The SF-36 physical and mental health summary measures: an example of how to interpret scores

OBJECTIVES: To provide normative data, in the form of percentile scores from a community sample, for the Physical Component and Mental Health Component Summary scores derived from the SF-36, and to provide an example of how to interpret scores on these measures, comparing normative results with data from a clinical sample. METHOD: Normative data were gained from a postal survey using a questionnaire, containing the SF-36 and a number of other items concerned with lifestyles and illness. The questionnaire was sent to 13,042 randomly selected subjects between the ages of 18 and 64 years, drawn from Family Health Services Authority computerised registers for four English counties. The clinical sample comprised 84 patients aged 18-64 years diagnosed with obstructive sleep apnoea (OSA) who were asked to take part in the study. The Physical Component Summary (PCS) score and Mental Health Component Summary (MCS) score gained from the SF-36 health status measure were the outcome measures. RESULTS: The community survey achieved a response rate of 72% (n = 9332). All 84 patients in the age range 18-64 years approached to take part in the OSA study agreed to do so; complete data were available for 60 patients. Results indicated that, prior to treatment, 75% of OSA patients' scores on the PCS/MCS were less than the standardised mean score of 50 and fell in the lowest 34% of scores in the general population. However, after treatment, over 50% of patients scored above the standardised mean score on both the PCS and MCS and more closely mirrored the distribution of the normative sample. CONCLUSION: The data provided here should enable a more meaningful presentation of data than is generally provided in research papers presenting SF-36 summary scores.     

Correlated physical and mental health summary scores for the SF-36 and SF-12 Health Survey, V.1

Background  

The SF-36 and SF-12 summary scores were derived using an uncorrelated (orthogonal) factor solution. We estimate SF-36 and SF-12 summary scores using a correlated (oblique) physical and mental health factor model.     

Do differences in methods for constructing SF-36 physical and mental health summary measures change their associations with chronic medical conditions and utilization?

Background and objective: Various approaches have been employed to derive physical health and mental health summary scores for the SF-36 and the RAND-36, but head-to-head comparisons of alternative scoring algorithms are rare. We determined whether the associations of the physical and mental health summary scores with chronic medical conditions and utilization would differ depending on the scoring algorithm used. Methods: We examined 5701 patients receiving medical care from an independent association of 48 physician groups located primarily in the western United States and compared SF-36 and RAND-36 scoring of physical health and mental health summary scores. Associations with the presence of diabetes, heart disease, and kidney disease, as well as with utilization of medical care and mental health care were compared using bivariate and multivariate analysis. To examine the relationship between SF-36 and RAND-36 scores, we regressed the SF-36 physical and mental health composite scores on the RAND-36 physical and mental health summary measures and vice versa. Results: We found that the SF-36 and RAND-36 summary scores generally yielded results similar to one another across measures of heart disease, diabetes, and kidney disease, as well as measures of utilization. However, for each chronic medical condition, the RAND-36 showed a slightly larger decrement in mental health than did the SF-36. Conclusions: Differences between the two sets of summary scores were consistent with their respective conceptual and analytic approaches. Where comparisons of results between the SF-36 and RAND-36 summary scores are desirable in future studies, they can be estimated using the regression equations derived in this study.     

Deriving SF-12v2 physical and mental health summary scores: a comparison of different scoring algorithms

Purpose  

Summary scores for the SF-12, version 2 (SF-12v2) health status measure are based on scoring coefficients derived for version 1 of the SF-36, despite changes in item wording and response scales and despite the fact that SF-12 scales only contain a subset of SF-36 items. This study derives new summary scores based directly on SF-12v2 data from a recent U.S. sample and compares the new summary scores to the standard ones. Due to controversy regarding methods for developing scoring coefficients for the summary score, we compare summary scores produced by different methods.     

