Chapter 7 170 instead of 3–30. Specificity increased even though the control group became more like the disease group; namely, the three groups (i.e. MD – MCI – NoCI) were all suspected of having cognitive symptoms. Referred patients in whom there was an obvious other cause were apparently excluded from the group ‘suspected of having cognitive symptoms’ after the initial assessment even if there was a (very) abnormal MoCA (in retrospect). This indicates that the clinician does not worry about cognitive impairment, for example, as it was temporary, or sees other (reasons why these patients present) symptoms that could affect the MoCA score, such as lack of motivation. Conversely, the effects of having few selection criteria for a cohort and without correction (by clinicians) due to individual factors (of the patient) is highly objective but introduces outliers. Although taking account of individual circumstances may be the obvious thing to do in clinical practice, not being able to do this is has consequences for a cohort study, as designed in Chapter 3. The advantage of not excluding anyone, automatically entails that everyone also participates and resembles more of a ‘blind’ screening situation. However, this will include the obviously unmotivated or manic patients. With an ‘open’ screening, that is, taking the clinical situation into account, the test assessment would perhaps have been delayed or at least the score would be reviewed in that light. However, to avoid arbitrariness or the influence of the examiner and thus introduce unwanted subjectivity, inclusion and exclusion criteria are used to prevent subjectivity. Due to the study design, this will then create false negatives and false positives that would normally be filtered out in daily practice. Subjectivity and clinical perspective seem to be the same here but are respectively feared (in research) or desired (in clinical practice) depending on the reason for taking the test. Our study attempted to maintain objectivity with respect to included patients, despite the exclusion criteria. However, there is room for comment on this. For example, referred patients who were already known to have moderately severe dementia were excluded. In doing so, we made it more difficult for the test because, again, the groups to be distinguished became more similar: patients with obvious dementia were excluded and (most likely) with them, MoCA scores in the lower ranges. Even though this met the STARD-D criteria (Noel-Storr et al., 2014), avoiding the extremes of the spectrum. Thus, healthy controls should also be considered an extreme of the spectrum. All the above considerations can be summarised as follows: Do not just implement or rely fully on a cut-off score provided by a researcher. This is especially true for a positive MoCA. Judge a bedside test by its total score but always include your clinical knowledge (of clinical and demographic characteristics of the individual patient) and observations, including how the MoCA score was obtained. Use the strengths of the test and know its weaknesses.
RkJQdWJsaXNoZXIy MTk4NDMw