Introduction to R-RCT

What are registry-based randomised clinical trials (R-RCT)?

Registry-based randomised clinical trials (R-RCTs) are defined as prospective randomised trials (RCT) that to some extent use registries for their implementation. The most common type of registries used for R-RCT are quality registries. A quality registry is a structured collection of personal data established specifically for the purpose of systematically and continuously developing and ensuring the quality of healthcare, and not primarily for research purposes.

An R-RCT can use a registry in different ways. What is appropriate in each case depends, for example, on the design of the registry and whether there is any possibility of adapting the registry with a technical solution/trial application for the needs of the trial and also how the registry is used by the staff in everyday clinical practice. For example, in an R-RCT, the registry can be used to find suitable trial participants for inclusion, background and outcome measures, or for collection of existing patient reported measures (PROMs). If trial-specific PROMs are to be collected, this is best done via a complementary Electronic Data Capture (EDC) system such as an eCRF.

Tasks/elements that can be performed with quality registries

This overview is mainly intended to support the planning and implementation of R-RCTs where the trial is conducted through one or more elements of data collection through an existing quality registry.

• Through a developed technical solution/trial application, selected inclusion criteria for the trial can be monitored and the clinic staff can be alerted in the quality registry if a patient would be suitable for the trial.
• Screening/inclusion and possible randomisation of trial patients in connection with registration in the quality registry through technical solution/trial application.
• Trial-specific variables for screening/inclusion and possible randomisation are collected in specific trial database.
• Using existing registry variables as trial variables. The variables of the registries can be found on the registry's websites and on the Swedish Research Council's metadata service RUT. External link.
• Manage participating clinics, display the inclusion in the R-RCT on the registry website, display detailed information on screened patients in logged-in mode for each clinic.
• Export screening and inclusion information for each clinic through technical solution/trial application.
• Data extraction for reports/monitoring/interim analysis. Extraction takes place according to the respective registry centre's data extraction processes for quality registries.

When is R-RCT suitable?

R-RCT is best suited for questions about treatment strategies or method comparisons. Questions involving the comparison of already approved medicines or CE-marked medical technology products within the intended use are also possible but require permission from the Swedish Medical Products Agency (please read more under the section Consequences of trials that belong under GCP).

A simple primary question with few secondary questions based on existing registry variables is recommended. The more secondary questions (outcome measures) and/or the desire to collect variables that are not in the registry, the more work and the higher the cost.

The simplest form of R-RCT compares established treatments for an outcome measure that can be taken from information already collected in the registry (or by cross-checking with another registry). Mortality, hospitalisation and/or readmission are common outcome measures that can be monitored without affecting the regular healthcare process.

Please take this into account for a successful R-RCT:

• The trial is planned together with representatives of the registry.
• Do trial-relevant variables have good coverage in the registry?
• Is it mandatory to fill in the critical variables of the trial?
• The continuous development of the quality registry, which may involve changes in variables during the trial period.
• What is the coverage of the trial's target population in the registry?
• Is it a direct registration in the register or is it a routine to register later. This must be considered during planning.
• How fast do clinics record data in the registry? What is the time delay between an event (e.g. a bleeding/myocardial infarction/other complication) and its actual registration in the registry?
• Are the data in the registry of sufficient quality, in terms of external and internal validity, to form the basis of research on?
• How is the ongoing monitoring of the registry conducted?

System support

Much of what makes an R-RCT cost-effective is that it can be integrated into the healthcare system and that a registry that already exists in the healthcare system is used. However, it is not evident that all healthcare registries are designed, implemented and used in a way that makes them suitable as a basis for an R-RCT. Whatever the design of the registry, specific technical solutions/trial applications must be adapted and quality assured for each individual trial. The fact that the registry contains data on patients in care is only of value if it is the same data that the trial needs and if it is maintained with sufficient quality.

Make careful analysis with the technical management

Once there is an idea for an R-RCT, a careful analysis of how the registry will be used in the trial is needed. It should be analysed with the technical management of the registry whether it is possible to adapt the registry with the technical solution/trial application required for the trial and at what cost. Other options for data collection that do not involve adaptation of the registry should be explored. A combination of registry data and eCRF is common in an R-RCT.

