Background and purpose The Stroke Specific Quality of Life Scale (SS-QoL) is a well validated measure of health related quality of life in patients with stroke, but with 49 items its length is a disadvantage. A short version of the SS-QoL was developed and tested here.
Methods Secondary analyses of three different studies. The short version was developed using data from 141 patients with aneurysmal subarachnoid haemorrhage (SAH) and tested on data from independent samples of 97 patients with SAH and 105 patients with ischaemic stroke or intracerebral haemorrhage. The item with the highest item domain correlation from each of the SS-QoL domains was selected to obtain a 12 item SS-QoL (SS-QoL-12) with a total score and physical and psychosocial subscores. Criterion validity of the SS-QoL-12 scores was tested in each sample with the original SS-QoL as reference.
Results All three scores of the SS-QoL-12 showed good internal consistency (Cronbach's alpha 0.77–0.89). The SS-Qol-12 scores predicted 88–95% of the variance of the original SS-QoL. Mean differences between the SS-QoL-12 and SS-QoL and their 95% CI were generally within 0.1 points on a 1–5 scale. The limits of agreement were generally within 0.4 points.
Conclusion The SS-QoL-12 has good criterion validity for all subsets of stroke. Because it consists of only 12 questions, this short form will be easy to use in research and clinical settings.
- QUALITY OF LIFE
- SUBARACHNOID HAEMORRHAGE
Statistics from Altmetric.com
Patients who survive a stroke often experience a substantial decrease in their health related quality of life (HRQoL).1 The Stroke Specific Quality of Life Scale (SS-QoL)2 3 is a well known, standardised, disease specific measure to assess HRQoL after stroke that has been validated in patients with different types of stroke.2 4–6 The SS-QoL consists of 49 items in 12 domains and takes about 15 min to complete. This length of the SS-QoL is a disadvantage, as patients with stroke often experience attention and concentration problems7 and measurement of HRQoL is usually only one part of a larger measurement battery. A further drawback of the SS-QoL is that previous studies did not confirm its proposed structure of 12 domains.4 8 This number of domains is less practical for research into correlates of HRQoL, leaving a choice to use all 12 domain scores as dependent variables or to use only the total SS-QoL score with a risk of concealing differences between HRQoL domains. In a previous study in patients with aneurysmal subarachnoid haemorrhage (SAH), we concluded that the 12 domains of the SS-QoL could be merged into two subtotal scores representing the dimensions of ‘physical’ and ‘psychosocial’ HRQoL.9 Both subtotal scores and the total score showed very high internal consistency (Cronbach's alpha 0.95–0.97). A 0.73 correlation between the physical and psychosocial subtotal scores however showed that one subtotal score explained only about half of the variance of the other.9 Using the two SS-QoL subtotal scores for physical and psychosocial HRQoL might therefore be a good compromise between simplicity and a need to provide a profile of different aspects of health. The very high internal consistency of these physical and psychosocial subtotal scores suggested that these scores can be accurately reproduced with fewer items, and therefore that it could be possible to develop a short version of the SS-QoL that reveals one total score and two subtotal scores that are equivalent to those of the original 49 item version. Such a short form would be more practical to use in a clinical setting and as an outcome measure in clinical studies.
The aims of this study were: (1) to develop a short form of the SS-QoL and (2) to test this short version in independent samples of patients with ischaemic stroke, intracerebral haemorrhage and SAH.
Materials and methods
We used data from three previous studies. The largest available sample was used to develop the short SS-QoL. This sample consisted of all patients who had been treated by clipping or coiling after SAH between January 2003 and July 2005 in the University Medical Centre Utrecht (UMCU). SAH was diagnosed by CT and aneurysms by CT angiography or conventional angiography.9 10Patients living in a nursing home and patients with severe comorbidity or insufficient command of the Dutch language were excluded from the study. All eligible patients received a mailed questionnaire 2–4 years post-SAH.
One validation sample consisted of another group of patients with SAH, also treated at the UMCU, who had been discharged home and who had visited the SAH outpatient clinic of the UMCU between September 2006 and September 2008. They completed a mailed questionnaire including the SS-QoL 1 year post-SAH (unpublished data).
The other validation sample consisted of patients with first ever ischaemic stroke or intracerebral haemorrhage (IS-ICH) admitted to the stroke units of three hospitals in The Netherlands.11 Inclusion criteria were: age <85 years, no comorbidity that might affect outcome and testable within the first 21 days after stroke. Stroke was diagnosed based on the presence of both an acute focal deficit and an associated lesion on CT or MRI scans. Exclusion criteria were: (1) pre-existing drug abuse/depression/activities of daily life dependence or cognitive impairment, (2) disturbed consciousness or inability to comprehend task instructions, (3) recurrent stroke and (4) comorbidity that might affect outcome. These patients completed the SS-QoL 6–11 months post-stroke.
The medical ethics committee of the UMCU approved all study protocols and all patients gave written informed consent.
