Carolien Zeetsen

133 CHAPTER Summary and general discussion 7 Self–reported and clinician–observed everyday cognitive functioning was also measured in Chapter 5. It was found that patient ratings were higher than clinician ratings, and that patient ratings did not differ between groups, where the clinician ratings did. This finding is in line with the literature in which patients do not always report subjective complaints because of a lack of insight into their own cognitive deficits (Walvoort et al., 2016). On the contrary, patients reported a significant improvement in everyday cognitive functioning, where the clinicians do not report significant changes. This may be influenced by the fact that patients were probably in a better emotional state when completing the final assessment, as they were abstinent and (soon to be) clinically discharged. The changes in cognitive performance were related to changes in everyday cognitive functioning, albeit with only small to medium effect sizes. This supports the literature that cognitive performance on objective measures does not predict cognitive impairments in the absence of subjective experiences of everyday cognitive problems (Horner et al., 1999). Strengths and limitations Each study had its own strengths and limitations, as described in detail at the end of the chapters. There are also some general remarks to be made taking all studies together. This research project found its origin in clinical practice, where several esteemed colleagues put their hands and minds together to formulate a plan to close a gap. This collaboration between addiction health care centres resulted in a multicentre project where large groups of patients could be included. There is an important note when interpreting our findings, specifically those regarding the prevalence of substance–induced NCD. As stated in the introduction, the number of patients entering addiction health care are likely only the tip of the ‘users’–iceberg. Epidemiologically, it is unknown what substance users that do not enter treatment use and if they have cognitive impairments or not. Also, as the predictive and concurrent validity of the MoCA were found to be low to moderate, the actual prevalence of substance–induced NCD may well be different than that currently found. However, as a result of well characterising our groups of patients in terms of which drugs they used, how much of it and for how long, they are representative for the population of substance (ab)users, making results highly generalisable. It is always the question (as also stated above regarding cognitive complaints) to what extent substance use can be objectively measured via self–report. We aimed to overcome this question by using a validated and evidence–based interview for relevant patient characteristics in addiction health care: the MATE (Schippers et al., 2011).