The SF-36 summary scores and their relation to mental disorders: physical functioning may affect performance of the summary scores

BACKGROUND AND OBJECTIVES: The Medical Outcomes Study 36-item Short-Form Health Survey (SF-36) has been widely used as a generic measure of health status. It can be scored to provide either a profile of eight scores or two summary measures of health. Several studies demonstrated shortcomings of the summary scores in accurately reflecting patients' physical and mental health on the basis of subscale scores. The objective of this study was to compare and evaluate different scoring algorithms for the summary scores. METHODS: The analysis was based on data on 4,052 respondents from the German National Health Interview and Examination Survey. Mental disorders were assessed using a structured clinical interview. Logistic regression and receiver operating characteristic analyses were used to evaluate the association between the mental component scores and mental disorders. RESULTS: Subjects with mental disorders reported poorer quality of life on all SF-36 subscales and component scores compared to those without mental disorders. The presence of physical disorders resulted in different summary scores. The screening accuracy in detecting subjects with mental disorders was satisfactory for both mental summary scores. CONCLUSIONS: The summary scores should be evaluated in relation to the profile of the eight subscales. Physical functioning should be evaluated carefully when comparing health status using summary scores.     

Comparing SF-36 scores across three groups of women with different health profiles

Background: The widespread use of the Medical Outcomes Study (MOS) 36-item Short-Form Health Survey (SF-36) facilitates the comparison of health-related quality of life (HRQL) across independent studies. Objectives: To compare the scores of eight scales and two summary scales of the SF-36 across participants in the Womens Healthy Eating and Living (WHEL) trial, the Womens Health Initiative-Dietary Modification trial (WHI-DM), and the MOS, and to illustrate the use of effect sizes for interpreting the importance of group differences. Methods: WHEL and WHI-DM are both multi-center dietary interventions; only data from the UC Davis sites were used in our study. WHEL participants had a recent history of breast cancer, WHI-DM participants were healthy, postmenopausal women, and women in the MOS had a history of hypertension, diabetes, heart disease, or depression. General linear models were used to identify statistically significant differences in scale scores. Meaningful differences were determined by effect sizes computed using a common within-group standard deviation (SD) and SDs from normative data. Results: After adjusting for age and marital status, SF-36 scores for the WHI-DM and WHEL samples were similar and both had statistically significantly higher scores than the MOS sample. Relative to the WHEL or WHI-DM studies, MOS scores for scales related to the physical domain were clearly meaningfully lower whereas scale scores related to the mental health domain were potentially meaningfully lower. Conclusions: The HRQL of breast cancer survivors is comparable to that of healthy women and better than that of women with chronic health conditions, particularly with respect to physical health. This study illustrated the use of ranges of effects sizes for aiding the interpretation of SF-36 scores differences across independent studies.     

Comparison of UK and US methods for weighting and scoring the SF-36 summary measures

BACKGROUND: The SF-36 is a widely used measure of health status that can be scored to provide either a profile of eight scores or two summary measures of health, the Physical Component Summary and Mental Component Summary (PCS and MCS). Scoring of the summary scales is undertaken by weighting and summing the original eight dimensions. These weights are gained from factor analysis of data from a general population and have been assumed to be country specific. However, it has been suggested that the weights gained from the US developers could be applied to all datasets, throughout the world, for purposes of comparability and simplicity. The purpose of this study is to evaluate US and UK scoring schemes in a UK population dataset, and in a cohort study of elderly congestive heart failure patients receiving standard therapy and a trial of open vs laparoscopic surgery for hernia repair. METHODS: This paper compares algorithms developed in the USA and the UK for the calculation of the Physical and Mental Health Summary scores (PCS and MCS) for the SF-36 health status measure. In this study the PCS and MCS were calculated using a weighting scheme recommended by the developers and derived from a US population sample dataset, as well as being calculated from weights derived from an UK population sample dataset. RESULTS: The two methods produced similar results, both cross-sectionally and in the assessment of change. CONCLUSIONS: It is suggested that it may be necessary to weight the PCS and MCS using only the original US algorithms, which will lead to more uniform analysis of datasets and may also lead to greater uptake of the summary measures. Furthermore, the results would suggest that in international trials the SF-36 can be adopted and summary scores calculated for countries where no large-scale normative dataset is available. However, further research is needed to determine that the similarity of results gained using UK and US algorithms is not an idiosyncratic feature of the UK data. Studies to verify the findings reported here are required from other countries.     

