Добавил:
kiopkiopkiop18@yandex.ru t.me/Prokururor I Вовсе не секретарь, но почту проверяю Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:
Ординатура / Хирургия / Библиотека им академика М.И. Перельмана / Книга_1116_Библиотеки_им_академика_М_И_Перельмана.pdf
Скачиваний:
0
Добавлен:
02.09.2026
Размер:
20 Мб
Скачать
P. K. H. Walton50
2 Definition of a Registry
“A Surgical Registry is a collection of observational data on consecutive patients undergoing a particular surgical procedure (or procedures) or for a given condition to enable systematic audit”.

3 The Value of Registries

When looking at classical hierarchies in scientific evidence (Fig. 1) as first described by the Canadian Task Force on the Periodic Health Examination [2], typically registries would appear as a “cohort study” in terms of the value of evi­dence that they provide—hence with more observer bias than randomised control trials (RCTs).
Registries are very distinct from clinical trials which are designed to test hypotheses and require power calculations to determine the appropriate number of cases that need to be recruited in order to show statistical differences. Registries are not bound by power calculations and in the ideal world are never ending par­ticularly because they can provide very useful trend data on patient demograph­ics, surgical practice and outcomes which can change quite dramatically even over relatively short periods of time. This is well illustrated in the series of IFSO Global Registry Reports, the last of which—the Fifth Report [3] which shows a rapid uptake in sleeve gastrectomy worldwide in over just a few years, with a dip in the number of Roux-en-Y cases, and a new upsurge in One Anastomosis Gastric Bypass procedures especially in certain countries.
Fig. 1 Hierarchy of Research Designs and Levels of Scientific Evidence
Sleeve Gastrectomy Registries
51
It is probably a common mistake to simply think of clinical trials being more valuable or better than registries as trials and registries are really complementary and are not trying to compete with each other. It is better to think of registries being “Hypothesis Generators” and clinical trials as “Hypothesis Confirmers”. This hypothesis generation component is very important and is well illustrated by the Swedish Obesity Surgery Registry [4], which has spawned an ever increas­ing number of scientific clinical papers over time, none of which could have been imagined at the starting point of the registry in the very first place.
Instead think of registries as helping to reset understanding about outcomes. They can report on real world practice instead of the best outcomes reported in the medical literature and can give surgeons confidence that they can compare their practice with others and share this information with patients.
Clinical Trials come to a natural end when the recruitment of patients and the collection of initial and follow up data has been completed and the data have been analysed, whereas registries can go on and probably should go on forever. Indeed, typically the functionality of high quality registries is that the inbuilt analysis capability can be extended and expanded over time. This is helped by the ever increasing speed of registry software development and innovation which leads to faster and more extensive reporting and analytic capability. The British National Bariatric and Metabolic Surgery Registry (NBSR) now offers instant dashboards and extensive one-button-push composite reports, neither of which could have been conceived of when the registry was first constructed. These reports have enhanced the functionality of the registry to provide surgeons with reports that can go straight into Appraisal files and can be created in seconds, whereas manually assembling the equivalent information even a year ago might have taken many days of effort for an individual surgeon to gather all the neces­sary data together. The real beauty is that registries can provide operational func­tionality that goes way beyond just addressing particular scientific questions, the real strength of registries is the capacity to provide a comprehensive suite of outputs:
Ad hoc or automated analysis of:
– Patient population demographics – BMI distributions and trends (by gender) prior to surgery – Levels of access-to-care on a geographical basis – Trends in comorbidities/obesity related diseases – Details of operative techniques – Interoperative complications – Post-operative and long-term complications – Volume/Outcome relationships – Long term outcomes by procedure type—weight loss, changes in the rates of
obesity related diseases over time
– Revision surgery outcomes
Dashboard analyses
P. K. H. Walton52
Composite Reports
Automated Patient Reports and automated generation of follow-up letters
National and international benchmarking
Links to Global Registries
Outputs to public portals e.g. (https://nbsr.e-dendrite.com)
Therefore, working on the basis that registries are a good thing, how do you get one started and what are the secrets that makes a National Registry a success? There are ten key steps to building a successful registry, miss one or two out and a registry is at risk, miss three or more out and a registry will more than likely fail.
4 Key Step 1—Identify the Most Suitable Data
Controller
Starting a National Registry is usually driven by either Specialist Societies wish­ing to provide a service to their surgical and patient community or by Ministries of Health and Governments wishing to monitor the performance of the bariatric service provision within a country.
The most successful registries are most often formed by specialist Surgical Societies. Why is this the case?
Societies tend to have a long-lasting & uniform mission & “direction of travel”,
which can be independent of government policy which can change with each
new administration Societies provide the safest legal haven for outcome data and offer exemption
from Freedom of Information enquiries—again surgeons will often feel safer
with a Society rather than an external body having ownership and oversight of
their data Societies set standards of care and treatment protocols which are respected by
their members Society Registry benchmarks can offer direct links with revalidation & re-
accreditation processes Societies offer the strongest incentives for data submission (e.g. society mem-
bership requirements)
Societies are the most “credible” location for registries and this concept really works and is very “replicable” and has been well tested
over time
5 Key Step 2—Recruit a Respected Database Chairman
with Long Tenure
Working on the basis that a Society is the most common Data Controller; the next challenge is to identify the best person to lead the development and implementa­tion of a National Registry. Do not be tempted to appoint a surgeon for just a year
Sleeve Gastrectomy Registries
53
or even two years for such a post. The development of a registry is a long pro­cess and it is vital that there is careful stewardship of the process over an extended period of time. Some would recommend a tenure for a Database Chairman of a minimum of 5 years with the option to extend between a further 3 and 5 years. The danger of rolling the position every one or two years is that by the time the individual has worked out how to do the job, it is being handed straight over to the next person who has to go through the same learning curve to get up to speed to manage the project—this roll-over process leads to too many “stops & starts” and can put a long term registry project at huge risk of failure.
When seeking to identify and recruit an individual to take on the role of Database Chairman there are a number of key attributes that are highly desirable. Namely the person should:
Be statesmanlike and a diplomat by nature Be well respected and impartial Be regarded as a “safe pair of hands” Have a long term vision Be a proven “deliverer” Be able to demonstrate careful stewardship Be an excellent communicator
6 Key Step 3—Define Clear Objectives for a Registry
Just in the same way that writing the objectives or mission statement for a surgi­cal society or association is a necessary challenge, doing the same for a surgical registry is also an essential but not easy task. Indeed the two are often interlinked. Setting up a Surgical Society involves creating a constitution and key roles and responsibilities for both Executive Council members and for members of the Society. Setting up a registry also requires constitutional considerations, decid­ing who can sit on the Database Committee, how long the tenure should be, what determines quorate decisions and so on.
Writing down the objectives is so critically important as these then drive the dataset design and reporting requirements along with steering the required activi­ties of the Committee and its chosen data management partner. The objectives may start off very simply indeed and may centre around feasibility of enrolling all cen­tres within a country and demonstrating that basic data can be merged, analysed and reported. Down-the-road, second tier objectives are likely to be included, such as providing benchmarks of activity and performance. The objectives of a Registry should ideally be reviewed on an annual basis and should include sign-off from a Society Executive Board.
If the desire for a Society is to develop a Quality Improvement Programme, set up a mechanism to gain public trust, develop an education and training programme and provide a suitable regulatory background and public release of data…..it all starts with collecting and analysing data as seen in Fig. 2.
P. K. H. Walton54
Fig. 2 Algorithm for the Development of a Quality Improvement Program
Along with defining objectives, the Database Committee should define roles and responsibilities for the major players in the Registry—for the Data Controller, the Data Processor & the Data Contributors.
7 Key Step 4—Contract with the Right Software
Partner for You
The simple recommendation here is to choose an innovative and professional Data Management Company with long established experience in helping to set up and run national and international registries. Naturally they must have suitable security and information governance certification. In addition, ensure you choose a soft­ware company that has an established reputation for providing prompt support and fast turnaround telephone help whenever it is required.
There are three basic data flow “Models”.

