How far do alternative capability and health-related quality of life instruments capture differing aspects of mental wellbeing for schizophrenic patients with depression?

doi:10.21203/rs.3.rs-44326/v4

Background

There is increasing evidence that assessing outcomes in terms of capability provides information beyond that of health-related quality of life (HRQoL) for evaluation of outcomes in mental health research and clinical practice.

Aims

This paper aimed to assess similarities and differences in the measurement properties of the ICECAP-A and OxCAP-MH instruments in schizophrenic patients with depression, and compare these measurement properties to those of (i) EQ-5D-5L / EQ VAS instruments and (ii) mental health specific measures.

Method

Using data for 100 patients from the UK, measurement properties were compared using correlation analyses, Bland Altman plots and exploratory factor analysis (EFA). Responsiveness was assessed by defining groups who worsened, improved or remained unchanged, based on whether a clinically meaningful change in the instrument scores between baseline and 9-month follow-up assessments.

Results

The two capability instruments had stronger convergent validity with each other (0.677) than with the HRQoL (0.354-0.431) or the mental health specific (0.481-0.718) instruments. The OxCAP-MH tended to have stronger correlations with mental health specific instruments than the ICECAP-A, while the ICECAP-A had slightly stronger correlation with the EQ VAS. Change scores of the capability instruments weakly correlated with change scores of the health-related HRQoL scales (0.131-0.269), but moderately with those of mental health specific instruments for the ICECAP-A (0.355-0.451) and moderately/strongly for the OxCAP-MH (0.437-0.557).

Conclusions

Assessing outcomes in terms of capabilities for schizophrenic patients with depression provided more relevant, mental health specific information than the EQ-5D-5L or the EQ VAS. The ICECAP-A and the OxCAP-MH demonstrated similar psychometric properties, but the OxCAP-MH was more correlated with disease-specific instruments.

Health Economics & Outcomes Research

Health Policy

capability approach

depression

EQ-5D-5L

ICECAP-A

measurement

mental health

OxCAP-MH

psychometric validation

quality of life

schizophrenia

UK

wellbeing

The capability approach is a theoretical framework with a range of potential applications. It has been highlighted as a way of facilitating people with a lived experience of mental health difficulties to engage with their values and priorities, for instance, through influencing the design and delivery of mental health services [1]. The capability approach was developed by Amartya Sen with a core focus on what individuals are free and able to do and be (i.e., capable of) [1]. This approach places emphasis on promoting wellbeing through enabling people to realise their capabilities and engage in behaviours that they value [1].

Economic evaluations are comparative analyses of alternative courses of action based on the costs and consequences of the alternatives being considered [2]. There is increasing interest in the use of the capability approach for the economic evaluation of health-related interventions including mental health [3]. One reason for this is the wider evaluative space this approach offers for the measurement of wellbeing in comparison to the commonly used methods of assessing impacts in terms of health-related quality of life (HRQoL) [4]. Currently the EQ-5D instruments (including a self-rated health assessment recorded on a vertical visual analogue scale, the EQ VAS) are the most commonly recommended and used HRQoL instruments for economic evaluations in healthcare [5]. However, HRQoL instruments such as EQ-5D may not capture important impacts of interventions when these go beyond health change and, therefore, underestimate their full welfare impact [6]. Mental health care interventions and services usually target both health and social impairments because many people with severe and enduring mental illnesses experience significant functional and social challenges which may also spill-over into effects on education, employment, justice and families [4]. There is evidence of these impacts beyond HRQoL in mental health, and particularly in the case of schizophrenia [7]. People living with schizophrenia are more likely to be homeless, unemployed, or living in poverty compared with the general population; moreover, their disease is progressive and causes increasing disability, dependence on care, and need for assistance from surrounding people in daily life activities [8]. The prevalence of depressive disorder in schizophrenia has been reported to be around 50% [9]. Evidence suggests that depression is linked to poorer outcomes in schizophrenia, as well as with particularly high levels of health care use and suicide [10].

A recent literature review of capability instruments in economic evaluations of health-related interventions identified 14 such instruments which differ substantially in their development methods, items, item levels, target populations and the interventions that were evaluated using these instruments [3]. Two of the most commonly used and validated instruments, particularly for economic evaluations among adults with mental health problems [3], are the ICECAP measure for Adults (ICECAP-A) and the Oxford CAPabilities questionnaire-Mental Health (OxCAP-MH). While both instruments are grounded in the capability approach, their conceptual approaches differ. The ICECAP-A belongs to a broader group of ICECAP capability instruments, each focusing on important capabilities according to a particular life stage [11]. It was developed in the UK, using bottom-up participatory methods as recommended by Sen [12] to generate contextual capability attributes for the whole adult population. The OxCAP-MH was developed also in the UK, originally for capability wellbeing measurement for mental health, using an alternative ‘top-down’ approach [6]. This approach is rooted in Nussbaum’s central human capabilities deemed less limited by geographical and cultural contexts [13].

Empirical comparisons between HRQoL and capability instruments have been conducted in a variety of disease areas, but there is limited information available for mental health conditions [3]. Furthermore, comparative studies of alternative capability instruments in the area of mental health are lacking including empirical studies that focus on the use of the ICECAP-A and OxCAP-MH instruments for the same patient population. So far, suggestions for choosing between these measures tended to rest only on the underlying properties of the descriptive and valuation systems [3]. Therefore, questions remain about how far different applications of the same broad theoretical concept of the capability evaluative space result in different measurement properties, and whether this influences the capability instruments’ properties when compared to HRQoL instruments for optimized research design in mental health. This study aimed to compare the measurement properties of two most commonly used capability instruments in the mental health context, the ICECAP-A and OxCAP-MH [3], and compare them to the EQ-5D-5L and the EQ VAS alongside some mental-health-specific instruments in schizophrenic patients with depression.

