This study presents data of hospital healthcare utilisation in the Netherlands, demonstrating that AF patients present with more comorbidities than non-CVD patients. Compared to non-CVD patients, AF patients had significantly more outpatient visits, emergency department visits and a longer in-hospital stay. Additionally, patients with AF consulted more specialists. Multimorbidity played a significant role in both AF and non-CVD patients, though its impact was more pronounced in patients with AF, being associated with more outpatient visits and longer hospital stays. Finally, age was an important factor associated with hospital healthcare utilisation, also more pronounced in patients with AF.

In our study, patients with AF had significantly higher hospital healthcare utilisation than non-CVD patients, including twice as many outpatient visits and more than double the length of in-hospital days. Although the available data are limited, our findings are comparable to previous data [4, 11]. In patients with incident AF compared to a propensity-matched cohort, patients with AF had a significantly higher all-cause inpatient visits and cardiovascular-related emergency room visits [11]. The reasons for hospitalisations in patients with AF have been studied, showing that CVD, including AF itself, are predominantly the primary cause [4, 12,13,14].

Patients with AF are complex, and their care often involves multiple specialists. A study assessing the feasibility of aligning medical specialty appointments in patients with multimorbidity in the hospital outpatient setting [15], showed that it was possible, but remains challenging. Clearly, the development of multidisciplinary clinics and care managed by single case-holder or care coordinator may improve patient care and reduce healthcare costs.

Multimorbidity increases the burden on healthcare systems, leading to higher costs and polypharmacy [12]. In elderly patients (mean age 75 years), 64% had five or more comorbidities [16]. In our study, multimorbidity was associated with greater healthcare utilisation, including more outpatient visits, emergency department visits and longer in-hospital stays. Prior studies in non-AF cohorts have demonstrated the impact of multimorbidity on healthcare utilisation [17]. However, there is limited data on the impact of hospital healthcare utilisation specifically in relation to multimorbidity in AF patients.

Interestingly, we observed a high prevalence of non-cardiovascular comorbidities, particularly cancer which was present in 24% of patients with AF. The association between AF and cancer is known, however this is a complex relationship due to shared risk factors, but only limited data are available [18]. A recent study also found an association, albeit with lower percentages for both incidence and prevalence of cancer [19]. This association may clearly differ depending on the age and type of patients being studied. Nevertheless, in patients with AF, undiagnosed cancer should be recognised as one of the underlying comorbidities.

It is important to recognise the economic burden multimorbidity can pose on the healthcare system, highlighting the need to focus on ways of reducing this burden [20]. The importance of addressing comorbidities is currently studied in the European Heart Rhythm Associations (EHRA-PATHS) project, which aims to increase the diagnosis and treatment of comorbidities through creating new and effective care pathways in patients with AF [21].

Age is a well-established risk factor for AF [22]. Nevertheless, the incidence of AF is also increasing among young individuals with associated risk factors and comorbidities [22, 23]. In a study of patients with a mean age of AF onset at 46 years, 9 out of 10 had risk factors or comorbidities [23]. Data from a matched cohort study comparing those aged < 65 years to > 65 years, showed a similarly increased risk of all-cause inpatient, outpatient, and emergency room visits for both age groups for patients with AF compared to non-AF controls [11]. This is contrasts with data from the Nationwide Inpatients Sample database, which also compared those aged < 65 years to > 65 years and observed that older patients were more likely to have a longer in-hospital stay and, following discharge, were less likely to return home [24].

Hospital healthcare utilisation may also be influenced by differences in management strategies based on age. Younger patients more often undergo rhythm control strategies, such as cardioversion or ablation [25] This could explain the higher hospital healthcare utilisation of patients in our study, with this population requiring more interventions such as cardioversion or rhythm control management. We acknowledge that healthcare burden in patients with AF may be driven not only by AF itself but also by associated comorbidities. Our analysis reflects this real-world complexity. While we used calendar age and a non-CVD comparator group, biological age or frailty may better capture underlying drivers of burden, this is an important direction for future research.

The strengths of our study included the use of electronic health records from 226,991 patients, all of which have been extracted uniformly, indicating the methodology utilised could be replicated across hospitals in the Netherlands. This data collection method enabled the inclusion of typically underrepresented individuals in research allowing for more diversity in the population studied [26].

However, there are several limitations. First, some comorbidities are underrepresented such as hypertension. Only 2% of the AF patients reported having hypertension, which is contrary to existing knowledge, as hypertension is recognised as one of the most prevalent comorbidities [27]. Second, we were restricted of data, limiting our ability to perform longitudinal analysis. Third, this cohort was taken from hospital data therefore representing a potentially sicker population, as we did not have access to patients with AF in the general community for comparison. However, this cohort does highlight the complexity of AF patients who present frequently to our healthcare system. Fourth, to isolate the association of AF with healthcare use, we compared AF patients with comorbid CVD to a control group without AF or CVD, excluding those with CVD alone. While this approach strengthens internal validity, we acknowledge that excluding CVD-only patients may underestimate the broader burden of non-AF cardiovascular conditions. Additionally, as all patients with AF also had comorbid CVD, it was not possible to isolate the effect of AF alone on healthcare utilisation. While this reflects real-world comorbidity patterns, it limits our ability to distinguish whether observed differences in healthcare utilisation are specifically attributable to AF or to associated cardiovascular conditions. We acknowledge this limitation and suggest that future studies explore stratification within the AF population where feasible. Finally, we did not have access to complete patient-level data such as medication use, or pathology, or procedures undertaken.