Introduction

Contraception stands as one of the family planning pillars on preventing unwanted pregnancies, enabling couples to enjoy safer and more satisfying sexuality¹. Although the use of oral contraceptives has been decreasing, it remains the contraceptive method of choice among Portuguese women¹. About 20% of these women admit forgetting to take the pill every cycle or more than once a month, however, they do not report it to their physician¹, attesting to the importance of long-acting methods. These methods tend to give the user more independence allowing them not to have the inconvenience of taking a daily pill¹. On the other hand, a significant number of women present contraindications to estrogen use and/or do not tolerate its side effects, leading to poor compliance or discontinuation².

The use of long-acting reversible contraceptive me-thods (LARCs) is gaining relevance in the current literature given their high efficacy, non-dependence on use, high satisfaction and continuation rates^1-3. According to the American College of Obstetricians and Gynecologists, LARCs are defined as contraceptive me-thods suitable for all women, including young and nu-lliparous women, with a periodic administration of more than 12 weeks that, together with sterilization, present 99% efficacy on preventing pregnancy^4,5. The National Institute for Health and Care Excellence defines that all contraceptives with an administration of more than four weeks, including Progestogen-only injectable, are included in the LARCs group.

The present study was based on the American Co-llege of Obstetricians and Gynecologists classification and addressed subcutaneous (Implant) and intrauterine (IUC: intrauterine contraception) LARCs. In Portugal, four types of IUC are available: intrauterine systems (IUS) with 13.5 mg, 19.5 mg, 52 mg of levonorgestrel and the non-hormonal copper intrauterine device (IUD) ⁽².

Although the technical procedures required for LARCs application integrate the Portuguese family physicians competencies, European studies reveal a scarce use^{5) -(7}. The reluctance observed in the choice of LARCs, especially in intrauterine devices, falls on users and health professionals^{6)- (8}. The need for a medical procedure and the professionals lack of experience, may explain the hesitancy on using these devices, leading to the adoption of other contraceptive methods and/or referral to a gynecology hospital appointment⁹. Greater accessibility to these methods and better education and training of family physicians would minimize the obstacles associated with their placement, especially the waiting time⁸.

There are currently no studies in Portugal describing either family physicians contraceptive practices or the use of LARCs and techniques associated with their use. Therefore, this study aims to characterize contraceptive practices and LARCs prescribing and placement in Northern Portuguese Primary Health Care.

Methods

A cross-sectional observational study was conducted using an online survey. A non-random, “snowball” sample of family physicians specialists and trainees from the Northern Regional Health Administration of Portugal was defined. All family physicians from this region were eligible. There were no exclusion criteria.

The questionnaire (Appendix I) was designed by the researchers based on previous studies^10,11. A pre test of 10 interviews was carried out to assess language issues, comprehension of the questions, and time required for the application of the questionnaire. It was created in Google Forms®, disseminated via social networks and supported by health care institutions and professional organizations. The study was conducted between July 6th and November 1st, 2020.

The questionnaire included sociodemographic characteristics (gender, age, education, years of career), workplace characterization (Groups of Primary Care Centers, type of health unit, practice location defined as urban or rural by the participant, relative distance to the reference hospital), main contraceptive methods, counseling, clinical practice, views and opinions on the use of LARCs. The evaluation of contraception me-thods was based on three main key points: (1) which contraceptive methods are most used; (2) which are most associated with misuse and adverse effects; and (3) advice and clinical practice related to each type of LARC. In the analysis of clinical practice, we evaluate the degree of agreement of family physicians regarding clinical situations in which the use of LARCs is generally indicated. The effectiveness, the frequency of counseling and the influence of parity on their choice was investigated. We analyze the different types of LARCs (Implant and IUC) separately and explored the family physicians formation, placement, removal, referral to hospital appointment, and the main reasons for not performing the respective procedures. Views and opinions about the use of LARCs were evaluated. The participants were questioned about placement in health care facilities, importance of training, access to necessary materials, willingness, institutional barriers, time, family physicians competence and need for ultrasound guidance.

This study was approved by the Ethics Committee of the São João Hospital Center/Faculty of Medicine of the University of Porto. No compensation was attributed to the participants. Informed consent was obtained from all participants at the beginning of the questionnaire, and the questionnaire only progressed after consent had been accepted.

Statistical analysis was performed using IBM SPSS® Statistics software, version 26 for Windows 10. Absolute and relative frequencies (n, %) were used to describe categorical variables. For quantitative variables with non-normal distribution, median and interquartile range (IQR) were used. Normality was assessed by observing histograms. The estimated prevalence of the population and the respective 95% confidence intervals were calculated. Likert Scale values were presented in horizontal bar charts, built in Tableau®, 2020.4 software, with three values: the neutral central value and the sum of the two extreme values. Pearson’s Chi-square Test was used to evaluate the association between categorical variables. In the analyses, a p ≤ 0.05 level was considered significant.

