Many people with cancer continue using blood thinners in the final stage of life to prevent blood clots. These clots can lead to thrombosis or strokes, but the medication also increases the risk of bleeding. New international research involving UMC Utrecht shows that such bleeding complications occur more often than the conditions the medication is meant to prevent. This raises the question of whether blood thinners are still beneficial in this phase of life.
Around half of all people with cancer use blood thinners near the end of life. For example, because they previously experienced a heart rhythm disorder, heart attack, stroke or thrombosis. Blood thinners reduce the risk of these events recurring and help prevent new blood clots from forming. People with certain types of cancer face an even higher risk, partly due to changes in their blood and reduced mobility. In such cases, continuing treatment may still be beneficial.
However, blood thinners also have a downside: they increase the risk of bleeding. People with cancer are particularly vulnerable to this because their bodies are more fragile, which raises the likelihood of bleeding complications. These can range from bruising and blood in the urine to severe bleeding events such as vomiting blood or brain haemorrhages.
Bleeding can be physically and emotionally distressing. “Even minor bleeding can have a major impact on someone’s quality of life. A nosebleed alone can be very burdensome, especially for people with a limited life expectancy,” says Denise Abbel, PhD candidate at Leiden University Medical Center. For this study, she collaborates with professor Geert-Jan Geersing and associate professor Carline van den Dries from UMC Utrecht.
In addition, people nearing the end of life often have limited time left to benefit from the preventive effects of blood thinners. The balance between benefits and risks may therefore shift. Taking medication every day and undergoing related monitoring can also become an additional burden in this final phase of life.
Although these disadvantages were already known, one important question remained unanswered: how often do patients with cancer stop using blood thinners near the end of life, and how often do thrombosis and bleeding occur? These were the central questions of the study.
To answer these questions, the researchers used data from the Julius network of general practitioners. This is a large network of GP practices in the UMC Utrecht region containing anonymised patient records. Abbel explains: “Many other studies rely on hospital data. But people in the final stage of life are often no longer referred to hospital. As a result, hospital data do not provide a complete picture of this group. That is why we chose GP data.”
The researchers analysed anonymised GP records from nearly 2,900 people with cancer. These patients received palliative care from their GP between 2018 and 2022. On average, they lived another 42 days after palliative care was initiated. Around one third were using blood thinners at the start of palliative care. Most patients continued taking the medication until death. Only one in five stopped, usually in the final days of life. On average, they stopped eight days before death.
The researchers also analysed the so-called “free text” notes written by GPs in patient records. This approach revealed many more reports of bleeding and thrombosis than previous studies had identified.
Abbel explains: “Doctors assign a billing code to each consultation, but these codes do not tell the full story. For example, if someone with oesophageal cancer vomits blood, the consultation is often coded simply as ‘cancer’ because the bleeding is related to the tumor. Earlier studies relied on these codes and therefore missed much of the relevant information.”
The free-text notes often contained much more detail about symptoms, events and clinical decisions. One striking finding was how common bleeding events were: they occurred slightly more often among users of blood thinners (28.5%) than among non-users (22%). By contrast, thrombosis occurred in only 3% of patients and strokes in only 2%. This applied both to people using blood thinners and those who were not.
“We see that blood thinners are often among the last medications to be stopped, sometimes only in the final days before death,” Abbel says. “This may be because doctors fear patients could still develop thrombosis or suffer a stroke. But our data show that bleeding is very common. For some patients, it may therefore be worth considering stopping blood thinners earlier. The balance between risks and benefits has changed.”
According to Abbel, it is important for doctors to discuss the risks of blood thinners with patients earlier and more openly. She also believes physicians should actively reconsider prescribing blood thinners in the final stage of life. “Many people are unaware of the disadvantages of blood thinners near the end of life. Yet the decision to stop treatment is often considered very late, or not at all.”
“We are not saying that everyone should always stop using blood thinners. That decision depends not only on the risk of bleeding or thrombosis, but also on what the patient values and wants,” Abbel adds.
To make these conversations easier, the researchers developed the CoClarity app. This digital tool helps GPs initiate discussions about blood thinners and provides information about their use and possible consequences. According to Abbel, the app may support earlier and better-informed decision-making, even before bleeding occurs. “Ultimately, it is about quality of life.”
The app is currently not publicly available and is only being used within a research setting.
This study is part of the international SERENITY collaboration. In the Netherlands, several hospitals are involved, including UMC Utrecht, Leiden University Medical Center, Erasmus MC, Reinier de Graaf Hospital and Groene Hart Hospital. The Julius General Practitioners’ Network in the UMC Utrecht region also played an important role by providing GP data, including free-text notes from patient records. This enabled researchers to assess how often patients stopped using blood thinners and how frequently bleeding and thrombosis occur in practice.
* Existing medical records were used for this study. Only data from patients who had not objected to the use of their data for scientific research were included. All directly identifiable personal data, such as names, addresses and telephone numbers, were removed beforehand to protect patient privacy.
