We show that the majority of people acting as analogue MCI patients are willing to undergo additional testing with amyloid PET in order to receive more specific information on the cause of their cognitive decline. Even in the absence of curative treatment options, people value such information. If an additional amyloid PET resulted in better patient management alone, participants were willing to pay a price of €2000. If additional amyloid-PET testing also resulted in a delay in institutionalization, participants were willing to pay roughly €500 more.

In the current study, we asked participants to identify with a patient who just received a MCI diagnosis. Based on the case we presented, most participants preferred additional testing to learn more about the cause of MCI. In discrete choice experiments and contingent valuations, income and a participant’s socio-economic background may influence whether someone is likely to pay a (high) amount of money for a service or intervention [23]. Although we found differences in the willingness to undergo additional testing between men and women, and between people who worry more or less about future dementia, the preference for additional testing was not associated with income, education level, or participants’ quality of life. While we can reason why someone who is not worried about future dementia is less likely to prefer (expensive) additional diagnostic testing, the sex difference seems less obvious. However, this is in line with a previous study that showed that women are willing to pay less for goods with uncertain revenues [24].

From earlier studies in a clinical setting, we know that amyloid-PET contributes to higher clinician confidence in the diagnosis and better patient management [9, 25]. In addition, using amyloid PET to obtain a more accurate diagnosis resulted in lower healthcare costs and a lower rate of institutionalization [12]. In particular, we observed a delay in institutionalization of more than one year, which is larger than we described in the scenarios in the current study. While it is unlikely that such a delay would be attributable to the amyloid PET scan per se, we reason that the PET scan may contribute to a more accurate diagnosis, which in turn leads to better fitting care, and less crises and hospitalizations further in the disease trajectory. In line with this reasoning, MCI diagnostic guidelines increasingly acknowledge the value of an accurate and accurately communicated diagnosis at the level of the patient and care partner, emphasizing the personal value of accurate information [1, 8, 21]. The current study adds to this previous research by showing how much healthy participants are willing to pay for better patient management and delay in institutionalization as a result of additional diagnostic testing.

The monetary value strongly depends on how detailed the information on costs in the questionnaire is. Where in an earlier study the monetary value for AD testing, either with cerebrospinal fluid or imaging, has been estimated at €700 the estimates of the current study are substantially higher [17]. However, the former study did not provide participants with information on what such a test could potentially cost. We think it is important to provide participants with information on what a realistic price of such diagnostic interventions could be, so that they can make an informed decision. However, cost estimates may differ from country to country and the relatively high cost estimate from the current study may very well be influenced by the fact that all participants were Dutch citizens with health insurance. In the Netherlands, out-of-pocket costs vary according to the chosen reimbursement package, between a minimum of €385 and a maximum of €885 per year. In other countries, the healthcare systems are arranged differently, as are the personal costs for healthcare. In the Netherlands, people might think that certain health care services are free, because a very large share is paid by the health insurer. A consequence of this may be that they report a higher monetary value than in other countries with different healthcare systems. Therefore, the results of this study have limited generalizability. Still, we show that people value additional information with regard to the underlying pathology of their complaints, and even more so if this information results in health benefits.

In our study, we focused on amyloid PET, rather than biomarkers in CSF or blood. Amyloid PET has some specific characteristics — i.e., quantifies and visualizes amyloid burden and comes at rather high monetary costs — that make it of interest for our study. Similar studies could be done for CSF or blood-based biomarkers, and in the future, there could also be designs asking participants to weigh pros and cons of the different modalities. In this first step, however, we chose for a straightforward design focusing on only one test modality, to keep the number of choices participants were faced with limited. We feel that rather as exact monetary value of a specific test (i.e., amyloid PET scan), the results should be interpreted in a more generalized way; people are willing to pay for an etiological diagnosis, even in the absence of curative treatment. With market access of the first generation of disease-modifying treatments, interest in etiological diagnosis may increase even further. To keep healthcare accessible and scalable, it will be of the utmost importance to make use of blood-based biomarkers, whose swift development allows future role out in clinical practice.

The AD field is moving fast and with the results of lecanemab and donanemab, the landscape is about to change. In Europe, these drugs are being assessed but there will be no advice from the European Medicines Agency before 2025, which is the target group of the current study. This means that in the near future, patients will not (yet) have access to these therapies. Demonstrating the health benefits of diagnostic tests when curative treatment options are not available is challenging. In case of Alzheimer’s disease, such health benefits are likely to take place later in the disease process, potentially years after a diagnosis has been made. This is further complicated by the fact that potential cost savings will most likely be observed in the social care setting, for example by a delay in institutionalization, while the costs of diagnosis are payed in the healthcare setting. This shift of costs between settings is a major challenge for the healthcare sector, since this sector needs to bear the costs but will not gain from the benefits [26]. In the current study, we clearly show that in scenarios where the health benefit is larger, people are willing to pay more for a diagnostic test. Moreover, even without specified health benefit, participants were already willing to pay substantially for the test.

Among the strengths of the current study is that we recruited cognitively healthy participants via the Dutch Brain Research Registery [20] and the use of a professionally designed case vignette to introduce the case of MCI [21]. By doing so, we believe that the participants of the current study could realistically identify with the situation in which someone is diagnosed with MCI. Participants in this study were particularly highly educated. Although this might have been an advantage, as contingent valuation questions are complex, it limits the generalizability of the findings of this study. Family income of participants was relatively high, which might have inflated the price values that we have found. When presenting the answer options for costs for amyloid-PET, patients were randomized into an ascending and a descending sub-scheme in order to cancel out the methodological issue of ordering effects in bidding game experiments [27]. There might have been some suggestiveness in the arguments presented, and the arguments in favor and against amyloid-PET scans were not randomized, and therefore, we cannot rule out primacy or recency effects. Given the complexity of the information and the task at hand (i.e., remember information and act as analogue patients), a primacy effect might have been more pronounced resulting. Since arguments against amyloid-PET were always given first, people might have been more inclined to answer “no” to additional testing. Nonetheless, the vast majority of participants indicated to be interested in additional testing.