Results

One hundred and seventy questionnaires were collected and three were discarded due to lack of data. A sample of 167 participants was analyzed, corresponding to 5.9% of the Northern Regional Health Administration family physicians¹². This represented 26.2% of the health units (95 of the 363 health units) ^(17-18 and 23 of the 24 Health Centre Grouping of Northern Regional Health Administration. Participants were mainly specialists, female, with ages ranging between 26 and 66 years (median 34 years). Most participants worked in model B family unit, in an urban environment, close to the reference hospital and had less than 10 years of professional experience. (Table I)

Contraceptive methods in Primary Health Care

Family physicians had a median of 6.0 (IQR: 4) family planning appointments per week. There was a diversified use of contraceptive methods, with a clear predominance of Estroprogestative pills. LARCs proved to be a highly considered option, with the Implant as the most used (64.0%) followed by IUS (59.3%). The least used contraceptives were the Progestogen-only injectable and Transdermal estroprogestative. (Figure 1)

Most family physicians recognize that LARCs are not user dependent. Oral contraceptives were the method most frequently associated with misuse, with a predominance of Estroprogestative pills over Progestogen-only pills (73.6% vs 59.8%). (Figure 2.A)

Family physicians considered that the contraceptive methods available in Primary Health Care have little association with adverse effects. Progestogen-only injectable (37.7%), Implant (34.1%) and Estroprogestative pills (32.9%) were the most associated with adverse effects, while IUC and Vaginal contraception were the least associated with adverse effects. (Figure 2.B)

Counseling and clinical practice regarding LARCs

The family physicians considered all LARCs a highly effective method. (Figure 3.A) The Implant was the most recommended LARC (74.8%). The frequency of LARCs counseling appears to decrease progressively with the decrease in the hormonal load of each device. Thus, lower hormonal load IUS and IUD are the least counseled. (Figure 3.B)

The most frequent indicators for LARCs use were estrogens contraindication (89.2%; 95%CI: 84.4% - 94.0%), patients request (83.8%; 95%CI: 78.4% - 89.2%), menorrhagia treatment (71.3%; 95%CI: 64.1% - 79.0%) and after voluntary interruption of pregnancy (60.5%; 95%CI: 53.3% - 68.3%). Five family physicians reported two additional clinical circumstances as a justification for LARCS preference: frequent forgetting to take oral contraceptives and the presence of a cognitive impairment/mental disability.

We found that LARCs recommendation was more dependent on the parity (51.5%; 95%CI: 43.1% - 59.3%) than on the age (32.9%; 95%CI: 25.7% - 40.7%). When asked about the use of LARCs according to parity, we concluded that family physicians tend to advise Implant for nulliparous (p < 0.001) and IUC for multiparous women (p < 0.001). Regarding multiparous women, higher hormonal load or copper devices tend to be specially advised. A quarter of family physicians recommend using IUC to nulliparous women, preferring the IUS with lower hormonal load. (Table II)

Procedures associated with the use of LARCs

The LARCs placement was assessed in all Health Centre Grouping participating in the study.

At least one family physician in 46.3% of the health units placed both types of LARCs and in 17.9% none of the professionals placed any of them. The Implant was the most placed LARC, being the only placed LARC in 32.6% of the health units.

Most family physicians (70.7%, 95%CI: 62.9% - 77.8%) placed the Implant, with a median of 10 placements per year (IQR: 14), as well as removed it (78.4%, 95%CI: 71.9% - 84.4%), with a median of 7 removals per year (IQR: 5). We concluded that younger family physicians (p < 0.001), with fewer years of professional experience (p < 0.001), and trainees (p = 0.014) were more apt on Implant placement. Factors such as gender or characteristics of the health care units (Groups of Primary Care Centres, type, environment, distance to the referral hospital) were not statistically significant for Implant placement or removal. (Table III)

Regarding IUC, we found that 40.1% (95%CI: 32.3% - 47.3%) of the family physicians placed the device, with a median of 9 insertions per year (IQR: 10). In contrast, 73.7% (95%CI: 67.1% - 80.2%) of family physicians removed these devices, with a median of 5 removals per year (IQR: 4). No sociodemographic factors were found to be associated with these procedures. (Table III)

