The Ethics of Using Caregivers as Cognitive Testers

In this issue, O’Caoimh et al. [1] pioneer a novel approach to cognitive testing by evaluating the acceptability and diagnostic accuracy of a caregiver-administered cognitive assessment. The Quick Memory Check, which consists of orientation, categorical verbal fluency for animals, and logical memory (immediate recall of a story story), was administered and scored by caregivers or other family members for participants with subjective memory complaints, mild cognitive impairment, dementia, and for normal controls. The assessment results detected cognitive impairment with comparable accuracy to similar assessments administered in person by trained staff.

Timely diagnosis of dementia is a global healthcare priority [2, 3], yet dementia remains widely undiagnosed [3–6]. A major barrier to diagnosis is a lack of time for test administration in busy primary care practices [7]. There is increasing interest in trying solutions that position testing outside the provider office. Testing at home may have even greater value when repeated assessments are needed, such as when monitoring a patient for delirium post-hospitalization or after starting a new CNS-acting medication. For some patients, medical instability may make coming to the clinic difficult, and families and the clinical team may need a minimally burdensome approach for in-home screening for cognitive dysfunction.

Telehealth and digital assessments are potential solutions. Traditional cognitive tests such as the Mini-Mental State Examination and Montreal Cognitive Assessment administered in office have been shown to produce similar results to the same or slightly modified versions administered via telehealth, including high test-retest across modalities and comparable accuracy for detecting cognitive impairment [8, 9]. Furthermore, telehealth cognitive assessments have been shown to be acceptable to patients [10]. Yet, with telehealth, the barrier of limited staff time persists; in fact, additional set-up time and technology support can be required for some patients. Self-administered digital assessments have great appeal because they could alleviate burdens on staff time and there is potential to administer in any language without an interpreter. These smartphone and other “bring your own device” assessments are increasingly used in research settings, but more development and evaluation are needed to ensure they are adequately reliable and valid to support clinical diagnosis and decision-making [11–13]. In sum, of these two potential solutions, only telehealth is ready for clinical use.

O’Caoimh and colleagues [1] offer a third solution: engaging caregivers and family members as testers. This novel approach is exciting to consider as a solution given its potential advantages including efficiency for providers or staff and comfort for patients. This new option also challenges us to consider the boundaries of how caregivers or other family members should be used in remote testing. We present here three key ethical considerations.

It is important to consider whether involving a family member in testing may create administration bias. Patients have complex pre-existing relationships with their caregiver or family member that may influence how they respond when that person quizzes them about the date or asks them to remember a story. The patient may respond differently to their spouse than to a neutral, trained clinical or research staff. The pre-existing relationships may also influence how the caregiver or family member administers the tests, as they may have conscious or subconscious reasons to want the patient to score well or to score poorly. In addition, caregivers may also have variable competence to complete testing. We are reassured that the Quick Memory Check produced good results for classifying the different impairment groups, which suggests that this bias was not a major factor in performance. We do think it is important to examine how this bias may affect results for some caregiver-patient dyads, especially if clinical decisions are to be made based on this type of paradigm.

A second consideration is that exposing caregivers and family members to test materials could compromise the integrity of those test materials [14, 15]. The verbal fluency and orientation components of the Quick Memory Check are common to many neuropsychological batteries and brief clinical assessments. If a caregiver is trained to administer these tests, there is likely to be learning effects that could compromise any future evaluation of the caregiver on the same or similar tests.

Thirdly, more research, including qualitative approaches, is needed to evaluate whether and how using the caregiver or family member as a tester could compromise the patient’s autonomy, negatively impact the patient–caregiver dynamic, or increase burdens for the caregiver. Lack of awareness regarding the disease or its deficits is a common symptom of mild cognitive impairment and dementia, often increasing with progression of objective deficits [16]. For these patients, completing memory tests can be an upsetting experience when the tests are harder than anticipated; tears are not uncommon. Asking the caregiver or family member to maintain standardization in an emotionally charged clinical interaction could be awkward or even harmful for both parties. Acceptability may also differ by cultural contexts; for example, it may be less acceptable in contexts where questioning the authority of elders is considered disrespectful. Dr. O’Caoimh and colleagues [1] took a preliminary look at acceptability by asking their participants via survey whether they were agreeable to doing the testing again. Of the 27% who responded, 86% were agreeable, and 14% were not.

While the scenario of using caregivers or family members as testers is novel, the related scenario of using family members as interpreters in cognitive testing has been more extensively examined from an ethical perspective. In general, ethicists recommend that this practice should be avoided due to the multiple relationship [17]; rather, a formal interpreter should be used [18]. Family members as interpreters may aid the patients in their responses [19] and push the boundaries of confidentiality [20].

In conclusion, O’Caoimh and colleagues [1] have challenged the field of brief cognitive screening to consider what role caregivers and family members can appropriately play, which is timely as the COVID-19 pandemic has catalyzed a shift to care at home [21]. At the present time, it is our opinion that it is appropriate for caregivers or family members to set up a telehealth assessment or digital self-assessment but asking them to serve as testers requires deeper examination into the potential risks of bias, test integrity, and impacts on the patient and caregiver or family member. Any use of caregivers in testing must do so with clear explanation of the known risks and benefits, and informed consent from both caregivers and patients should be obtained with alternative options for telehealth or in-person testing by trained staff provided.