Data

The analysis in this paper was based on data from the PoMeT [14], which investigated the impact of Positive Memory Training on depression symptoms of schizophrenia patients (n=100) in the UK between 2014 and 2016. Patients were eligible for inclusion if they were between 18-65 years of age, had a DSM-V diagnosis of schizophrenia or schizoaffective disorder, and had at least a mild level of depression as measured by scoring 14 or more on the Beck Depression Inventory-II (BDI) [15]. Participants were identified by trial research assistants who worked in collaboration with care coordinators based within community mental health teams. Patients were assessed at four time points through the 9-month study period: baseline, 3 months, 6 months and 9 months. The assessors, who were blind to treatment allocation, administered the questionnaires to the participants. The PoMeT study was performed in accordance with the Declaration of Helsinki and was approved by Berkshire Research Ethics Committee (REC ref 13/SC/0634). All adult participants provided written informed consent to participate in this study. More details about the PoMeT trial can be found in Steel et al. [14].

Instruments

Two capability measures (ICECAP-A and OxCAP-MH), a generic HRQoL instrument (EQ-5D-5L including the EQ VAS), and four mental health specific instruments (Beck Depression Inventory (BDI), General Anxiety Disorder (GAD), Rosenberg self-esteem scale (RSES), and the Warwick-Edinburgh Mental Wellbeing Scale (WEMWBS) were used in the study. The focus of this paper is on the comparison of the capability and generic HRQoL instruments, with the mental health specific instruments used as anchors.

Capability instruments

ICECAP-A

The ICECAP-A is a brief self-reported measure for the general adult population with five items, each of which can take one of four levels ranging from full capability to no capability. The items include Stability (being able to feel settled and secure), Attachment (being able to have love, friendship and support), Autonomy (being able to be independent), Achievement (being able to achieve and progress), and Enjoyment (being able to have enjoyment and pleasure). The ICECAP-A value set used in this study was derived from the UK general population and ranges from 0 (no capability) to 1 (full capability) [16].

The instrument has shown validity (e.g. [17]), reliability (e.g. [18]), responsiveness [19] and feasibility [20] in different populations, including depression [21], and it is increasingly used in economic evaluations [22].

OxCAP-MH

The OxCAP-MH is a self-reported, 16-item instrument developed in the context of mental health outcome measurement. The 16 items cover a broad range of individual capability wellbeing aspects including: Overall health, Enjoying social and recreational activities, Losing sleep over worry, Friendship and support, Having suitable accommodation, Feeling safe, Likelihood of discrimination and assault, Freedom of personal and artistic expression, Appreciation of nature, Self-determination and Access to interesting activities or employment [4]. The OxCAP-MH items are rated on a 1–5 Likert-scale where each question provides an equal contribution to the overall score. The standardised score ranges between 0-100 with higher scores reflecting higher levels of capabilities [23]. The OxCAP-MH has shown validity [23, 24], responsiveness [23, 24] and feasibility [4] in several settings and mental health disease areas, including schizophrenia and depression (e.g. [25]).

HRQoL instruments

EQ-5D-5L including EQ VAS

The EQ-5D instruments are one of the most commonly used self-reported generic health status measures, and their validity and reliability have been reported in various health conditions and populations [26]. The more recent EQ-5D-5L version comprises five dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression with 5-level answer options and with preference-based value sets developed in multiple countries [27]. This study used the UK crosswalk mapping value set developed for the EQ-5D-5L [28], following recommendations from the National Institute for Health and Care Excellence [5]. Scores range between the worst (-0.594) and best (1.000) imaginable health state, whilst zero refers to death. As part of the EQ-5D-5L, respondents’ self-rated health is also recorded on a vertical visual analogue scale (EQ VAS), where scores range between 0-100 referring to worst imaginable health state and best imaginable health state, respectively.

Mental health specific instruments

Four mental health-specific instruments used in the PoMeT trial were applied as anchors in the correlation and responsiveness analyses. This was due to their ability to capture the clinically relevant status of the patients together with any changes in these, thereby enable the responsiveness assessment of the OxCAP-MH, ICECAP-A, EQ-5D-5L and EQ VAS instruments.

BDI

BDI is a self-reported measure of depressive symptoms and their severity in adolescents and adults according to the Diagnostic and Statistical Manual for Mental Disorder [15]. It has 21-items scored on 4-point polytomous response scale ranging from 0 to 3. Scores range between 0 and 63 with higher scores representing more severe depression [15].

GAD

GAD is a self-reported measure of anxiety symptoms over the last two weeks. It consists of seven items scored on a 0-3 scale with higher scores indicating more severe symptoms (range from 0 to 21). The cut-off scores of 5, 10 and 15 reflect mild, moderate and severe anxiety symptoms, respectively [29].

RSES

RSES is a 10-item, self-reported instrument that measures global self-worth by measuring both positive and negative feelings about the self. Items are answered using a 4-point polytomous response scale format ranging from strongly agree to strongly disagree. Items 2, 5, 6, 8, 9 are reverse scored. The scale ranges from 0-30, with 30 indicating the highest score possible [30].