About one third of the family physicians (35,9%; IC95%: 28,7% - 44,3%) placed both types of LARCs and one fourth (25.1%; 95%CI: 18.6% - 32.3%) placed neither. Family physicians who placed LARC also reported its removal (p < 0.001). About half of the family physicians who placed the Implant also placed the IUC (p < 0.001). (Figure 4.A) Of those who inserted IUC, 89.6% also placed the Implant (p < 0.001). (Figure 4.B)

The presence of a colleague in the health unit responsible for placing/removing the Implant was the main reason for not performing these procedures. The main reason for not perform IUC, was to refer the user to a gynecology appointment, followed for an uncomfortable position regarding the procedure and the fact that there are few colleagues with this competence in the health units. Placement of IUC justified almost twice as many referrals to gynecology than Implant. (Table IV)

Views and opinions

Procedures related to LARCs placement were significantly associated with the participants training and sense of autonomy. Most family physicians who placed LARCs were trained (p < 0.001) and, thereby, felt able to perform the procedures autonomously (p < 0.001). Among family physicians who did not place LARCs, about half were non-trained (p < 0.001), and of those who were trained, only one-fifth felt able to perform the procedures autonomously (p < 0.001). (Figure 5)

The importance of training family physicians was unanimous for placing both types of LARCs. The majority considered that the LARCs placement is a competence of family physicians. However, the number of physicians who excluded the placement of IUC from their competence was about three times higher than that observed for the Implant (12.0% vs 4.2%). Participants agree that the Implant should be placed in health units, but the opinion differs for the IUC (92.2% vs 65.2%). (Figure 6)

In general, family physicians felt more comfortable on the Implant placement (23.4% vs 53.2%) and showed less time to dedicate to IUC placement (48.8% vs 37.2%). Access to materials needed for Implant placement appears to be easier when compared to IUC (82.0% vs 53.2%), with family physicians identifying more institutional barriers for IUCs placement (26.4% vs 11.4%). To mention that ultrasound-guided IUC insertion was considered necessary by 47.9% of the participants. (Figure 6)

The overall sample expressed interest in acquiring more skills in LARCs placement, more pronounced for IUC (77.2%, 95%CI: 70.7% - 83.2%) than for the Implant (67.1%, 95%CI: 59.9% - 74.3%).

Discussion

Estroprogestative pills, as in other countries, remain the first-line contraceptive method in the Northern Regional Health Administration of Portugal^13-15, although the opinion of family physicians is that this method is associated with a higher rate of misuse, adverse effects and disadvantages for users.

The Implant is the second most used contraceptive method and the most recommended and placed LARC in northern region health units, opposing the worldwide tendency^14-16. For example, compared to the United States, most family physicians from the Northern Regional Health Administration of Portugal place the Implant (70.7% vs 11.3%) and feel comfortable with its insertion (68.3% vs 11.0%)¹¹.

Regarding IUC, the authors consider a marked and transverse limitation of its placement in Primary Health Care. Still, the percentage of family physicians placing IUC in the northern region of Portugal is twice the reported in the United States (40.1% vs 19.7%). The comfort level experienced by the PFs when inserting the IUC is very similar in both countries (40.8% vs 42.0%)¹⁷.

Despite the very high contraceptive efficacy of the different LARCs⁴, the family physicians in the northern region of Portugal consider the implant to be more effective than IUC. In addition, these devices are more available in health care units and family physicians feel more comfortable placing them, which may motivate their increased recommendation and placement.

Of all LARCs available, the IUS with lower hormonal load is the least recommended. Besides being more recent, it is associated with a decreased lifetime of use and increased bleeding pattern than those with higher hormonal load¹⁸, then this could be the reason for your lesser advice.

A study by Água F et al, which assessed women’s contraceptive practices in Portugal in 2015, compared the contraceptive method counseling between the family physicians and the gynecologists. Despite the different data collection methods, we inferred a significant increase in the LARCs advisement by family physicians in the northern region of Portugal (Implant: 27.3%¹ vs 64.0% and IUC: 42.9%¹ vs 81.5%) in the last 5 years, a trend also reported in Europe, Asia and developed countries¹⁶.

Family physicians from the north of Portugal recognize the clinical circumstances inherent to the preferential use of LARCs. However, its recommendation seems to depend on parity and, less extensively, on age. Most family physicians recommend IUC to multiparous and Implant to young nulliparous, limiting the recommendation of IUC to these women. This limitation appears to be based on several beliefs globally reported: the restricted indication for multiparous women, who do not intend to have more children and seek contraceptive efficacy equivalent to definitive contraception; the increased probability of ectopic pregnancy; the difficulty and/or complications associated with its insertion, such as pain, increased risk of pelvic inflammatory disease; the possibility of intrauterine perforation in adolescent and/or nulliparous women^4,20,21. However, evidence shows that all IUC are safe for all woman and associated with a low incidence of pain and complications upon insertion^4,22,23. The difficulty of insertion seems to depend more on the education and training of physicians than on the parity of the women since it does not seem to be more difficult to insert in nulliparous women. Additionally, there are two types of IUC smaller in size and hormonal load and thus fully adapted to the uterine cavity of these women⁴. These results show that more training and evidence-based education is needed to dissipate these myths.

