The Bariatric Surgery Registry (BSR)

The Bariatric Surgery Registry (BSR) is a population based observational quality improvement study of patients with obesity undergoing bariatric surgery and collects identifiable data obtained from participants' medical records, the participants themselves and from surgeons or hospitals. When mature, the registry will collect data from all public and private hospitals in Australia and New Zealand that perform bariatric surgery (estimated to be around 152 sites in Australia and 12 in NZ). Contributing surgeons or public hospital clinics are required to make patients who will undergo bariatric procedures aware that information related to their procedure will be provided to the BSR. Surgeons and public hospital clinics are required to display a poster that advises their practice/clinic contributes to the Bariatric Surgery Registry and give all bariatric patients a registry information flyer. Upon the registry receiving patient demographic and procedure information from the surgeon or the hospital, bariatric patients are posted a Participant Fact Sheet (PFS) further advising that participation in the registry is voluntary and their decision not to participate in the registry or to participate and later opt out, will not affect their relationship with their treating surgeon or the hospital, and their surgeon will be unaware of their participation status. After the PFS are posted, bariatric patients have two weeks to contact the Registry staff via the free call telephone number to opt out. Their participation will be presumed if no contact is made within that period. There are three options for participation: full participation; partial opt out - data is included on the registry but registry staff will not contact the participant during the follow-up period; or, complete opt out - no clinical data is included on the registry. All participants retain the right to opt out at any time. For those who choose to opt out, their information is disposed of in accordance with the National Statement for Ethical Conduct in Human Research 2007 (2015). Data are provided using one of the following options: web browser with secure authorised entry using the Bariatric Surgery Registry Interface (BSR-i); paper based data forms securely faxed or posted; or, secure electronic record transfer from surgeons' or hospitals' electronic medical record. In New Zealand, data will only be collected through direct entry through the BSR-i. Completeness and accuracy of the eligible population captured by the Registry will be assessed using hospital International Classification of Disease 10-Australia Modified (ICD-10-AM) codes. Cross-checking with ICD-10-AM codes allows for complete data capture at each hospital site and provides data validation for each procedure submitted by the surgeon. Hospital Information Services (HIS) at each hospital site or of a private hospital group provide regular ICD-10-AM coding reports for bariatric procedures performed by surgeons who participate in the Registry. These reports are sent to the registry at regular intervals by the administering organising and will only contain data of contributing surgeons. The reports are sent to the registry using a secure file transfer platform (SFTP). In the case where the surgeon or hospital had not previously provided information of a bariatric patient, the reports will be the primary source of data. When ICD-10-AM coding is the primary source, surgeons are asked to complete the missing data elements not made available from the hospitals (e.g. device/stapling information, whether it is a primary or revision operation, height/ weight information and diabetes treatment). If the surgeons are providing the missing operation data with the ICD-10-AM report verification process, they will also be asked to provide their perioperative follow-up data at the same time to enable efficiency. Patients whose details are obtained from ICD-10-AM reports will be sent a PFS and the opt-out approach is followed. If the surgeons do not supply the missing data elements, registry staff will contact the patient to confirm whether the procedure was a primary or revision and to ascertain the participant's height, weight, diabetes status and treatment prior to surgery. If surgeons or public data collectors indicate they have not seen the participant for follow-up and are unable to provide post-operative data, registry staff will contact the participant for a brief phone call to collect the information related to the perioperative period and/or 12 month intervals after surgery. This call may also include the missing data elements about their procedure if the surgeon or data collector had not provided them. Five attempts will be made to contact the patient before they are allocated to "Lost to Follow-Up." A waiver of consent has been granted by 40 ethics committees for the time during the data acquisition process when patients' data may be held without providing them the opportunity to opt out. The waiver covers from the time when the surgeons and hospitals provide patients' information before the PFS is posted and two weeks have passed; and, from the time when the BSR receives the Health Information Services discharge coding (ICD-10-AM reports) and when patients are identified whose data was not provided by surgeons or hospitals until the PFS is posted and two weeks has passed. Further, a waiver of consent has been also granted by the same ethics committees in the event that the patient dies following bariatric surgery so as not to burden next of kin. Information related to a mortality is essential to determine if the death is likely to have been caused by the bariatric surgery. Data collected are defined in the BSR's data dictionary and governed by the Steering Committee with any changes managed through a data element variation process. The data elements collected include: Patient Identifiers-full name, address, date of birth, unit record (UR) number/hospital, Individual Healthcare Identifier (IHI) number when available, email address; Patient Demographic Information-Medicare, or Department of Veterans' Affairs (DVA) number, National Health Index number (NHI) in New Zealand, phone numbers, gender, and indigenous status/ethnicity; Procedural information-surgeon name, procedure date, procedure type, primary or planned/unplanned revision status, reason for revision, bariatric procedure prior to revision, procedure abandoned or not, length of hospital stay, mortality status and if death likely related to bariatric procedure; Clinical details: start weight (for primary patients only), day-of-surgery (DOS) weight, height, diabetes status and treatment, and whether there was a concurrent renal and liver transplant; Device or staple details: type, brand, model, port fixation method, and buttressing; Peri-operative (up to 90 days :post-surgery) outcome data: date patient examined for follow-up; defined adverse events (unplanned return to theatre, unplanned admission to an intensive care unit, unplanned re-admission to hospital, prolonged hospital stay and reason or complication that caused the adverse event); mortality status and if death likely related to bariatric procedure; Annual Outcome data (for primary patients only) up to 10 years after surgery: weight, diabetes status and treatment, re-operation and reason; mortality status and if death likely related to bariatric procedure; and, Participant Feedback For quality assurance purposes, and as a basis for the potential development of the SMS/Secure Portal platform as described above, participants will be given the opportunity to provide feedback on their experience with the Registry and to advise how the BSR could better support their journey as bariatric patients. Perioperative follow-up data refers to the period between 20 and 90 days post-surgery. Annual follow-up refers to the period between 9 months prior to anniversary or 3 months following. Participants are identified as "primary" if they are entered the Registry with their first bariatric procedure or as "legacy" if they are entered with a revision bariatric procedure. The Bariatric Surgery Registry data dictionary defines each of the items on which data is collected and provides information on how the required data is sourced and entered in the database. The dictionary consists of a contents page that groups data items according to data that is collected from same source and refers to a particular phase of patient's care. Data completeness and accuracy are optimized through in-built validation and completion checks in the registry database to minimize data entry error. These include data entry controlled by form logic and limited to feasible data; the use of built in edit checks to ensure data meet required formats and ranges; accuracy enhanced by the use of exhaustive drop down lists providing all possible answers to minimize free text entry where applicable; the use of hide and show mechanisms to guide data entry to required fields; use of explanatory texts to assist data entry; validation rules applied at the time of submission with alerts to assist with errors and missing data; and the use of a participant management system to list incomplete data and other actions required. Additional quality checks post data entry includes checks for duplicate data, missing data, and data consistency. Data quality and completeness are checked at a number of stages of the data management process. Errors in data quality, when identified, may be referred back to contributing surgeons or sites for review. Case ascertainment as well as both brief and comprehensive audit processes will be performed on site by Monash University registry staff (Australia only). Registry staff will perform regular remote audit checks to verify the accuracy of data received from surgeons. This will be done during the course of a follow-up phone call from the call centre or by registry staff directly contacting surgeons' rooms. In Australia, it is intended that the BSR will periodically link to other repositories, or datasets, of health information which can validate registry data. This is to verify critical quality data such as patient mortality, defined adverse events and other outcome measures are completely captured, minimising the risk of surgeons selectively choosing whom they enter on the BSR. Requests to link New Zealand BSR data to national data sets will be considered by the Steering Committee on a case by case basis. For data linkages in Australia and New Zealand, the necessary approvals from an ethics committee will be sought when required. All statistical analyses of this registry will be undertaken with the support of the Biostatistics unit at Monash University. Non-identifiable data for both Australia and New Zealand will be provided to an independent biostatistician employed by the Monash University, Department of Epidemiology and Preventive Medicine by secure file transfer for review on at least a six monthly basis. Participants, hospitals and surgeons will not be identified in this dataset. The results will be provided to the Monash by the registry's Project Manager. Only non-clinical registry staff will have access to identifiable data. All data to be used in reports or publications are to be cleaned and verified prior to analysis. Routine, re-analysis of data will be undertaken by the Project Manager to ensure findings are reproducible. The governance of how the data activities are undertaken lies with the Steering Committee. The Terms of Reference of the BSR Steering Committee outline specific roles of the Steering Committee in relation to data as follows- monitor the registry's data quality management processes and timeliness of reporting; develop and monitor policies for access to data and responses to quality of care issues identified; review and advise on outputs from the registry; and review all research and data requests for identifiable data.