WEMWBS

The self-reported WEMWBS instrument was developed in the UK to assess mental wellbeing including affective-emotional aspects, cognitive-evaluative dimensions and psychological functioning. It is a 14-item scale with 5 response categories (‘none of the time’, ‘rarely’, ‘some of the time’, ‘often’, ‘all of the time’), with a total score ranging from 14 to 70. A higher score indicates a higher level of mental wellbeing [31].

Analyses

We conducted analyses to assess the strength and statistical significance of correlations between instruments using statistical and graphical methods; exploratory factor analysis (EFA) to assess the underlying construct of the measures; assessment of responsiveness, based on Spearman’s rank correlations between change scores of instruments; and agreement analysis using Bland-Altman plots. All analyses were performed using the general population value sets for the ICECAP-A and EQ-5D-5L instruments, with the summed ranges for other instruments. Sensitivity analysis was performed using level sum scores (results are presented in Appendix 1).

For all analyses, the level of significance was set at p < 0.05, unless stated otherwise. EFA was conducted with the FACTOR software, and STATA Version 16 was used for all other analyses. Analysis was conducted on complete cases, excluding missing items at the relevant time point, unless stated otherwise.

Convergent validity

We explored convergent validity through correlations across the instruments to test the expectation that capability instruments would be more highly correlated (converge) with each other than with the HRQoL instruments. Convergent validity indicates the degree to which two measures of constructs, that are theoretically related, are in fact related. Convergence between the instruments was tested through exploring the correlation (Spearman’s rank) between their baseline scores and assessed based on Cohen's effect size classification, namely < 0.3 is small, 0.3 - < 0.5 is moderate and ≥ 0.5 is large [32]. Group comparisons of mean baseline scores were conducted using the Wilcoxon rank-sum test [33] for two-group comparisons and Kruskal-Wallis one-way ANOVA for multiple group comparisons [34].

Exploratory factor analysis

We used EFA on the ICECAP-A, OxCAP-MH and EQ-5D-5L’s baseline scores to explore the instruments’ factor structure. We conducted EFA for all instruments on the same sample to determine potential overlaps of factors. Making the factor structure explicit contributes to the validity of an instrument [35]. The factor solution was chosen according to the Kaisers criterion based on a scree plot and the eigenvalues, as described in Appendix 2, together with further details on the methods of the EFA.

Responsiveness

The responsiveness of OxCAP-MH, ICECAP-A, EQ-5D-5L and EQ VAS scales was assessed using external anchor instruments. The process started with the definition of two instruments which could be used as autonomous anchors because they identify change that is unlikely to have arisen by chance (7). The level of responsiveness was evaluated by defining groups who worsened, improved or remained unchanged, based on whether a clinically meaningful change in instrument scores between baseline and 9-month follow-up assessments was measured for individuals both by the anchor instruments. The standard error of measurement (SEM) was calculated based on the difference between the baseline and 9-month values using the following formula [23, 36]:

There is no consensus about how many SEMs an individual's score must change for that change to be considered clinically meaningful. This paper used the threshold of one SEM, which is known to frequently correspond to a minimally important difference [23, 36]. Next, the percentages of the study respondents who improved, worsened or showed no change according to the capability, HRQoL and anchor questionnaires were calculated to explore the detected congruence in changes at the individual patient level. An additional analysis of responsiveness in terms of correlation between change scores (baseline to endpoint) of the instruments can be found in Appendix 5.

Agreement analysis

The pattern and extent of agreement between ICECAP-A, OxCAP-MH, EQ-5D-5L and EQ VAS scores were plotted on Bland and Altman diagrams because they are competitor measures for economic evaluations [37]. The difference between the instruments is shown on the vertical axis against the mean of the pair on the horizontal axis. For the calculation of the values for the Bland and Altman plots, OxCAP-MH, ICECAP-A and EQ-5D-5L and EQ VAS scores were standardised to 0-1 intervals whenever necessary.

Patient characteristics

The mean age of participants was 43 years and 75% were males. About half of the participants had received higher education and 86% were unemployed at the time of data collection. Schizophrenia was the primary diagnosis in 70% of the participants, and the rest suffered from schizoaffective disorder or psychosis not otherwise specified. Overall, 83% of participants reported hospitalisation due to psychiatric reasons prior to the trial with a mean number of 4.5 admissions. The severity of participants’ depression was high for 59% of participants. There were significant differences between those with mild-moderate and those with high levels of depression in the baseline score for each instrument. Further patient characteristics are presented in Table 1.

(Insert Table 1)

Convergent validity

Mean baseline scores for all instruments and their correlations are presented in Table 2. Graphical presentation of baseline correlations is included in Appendix 3.

(insert Table 2)

Correlations between the capability and HRQoL measures (0.354-0.431) were lower than those between ICECAP-A and OxCAP-MH (0.677). The ICECAP-A was slightly more correlated with EQ VAS (0.431) than with the EQ-5D-5L (0.363), whilst the OxCAP-MH was somewhat higher correlated with the EQ-5D-5L (0.389) than with the EQ VAS (0.354). The baseline scores of both capability instruments were more highly correlated with the mental health-specific (0.481-0.718) instruments than with the generic HRQoL (0.354-0.431) instruments.

(insert Figures 1a-f)

Exploratory factor analysis

A four-factor solution was selected for EFA. It found that all items of the instruments had communalities greater than 0.35, i.e. none of the items struggled to load significantly on any factor. Factor loadings are shown in Table 3 for any factor >0.35.