We recognize that Implant placement has become a transversal competence among family physicians from the north of Portugal (70.7%, 95%CI: 62.9% - 77.8% family physicians from the north of Portugal placed it). These devices tend to be significantly more placed by younger doctors, trainees and family physicians with less than 10 years of experience. Therefore, a future paradigm shift in the contraceptive practices of Portuguese family physicians is expected. Most family physicians consider that the Implant should be placed in the health units, where easy access to this method is guaranteed. However, a significant fraction expressed a lack of time to dedicate to this area and the presence of a colleague in the health unit responsible for placing/removing the Implant was the main reason for not performing it.

We highlight the number of family physicians who distance the IUC insertion from the family physicians competencies. Our study found that the limited use of IUC is related to the uncomfortable position regarding this procedure, the lack of time to dedicate to this area, institutional barriers (such as inherent bureaucracy), lack of equipment and a “false need” for ultrasound control. Ultrasound-guided IUC insertion is reserved for insertion difficulties, obesity that limits bimanual examination or suspicion of uterine anomalies with cavity distortion^4,24. Ultrasound control after insertion is optional and only indicated when the wires are not visible or in case of suspicion of uterine perforation^4,24. Many family physicians are trained for IUC placement during their Gynecology and Obstetrics internship at the referral hospital. In addition to this training being often only observational, they view their placement primarily under ultrasound supervision. Consequently, most of these physicians choose to refer patients to a gynecology hospital appointment. We believe that this scenario may have a negative impact on health and wellbeing of Portuguese women by: (1) unnecessary increase in the number of gynecology appointments; (2) additional travel of the user; (3) delay at the beginning of contraception, which may negatively influence its adherence; (4) adoption of less effective contraceptive methods.

The fact that there are family physicians with the LARC insertion competence in all Health Centre Grouping who participated in the study, and in about half of the Northern region health units, the authors believe that Primary Health Care could respond to these users without the need for a hospital referral. The creation of specific medical appointments for the placement of LARCs would allow internal referrals as well as the training and experience acquisition centers for other physicians because we are of the opinion that all family physicians should have this competence.

This study shows that family physicians training is effective and a key element for the successful use and placement of LARCs and that there should be an investment in this area, especially when it comes to IUC. Regardless of the competence in the placement of LARCs, family physicians demonstrate interest in acquiring more skills, representing valuable data for the consolidation of the change that this study highlights in the training of specialists in Family Medicine.

The paradigm shift in LARC placement in Northern Primary Health Care seems to be more dependent on external barriers than on the knowledge and competence of family physicians in the northern region. However, these data deserve further attention and investigation to objectify these barriers. Portuguese legislation provides that all LARCs are free of charge and must gather all the conditions to place them in Primary Health Care. Additionally, their use may only be refused on “duly justified medical grounds”²⁵. The optimization of conditions in health facilities could improve not only the satisfaction of professionals but also the sexual health of users.

We found that in Northern Regional Health Administration of Portugal, LARC removal procedures are more performed than LARC placement procedures (78.4%, 95%CI: 71.9%-84.4% of family physicians remove the implant and 73.7%, 95%CI: 67.1% - 80.2% remove the IUC). We find these data interesting and merit an investigation in the future.

Our research portrays the contraceptive practices of family physicians from the north of Portugal and performs a psychometric analysis of the use of LARCs in this region. Given the absence of Portuguese studies in this area, this research remains a pioneer in the Portuguese health system. The limitations to this study include the low response rate. However, this potential bias was minimized by the good representativeness of the health units and Health Centre Grouping of Northern Regional Health Administration of Portugal. The “snowball sample” may cause a selection bias, as participants may have more interest or knowledge in contraception and/or use of LARCs. The current SARS-Cov-2 pandemic may have contributed to this low adherence and the decrease in the number of users of women of childbearing age, Family Planning appointments, and clinical practices evaluated in this study. Nevertheless, more studies concerning this new reality and its impact on Primary Health Care would be of great interest. A non-validated survey may have contributed to an information bias, despite having undergone a pre-test³.

Author’s contributions

MJR and LS contributed to the conception and design of the study and the survey, data collection and analysis, and writing the manuscript. LS contributed to the revising of the paper.

Conflicts of interest

The authors have no conflicts of interest.