A Systematic Review of the National Breast Implant Registry for Application in Korea: Can We Predict “Unpredictable” Complications?

Background and Objectives: Since silicone breast implants were introduced to the market several decades ago, the safety of breast implants has remained controversial. Recently, several studies have explored breast implant-associated anaplastic large-cell lymphoma (BIA-ALCL) and breast implant illness (BII). Several countries have developed national breast implant registries to improve the safety and quality of breast implant surgery. We performed a systematic review of the current status of national breast implant registries and propose a pilot form of an appropriate breast implant registry model for Korea. Materials and Methods: The systematic review was conducted in accordance with the “preferred reporting items for systematic reviews and meta-analyses (PRISMA) pro forma”. PubMed and Google Scholar databases were searched to identify all articles containing information on national breast implant registries. We limited the search to articles written in the English language from 2010 to 2020. Articles were reviewed by two independent authors. Results: A total of 63 articles related to national breast implant registries, registry principles and national breast implant registry annual reports were identified. After reviewing the literature, 25 national breast implant registry-related articles were included in the full-text synthesis. Currently, four countries, The Netherlands, Australia, Sweden, and the UK, have breast implant registries with well-formed sources for big data. Overall, similarities in data points were detected for three categories: implant-related complications, operation details, and device information. However, there were differences for each registry in terms of governance, funding, and capture rate. Conclusion: After reviewing other countries’ experiences, tentative datasets for the Korean Breast Implant Registry (K-BIR) were developed. The K-BIR can improve the quality of breast implant surgery in Korea by providing datasets on overall processes and outcome measures with quality indicators and risk adjustment factors. This approach will register characteristics of patients and monitor breast implants, complications, and surgical procedures to improve the outcomes of breast implant surgery in Korea. In addition, it can be used as a track-and-trace system with automated notifications to patients in the event of a product recall or other safety concerns related to a specific type of implant. Background and Objectives: Since silicone breast implants were introduced to the market several decades ago, the safety of breast implants has remained controversial. Recently, several studies have explored breast implant-associated anaplastic large-cell lymphoma (BIA-ALCL) and breast implant illness (BII). Several countries have developed national breast implant registries to improve the safety and quality of breast implant surgery. We performed a systematic review of the current status of national breast implant registries and propose a pilot form of an appropriate breast implant registry model for Korea. Materials and Methods: The systematic review was conducted in accordance with the “preferred reporting items for systematic reviews and meta-analyses (PRISMA) pro forma”. PubMed and Google Scholar databases were searched to identify all articles containing information on national breast implant registries. We limited the search to articles written in the English language from 2010 to 2020. Articles were reviewed by two independent authors. Results: A total of 63 articles related to national breast implant registries, registry principles and national breast implant registry annual reports were identified. After reviewing the literature, 25 national breast implant registry-related articles were included in the full-text synthesis. Currently, four countries, The Netherlands, Australia, Sweden, and the UK, have breast implant registries with well-formed sources for big data. Overall, similarities in data points were detected for three categories: implant-related complications, operation details, and device information. However, there were differences for each registry in terms of governance, funding, and capture rate. Conclusion: After reviewing other countries’ experiences, tentative datasets for the Korean Breast Implant Registry (K-BIR) were developed. The K-BIR can improve the quality of breast implant surgery in Korea by providing datasets on overall processes and outcome measures with quality indicators and risk adjustment factors. This approach will register characteristics of patients and monitor breast implants, complications, and surgical procedures to improve the outcomes of breast implant surgery in Korea. In addition, it can be used as a track-and-trace system with automated notifications to patients in the event of a product recall or other safety concerns related to a specific type of implant.


Materials and Methods
The systematic review was performed to identify all studies related to national breast implant registries and databases by following the "preferred reporting items for systematic reviews and meta-analyses (PRISMA) pro forma". The PubMed (accessed 20 March 2020) and Google Scholar (accessed 20 March 2020) interfaces were used for a systematic literature search. We limited the search to articles written in the English language and published between 2010 and 2020 to include only the most recent papers. The PubMed search was performed using medical subject headings (MeSH) and keywords. The following keywords to be included in journal title were "Breast" AND "device" OR "implant" AND "registry" OR "database". The Google Scholar search was performed using free terms. The search terms were "breast implant registry" and "breast device registry". The key words were used as MeSH terms or as a free text appearing in the title or abstract. In addition, we consulted academic literature sources related to registry principles and the annual reports of each of the various national breast implant registries based on the data found on registry websites and Google search results. Due to the observational nature of the data and heterogeneity of the methodologies, it was not possible to perform conventional bias analysis such as through the use of funnel plots. Instead, the registry output differed slightly by country. For example, in the case of the Breast and Cosmetic Implant Registry (BCIR), the number of breast implants in each category of operation was not registered. Furthermore, in the case of the Swedish Breast Implant Registry (BRIMP), the number of procedures was not stated in the annual report. Our primary focus was on the nation, year established, governance, funding, data collection, registry participation, and registry output by subgroup analyses.

