​

Identifying low-risk patients who can safely discontinue specialised follow-ups at hospital would benefit everyone. Patients would be spared the worry that often accompanies follow-ups and they would not have to spend unnecessary time visiting the hospital. It would also allow hospitals to allocate resources to patients with greater needs.

This could now become a reality for patients with chronic lymphocytic leukaemia (CLL). A group of researchers, including researchers from Rigshospitalet, have shown that asymptomatic, low-risk CLL patients can end specialised follow-up without any negative consequences. The results have been published in the prestigious scientific journal Blood Advances.

​The new findings have been published alongside the development of a data-driven algorithm for CLL. The algorithm has been included in a clinical trial and as an operational tool for physicians at the Department of Haematology at Rigshospitalet. The algorithm makes it possible to identify the 20%-40% of CLL patients at such low risk of disease progression that they can safely end follow-ups, explains Consultant Surgeon Carsten Utoft Nieman from the Department of Haematology and head of the research project.

“The results show that the algorithm and new approach can reduce the psychological strain of hospital visits for a large group of patients, can help us put the time of specialist physicians to better use, and can even reduce the number of infections. We believe we've found a data model that can improve the allocation of healthcare resources and which can also be used as a decision-making tool when we need to find out which patients actually benefit from follow-ups at the hospital," says Carsten Utoft Niemann.

Dropping cancer follow-ups at hospital

CLL is the most common type of blood cancer in adults. While some patients require immediate treatment, others are able to remain stable for years without intervention.

In 2019, the Department of Haematology decided to stop follow-ups of low-risk patients with CLL. The decision to end follow-ups for individual patients is based on the rule-based algorithm in combination with a medical assessment, which has now been tested in a three-year trial for patients with a low risk of relapsing with CLL.

The study included 200 patients in Denmark. The algorithm identified 112 patients who could be sent home without specialised follow-up, while the remaining 88 patients continued follow-up at the hospital as usual. Of the 112 patients, 84% never showed signs of disease progression and were therefore not referred back to the hospital within three years. The remaining 16% were re-referred to the department due to concern by their general practitioner or because they showed signs of disease progression, and there was no delay in their follow-up or treatment, according to Carsten Utoft Niemann.

The research results showed that patients without specialised follow-up had fewer hospital visits, fewer infections and had the same overall survival rate after three years as patients who continued attending follow-ups at hospital.

Data to aid decision-making

Data-driven algorithms for CLL can now be integrated with the electronic patient record system (Sundhedsplatformen), and physicians at the Department of Haematology can automatically see if a patient meets the criteria for a higher risk of infection or CLL progression. The algorithms can also help physicians in clinical decision-making to decide safely which patients can stop specialised follow-up at the hospital. The patient and physician can together decide whether to discontinue follow-up or choose follow-up as an annual questionnaire and blood sample. The practicalities in connection with the latter choice are managed by a secretary.

“We can reduce the resource consumption of specialist physicians and nurses by the 20-40% of patients who are low risk," said Carsten Utoft Niemann.

He points out that these types of data-driven solutions are necessary to meet the requirements of the healthcare system in the future.

“We believe that data solutions are a way to free up specialist physicians and nurses, which is something we really need to do. In a broader perspective, hospitals will increasingly need the help from algorithms in all areas to sort through the thousands of megabytes of data we collect for each patient group and within all disease areas. By doing so, we can generate data-driven evidence to help us make the best decisions for our patients," said Carsten Utoft Niemann.