(insert Table 3)

Factor one is an undoubtedly HRQoL-related factor, which consisted of the five EQ-5D-5L and some ICECAP-A and OxCAP-MH items. None of the EQ-5D-5L items loaded on factors two, three and four, thereby suggesting that these factors capture the broader capability-related concepts. Multiple items of the ICECAP-A and OxCAP-MH loaded on to these factors, however, factor four consisted of only two OxCAP-MH items (Appreciation of nature and Respect for people around), with remarkably high communalities.

Responsiveness

The GAD and WEMWBS measures were selected as suitable reference anchor instruments for the analysis of responsiveness since they had the highest correlation with the four scales under investigation in this paper. Table 4 shows the number of patients who improved, showed no change or deteriorated based on the capability, HRQoL and anchor instruments. The OxCAP-MH, ICECAP-A, EQ-5D-5L and EQ VAS measures identified different proportion of improved (48-52%, 48-52%, 29-35% and 39-52%, respectively) and deteriorated (35-53%, 27-35%, 21-33% and 21-33%, respectively) patients in agreement with the anchor instruments. Out of the four scales, the OxCAP-MH identified the largest proportion of patients who had a clinically meaningful change in their capability status over the 9 months (44-54%), and had the overall best congruency with the anchor measures, especially in terms of identifying deterioration of clinical status (53%), indicating better sensitivity to change than the rest of the measures.

(insert Table 4)

Agreement analysis

The Bland and Altman plots (Figures 1a-f) showed that the ICECAP-A and OxCAP-MH have poorer agreement with EQ-5D-5L than with each other or the EQ VAS. More specifically, there was small average discrepancy between the four instruments, however, the limits of agreement were wider and therefore more ambiguous in the comparisons with EQ-5D-5L descriptive system than in the direct comparison of OxCAP-MH, ICECAP-A and EQ VAS.

(Insert Figure 1a-f)

Summary and interpretation of the findings

This paper aimed to contribute to the utilisation of the capability approach in the context of mental health by assessing how far alternative capability instruments capture broader outcome information in schizophrenic patients with depression compared to HRQoL and in relation to each other with some mental health specific measures serving as anchors. At baseline, the two capability instruments were strongly correlated with each other, but moderately correlated with the EQ-5D-5L and EQ VAS. Both capability measures correlated with mental health specific instruments stronger than the two HRQoL measures and. proved superior to the HRQoL scales in terms of responsiveness. Together with the results of the EFA and the Bland-Altman plots, our findings are confirmative that the capability evaluative space captures aspects of what might be important for individuals with severe mental health conditions beyond the aspects that commonly used HRQoL scales are able to measure. This is also suggestive to their ability to measure important impacts of mental health interventions beyond health. In general, our research shows that capability instruments may be seen as complementary to HRQoL instruments in their measurement properties in this context.

This paper also compared the psychometric properties of the two capability instruments with each other in the context of the given severe mental health condition. The study empirically demonstrated that these two instruments show somewhat different psychometric properties when deployed on the same patient cohort despite their common theoretical framework. The ICECAP-A showed a slightly weaker anchor-based responsiveness with depression, anxiety and mental health wellbeing as measured by the GAD and WEMWBS scales than the OxCAP-MH. Furthermore, the EFA and responsiveness analyses suggested a somewhat broader evaluative space and better sensitivity to change for the OxCAP-MH in comparison to the ICECAP-A. These differences may be explained by: (i) their different approaches to development, (ii) the number of items in the measures (5 items for the ICECAP-A, 16 items for the OxCAP-MH), and (iii) the more mental health-specific nature of the OxCAP-MH compared to the fully generic nature of the ICECAP-A.

Relationship between the findings and the existing literature

This analysis found slightly weaker correlations between the capability instruments and the EQ-5D-5L than some of the previous studies conducted in the area of mental health. The OxCAP-MH was compared to the EQ-5D-3L and -5L instruments in a mixed mental health population context and found correlation coefficients between 0.45-0.66 [23, 24]. Similar correlations were observed between the ICECAP-A and the EQ-5D instruments when they were compared for opiate dependent patients. The study by Goranitis et al. found that ICECAP-A and EQ-5D-5L have similar construct validity when compared to other clinical measures [38]. The slightly different results of the current study seem confirmatory of the previously identified weaknesses of the EQ-5D-5L instrument to measure HRQoL in severely ill mental health patients [39]. Our results are also similar to the findings of a study comparing ICECAP-A and EQ-5D-5L instruments in the area of depression, which concluded that instruments designed specifically to measure depression and mental health explained a greater proportion of the variation in ICECAP-A than the EQ-5D-5L [21]. The slightly weaker correlations between the ICECAP-A and EQ-5D-5L could be explained by the specific area of depression, where evidence has shown that this disease is associated with a higher level of disability in the item of interpersonal activities [40]. Moreover, previous factor analyses comparing the ICECAP-A with the items of EQ-5D-5L and EQ-5D-3L found that these instruments measure two different constructs and therefore provide potentially different information. A recent systematic literature review found inconsistencies between the ICECAP-A and EQ-5D instruments, suggesting that the ICECAP-A is most appropriately regarded as a complement for and not a substitute to the EQ-5D-3L and EQ-5D-5L [36].