Results
A total of 63 articles were reviewed. Forty-eight articles were identified from the two search interfaces using the previously mentioned search terms. Fifteen additional articles about registry principles and national breast implant registry annual reports from four countries were included in the review process. After removing 24 duplicates, 39 records were screened for national information on breast implant registries. Non-English articles, unrelated articles, and articles published before 2010 were excluded. In addition, articles containing only discussion or commentaries were removed. A total of 25 national breast implant registry-related articles were included in the full-text synthesis ( Figure 1).
Medicina 2020, 56, x FOR PEER REVIEW 3 of 12 were used as MeSH terms or as a free text appearing in the title or abstract. In addition, we consulted academic literature sources related to registry principles and the annual reports of each of the various national breast implant registries based on the data found on registry websites and Google search results. Due to the observational nature of the data and heterogeneity of the methodologies, it was not possible to perform conventional bias analysis such as through the use of funnel plots. Instead, the registry output differed slightly by country. For example, in the case of the Breast and Cosmetic Implant Registry (BCIR), the number of breast implants in each category of operation was not registered. Furthermore, in the case of the Swedish Breast Implant Registry (BRIMP), the number of procedures was not stated in the annual report. Our primary focus was on the nation, year established, governance, funding, data collection, registry participation, and registry output by subgroup analyses.

Results
A total of 63 articles were reviewed. Forty-eight articles were identified from the two search interfaces using the previously mentioned search terms. Fifteen additional articles about registry principles and national breast implant registry annual reports from four countries were included in the review process. After removing 24 duplicates, 39 records were screened for national information on breast implant registries. Non-English articles, unrelated articles, and articles published before 2010 were excluded. In addition, articles containing only discussion or commentaries were removed. A total of 25 national breast implant registry-related articles were included in the full-text synthesis ( Figure 1).  Table 1 [9,10,12,13]. Overall, similarities in data points were detected for three categories: implant-related complications, operation details, and device information. The US (National Breast Implant Registry; NBIR, since 2019) and Denmark (Danish Registry for Plastic Surgery of the Breast; DPB, since 1999) were excluded due to the start-up status and resultant short trial period and the lack of recent annual reports, respectively [16][17][18]. The DBIR register included characteristics of patients, procedures, and breast implants since April 2015. Recently, 96% of the hospitals and 69% of private clinics eligible for breast implant surgery participated in DBIR. From the start of the registry in April 2015 to the end of 2018, approximately 30,000 patients, 31,000 procedures, and 62,000 breast implants were registered. In 2018, approximately 3000 patients, 3500 procedures, and 4500 breast implants were registered for reconstructive indications, and 6300 patients, 6500 procedures, and 12,500 breast implants were registered for aesthetic indications. Overall, in 2018, the data capture rate was >90%. Registry output included indications for breast implant surgery, laterality, age, smoking, body mass index (BMI), intraoperative techniques (timing of reconstruction, incision site, plane, mastopexy, capsulectomy, autologous flap cover, fat grafting, drains, mesh/acellular dermal matrix use), infection control measures, revision surgery, and device characteristics (shape, texture, coating, fill, volume) [13,20]. With 2 full registration years, the data completeness of DBIR still needs to improve [21].
The ABDR commenced national rollout of data collection in June 2015. As of the end of 2018, the ABDR had collected data on 37,603 patients with 41,921 procedures involving 78,024 breast implants. Of all registered implants, 74% had a textured surface and 20.8% had a smooth surface in reconstructive indication, while 64.0% had a textured surface and 29.5% had a smooth surface in aesthetic indications. Capsular contracture was the most common complication in both reconstructive (39.6%) and aesthetic (40.9%) indication. The opt-out rate remained low, with only 1.1% of patients choosing to opt out of participating in the ABDR. In 2018, approximately 3300 patients, 3544 procedures, and 5200 breast implants were registered for reconstructive indications, and 8690 patients, 9337 procedures, and 18,000 breast implants were registered for aesthetic indications. Registry output included procedure numbers, procedure type, laterality, age, site type, intraoperative techniques (antibiotics, antiseptic rinse, glove change for insertion, antibiotic dipping solution, sleeve/funnel), device characteristics, (texture, shape, fill), acellular dermal/synthetic matrix use, complications, and revision incidence. Collection of patient-reported outcome measures (PROM) was rolled out nationally, showing a 78% response rate at 1-year follow-up in patients with reconstructive indications and 61% response rate with aesthetic indications [22]. Their work is being undertaken with the International Collaboration of Breast Registry Activity (ICOBRA) registries including The Netherlands, Sweden, UK, and US, towards a large combined annual report examining breast implants across these countries. This will ensure that when analyses of breast implants are undertaken in different countries, implants will be compared to other similar implants [9,23].