7.1 Direct-Data Entry Only

If the intention is to create a new registry with just Direct-Data-Entry then the web-database should be designed so that it is as easy to enter an operation record into a bariatric surgery registry as it is to book an airline ticket on-line—the soft­ware should be intuitive, navigation controls should simple and the process of
Sleeve Gastrectomy Registries
data entry should be logical and easy with suitable onscreen prompts and/or hover tip messages available where additional user guidance is required. The SOReg Swedish National Bariatric Surgery Registry works on this model.
The simple rule of thumb is to ensure that it should take no more than 5 min per case to complete data entry into a national registry. Good database software design will ensure that there is on-line data validation to prevent inappropriate data entry e.g. to ensure there are date controls to prevent negative lengths of stay or to ensure that a balloon entry date cannot be subsequent to a definitive bariatric surgery procedure or to stop any dated data entry that is subsequent to the date of a patient’s demise. Ideally a system will allow for detailed entry for complex cases but provide a very quick run through for simple cases. Limits should be set on integer or decimal answers to alert for an entry being an abnormal result and to stop the entry of answers that are physiologically impossible (i.e. normal ranges and absolute ranges). Good registry software will include rare event trig­gers to ensure that entry of such events (e.g. death) is not accidental and must be confirmed several times. A further step is to trigger an e-mail alert to a central administrator whenever such rare events are logged so that they can be checked and confirmed.
Inbuilt security measures must ensure that a given contributor can never see or access data that belongs to another surgeon or centre unless specific permissions have been granted.
55

