| Sign In to gain access to subscriptions and/or personal tools. |
DOI: 10.1016/J.JHSB.2006.11.002
The SF-6D Health Utility Index in Carpal Tunnel SyndromeFrom the Department of Orthopedics, Hässleholm and Kristianstad Hospitals, Hässleholm, Sweden, the Department of Physiotherapy, Lund University, Lund, Sweden, and the Department of Bioinformatics and Biostatistics, School of Public Health and Information Sciences, University of Louisville, Louisville, Kentucky, USA Correspondence: Dr Isam Atroshi, Associate professor, Department of Orthopedics, Hässleholm and Kristianstad Hospitals, SE 28125 Hässleholm, Sweden. Tel.: +46 44 3091260; fax: +46 44 3091264. E-mail:Isam.Atroshi{at}skane.se
Cost effectiveness is an important factor to consider when choosing between various hand surgical interventions. Health utility measures can be used to determine cost effectiveness. The SF-6D is a health utility index derived from 11 items of the SF-36 quality of life questionnaire; values range from 0.296 to 1.0 ("perfect" health). We evaluated the validity of the SF-6D in patients with carpal tunnel syndrome (CTS) who completed the SF-36 and the CTS symptom severity and functional status questionnaire before and 3 months after carpal tunnel release. Complete responses to the SF-6D items were available for 100 patients at baseline and 95 patients at baseline and follow-up. The mean SF-6D health utility index was 0.69 (SD 0.13) before surgery and 0.77 (SD 0.13) after surgery (moderate effect size). The SF-6D could discriminate between patient groups differing in self-rated global health and in whether, or not, they had a minimal clinically important improvement in CTS symptom severity after surgery. The SF-6D appears to be a valid measure of health utilities in patients with CTS and can be used in cost effectiveness studies.
Key Words: carpal tunnel syndrome • utility measures • SF-6D • quality of life The assessment of patients with carpal tunnel syndrome (CTS) commonly includes patient-reported outcome and health-related quality of life measures (Amadio et al., 1996; Hobby et al., 2005). Although valuable in assessing treatment effectiveness, these measures are not designed to be used for economic evaluation, such as comparing cost effectiveness of different treatments, because they are not based on people’s health preferences (Brazier et al., 2002). A large number of surgical and non-surgical methods have been proposed as being effective in CTS. Many of these treatments may not be costly but, because CTS is such a common condition, even small differences in cost effectiveness may be considered important when making decisions regarding health care policy (Atroshi et al., 2006). Health utility measures are gaining increasing importance in cost effectiveness analyses (Maniadakis and Gray, 2000). A health utility index can be used to estimate quality-adjusted life years gained with a specific treatment which, in combination with the costs of the treatment, provide a measure of cost effectiveness. Conventional (direct) utility measures involve rather complex techniques such as the "standard gamble" and "time tradeoff" methods, which generate preference values for different health states (Gold et al., 1996; Lenert and Kaplan, 2000). Multi-attribute (indirect) utility measures are multi-item health-related quality of life measures for which valuations of different health states (i.e., combinations of item responses) have been determined from general population samples, using one of the direct utility measurement methods (Lenert and Kaplan, 2000; Petrou and Hockley, 2005). Thus, they reflect the preferences of the general population for different health states. They can then be used to measure utilities for patient groups based on their responses to the items and the predetermined valuations of health states. The SF-6D health utility index has been introduced as a preference-based multi-attribute health utility measure derived from 11 items of the SF-36 quality of life questionnaire (Brazier et al., 1998). Because cost effectiveness has become an increasingly important factor when choosing between different treatments, the SF-6D may be useful in CTS and other upper extremity conditions. The purpose of this study was to assess the SF-6D health utility index with regards to validity and sensitivity to change in patients with CTS and determine health utilities before and after surgical treatment.
Patients The patients included in this study were among patients with CTS who participated in a large prospective cohort study of outcomes of carpal tunnel release at our hospital. The diagnosis of CTS was established by the examining orthopaedic surgeon, based on characteristic symptoms, viz. recurring numbness and/or tingling, with or without pain, in at least two of the four radial fingers. Presence of sensory loss in the median nerve distribution in the hand, thenar muscle weakness, or atrophy, and a positive Tinel sign, or Phalen, test were considered supportive of the diagnosis. Nerve conduction testing was performed when considered necessary to support the clinical diagnosis. All patients had failed treatment with a wrist splint. The patients were treated surgically with two-portal endoscopic release. The inclusion criterion for this study was that the patients had completed the SF-36 before and after surgery for CTS. The study was conducted in agreement with the local ethical guidelines for clinical studies and informed consent was obtained from the participants.
Outcome measures The SF-36 consists of eight scales; physical functioning (ten items), bodily pain (two items), role limitations because of physical health problems (four items), vitality (four items), general health perceptions (five items), social functioning (two items), role limitations because of emotional health problems (three items), mental health (five items) and one independent item on health transition (Ware and Sherbourne, 1992). Each scale is scored from 0 (worst) to 100 (best).
Utility measures
Analyses Validity of the SF-6D was also evaluated by assessing its ability to discriminate among patient groups that differed in health, using two external criteria, viz. self-rated global health and whether, or not, they had achieved a minimal clinically important change in CTS symptoms after surgery. The self-rated global health item of the SF-36 (not part of the SF-6D) was dichotomised into (excellent, very good, good) versus (fair, poor). An improvement in CTS symptom severity score of 0.5 was considered as a minimal clinically important change and patients’ change scores were dichotomised according to this cut-off. This was a somewhat conservative estimate because half a standard deviation, commonly used as a minimal clinically important change (Norman et al., 2003) would correspond to a smaller change on the symptom severity score. The SF-6D scores for the two groups were compared with the t-test. The correlations between the SF-6D index and the five-level responses to the self-rated global health item and the CTS symptom severity score also were assessed using the Spearman correlation coefficient (r). Differences were considered statistically significant when the P value was less than 0.05.
