Short communication
Validation of an abbreviated quality of life scale for schizophrenia

https://doi.org/10.1016/j.euroneuro.2012.11.009Get rights and content

Abstract

The field of therapeutics in schizophrenia is redefining optimal outcome, moving beyond clinical remission to a more comprehensive model that also includes functional recovery. The Quality of Life Scale (QLS) has been adopted by many large clinical trials, including CATIE and CUtLASS, as a measure of functioning. The QLS is a 21-item semi-structured interview that takes approximately 45 min to administer. Although the QLS is considered comprehensive, its length limits its applicability across studies. To circumvent this issue, short scales of the QLS have been created that estimate total scores with high accuracy. However, these abbreviated measures have not been adequately cross-validated in a large enough sample to allow for subsample estimations nor has its predictive ability been compared to the full scale. Here, we used data from the CATIE trial (n=1460) to demonstrate the validity and utility of an abbreviated 7-item QLS. The shortened QLS was robust in estimating total scores (r=0.953, p<0.001) across subsamples and demonstrated predictive ability similar to the full QLS in multiple regression models. The abridged QLS is recommended as a surrogate measure of psychosocial functioning, especially in cases where functioning is not the primary outcome.

Introduction

The field of therapeutics in schizophrenia is redefining optimal outcome, moving beyond clinical remission to a more comprehensive model that also includes functional recovery (Remington et al., 2010). There are now clinical trials that have adopted a measure of functioning as their primary outcome (Grant et al., 2012), and in recent years scales such as the Global Assessment of Functioning (American Psychiatric Association, 2000) have been embraced as part of routine clinical assessment (Tungstrom et al., 2005). However, recent consensus meetings have suggested that more comprehensive instruments may be better suited for assessing “real-world” functioning, partially due to their superior psychometric properties (Leifker et al., 2011).

Two of the recent major antipsychotic efficacy trials, CATIE (Lieberman et al., 2005) and CUtLASS (Jones et al., 2006), both adopted the Heinrichs–Carpenter Quality of Life Scale (QLS) (Heinrichs et al., 1984) as a measure of functioning. The QLS is a 21-item semi-structured clinical interview that was originally designed to assess deficit symptoms in schizophrenia. Although the QLS is a comprehensive measure, the extensive time required to administer the scale (approximately 45 min) limits its applicability across studies and in clinical practice. To circumvent this issue, shortened versions of the QLS have been developed that take a fraction of the time to administer yet predict total scores with high accuracy (Bilker et al., 2003, Ritsner et al., 2005). Employing such abridged scales may be quite fruitful, especially in cases where functioning is not the primary outcome, and several studies have already adopted these shortened scales for this purpose (Fisher et al., 2009, Grant and Beck, 2009).

Having an abridged QLS is appealing, although the shortened version should predict total scores with high accuracy and this prediction should be adequately validated. This is crucial if one is to retain the superior psychometric properties of the original scale, else the shortened scale bares questionable resemblance to the standard version. The 7-item QLS proposed by Bilker et al. (2003) demonstrated good predictive values in their model construction sample as well as in their re-test sample (r>0.97); however their new sample was quite limited (n=37) and restrictive (first-episode patients). Hence, we sought to extend the validation of this shortened QLS using the large heterogeneous sample from the CATIE trial. Other analyses of the CATIE data have affirmed that QLS scores are linked to symptoms and that antipsychotic treatment has a modest effect on functional changes in schizophrenia (Mohamed et al., 2008, Rabinowitz et al., 2012, Swartz et al., 2007). Here, we extend these findings using the shortened QLS, and suggest that statistical models predicting variance in QLS scores are nearly identical regardless of whether total scores or estimated total scores are used.