7.2 Electronic Upload Only

It may be that when setting up a national registry, all the contributor hospitals already have local database systems in place, in which case a central database must have the capability of uploading data files, processing them to ensure they meet a defined upload specification and reporting back to contributors if there are any deficiencies in the upload files (e.g. fields that are out of range, incompati­ble data formats, missing desirable or mandatory fields) so that the uploader can constantly refine the source data file for re-upload of high quality data. The cen­tral registry must then have the capacity to merge all uploaded data so that is then available for data analysis and reporting. The Kuwait National Bariatric Surgery Registry works on this model.
7.3 Hybrid ModelCombined Direct-Data-Entry and
Electronic Upload
The most common environment that is encountered when wishing to set up a new national registry is that are a mix of centres where the more established centres will already have a home grown or proprietary database system in place, whereas newer smaller centres may have not yet set up registry systems within their clin­ics/hospitals. In this scenario it is necessary to offer a hybrid system where data
P. K. H. Walton56
submissions can be en-bloc via electronic upload or by entering records on-line patient-by-patient. The UK National and IFSO Global Registries work in this fashion.

8 Key Step 5—Create a Suitable Minimum Dataset

With so many national bariatric and metabolic surgery registries already up and running around the world (there are examples from Australia/New Zealand to Austria, from Sweden to South Korea and from the USA to UK) it is now relatively easy to review minimum dataset that have been successfully used around the globe. Generally, these datasets are available in the public domain—as in the 2nd UK National Bariatric Surgery Registry Report [5], and these existing designs make a good starting point before adding in additional fields to suit local patient demo­graphics, practice and both research and management or sponsor needs. The key is to make the dataset comprehensive enough to permit suitable analysis and reports but to avoid making the registry design too long and onerous for contributors to complete. As a rule, it is better to have a smaller but more complete registry than an extensively detailed dataset that nobody can ever complete on a consistent basis.
Datasets should include:
Demographics and medical/surgical history Information on obesity related diseases (formerly described as comorbidities) (Possibly laboratory tests e.g. HbA1C level) Use of medical treatment or Balloons pre-operatively Height and Weight on entry to the weight loss programme and weight at the time
closest to the date of operation Operative details Peri/Post operative complications (if any) Long term outcomes & details of any revision surgery PROMS data
The International Federation for the Surgery of Obesity and Metabolic Disorders is proposing a full Delphi study to review existing datasets and to ensure that data col­lection is designed not just to track surgical outcomes, but also to consider patient per­spectives and input from all stakeholders involved in bariatric and metabolic medicine and surgery. Likewise, there are attempts underway to develop a Patient Recorded Outcome Measure (PROM) that is very specific and tailored to obesity management.