For the OxCAP-MH, a recent study found that the capability measure had strong correlations with the EQ-5D-5L, EQ VAS and mental health specific instruments [23, 24]. This indicates that the OxCAP-MH indeed measures some aspects of HRQoL and can be considered enhanced rather than fully complementary to the EQ-5D-5L. The same study deployed a two-factor EFA and found that all EQ-5D-5L items and seven OxCAP-MH items loaded on one factor, while nine remaining OxCAP-MH items loaded on a separate factor also suggestive to the enhanced nature of the instrument when compared to the EQ-5D. The results of the responsiveness analyses from the current study and those from Laszewska et al. are also in line, both concluding that the OxCAP-MH is responsive to the changes in individuals’ health states over time measured by anchor questionnaires, and it discriminates among defined patient groups with high sensitivity [24].

Strengths and limitations

The study benefited from a dataset that enabled the comparison between the two most commonly used capability instruments in the area of mental health with generic HRQoL and mental health specific scales and with each other. The scope of the paper included differing aspects of mental wellbeing and provided the opportunity for a comprehensive investigation of the capability instruments’ measurement properties in the same context and under the same conditions.

Limitations of this research included a restricted number of data points compared to the number of items which may somewhat impact the robustness of the EFA and the responsiveness analyses, the two most sensitive methods to sample size issues. The study used value sets for ICECAP-A and EQ-5D-5L but not for OxCAP-MH, which can introduce an exogenous source of variance into statistical inference. Nevertheless, the sensitivity analysis reported in Appendix 1 confirmed that this had insignificant impact on the findings on this study. Finally, the study conducted the comparison and established the psychometric properties of the instruments in a specific context, hence, the results should be generalised with caution to other disease areas.

Implications for practice/policy

To our knowledge, this is the first paper to empirically compare the two most commonly used capability instruments in the area of mental health and compare them simultaneously to HRQoL and mental health-specific scales. The findings of this study provide an in-depth analysis and understanding of how the method of operationalizing the capability approach and the items included in the instruments may eventually influence the interpretation of an economic evaluation based on these instruments.

Our findings confirmed that capability instruments capture complementary information compared to HRQoL instruments. Therefore, consideration needs to be given for the inclusion of a valid capability instrument in future research studies in the area of (severe) mental health conditions. This is particularly relevant for economic evaluations that are conducted from a societal perspective where outcomes beyond health play an important role. Currently, a context-specific choice between ICECAP-A and OxCAP-MH should be based on the trade-off between the higher responsiveness and broader evaluative space of the OxCAP -MH, against the somewhat lower participant burden and the currently available value set for ICECAP-A.

Implications for future research

Establishing the psychometric properties of an instrument is context specific; therefore, these conclusions must be strengthened in other disease areas. Furthermore, comparisons of the investigated two capability instruments with other, newer wellbeing ones such as the Achieved Capabilities Questionnaire for Community Mental Health) or wellbeing (e.g. ReQol) instruments developed for the area of mental health would further contribute to our understanding of their comparative measurement characteristics and potentially differing outcome results if measured simultaneously. Future research should explore further the impact of value-based scoring once this becomes available also for the OxCAP-MH instrument.

BDI; Beck Depression Inventory

EFA: Exploratory Factor Analysis

EQ-5D-5L: Descriptive system of the 5-level EuroQol instrument

EQ VAS: Visual Analogue Scale of the EuroQol instrument

GAD: General Anxiety Disorder

HRQoL: Health-Related Quality of Life

ICECAP-A: ICEpop CAPability measure for Adults

LOWESS: Locally Weighted Scatterplot Smoothing

OxCAP-MH: Oxford CAPabilities questionnaire-Mental Health

RSES: Rosenberg Self-Esteem Scale

WEMWBS: Warwick-Edinburgh Mental Wellbeing Scale

1. Sen, A., Commodities and capabilities. 1985, Amsterdam New York New York, N.Y., U.S.A: North-Holland Sole distributors for the U.S.A. and Canada: Elsevier Science Pub. Co. .

2. Drummond, M.F.S., M.J.; Torrance, G.W.; O'Brien, B.J.; Stoddart, G.L., Standard Methods for the economic evaluation of health care programme. Third Edition ed. 2005, Oxford: Oxford University Press.

3. Helter TM, C.J., Laszewska A, Stamm T, Simon J., Capability instruments in economic evaluations of health-related interventions – A comparative review of the literature. Quality of Life Research, 2020. 29: p. 1433–1464.

4. Simon, J.A., P.; Gray, A.; Rugkasa, J.; Yeeles, K.; Burns, T., Operationalising the capability approach for outcome measurement in mental health research. Social Science & Medicine, 2013. 98: p. 187-196.

5. NICE, Developing NICE guidelines: the manual 2014 (April 2017 update) - Incorporating economic evaluation, in [PMG20]. 2017, National Institute for Health and Clinical Excellence.

6. Francis, J.B., S., SCIE's approach to economic evaluation in social care. 2011, Social Care Institute for Excellence: London.

7. Halling Hastrup, L., et al., Does the EQ-5D measure quality of life in schizophrenia? The journal of mental health policy and economics, 2011. 14(4): p. 187-196.

8. Wander, C., Schizophrenia: opportunities to improve outcomes and reduce economic burden through managed care. Am J Manag Care, 2020. 26(3 Suppl): p. S62-s68.

9. Buckley, P.F., et al., Psychiatric comorbidities and schizophrenia. Schizophrenia bulletin, 2009. 35(2): p. 383-402.

10. Palmer, B.A., V.S. Pankratz, and J.M. Bostwick, The lifetime risk of suicide in schizophrenia: a reexamination. Archives of general psychiatry, 2005. 62(3): p. 247-253.