Patients The SF-36 was administered to 124 consecutive patients, of whom 21 did not return the follow-up questionnaires. Of 103 eligible patients who had completed the SF-36 before and after carpal tunnel release, 100 patients had complete responses to all 11 items necessary for computing the SF-6D index at baseline and 95 patients had complete responses at both evaluation times. The mean age of the 100 patients was 52 (SD 15, range 21–83) years and 70 were women. The dominant hand was involved in 69 patients. Six patients had diabetes and three had inflammatory joint disease. Sixty patients were employed. The median duration of symptoms was 21 (range 2–120) months. Clinical examination revealed a positive Phalen test and/or Tinel sign in 91 patients, thenar weakness in 28 patients and abnormal 2-point discrimination (>6 mm) in at least one of the three and half radial digits in 24 patients. The mean age of the 21 patients (nine men) who did not return the follow-up questionnaires was 53 (SD 17) years.
Health utility index and outcome measures
The SF-6D could discriminate between groups that differed in self-rated global health and in whether, or not, they had a minimal clinically important improvement in CTS symptom severity score after surgery (Table 2). Of the five patients for whom follow-up SF-6D index could not be calculated because of at least one missing item, four had improvement in the CTS symptom severity score greater than 0.9 and one had an improvement of 0.3. Moderate to strong correlations were found between the SF-6D index and the five-level responses to the self-rated global health item (r = 0.60, P<0.001) and the CTS symptom severity score before surgery (r = 0.49, P<0.001) and at 3 months after surgery (r = 0.60, P<0.001).
This study has estimated health utility values with the SF-6D for CTS before and after surgery. It has also shown that the SF-6D is a valid measure of health utilities in patients with CTS and that it is sensitive to change in health after surgery. The SF-6D utilities can be used in cost effectiveness analyses in CTS, to compare utilities of various conditions or to compare the degree of change in utilities produced by different treatments. The mean SF-6D health utility index in patients with CTS in this study was 0.69 before surgery and the mean change at 3 months after carpal tunnel release was 0.08, an improvement corresponding to an effect size of 0.59, indicating moderate health change. The mean SF-6D index reported for patients with hip osteoarthritis was 0.61, the mean change at 3 months after hip arthroplasty was 0.10, and the effect size was 1.1, indicating large health change (Feeny et al., 2004). Except for the bodily pain scale and, possibly, the physical role limitation scale, the other SF-36 scales have consistently shown low sensitivity to change in CTS (Amadio et al., 1996; Atroshi et al., 1999). In the absence of a specific purpose for using one of the less sensitive scales, there does not seem to be any additional value to administering the entire SF-36 in upper extremity conditions (Gummesson et al., 2003). A utility measure derived from the SF-12 has also been reported. However, the SF-12 generates a physical health summary score, which is less sensitive to change than the bodily pain scale in CTS (Atroshi et al., 1999). A measure of pain is an important component of the outcome assessment of upper extremity disorders and the SF-36 bodily pain scale is appropriate for this purpose. Thus, the use of the 11 items that generate the SF-6D will not only provide a multi-attribute measure of utility and quality of life, but will also generate a pain score. The use of SF-6D may, thus, improve the efficiency of health-related quality of life measurement, decreasing responder and administrative burden, as 11 items provide information that otherwise would require the use of several types of scales. Another multi-attribute utility measure that has been widely used in many conditions, but rarely in upper extremity disorders, is the EQ-5D, which is a brief five-item questionnaire covering five dimensions (mobility, self-care, usual activities, pain/discomfort and anxiety/ depression) (Hurst et al., 1997). The EQ-5D index describes a total of 243 possible health states and generates values ranging from –0.109 to 1.0 (full health). The EQ-5D health state valuations have been derived from random general population samples (Dolan, 1997; Shaw et al., 2005) using the time trade-off method (persons decide the hypothetical length of time they are willing to give up in return for a better health state). The EQ-5D has the advantage of being short but it does not generate an independent pain score. Also, the EQ-5D has produced very low utility scores, indicating heath states equal to, or worse than, death in a substantial number of patients in conditions such as lumbar disc herniation and hip osteoarthritis (Jansson et al., 2005; Ostendorf et al., 2004), a phenomenon that does not occur with the SF-6D. Varying degrees of disparity between the EQ-5D and SF-6D utilities have been shown (Brazier et al., 2004; Marra et al., 2005), which may be caused by differences in the descriptions of lower health states and in the valuations of similar health states (Bryan and Longworth, 2005). Thus, estimates of cost effectiveness based on these two utility measures may not be directly comparable. This study confirmed previous findings that the CTS questionnaire, similar to other disease-specific measures, had the highest sensitivity to change. However, the choice of outcome measures should depend on the purpose for their use. If a study aims only to compare the efficacy of two treatments in CTS, then the choice should be the CTS symptom severity scale, because the other measures can only detect larger differences. However, if the purpose is also to compare the cost effectiveness of established, or new, surgical or non-surgical treatments, then the 11-item SF-6D would be an appropriate choice.
This study was supported by grants from Skane County Council’s Research and Development Foundation. Received for publication November 8, 2005. Accepted for publication November 2, 2006.
  CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati What's this?
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
Abstract
PATIENTS AND METHODS