9 Key Step 6—Layer in GDPR Compliance

In very recent years, Data Protection has become an important buzzword which has resulted in registries being required to comply with new legal standards, in particular the General Data Protection Regulation (GDPR) 2016/679 European
Sleeve Gastrectomy Registries
Union laws [6] on data protection and privacy. Every registry must now have in place not just a designated data controller but also the right documentation and processes around data management and patient confidentiality. For properly anonymised registries that are better described as “audit”, consent is not a require­ment, but it is nevertheless generally “advised” and indeed is mandatory if a national registry is collecting identifiable personal data.
Key documentation that needs to be in place for all registries:
1. Data Processing Agreement (DPA)—All “Data Controllers” are required to
have a DPA with any and each organisation who will be processing their data
(The Data Processors).
2. Privacy Notice/Fair Processing Statement—This document basically
explains to the general public why you’re collecting the data, what data you’re
collecting, what you intend doing with it, who it will be shared with etc. and
also includes processes how to request what data the registry is holding and
how patients can opt-out if they wish to remove their consent.
3. Subject Access Request Page—If a database is holding identifiable patient
data of any kind, it is necessary to provide a means for the public to request
what data may be held about them…. and also provide them with an opt-out
mechanism. This should also be noted in the Privacy Notice/Fair Processing
Statement which should describe the process that is followed/detailed in
the Data Processing Agreement. Some databases also collect patient email
addresses for PROMs—in this instance there needs to be TWO consent ques-
tions with an opt-out for both. The first for holding personal data on the regis-
try, and another for holding their email address for PROMs.
4. Data/Information Sharing Agreement—Some Data Controllers implement
Information/Data Sharing Agreements with the end-user data contributors so
that they are aware of their own data collection responsibilities. This is not a
mandatory requirement but it is best practice.
57
It is essential that any partnering data management company that is acting as the Registry “Data Processor” can demonstrate compliance with the highest levels of data security.
10 Key Step 7—Recognise that There are Multiple
Stakeholders with an Interest in National Registries
It might be tempting to think that bariatric surgery registries are primarily for sur­geons. In reality there are multiple other stakeholders who take a deep interest in the analyses and the reported outcomes coming from a registry including:
– Patients and Patient Advocacy Groups – Governments because of focus on Cost & Quality and Healthcare Rationing – Colleges of Surgeons
P. K. H. Walton58
– Other National Specialist Medical & Surgical Societies – Specialist Commissioners of Care – Epidemiologists & Public Health – Institutions/Hospitals (CEOs) – Medical and Quality Assurance Directors – Referring doctors—General Practitioners and other physicians – Medical Device Companies – Pharmaceutical Companies – Health Observatories – The Press – International Audiences
With this long list in mind, it is wise to review datasets to ensure that all legitimate stakeholder interests are accounted for. Likewise when producing any reports the content, analyses and accompanying commentary should be carefully tailored to accommodate all pertinent audiences.
11 Key Step 8—Create a Suitable “Carrot and Stick”
Environment to Recruit Contributors
As with the adoption of any new technology, registries are subject to the laws of Diffusion of Innovation (Fig. 3) as described by Everett Rogers, a Professor of Communication Studies which was first published in 1962, and is now in its fifth edition [7], which describes at what rate new ideas progress The important aspect to the observation is that the speed of take up of innovation determines the point of critical mass and/or success of a registry project.
There will always be a group of surgeons who are the innovators followed closely by early adopters. Fairly rapidly there will be an adoption swell of early majority and late majority users followed finally by the laggards who are reticent to adopt new technology unless they are forced to join, become too embarrassed by not taking part or just simply wait until they can see everybody else regards the project as a real success and being involved can no longer be avoided.
Fig. 3 An Illustration of the Laws of Diffusion and Innovation
Sleeve Gastrectomy Registries
There is constant discussion at surgical Scientific Congresses about who should shoulder the responsibility of Data Collection. Some will say that the data col­lector should be completely independent of the clinical team (in order to elimi­nate bias), often a North American viewpoint, whereas in other cultures e.g. the British environment there is a different mantra, namely: “The operation is not fin­ished when you, or your assistant, puts the last stitch into the patient, the operation is finished when the data has been entered into a database”. Professor Sir Bruce Keogh (former Medical Director of NHS England) in his introduction to the 1st UK National Bariatric Surgery Report [8] stated: “After all, in my view, if you can’t describe what you’re doing and define how well you’re doing it, you have no right to be doing it at all”.
59
12 Key Step 9. - Produce Regular Reports and Analytics
and Other Outputs
It is an absolute imperative to ensure that registries do not become data cemeteries where data are never seen again. In order to encourage continued data collection and the success of any registry, regular outputs must be generated. These should span:
Individual patient reports—automated operation notes and discharge summaries Individual surgeon/hospital dashboards Ad hoc queries Data output for research studies National Reports [1, 35, 8] On-line patient portals such as the UK Surgeon Specific Outcome Reports for
NHS Bariatric Surgery [9]
Only by producing outputs to the benefit of surgeons can administrators of regis­tries expect continued commitment to data entry.
13 Key Step 10—Recognise that Each Registry Has Its
Own “Journey”
Developing a Registry is never a “single point action”, it is never finished and will always be moving. The key is to remember that a registry is always “work in pro­gress” and will evolve and mature over time. The great beauty of all registries is that as time passes, the historical data that has been entered increases in value, because, (a) trends appear and (b) long term follow up data evolves.
The very good news is that with any registry there is a possibibility of a Hawthorne Effect [10], whereby the process of auditing itself helps drive improve­ments in the quality of care and brings reductions in all kinds of adverse outcomes: post-operative complication rates, long and short-term morbidity and mortality.