The SS-QoL consists of 49 items encompassing 12 domains: self-care, mobility, upper extremity function, language, vision, work, thinking, family roles, social roles, personality, mood and energy. Each item is ranked on a 5 point scale, with higher scores indicating better function. Domain scores are the unweighted averages of the items scores and the total score is the unweighted average of the domain scores. All summary scores thereby also range from 1 up to 5.3 In an earlier study, we showed that the 12 domain scores can be merged in two subtotal scores, the first six domains in a physical subtotal score and the last six domains in a psychosocial subtotal score.9
Development of the short SS-QoL
To ensure that the short form represents the full scope of the original SS-QoL, one item was selected from each of the 12 domains of the SS-QoL. These domains showed high internal consistency in earlier studies, indicating conceptual homogeneity within each domain.2 4 9 The item with the highest item total correlation, thereby being most representative for the domain score, was selected. These 12 items were grouped into physical and psychosocial dimensions according to the merging of the 12 domains of the SS-QoL into a physical and a psychosocial dimension, as described above. The dimension scores and total score of the SS-QoL-12 are the unweighted averages of the item scores and range from 1 up to 5. A score of, for example, 3.5 on the short version thereby should reflect the same level of HRQoL as a score of 3.5 on the original version.
SS-QoL-12 subtotal and total scores were computed. Cronbach's alpha was used to assess internal consistency. Internal consistency requires a Cronbach alpha coefficient of at least 0.70.12 Criterion validity was examined using the 49 item SS-QoL as a reference. The percentages of variance of the SS-QoL scores that could be explained by the 12 item version scores were computed. Bland–Altman plots were used for visual inspection.13 Agreement between the 12 item and 49 item versions was examined at group level by computing the mean difference and its 95% confidence interval, and at the individual level by computing the limits of agreement (±1.96×SDdifference).13 To substantiate these figures, they were expressed as effect sizes by dividing them by the SD of the corresponding 49 item SS-QoL scores. The conventional interpretation of effect sizes is: 0.2 is small, 0.5 is medium and 0.8 is large.14
The development sample consisted of 141 patients with SAH (response 81%). The SAH validation sample consisted of 97 patients (response 82%). The stroke validation sample consisted of 105 patients (response 71.7%). Respondent characteristics are displayed in table 1.
Development of the SS-QoL-12
In the development sample, the item domain correlations of the selected items were very high (0.85–0.95) (table 2). In two domains, there were two items with equal highest item domain correlations and we arbitrarily chose one of these two items. The selected items are also displayed in table 2.
Criterion validity of the SS-QoL-12
In all three cohorts, the subtotal and total scores of the SS-QoL-12 showed good internal consistency and SS-QoL-12 scores explained high percentages of variance of the long version (table 3).
The mean differences between scores on the short and long versions were negligible. The percentages of explained variance were very high in all samples, with the lowest percentage (91%) in the IS-ICH sample. A sensitivity analysis showed that an SS-QoL-12 version with the two other items would have resulted in nearly identical Cronbach's alpha and explained variance figures (maximum 0.01 point or 1% lower). The limits of agreement were also widest in the IS-ICH sample (−0.46 to 0.39). The effect sizes of the individual differences between the 12 item and 49 item total scores in the IS-ICH sample were however very small for most patients: between 0 and 0.2 for 53.4% of the sample, between 0.2 and 0.5 for 35.2% of the sample and between 0.5 and 0.8 for the remaining 11.4% of the stroke sample.
All nine Bland–Altman plots showed a similar relationship between the mean of the 12 item and 49 item scores (x axis) and the difference between these scores (y axis). The plot of the total scores in the stroke sample is displayed in figure 1. Figure 1 shows that in patients with low SS-QoL scores, scores on the 12 item version were slightly lower than scores on the 49 item version. For the 10 stroke patients with the worst SS-QoL scores (SS-QoL 49 <3.25), their mean SS-QoL-12 total score was only 0.16 (SD 0.21) points lower than their score on the original SS-QoL (effect size=0.25).
We developed a short version of the SS-QoL, the SS-QoL-12, which appears a valid summary of the original 49 item SS-QoL for patients with ischaemic stroke, intracerebral haemorrhage and SAH. This short form questionnaire differs from the long form by as little as 0.5 points on a 5 point scale. The major advantage of this short version is that it minimises administration time with 37 questions or an estimated 10 min. This will make the shorter form easier to administer and thereby a more practical tool in research and clinical practice
Our results appear robust as the figures from both validation samples were similar to each other and to the development sample. A few limitations however apply to this study. Firstly, the SS-QoL-12 short form was developed in an SAH population. If we had used a sample of patients with ischaemic or haemorrhagic stroke to develop the SS-QoL-12, some different items would have been selected. However, all item domain correlations were high, as expected, because this was the way the SS-QoL was developed, and differed only slightly between items, and the results of the stroke validation sample were similar to those of the SAH validation sample. Secondly, we used existing SS-QoL data to validate the short version by selecting the SS-QoL-12 items from the database. In theory, actual answers on the SS-QoL-12 may deviate from answers retrieved for these items from the long SS-QoL, as a patient's answers on these questions might also be shaped by the other questions. Our approach is however common,15 as it is less feasible to include short and long versions of the same measure in the same questionnaire, and even then bias by this kind of shaping cannot be excluded. Finally, since we used the Dutch SS-QoL, replication of this study using other language versions is recommended.
In conclusion, pending validation in prospective studies in other countries, we feel the SS-QoL-12 can replace the original SS-QoL in clinical and research settings if only total or subtotal scores are required. If the study goal is to report all 12 individual domain scores, the use of the original SS-QoL is recommended, as these domain scores would be based on only one item for each domain using the SS-QoL-12.
Competing interests None.
Ethics approval This study was conducted with the approval of the medical ethics committee of the University Medical Centre Utrecht.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.