Background
Mental and substance use disorders contribute substantially to the global burden of disease, being, directly and indirectly, responsible for a significant amount of mortality and morbidity [
1]; to a large extent through delayed access to treatment and services [
2]. Furthermore, the impact of mental health on disability and incapacity is on the rise [
3‐
5]. Mental disorders are a huge challenge for public health, compounded by the challenges of psychiatric diagnosis itself [
1]. In contrast to most other fields of medicine, psychiatric diagnosis remains primarily restricted to subjective symptoms and observable signs [
6]; furthermore, there is a poor correspondence between impairment and symptom load [
7‐
10]. This has led to ongoing controversy about the usefulness of psychiatric diagnoses, as well as their reliability and validity [
1,
10,
11].
Accordingly, the WHO (World Health Organisation) introduced the ICF (International Classification of Functionality), designed to assess the particular mix of strengths, weaknesses, and other circumstances which may affect the course of a patient’s illness and treatment response, independent of actual symptom levels [
12]. However, the ICF is far too complex and time-consuming to be implemented in everyday clinical practice [
7,
13]. The mini-ICF-APP was developed in accordance to the WHO-ICF parameters; it is a 13-item short observer-rating instrument to assess functioning and impairment in persons suffering from mental health problems, irrespective of diagnosis [
13].
The original version, as well as the first studies involving the mini-ICF-APP were validated on patients with a non-psychotic mental disorder hospitalized for rehabilitation treatment [
13‐
15]. The subsequent studies of the English and Italian version of the scale covered a broader diagnostic spectrum, including patients with a psychotic disorder [
16‐
18]. Recent studies focused on patients with a chronic mental disorder [
19] as well as with posttraumatic stress disorder [
20]. Since its introduction, the mini-ICF-APP has been progressively implemented by health care providers, insurance companies and pension funds to assess disability and work impairment [
16‐
18,
21,
22].
The evaluation of functioning and disability has vast implications for those affected since it may impact access to services and financial support (e.g., disability pension). Despite a number of validation studies [
13,
16‐
18], there is limited support for the clinical usefulness and interpretation of the mini-ICF-APP. Although higher scores imply greater severity, its clinical relevance is not fully understood. The Clinical Global Impression Scales (CGI), in contrast, describes the overall clinical impression regarding severity or improvement [
23]. Therefore, the CGI has been previously used to establish the usefulness of other clinical measures [
24‐
26]. This study aims to provide clinically meaningful and valid scores for the mini ICF which correspond statistically to CGI levels in a sample of patients hospitalized for treatment.
Discussion
This approach of equipercentile linking allows the definition of severity cut-off values for the mini-ICF-APP according to clinical judgment. Taken together this provides a better understanding and interpretation of the mini-ICF-APP in day-to-day clinical practice. This is of singular importance since the mini-ICF-APP is increasingly being used by health insurance companies and pension funds to determine disability or incapacity benefits caused by a psychiatric condition [
21]. Our large clinical sample included diagnoses underrepresented in prior studies with the inclusion of: alcohol and substance use disorders; personality disorders; neurocognitive and neurodevelopmental disorders.
The results of the multivariate analysis did not show a statistical correlation between the mini-ICF-APP and CGI scales and demographic and clinical variables (age, gender, education and main diagnosis). Between the mini-ICF and CGI we were able to show a strong statistical correlation. The correlations found show that illness severity at large is associated with capacity limitations in general. The pairwise comparison by of the scales according to the main diagnosis, produce similar associations between both scales for practically all diagnostic groups. In our analysis age, gender, education and main diagnosis seemingly have no effects on the anchor points between the mini-ICF-APP and CGI scales. Furthermore, change and improvement were independent of the baseline severity. Therefore, we consider that the mini-ICF-APP can be viewed as a universally applicable scale for a diagnosis-independent judgment of functionality; broadening the applicability of the scale.
The mini-ICF-APP severity cut-off points we propose for the mini-ICF-APP are roughly comparable with the values reported in previous studies [
16‐
18,
21,
36]. A direct contrast is however not possible, since the approach to severity assessment differs; with either different gradings of severity [
36]; gross allocation of severity grades to mini-ICF-APP values [
16,
18]; or the use of psychosocial attributes of severity [
17,
21]. Our findings regarding the percentage change are in line with previous findings using the mini-ICF-APP to assess the effectiveness of treatment [
14]. Furthermore, our results also concord with the general (although controversially discussed) parameters used to judge improvement [
37,
38], were a reduction of 20–30% is considered an indication of response to treatment while a reduction greater than 50% is seen as a significant improvement [
39,
40]. The difficulties comparing severity rankings and improvement are not surprising since most available studies dealt with validation issues [
13,
16‐
18].
The severity of a disorder has waste clinical implications, it influences decisions regarding the type and intensity of treatment, whether to continue or stop treatment, but also affect the assess to assistance or disability benefits [
41]. The assessment of severity in psychiatric disorders is, however, an issue of ongoing debate, with no resolution insight [
41,
42]. The CGI is considered a pragmatic and intuitive scale [
23]; that is commonly used to approach severity and change [
41,
43‐
45]. The CGI is considered useful and necessary in routine clinical practice as well as research settings [
32]; as well as previously used for the determination of severity cut-off values [
24,
26,
46,
47].
Equipercentile linking, allows for a nominal translation from one scale to another, identifying those scores on both scales which have the same percentile ranks [
35]. It’s is used in testing programs that involve multiple test forms, like in the educational system [
34,
35]. Equipercentyle analysis does not require a specific distribution type and allows for possible measurement errors on both scales compared [
35], for this reason, it is considered the preferred linking method amongst other available methods [
47,
48]. With an increasingly large number of published studies using this method comparing scales in the fields of psychiatry and neurology [
49‐
52]; with the CGI widely used as a determinant of severity ranking [
24‐
26,
46,
47,
53,
54].
The main strength of our study is the large sample population, without inclusion or exclusion criteria, with a broad diagnostic range, including diagnoses underrepresented in previous studies. The similar correlation coefficients and linking values for the different diagnostic categories confirms the use of the mini-ICF-APP independent of diagnosis [
13,
16,
17]. The inclusion of one study site determines some peculiarities in the sample studied. In our case, the large number of patients with AUD is, due to a detoxification treatment program, with a high admission capacity. The over-representation of male patients might be a potential limitation; however, neither the regression model nor the post hoc analysis revealed an effect of gender on the results. One limitation of our study is the recruitment of the sample from patients hospitalized for treatment, which limits the interpretation of the data to these patients requiring treatment. In this context also the influence of the different health systems on treatment seeking behaviour and utilization of mental health services should considered [
55].
Our results should be interpreted with caution; since the mini-ICF-APP sum score does not reflect a global incapacity index. Patients with limitations in certain dimensions still could experience impairment; due to the repercussion in other domains. Furthermore, it also should consider how the individual human beings deal with incapacity and seek for solutions on their own. We consider that the CGI scales take account all of these phenomena; with the resulting rather exponential correlation between both scales.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.