In medical practice, a patient’s loss of competency is a major obstacle when choosing a treatment and a starting treatment program smoothly. A large number of studies have revealed the lack of medical competency in patients with dementia. However, there have been only a few reports focusing on the capacity of patients with mild cognitive impairment (MCI) to make a medical choice.

In this study, we evaluated the competency of 40 patients with amnestic MCI (aMCI) and 33 normal subjects to make a medical choice using the MacArthur Competence Assessment Tool-Treatment (MacCAT-T). We compared the judgement of a team conference using the recorded semi-structured interview with the clinical judgement of a chief clinician.

A team conference concluded that 12 aMCI patients had no competency, and the clinical judgement, without any special interview, judged that five aMCI patients had no competency. All subjects in the control groups were judged to be competent to consent to treatment by both clinicians and the team conference.

Without supplementary tools such as explanatory documents, not a few patients with aMCI were judged by a team conference to have no competency to consent to therapy even in a relatively simple and easy case. In contrast, clinical physicians tended to evaluate the competency of aMCI patients in a generous manner.

The use of assessment tools has been shown to improve the inter-rater reliability of capacity assessments. However, instrument-based capacity assessments of people with dementia face challenges. In dementia research, measuring capacity with instruments like the MacArthur Competence Assessment Tool for Treatment (MacCAT-T) mostly employ hypothetical treatment vignettes that can overwhelm the abstraction capabilities of people with dementia and are thus not always suitable for this target group. The primary aim of this study was to provide a standardized real informed consent paradigm that enables the dementia-specific properties of capacity to consent to treatment in people with dementia to be identified in a real informed consent process that is both externally valid and ethically justifiable.

The sample consisted of 53 people with mild to moderate dementia and a group of 133 people without cognitive impairment. Rather than using a hypothetical treatment vignette, we used a standardized version of the MacCAT-T to assess capacity to consent to treatment with cholinesterase inhibitors in people with dementia. Inter-rater reliability, item statistics, and psychometric properties were also investigated.

Intraclass correlations (ICCs) (0.951–0.990) indicated high inter-rater reliability of the standardized real informed consent paradigm. In the dementia group, performance on different items of the MacCAT-T varied. Most people with dementia were able to express a treatment choice, and were aware of the need to take a tablet. Further information on the course of the disorder and the benefits and risks of the treatment were less understood, as was comparative reasoning regarding treatment alternatives.

The standardized real informed consent paradigm enabled us to detect dementia-specific characteristics of patients’ capacity to consent to treatment with cholinesterase inhibitors. In order to determine suitable enhanced consent procedures for this treatment, we recommend the consideration of MacCAT-T results on an item level. People with dementia seem to understand only basic information. Our data indicate that one useful strategy to enhance capacity to consent is to reduce attention and memory demands as far as possible.