Reliable data on treatment outcomes are essential for continuously improving healthcare. Collecting these data, however, takes healthcare professionals a considerable amount of time. A team at Erasmus MC has developed a solution that automatically retrieves 90% of the mandatory data required for the national colorectal cancer registry from the electronic health record. Validation in other hospitals is the next step towards potential nationwide implementation.
.

Important data, time-consuming process

Quality registries provide insight into how treatments are delivered and what outcomes are achieved. Healthcare professionals can use this information to compare results, learn from each other and further improve care.

The Dutch Institute for Clinical Auditing (DICA) currently facilitates 26 quality registries covering a range of conditions and medical specialties. One of these is the Dutch ColoRectal Audit (DCRA), which collects data on the treatment of patients with colorectal cancer.

Although healthcare professionals recognize the value of these registries, collecting the required information creates a substantial administrative burden. Much of the data still has to be manually retrieved from the electronic health record (EHR) and entered into DICA’s online portal. At Erasmus MC, this work is carried out by a nurse specialist. In other hospitals, it may be done by surgeons themselves.

For each patient, collecting and entering the DCRA data takes an average of 25 minutes. Manually transferring information also increases the risk of data-entry errors.

.

Automatically reusing data that are already available

To reduce this administrative burden, the Erasmus MC Datahub and Department of Surgery jointly developed a technical solution that automatically retrieves the required DCRA data from the EHR.

This is done through the Health Data Platform (HDP), an additional data layer between the electronic health record and applications that use these data for other purposes, such as scientific research and quality registries. At Erasmus MC, data from the EHR are first made available within the HDP environment. The newly developed solution then automatically selects the fields required for the DCRA and can submit them to DICA in batches.

The principle is simple: information that is already available in the patient record should not have to be searched for and entered manually a second time.

.

90% of data fields retrieved automatically

For patients at Erasmus MC, 90% of the mandatory DCRA fields could be retrieved automatically from the EHR. As a result, the time required for registration fell from an average of 25 minutes to just 3 minutes per patient.

This time saving relates to a single quality registry in one hospital. There are dozens of other registries for which data are still collected in a similar, largely manual way. If the same approach proves suitable for these registries as well, the potential time savings could be much greater.

.

From a local solution to wider implementation

The next step is to validate the solution outside Erasmus MC. This will take place at Leiden University Medical Center (LUMC) and Reinier de Graaf Gasthuis.

Like Erasmus MC, LUMC has a comparable HDP architecture. This makes it possible to investigate whether the approach also works well in another hospital using the same technical foundation. Reinier de Graaf Gasthuis does not have the same HDP structure. There, the team will investigate whether the required data can be extracted directly from HiX and transferred automatically.

For this validation phase, the team is working with surgeon Jan-Willem Dekker, who is also involved in the DCRA and DICA.

If the approach proves successful in both settings, it will provide a stronger foundation for further scaling to other hospitals. The team then aims to work with DICA towards nationwide implementation of automated data submission, as well as potential application to other quality registries.

Wider implementation will require, among other things, agreement with DICA, funding and technical adaptations within individual hospitals. The team is already discussing these next steps with DICA. Differences between hospital information systems, HDP environments and arrangements for data exchange will be important factors to address.