The BRIMP started enrolling patients in May 2014; approximately 15,000 patients and 25,000 breast implants have been registered with the aim to decrease complications and increase patient safety. In 2018, approximately 300 patients (400 breast implants) were registered for reconstructive indications, and 2600 patients (5100 breast implants) were registered with data for plane, incision site, and other surgical elements for aesthetic indications. The number of procedures was not stated in the annual report. The current total level of coverage of the BRIMP is approximately 65%. Registry output included indications for breast implant surgery; the number of patients; the number of breast implants; BMI; device characteristics (manufacturer, shape, texture); patient-reported data regarding dissatisfaction with breast prior to operation, plane, incision site, and other surgical elements; antibiotics; reoperations; and risk for reoperation. The BRIMP database reported data regarding the number of implants according to a manufacturer. Mentor had up about 55% (3018 implants) of reported implants during 2018, Motiva had 37% (2050 implants), while Allergan had 8% (424 implants). The number of smooth implants increased from 639 to 948 in 2018 due to increased awareness of BIA-ALCL. The BRIMP also has displayed robust data through the years regarding the age and BMI distribution. The age distribution in primary operation shows that 80% of patients were younger than 40 years old, and normal BMI distribution was seen in 80% of patients in all age categories. The BRIMP holds data from a total of 7694 number of breasts undergoing reoperations between 2014 to 2018 with 33% of breasts augmentations undergoing reoperation within 2 years of the index operation. The patients' motivation for reoperation was mainly focused on desire for a change in breast volume or shape. The convergence of data from BRIMP with the ICOBRA data is planned [12].
The BCIR has been collecting data since the registry opened in October 2016. A total of 340 submitting organizations from England are currently registered to enter data in the registry. A total of 36,195 patients with 37,725 procedures were registered until June 2019. From July 2018 to June 2019, 2425 procedures were registered for reconstructive indications, and 9965 procedures were registered for aesthetic indications. The number of breast implants in each category of operation was not registered. Currently, the data suggest that the BCIR had an approximate case ascertainment of just over one-third. This suggests that nearly two-thirds of all people who had breast implant surgery in July 2017-June 2018 were not included in the registry. Registry output included the number of provider organizations, patients, operations, and breast implants; the category of operation; laterality, plane, incision site, other surgical elements; complications, intraoperative techniques (antibiotics, antiseptic rinse, glove change for insertion, antibiotic dipping solution, sleeve/funnel, nipple guards etc.), device, and mesh characteristics [10].
The results from four national breast implant registries showed that it can be used as an objective and transparent medical device evaluation system for post-marketing surveillance once the collected data have been effectuated.

Discussion
Compared with other implantable medical devices, breast implants have several unique characteristics. First, they are mainly used for improving quality of life for breast augmentation as well as reconstruction; i.e., patients can live without breast implants. For this reason, more accurate risk assessment is required, and tremendous anxiety or dissatisfaction is inevitable when unexpected complications occur. Second, although their use in breast reconstruction is covered by the National Health Insurance Service (NHIS) in South Korea, there is no NHIS benefit for breast augmentation. Hence, patients are reluctant to visit the hospital unless there are fatal complications from breast implants, and many complications can be under-recognized. In addition, early recognition of unexpected complications such as BIA-ALCL is not possible because patients cannot be informed of them preoperatively.
A medical device registry can provide safety systems for these conditions and has proven an integral part of the healthcare quality assurance system. By matching specific device data with patient data and recording surgical outcomes and complications, device safety or performance can be monitored, and early recognition of specific types of device failure becomes possible. In addition, with the breast implant registry, patients having high-risk implants can be tracked and informed of safety information when there are serious medical issues, such as BIA-ALCL or PIP fraud.