If the trial is to be conducted under Good Clinical Practice (GCP)

If the trial is to be conducted under Good Clinical Practice (GCP), it is a requirement that electronic systems used in the trial must have been formally validated. In such a case, integrating randomisation into the quality registry is not recommended, as validation involves a high workload for the registry management team and a higher cost for the trial. For these trials, an alternative technical solution should be used to build a stand-alone trial application from the registry where basic patient information is entered and randomisation takes place. Subsequently, manual or automatic data extractions are made from the registry. It could be beneficial to add a link to the trial application in the quality registry to simplify the process for clinical staff.

Time and resources

Building a working technical solution/trial application for an R-RCT requires careful planning. All the requirements needed to develop the technical solution/trial application must be carefully planned in collaboration between researchers, registry controllers and the “supplier” of the technical solution/trial application. The planning should also include management of the trial within the registry and a plan for data export and decommissioning of the trial within the registry. During the course of the trial, no variables used in the trial should be updated (either registry variables or trial-specific variables in the technical solution/trial application) without discussing the implications of these changes. The trial should be aware of the registry's long-term development plan and how it may affect the trial. A plan for communication between the trial and the registry should be established.

Time savings with R-RCT

The start-up period for an R-RCT is significantly longer and more time-consuming than for a traditional RCT. There is often a waiting period to start an R-RCT in a registry depending on the registry's regular priorities and other ongoing R-RCTs. The cost of preparing an R-RCT should be compared to the potential time gain estimated in the overall conduct of the trial.

The time savings are made in the clinics where research support staff get more time to include patients. The entry time in the eCRF is reduced as the information is partly entered directly into the registry. Patients can be randomised and included faster because the registry can remind you immediately when a patient is suitable for the trial, if the technical support is available. However, this requires clinic staff to be logged into the register and enter data directly when the patient is present. Usually this means that the registry detects whether the inclusion criteria for the R-RCT are met and then provides a notification that the patient meets the requirements for the relevant R-RCT. Considering the practices of the clinics can therefore be crucial for a successful R-RCT.

Cost of an R-RCT

The cost of an R-RCT depends on many factors and needs to be calculated for each individual trial. An R-RCT involves some new costs compared to a traditional RCT. Depending on the extent and complexity of the technical solutions and validation required for the trial, there will be an increased management cost in terms of technical support such as:

• Operation
• Bug fixes
• Adjustment of solution
• Support
• Data extraction

This should be compared to the time and cost for clinicians to enter trial data into a traditional eCRF. It is therefore important to take this into account at the planning stage of an R-RCT.

Quality

The possibilities to ensure data quality differ between quality registries and eCRFs. This is because in an eCRF there is the possibility to tailor logic controls to the specific trial. Data in registries are collected for a different purpose and under different legislation and it is therefore possible that there is a different practice on how and what to document. Another factor that may make a difference is that there may be a larger group of staff entering data into a register. If specific instructions or directions are required for data entry in the trial, an eCRF should be used.

Ongoing quality control

Ongoing quality control is most easily done by extracting data from the registry or by continuously saving the trial data in an EDC or trial database if a technical solution/trial application for this is available. The agreement for data extraction for quality control should be established between the trial and CPUA authorities before the start of the trial and mentioned in the application to the ethical review board. If the trial falls under GCP then please read more in the section Consequences of trials falling under GCP.

Advantages and disadvantages of R-RCTs and RCTs

As a rule of thumb, an R-RCT can be regarded as most effective when:

• The trial is integrated in only one registry
• All safety and primary outcomes are in registries.
• Minimal changes must be made to the clinic's normal workflow.
• No completion/follow-up needs to be done in an EDC system.
• The trial has a short recruitment and follow-up period.
• The patient population is large
• Patients are randomised to compare established and approved treatments.
• Training needs to comply with the protocol are low
• The registry where the inclusion will be made is used at the time of meeting the patient.

When choosing between designing the trial as an R-RCT or as a traditional RCT, the advantages and disadvantages of the concepts should be considered.

Advantages and disadvantages of an R-RCT