd intervention. Although the use of LARCs in our population of participants was lower at baseline (13%) than a recently reported population-based survey that involved a younger population (19%),
At 6 months, 44% of our intervention group and 29% of our control group were using LARCs, which reflects an increase in LARC use over both groups (but significantly higher in the intervention group) and a higher proportion of current LARC users than recently reported. At 12 months, the increase was sustained with 47% of women in the intervention group and 33% in the control group. Longer follow up would have allowed us to determine whether this rise in LARC uptake persisted beyond 1 year.
Our trial had several limitations. Masking of doctors and women during implementation was not feasible; because women’s outcomes were self-reported, there may have been some bias responding to the survey questions.
Withdrawal of both FPs (58% in the intervention group and 52% in the control group) and participants (29% across both groups) from the study was higher than the 10% anticipated, which may reflect the difficulty some FPs had completing a 6-hour online learning module, an inability of participants to spend the required time to complete the study, and/or poor incentives for both FPs and participants. Future research should focus on determining whether other approaches to the training of FPs that are less time consuming, such as academic detailing or involvement in an online community of practice to achieve the same outcomes.
We originally designed the study with 24 FPs in each group and each FP recruiting 24 women. However, once recruitment began, it was apparent that some FPs would not reach the target of 24 women in the required time. For some FPs, this was because their patient population did not include many women of reproductive age. This was particularly the case for male FPs and female FPs who were themselves >45 years old. To compensate, we decided to recruit more FPs, and we allowed FPs (who were able) to recruit >24 women.
Setting 1 of the primary outcomes as LARC insertion at 4 weeks was problematic for some women because there was a delay in returning to the FP for a contraceptive consultation and a further delay if LARC referral/insertion was instigated. A more clinically meaningful outcome may have been LARC use at 6 or 12 months, to reflect LARC insertion and retention over time.
Our sample of FPs and their patients were highly educated. We anticipated that FPs who were interested in contraception would be over-represented in our study, and indeed 25% of ACCORd FPs had undertaken additional training in contraception. This rate, however, was well-balanced across both intervention and control groups, making the effect of our intervention even more compelling. Noninclusion of women who spoke limited English may affect the generalizability of our findings to women of non–English-speaking backgrounds. Additionally, our sample of women was from the metropolitan area, and rural women may face greater challenges with access to LARC insertion. The small number of male FPs in our study may impact on the generalizability of the ACCORd intervention in general practice settings where there are larger proportions of male practitioners.
The probability value for the outcome insertion at 4 weeks differed when calculated by the χ2 test, adjusted for clustering and stratification, and binary regression model with generalized estimating equations. However, the χ2 test can be less powerful than binary regression and may not detect a difference if it exists; the binary regression model will provide an unbiased estimate with appropriate confidence interval coverage. Hence, we consider the results from the binary regression model to be more informative.
In conclusion, the provision of training to FPs in structured efficacy-focused contraceptive counselling together with providing FPs with a rapid referral pathway to LARC insertion clinics results in increased LARC uptake. Implementation of this approach more broadly in family medicine practice settings, particularly in contexts in which free contraception is not feasible and in which specific sexual and reproductive health services are either not available or accessible, could lead to reductions in unplanned pregnancies and abortion.
Acknowledgments
We thank the doctors and women who participated in this study; our clinical trial coordinators Catriona Rowe, Catherine Savage, and Jennifer Raymond for recruitment and conduct of the trial; the summer and winter scholars Vivien Le, Dennis Wan Hei San, Zoe Hutton, Mian Wu, Manogna Metlapali, Hannah Youn, Sophie Kinnear, Assia Commella, Sarah Ashman, and William Poh, who assisted with data collection, data entry, administrative tasks, and literature reviews; Edwina McCarthy, Xiaoping Lin, Jacinta Clements, and Ting Xia for conducting the fidelity checks, and Maria de Leon-Santiago for administrative assistance in trial set-up.