11. Coast, J., Assessing capability in economic evaluation: a life course approach? European Journal of Health Economics, 2019. 20(6): p. 779-784.

12. Al-Janabi, H.F., T. N.; Coast, J., Development of a self-report measure of capability wellbeing for adults: the ICECAP-A. Quality of Life Research, 2012. 21(1): p. 167-176.

13. Nussbaum, M., CAPABILITIES AS FUNDAMENTAL ENTITLEMENTS: SEN AND SOCIAL JUSTICE. Feminist Economics, 2003. 9(2-3): p. 33-59.

14. Steel, C., et al., Positive memory training for the treatment of depression in schizophrenia: A randomised controlled trial. Behaviour Research and Therapy, 2020. 135: p. 103734.

15. Beck, A.T., et al., An inventory for measuring depression. Arch Gen Psychiatry, 1961. 4: p. 561-71.

16. Flynn, T.N., et al., Scoring the Icecap-a Capability Instrument. Estimation of a UK General Population Tariff. Health Economics, 2015. 24(3): p. 258-269.

17. Al-Janabi, H., et al., An investigation of the construct validity of the ICECAP-A capability measure. Quality of life research : an international journal of quality of life aspects of treatment, care and rehabilitation, 2013. 22(7): p. 1831-1840.

18. Khan, M.A. and J. Richardson, Variation in the apparent importance of health-related problems with the instrument used to measure patient welfare. Quality of Life Research, 2018. 27(11): p. 2885-2896.

19. Keeley, T., et al., A longitudinal assessment of the responsiveness of the ICECAP-A in a randomised controlled trial of a knee pain intervention. Quality of Life Research, 2015. 24(10): p. 2319-2331.

20. Keeley, T., et al., A qualitative assessment of the content validity of the ICECAP-A and EQ-5D-5L and their appropriateness for use in health research. PLoS ONE, 2013. 8(12).

21. Mitchell, P.M., et al., Assessing the validity of the ICECAP-A capability measure for adults with depression. BMC Psychiatry, 2017. 17(1).

22. Afentou, N. and P. Kinghorn, A Systematic Review of the Feasibility and Psychometric Properties of the ICEpop CAPability Measure for Adults and Its Use So Far in Economic Evaluation. Value in Health, 2020. 23(4): p. 515-526.

23. Vergunst, F., et al., Psychometric validation of a multi-dimensional capability instrument for outcome measurement in mental health research (OxCAP-MH). Health and Quality of Life Outcomes, 2017. 15(1).

24. Łaszewska, A., Schwab, M., Leutner, E. et al. , Measuring broader wellbeing in mental health services: validity of the German language OxCAP-MH capability instrument. Qual Life Res, 2019.

25. Simon, J., et al., Cost and quality-of-life impacts of community treatment orders (CTOs) for patients with psychosis: economic evaluation of the OCTET trial. Social psychiatry and psychiatric epidemiology, 2020.

26. Payakachat, N., M.M. Ali, and J.M. Tilford, Can The EQ-5D Detect Meaningful Change? A Systematic Review. PharmacoEconomics, 2015. 33(11): p. 1137-1154.

27. Foundation, E.R. 2019; Available from: https://euroqol.org/eq-5d-instruments/eq-5d-5l-about/valuation-standard-value-sets/.

28. van Hout, B., et al., Interim scoring for the EQ-5D-5L: mapping the EQ-5D-5L to EQ-5D-3L value sets. Value Health, 2012. 15(5): p. 708-15.

29. Spitzer, R.L., et al., A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med, 2006. 166(10): p. 1092-7.

30. Rosenberg, M., Society and the adolescent self-image. 1965, Princeton, NJ: Princeton University Press.

31. Tennant, R., et al., The Warwick-Edinburgh Mental Well-being Scale (WEMWBS): development and UK validation. Health Qual Life Outcomes, 2007. 5: p. 63.

32. Cohen, J., Statistical Power Analysis for the Behavioral Sciences (2nd ed.). 1988.

33. Wilcoxon, F., Individual comparisons by ranking methods, in Breakthroughs in statistics. 1992, Springer. p. 196-202.

34. Kruskal, W.H. and W.A. Wallis, Use of Ranks in One-Criterion Variance Analysis. Journal of the American Statistical Association, 1952. 47(260): p. 583-621.

35. Thompson, B. and L.G. Daniel, Factor Analytic Evidence for the Construct Validity of Scores: A Historical Overview and Some Guidelines. Educational and Psychological Measurement, 1996. 56(2): p. 197-208.

36. Wyrwich, K.W., W.M. Tierney, and F.D. Wolinsky, Further evidence supporting an SEM-based criterion for identifying meaningful intra-individual changes in health-related quality of life. J Clin Epidemiol, 1999. 52(9): p. 861-73.

37. Martin Bland, J. and D. Altman, STATISTICAL METHODS FOR ASSESSING AGREEMENT BETWEEN TWO METHODS OF CLINICAL MEASUREMENT. The Lancet, 1986. 327(8476): p. 307-310.

38. Goranitis, I., et al., Measuring Health and Broader Well-Being Benefits in the Context of Opiate Dependence: The Psychometric Performance of the ICECAP-A and the EQ-5D-5L. Value in Health, 2016. 19(6): p. 820-828.

39. Brazier, J., Is the EQ–5D fit for purpose in mental health? British Journal of Psychiatry, 2010. 197(5): p. 348-349.

40. Zlotnick, C., et al., The relationship between quality of interpersonal relationships and major depressive disorder: findings from the National Comorbidity Survey. Journal of Affective Disorders, 2000. 59(3): p. 205-215.