Section snippets

Participants

Data were drawn from the CATIE trial for chronic schizophrenia (n=1460). Details of the study design and rationale (Stroup et al., 2003), as well as primary findings (Lieberman et al., 2005), have been presented elsewhere. The primary purpose of the CATIE study was to compare the effectiveness of atypical and conventional antipsychotics through a randomized controlled clinical trial conducted at 57 sites in the United States. Diagnoses were confirmed using the Structured Clinical Interview for

Results

Patient demographics are presented in Table 1. Bivariate correlations between the QLS measures are presented in Table 2. Both abridged versions of the QLS estimated the total score with high accuracy and were indistinguishable in their estimates (Cohen's q=0.011, 95% CI: −0.062, 0.084). Further, QLS-AS and QLS-AR were highly correlated, with little variance not shared between the two (Table 2); therefore, only the QLS-AS was used for further analyses (Figure 1).

Next, analyses were run to

Discussion

The present study sought to extend the validation of an abridged QLS using a large dataset that allows for adequate stratification. The QLS-AR and QLS-AS both predicted QLS-Total scores with high accuracy, and to a non-differential degree. Due to the use of regression weights, rather than simple summation, in computing the QLS-AR and the lack of added value over the QLS-AS, the QLS-AS was used in all subsequent analyses and we recommend its use in future studies. The QLS-AS predicted total

Role of funding source

None.

Contributors

GF and GR conceived the idea and wrote the protocol. GF performed the statistical analyses and wrote the first draft of the manuscript. GF and GR contributed to editing and have approved the final manuscript.

Conflict of interest

In the last 3 years, GR has received research support (Principal Investigator) from the following external funding agencies: Canadian Institutes of Health Research, Research Hospital Fund—Canada Foundation for Innovation, Schizophrenia Society of Ontario, and the Canadian Diabetes Association. GR has also received support from Novartis Canada, Medicure Inc., and Neurocrine Bioscience. GR has received consultant fees from Laboratorios Farmacéuticos ROVI, Novartis, and Roche, as well as speaker's

Acknowledgements

Data used in the preparation of this article were obtained from the limited access datasets (Version 1) distributed from the NIH-supported “Clinical Antipsychotic Trials of Intervention Effectiveness in Schizophrenia” (CATIE-Sz). This is a multisite, clinical trial of persons with schizophrenia comparing the effectiveness of randomly assigned medication treatment. The study was supported by NIMH Contract #N01MH90001 to the University of North Carolina at Chapel Hill. The ClinicalTrials.gov

References (25)

  • M. Fisher et al.

    Using neuroplasticity-based auditory training to improve verbal memory in schizophrenia

    Am. J. Psychiatry

    (2009)
  • P.M. Grant et al.

    Defeatist beliefs as a mediator of cognitive impairment, negative symptoms, and functioning in schizophrenia

    Schizophr. Bull.

    (2009)
  • Cited by (17)

    • Abbreviated quality of life scales for schizophrenia: Comparison and utility of two brief community functioning measures

      2014, Schizophrenia Research
      Citation Excerpt :

      First, although global measures of functioning are commonly employed, important information may be lost when not examining specific functional domains (Harvey et al., 2012). Subscale scores can however be estimated using single items (Fervaha and Remington, 2013). Second, the correlations between the abbreviated QLS and the full QLS may be inflated due to part-whole relationships (Kaufman, 1977).

    • Impact of primary negative symptoms on functional outcomes in schizophrenia

      2014, European Psychiatry
      Citation Excerpt :

      Moreover, examination of the standardized regression coefficients (i.e., beta weights) in the multiple regression model (Tables 2 and 3), suggests that negative symptoms had the greatest relative effect on each facet of functioning evaluated. This finding is consistent with previous work demonstrating that negative symptoms continue to have explanatory power in predicting functional outcome, even after a host of other variables have been accounted for [16,17,21,39,47,52,54], and highlights the central role of negative symptoms in the prediction of functional outcome [20,25]. One additional strength of the present study that should be mentioned was the examination of multiple domains of functioning.

    View all citing articles on Scopus
    View full text