When building a breast implant registry (BIR), various approaches can be employed depending on the components of quality indicators (QIs) and risk adjustment factors (RAFs) selected [24]. The four national breast implant registries mentioned in the Results section have shown the greatest overlap in QIs and RAFs [9,10,12,13]. In the outcome category of QIs, various local (capsular contracture, device rupture, device deflation, device malposition, hematoma/seroma, infection) and systemic (BIA-ALCL, newly diagnosed breast cancer) complications are recorded. The QI process category includes operation details (laterality, operation type, incision site, drain use, antiseptic rinse, etc.), preoperative antibiotics, drains, and unique device identifiers (UDI; device manufacturer, device serial number, texture/shell, fill). RAFs include records of various patient risk factors (age, BMI, smoking, diabetes, chemotherapy, radiotherapy, acellular dermal matrix, or mesh etc.) that can affect surgical outcomes.
The more complex the dataset in the registry, the lower the surgeon participation, which affects data completeness and quality. Therefore, simplifying the dataset can reduce the time requirement and increase surgeon participation [25]. The ABDR and ICOBRA improved the quality of breast implant registry datasets by developing a minimum dataset form on a single page that can be completed in a few minutes to decrease complexity (Table 2) [26].
After reviewing other countries' experiences with breast implant registries, tentative datasets for the Korean Breast Implant Registry (K-BIR) were developed by commission of the Korean Society of Plastic and Reconstructive Surgeons and the Ministry of Food and Drug Safety (MFDS). The final form was selected after an online survey of 14 active members of scientific committees in the Korean Academic Association of Aesthetic and Reconstructive Breast Surgery (Table 3 and Supplementary Material Figure S1). The resulting K-BIR pilot form includes patient demographic information, operation type, timing of reconstruction, device operation type, site details, elements of the operation, and revision details as QIs and medical history, smoking, and category of operation as RAFs. We collected the breast implant registry on a trial basis in consultation with the MFDS using this form and planned to adjust the QIs through feedback from various stakeholders in the future.  To maintain newly developed breast implant registries, several obstacles should be overcome. First, it is important to increase surgeon participation. To accomplish this, government organizations such as the MFDS must have strong initiatives and promote policies in a responsible manner, indicating that the obligation to patient safety lies not with the doctor who is the user of the medical device but with the government that regulates the medical device [25]. Completing and submitting the breast implant registry form remains complicated for doctors. It takes time and effort to obtain consent from patients and to enter and transmit data. Government agencies should secure a budget to operate the registry and establish an incentive system that benefits medical institutions participating in the registry. Financial incentives are known to be the most important factors motivating surgeon participation [27]. Although the MFDS does not currently have the authority to require registry participation, reimbursement, and coverage incentives through the government incentive program or from the manufacturer will encourage participation in the breast implant registry. Punishment or unreasonable incentive/disincentive systems should not be used, as they may lead to reports of false data.
Second, legislative support should be integrated, and an opt-out consent model should be mandatory for large population capture. An opt-in consent requires consent from surgeons and patients before the data can be included on a voluntary basis, whereas an opt-out consent allows surgeons and patients to enter into the registry at the time of the procedure and assumes consent unless they have indicated its withdrawal [28]. Typically, opt-in systems have a limited data capture rate ≤20% of all implants [29]. The opt-out model has proven effective in the national breast implant registries mentioned above, where the data capture rate reached approximately 70-90%, excluding the BCIR, which remains at 30% because entry into the BCIR was originally based on patient consent. However, the basis for collection has been changed from opt-in to opt-out, and consent for registry is no longer required for operations that took place from 14 January 2019; higher patient recruitment is anticipated [10].
Third, it is important to have a system in place to protect patients' personal information [30]. People undergoing cosmetic or reconstructive breast surgery are sensitive to privacy and are often reluctant to disclose that they have had the surgery. Therefore, to increase patient participation in the breast implant registry, the security of patients' personal information is extremely important, and there should be no risk of its being exposed. Furthermore, patients should be informed that if there is a problem with the safety of an implant in the future, these systems allow for faster detection and advanced warning.