Table 1: Patient characteristics and mean baseline OxCAP-MH, ICECAP-A, EQ-5D-5L and EQ VAS scores

Scale (Min-Max)		Overall data		OxCAP-MH (0-100)			ICECAP-A (0-1)			EQ-5D-5L (-0.594-1)			EQ VAS (0-100)
		n	%	n	Mean (SD)	p*	n	Mean (SD)	p*	n	Mean (SD)	p*	n	Mean (SD)	p*
Full cohort		100	100%	93	55.66 (12.90)		97	0.525 (0.211)		100	0.534 (0.299)		99	50.14 (21.19)
Gender	Male	75	75%	69	56.46 (12.25)	0.404	72	0.529 (0.224)	0.486	75	0.548 (0.274)	0.613	74	52.48 (2.39)	0.079
Gender	Female	25	25%	24	53.39 (14.65)		25	0.513 (0.168)		25	0.515 (0.313)		25	43.20 (4.40)
Higher education	Yes	49	49%	46	53.36 (1.93)	0.123	47	0.507 (0.201)	0.513	49	0.528 (0.315)	0.909	49	48.91 (3.16)	0.725
Higher education	No	51	51%	47	57.91 (1.81)		50	0.542 (0.220)		51	0.551 (0.252)		51	51.29 (2.89)
Living situation	Living with family	19	19%	19	51.48 (13.20)	0.440	19	0.488 (0.168)	0.856	19	0.609 (0.199)	0.439	19	57.89 (18.05)	0.263
	Renting a flat	12	12%	10	59.06 (16.38)		10	0.520 (0.226)		12	0.567 (0.249)		11	46.73 (20.24)
	Owning a flat	5	5%	4	51.95 (5.90)		5	0.524 (0.157)		5	0.393 (0.309)		5	43.40 (21.63)
	Other	64	64%	60	56.67 (12.41)		63	0.537 (0.226)		64	0.526 (0.307)		64	48.95 (22.03)
Employment	Employed full-time	3	3%	3	67.19 (14.32)	0.152	3	0.675 (0.122)	0.479	3	0.591 (0.227)	0.004	3	55.00 (18.03)	0.180
	Employed part-time	9	9%	8	59.18 (9.21)		8	0.580 (0.113)		9	0.589 (0.273)		8	52.13 (20.01)
	Unemployed	86	86%	80	55.22 (13.00)		84	0.516 (0.220)		86	0.549 (0.270)		86	50.60 (21.08)
	Other (Student/Retired)	2	2%	2	42.19 (8.84)		2	0.445 (0.100)		2	-0.162 (0.023)		2	15.00 (14.14)
Primary diagnosis	Schizophrenia	70	70%	66	56.72 (12.77)	0.201	68	0.521 (0.216)	0.859	70	0.545 (0.293)	0.580	70	51.19 (2.48)	0.411
Primary diagnosis	Schizoaffective or psychosis NOS**	30	30%	27	53.07 (2.52)		29	0.533 (0.201)		30	0.528 (0.263)		29	47.59 (4.15)
Depression severity	Mild/moderate	41	41%	40	63.44 (1.58)	<0.001	41	0.646 (0.152)	<0.001	41	0.645 (0.238)	0.002	41	56.56 (3.17)	0.011
Depression severity	High	59	59%	53	49.80 (1.61)		56	0.436 (0.204)		59	0.467 (0.291)		58	45.59 (2.73)
Intervention	Treatment	49	49%	46	55.10 (12.59)	0.741	49	0.534 (0.209)	0.634	49	0.571 (0.301)	0.110	49	52.37 (22.82)	0.482
Intervention	Control group	51	51%	47	56.22 (13.30)		48	0.516 (0.213)		51	0.510 (0.264)		50	47.95 (2.75)

*Wilcoxon rank-sum test for two-group comparison, Kruskal–Wallis one-way ANOVA for multiple group comparison; **NOS: Not Otherwise Specified

Table 2: Baseline scores of the relevant outcome measures used in the trial and the associated Spearman’s rank correlations

with OxCAP-MH

with ICECAP-A

with EQ-5D-5L

with EQ VAS

n

Coef.

p value*

n

Coef.

p value*

n

Coef.

p value*

n

Coef.

p value*

OxCAP-MH

ICECAP-A

92

0.677

<0.001

EQ-5D-5L

93

0.389

<0.001

97

0.363

<0.001

EQ VAS

93

0.354

0.001

97

0.431

<0.001

99

0.495

<0.001

BDI

93

-0.551

<0.001

97

-0.505

<0.001

94

-0.302

0.003

99

-0.327

0.001

GAD

93

-0.481

<0.001

97

-0.584

<0.001

94

-0.518

<0.001

99

-0.332

0.001

RSES

93

-0.619

<0.001

97

-0.558

<0.001

94

-0.260

0.011

99

-0.234

0.020

WEMWBS

93

0.718

<0.001

97

0.716

<0.001

94

0.248

0.016

99

0.377

<0.001

BDI: Beck Depression Inventory; GAD: General Anxiety Disorder; RSES: Rosenberg Self-Esteem Scale; WEMWBS: Warwick-Edinburgh Mental Well-Being Scale; Moderate correlations (0.3-0.5) in italic, Strong correlations (>=0.5) in bold; *Wilcoxon matched pairs signed rank test