Supplemental Material
Supplemental Table 1Subgroup analyses: insertion of long-acting reversible contraceptives at 4 weeks
Subgroup and variable
Women with information available, n
Intervention group, n (%)
Control group, n (%)
P value for interaction between intervention and subgroup variable
Yes
No
Women with information available
Yes n (%)
No n (%)
Age group, y
16–24
87
20 (23.0)
67 (77.0)
142
17 (12.0)
125 (88.0)
.61
25–34
84
17 (20.2)
67 (79.8)
153
23 (15.0)
130 (85.0)
35–45
77
11 (14.3)
66 (85.7)
83
5 (6.0)
78 (94.0)
Parity
No children
164
33 (20.1)
131 (79.9)
275
36 (13.1)
239 (86.9)
.08
1 Child
19
2 (10.5)
17 (89.5)
24
4 (16.7)
20 (83.3)
2 Children
44
7 (15.9)
37 (84.1)
63
5 (7.9)
58 (92.1)
≥3 Children
21
6 (28.6)
15 (71.4)
16
0
16 (100.0)
Marital status
Married/de facto
103
18 (17.5)
85 (82.5)
160
14 (8.8)
146 (91.3)
.23
Single
144
30 (20.8)
114 (79.2)
218
31 (14.2)
187 (85.8)
Household income
≤$600 per week
59
10 (16.9)
49 (83.1)
110
18 (16.4)
92 (83.6)
.31
>$600 per week
140
29 (20.7)
111 (79.3)
201
21 (10.4)
180 (89.6)
Highest level of education
Year 12 or below
84
18 (21.4)
66 (78.6)
127
18 (14.2)
109 (85.8)
.64
Beyond Year 12
164
30 (18.3)
134 (81.7)
251
27 (10.8)
224 (89.2)
Previous unintended pregnancy
No
200
38 (19.0)
162 (81.0)
319
33 (10.3)
286 (89.7)
.18
Yes
48
10 (20.8)
38 (79.2)
59
12 (20.3)
47 (79.7)
Previous abortion
No
214
40 (18.7)
174 (81.3)
340
36 (10.6)
304 (89.4)
.22
Yes
34
8 (23.5)
26 (76.5)
38
9 (23.7)
29 (76.3)
Using long-lasting reversible contraceptives at baseline
No
219
179 (81.7)
40 (18.3)
333
33 (9.9)
300 (90.1)
.82
Yes
29
8 (27.6)
21 (72.4)
45
12 (26.7)
33 (73.3)
Mazza et al. Increasing LARC uptake: the ACCORd cluster RCT. Am J Obstet Gynecol 2020.
Supplemental Table 2Probability values from chi-squared Mantel-Haenszel analysis and binary regression models with generalized estimating equations for outcomes
Outcomes
P value
Mantel-Haenszel analysis
Generalized estimating equation
At 4 wks after initial consult
Referred for long-acting reversible contraceptives insertion
.0001
.0002
Long-acting reversible contraceptives insertion
.033
.20
At 6 mos
Long-acting reversible contraceptive use at any time in 6 mos
<.0001
.00053
Currently using a long-acting reversible contraceptives
.0007
.003
At 12 mos
Long-acting reversible contraceptive use at any time in 12 mos
.0002
.0011
Currently using a long-acting reversible contraceptives
.0015
.0086
Mazza et al. Increasing LARC uptake: the ACCORd cluster RCT. Am J Obstet Gynecol 2020.
The prevalence of unplanned pregnancy and associated factors in Britain: findings from the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3).
Long-acting reversible contraceptive acceptability and unintended pregnancy among women presenting for short-acting methods: a randomized patient preference trial.
Current barriers and potential strategies to increase the use of long-acting reversible contraception (LARC) to reduce the rate of unintended pregnancies in Australia: an expert roundtable discussion.
Increasing the uptake of long-acting reversible contraception in general practice: the Australian Contraceptive ChOice pRoject (ACCORd) cluster randomised controlled trial protocol.
The cost-effectiveness of long-acting reversible contraceptive methods in the UK: analysis based on a decision-analytic model developed for a National Institute for Health and Clinical Excellence (NICE) clinical practice guideline.
Screening and counselling in the primary care setting for women who have experienced intimate partner violence (WEAVE): a cluster randomised controlled trial.
Supported by the National Health and Medical Research Council, (APP1081743).
This fund had no part in the conduct of the research and/or preparation of the article or any part of the in study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the article for publication.
D.M. has received research funding and sponsorship to attend conferences and has been involved in training and education activities and advisory boards outside this submitted work related to Bayer Australia and Merck Sharp & Dohme Corp (MSD); J.L. has been Chief Investigator on an ARC Linkage Grant that involves cash and in-kind support from Family Planning New South Wales and Bayer Australia and is the Director of The Australian Research Centre in Sex, Health and Society that receives funding from diverse sources listed in the annual report available from the website: http://www.latrobe.edu.au/arcshs; K.M. reports training and education activities outside this submitted work related to Bayer Australia and Merck Sharp & Dohme Corp (MSD); M.H. reports no conflicts of interest; J.F.P. has received research funding and support from CooperSurgical, Bayer, and Merck and serves on Advisory Boards for CooperSurgical and Bayer Healthcare Pharmaceuticals; K.I.B. has attended 1 international advisory board meeting for Bayer Australia for which no personal fees were received. The remaining authors report no conflict of interest.
Cite this article as: Mazza D, Watson CJ, Taft A, et al. Increasing long-acting reversible contraceptives: the Australian Contraceptive ChOice pRoject (ACCORd) cluster randomized trial. Am J Obstet Gynecol 2020;222:S921.e1-13.
32688-2&id=gr1.jpg){kind=link}