Fourth, when gathering data, it is important to develop an interface suitable for producing statistics in a time-efficient manner. The current K-BIR pilot project will generate paper reports and transmit the data to MFDS. However, this is time consuming, and the accuracy may be low. Based on multiple consultations, we concluded that an electronic data entry system that can be used immediately in the operating theatre can facilitate data capture and potentially improve completion rates and data accuracy. Therefore, we propose an application platform that the manufacturer, surgeon, and patient can access. Within the system, it is possible to identify where the implant was used and for which patients. When the manufacturer produces the breast implant and delivers it to the medical institution, the corresponding UDI is scanned and registered by the application. As this implant with the UDI is inserted, operative details can be registered by the surgeon. Furthermore, when patients access the application, they can check the type, manufacturer, and surgical information of the breast implant through the individually linked UDI ( Figure 2). Currently, South Korea has started the pilot study for national breast implant surgery, and about over 300 patients have been registered until now. Before registration, we have received informed consent from all patients. If the patient refuses to enroll, they are excluded from the registry. After registration, patients' personal information undergoes a de-identification process and cannot be accessed. Their personal information can only be accessed when there is a newly developed safety issue, such as BIA-ALCL crisis, and it is necessary for patients to be informed of it. Manufacturers can only trace the status of implants they produced, whether they are used for patients or not. In other words, when implants remain in stock or are returned, their information is removed out from the registry and only implants used are left in the registry. The hospital has authority to access surgical and implant information of the patients they operated on, not all patients. Patients can be logged in the registry after identification with their resident registration number. The patient has the authority to access their own surgical and implant information, not that of other patients on the registry. After tripartite coordination, the K-BIR application can eventually become transparent with manufacturers, institutions, and patients. Moreover, PROM is not yet included in the tentative K-BIR dataset to increase the compliance rate, but the K-BIR application intends to include PROM in the future. Recently, the importance of subjective evaluation by patients, as well as the judgment of doctors, has been supported when determining surgical outcomes. In addition, although significant scientific evidence is not yet available, studies on breast implant illnesses are increasing [31][32][33][34][35]. Therefore, a platform must be developed in which the patient can evaluate the surgical outcomes on their own and obtain more information regarding the quality of all registered breast implants by linking patient feedback to the clinical data.
used are left in the registry. The hospital has authority to access surgical and implant information of the patients they operated on, not all patients. Patients can be logged in the registry after identification with their resident registration number. The patient has the authority to access their own surgical and implant information, not that of other patients on the registry. After tripartite coordination, the K-BIR application can eventually become transparent with manufacturers, institutions, and patients. Moreover, PROM is not yet included in the tentative K-BIR dataset to increase the compliance rate, but the K-BIR application intends to include PROM in the future. Recently, the importance of subjective evaluation by patients, as well as the judgment of doctors, has been supported when determining surgical outcomes. In addition, although significant scientific evidence is not yet available, studies on breast implant illnesses are increasing [31][32][33][34][35]. Therefore, a platform must be developed in which the patient can evaluate the surgical outcomes on their own and obtain more information regarding the quality of all registered breast implants by linking patient feedback to the clinical data. A breast implant registry can be helpful for post-market safety surveillance. Current sources of nationwide data for post-market surveillance rely on passively reported data that have been collected retrospectively after the occurrence of an adverse event. However, if we combine breast implant registry data with a unique computerized automated safety surveillance tool, such as the Data Extraction and Longitudinal Trend Analysis system (DELTA), the breast implant registry will provide prospective and continuous monitoring of adverse events related to breast implants [36].
The limitation of this study is that it is a comprehensive systematic review without meta-analysis. As the four national breast implant registries are currently providing active services with slightly different data collection, the details could not be compared. Finally, our research is limited to articles retrieved from PubMed and Google scholar.

Conclusions
We have analyzed eligible functioning national breast implant registries that are using similar datasets. Previously, Wurzer et al. (2019) published a systematic review of breast implant registries to summarize the published data based on the available registers. They reported three countries currently operating national breast implant registries (Austria, Australia, and United States). The systematic review analyzed time span, number of implants, and number of patients of the UKBIR, the DPB, the Austrian Breast Implant Register, and the DBIR. However, no detailed information or results from the UKBIR were shown since 2004, and the DPB has not published another follow-up article since 2007 [14]. There are some differences and novelties in our study when compared to this. We limited our search to only data published from 2010 to analyze the current trends in national breast implant registries. To the best of our knowledge, currently four countries are active in operating national breast implant registries with annual report (The Netherlands, Australia, Sweden, and the UK). We also analyzed what is included in the data collection form of each national breast implant registry and demonstrated the process of making the K-BIR based on this. The building of this the new breast implant registry is considered necessary on a national level. Integration of these datasets is important to compare and pool data from registries. The Korean Breast Implant Registry (K-BIR) can improve the quality of breast implant surgery in Korea by providing datasets on overall processes and outcome measures with QIs and RAFs. This approach will register characteristics of patients and monitor breast implants, complications, and surgical procedures to improve the outcomes of breast implant surgery in Korea. In addition, it can be used as a track-and-trace system with automated notifications to patients in the event of a product recall or other safety concerns related to a specific type of implant.
Supplementary Materials: The following are available online at http://www.mdpi.com/1010-660X/56/8/370/s1, Figure S1: K-BIR format. Funding: This research was supported by a grant (19172MFDS301) from the Ministry of Food and Drug Safety in 2019 as part of the "Patient registration research system for investigating complications of breast implants" project. The funding body did not participate in the study design, writing of the manuscript, or in the decision to submit the manuscript for publication.