Table 3: Exploratory factor analysis of the OxCAP-MH, ICECAP-A and EQ-5D-5L items with 4 factors using promin rotation (n=78)

		Factor 1	Factor 2	Factor 3	Factor 4
OxCAP-MH
	Limit daily activities	0.453
	Meet socially with friends or family			0.646
	Less sleep over worries		0.523
	Enjoy free time activities			0.584
	Suitable flat situation	0.391
	Safety in neighbourhood		0.411
	Probability of assault		0.513
	Probability of discrimination		0.635
	Local decisions			0.517
	Freedom of expression			0.384
	Appreciation of nature				0.688
	Respect for people around				0.786
	Enjoy love and support		0.381	0.562
	Freedom of deciding for yourself		0.510	0.494
	Creativity			0.498
	Access to interesting activities/employment		-0.513	0.752
ICECAP-A
	Feeling settled and secure		0.499	0.404
	Love, friendship and support			0.725
	Being independent	0.385
	Achievement and progress			0.598
	Enjoyment and pleasure			0.812
EQ-5D-5L
	Mobility	0.928
	Self-care	0.806
	Daily activities	0.754
	Pain/discomfort	0.761
	Anxiety/Depression	0.389

Loadings ≤0.35 were removed

Table 4: Number of patients improved, deteriorated or remained unchanged as defined by the investigated and anchor questionnaires (based on SEM) (n=78)

Instrument	Change*	n	GAD				WEMWBS
Instrument	Change*	n	Improved	No change	Deteriorated	n	Improved	No change	Deteriorated
	n		23	41	14		31	32	15
OxCAP-MH	Improved	24	12 (52%)	12 (29%)	0 (0%)	24	15 (48%)	9 (28%)	0 (0%)
	No change	38	10 (44%)	19 (46%)	9 (64%)	38	13 (42%)	18 (56%)	7 (47%)
	Deteriorated	16	1 (4%)	10 (24%)	5 (35%)	16	3 (10%)	5 (16%)	8 (53%)
ICECAP-A	Improved	22	12 (52%)	9 (22%)	1 (7%)	22	15 (48%)	4 (13%)	3 (20%)
	No change	44	10 (43%)	26 (63%)	8 (57%)	44	13 (42%)	23 (72%)	8 (53%)
	Deteriorated	12	1 (4%)	6 (15%)	5 (35%)	12	3 (10%)	5 (16%)	4 (27%)
EQ-5D-5L descriptive system	Improved	18	8 (35%)	8 (20%)	2 (14%)	17	9 (29%)	6 (19%)	2 (13%)
	No change	47	14 (61%)	24 (58%)	9 (64%)	48	17 (55%)	23 (72%)	8 (53%)
	Deteriorated	13	1 (4%)	9 (22%)	3 (21%)	13	5 (16%)	3 (9%)	5 (33%)
EQ-5D VAS	Improved	22	12 (52%)	6 (15%)	4 (29%)	17	12 (39%)	4 (13%)	1 (7%)
	No change	46	11 (48%)	28 (68%)	7 (50%)	46	17 (55%)	20 (62%)	9 (60%)
	Deteriorated	10	0 (0%)	7 (17%)	3 (21%)	15	2 (6%)	8 (25%)	5 (33%)

*Changes in instrument scores between baseline and 9-month follow-up were categorised as improved, worsened or unchanged, definition of groups is based on the difference in SEM; numbers were rounded and do not always add up to 100%; values in agreement are in bold

Ethics approval and consent to participate: The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. All procedures involving human subjects/patients in the PoMeT trial were approved by the Berkshire Research Ethics Committee (REC ref 13/SC/0634). All participants provided written consent.

Declaration of interest: JC has led the development of the ICECAP measures. JS has led the development of the OxCAP-MH measure. The remaining authors declare that they have no conflict of interest.

Funding: The PoMeT trial was funded by the National Institute for Health Research (NIHR) under its Research for Patient Benefit (RfPB) Programme (Grant Reference Number PB-PG-0712-28021). Joanna Coast is supported by the Wellcome Trust [205384/Z/16/Z].

Acknowledgements: We would like to say ‘thank you’ to all contributors and participants of the PoMeT trial. We would also like to thank Georg Heinze for his comments on the statistical analyses used in this paper and David Mott who discussed a previous version of this paper at Winter 2020 meeting of the Health Economics Study Group in Newcastle, UK.

Author’s contributions: TH and JS conceived the research idea and developed the conceptual framework of this research. JS provided resources for the study. TH conducted the analysis supervised by JS and with input on different aspects of the study from JC, AL and TS. TH took the lead in writing the manuscript in close consultation with JC, AL, TS and JS. All authors provided critical feedback and helped shape the research, analysis and manuscript. All authors approved the final manuscript.

Availability of data and materials: The data that support the findings of this study are available on request from the corresponding author.

Supplementarymaterial.pdf

How far do alternative capability and health-related quality of life instruments capture differing aspects of mental wellbeing for schizophrenic patients with depression?

Status:

Version 4

Abstract

Figures

Introduction

Methods

Data

Instruments

Analyses

Convergent validity

Exploratory factor analysis

Responsiveness

Agreement analysis

Results

Patient characteristics

Convergent validity

Exploratory factor analysis

Responsiveness

Agreement analysis

Discussion

Summary and interpretation of the findings

Relationship between the findings and the existing literature

Strengths and limitations

Implications for practice/policy

Implications for future research

List Of Abbreviations

References

Tables

Declarations

Supplementary Files

Status:

Version 4