Interpreting SF&-36 summary health measures: A response

In response to questions raised about the “accuracy” of SF-36 physical (PCS) and mental (MCS) component summary scores, particularly extremely high and low scores, we briefly comment on: how they were developed, how they are scored, the factor content of the eight SF-36 subscales, cross-tabulations between item-level responses and extreme summary scores, and published and new tests of their empirical validity. Published cross-tabulations between SF-36 items and PCS and MCS scores, reanalyses of public datasets (N = 5919), and preliminary results from the Medicare Health Outcomes Survey (HOS) (N = 172,314) yielded little or no evidence in support of Taft's hypothesis that extreme scores are an invalid artifact of some negative scoring weights. For example, in the HOS, those (N = 432) with “unexpected” PCS scores worse than 20 (which, according to Taft, indicate better mental health rather than worse physical health) were about 25% more likely to die within two years, in comparison with those scoring in the next highest (21– 30)␣category. In this test and in all other empirical tests, results of predictions supported the validity of extreme PCS and MCS scores. We recommend against the interpretation of average differences smaller than one point in studies that seek to detect “false” measurement and we again repeat our 7-year-old recommendation that results based on summary measures should be thoroughly compared with the SF-36 profile before drawing conclusions. To facilitate such comparisons, scoring utilities and user-friendly graphs for SF-36 profiles and physical and mental summary scores (both orthogonal and oblique scoring algorithms) have been made available on the Internet at www.sf-36.com/test. This revised version was published online in June 2006 with corrections to the Cover Date.     

Health-related quality of life in chronic disorders: a comparison across studies using the MOS SF-36

The purpose of this report is to examine health-related quality of life (HRQoL) as measured by the Medical Outcomes Study Short Form-36, across patient populations with chronic disorders and to compare quality of life (QoL) in these subjects with normative data on healthy persons. Six studies, within the Center for Research in Chronic Disorders at the University of Pittsburgh School of Nursing, in patients with urinary incontinence, prostate cancer, chronic obstructive pulmonary disease (COPD), acquired immune deficiency syndrome (AIDS), fibromyalgia and hyperlipidaemia provided the data for analysis. The results demonstrated that not only did the prostate cancer and hyperlipidaemia patients have the highest QoL across the chronic disorders, but their QoL was comparable to normative data on healthy persons. Homebound, elderly, incontinent patients had the lowest QoL for physical functioning, whereas patients hospitalized with AIDS had the lowest QoL in general health and social functioning. Patients with COPD had the lowest QoL in role-physical, role-emotional and mental health. Patients with fibromyalgia had the lowest QoL in bodily pain and vitality. Compared to normative data, patients with urinary incontinence, COPD, AIDS and fibromyalgia generally had lower QoL. Prostate cancer and hyperlipidaemia patients had QoL comparable to normative data. Compared to normative data, patients with urinary incontinence, COPD, AIDS and fibromyalgia had more variability for role-emotional. AIDS patients had more variability on physical functioning, bodily pain and social functioning compared to the normative data. These data suggest that patients with various chronic disorders may have QoL that is lower in most domains compared to a healthy population. However, there may be differences in the domains affected as well as the extent of variation across specific chronic disorders.     

The SF-36 physical and mental health factors were confirmed in cancer and HIV/AIDS patients

BACKGROUND AND OBJECTIVE: This study assesses the extent to which the RAND-36/SF-36 items measure physical and mental health (PH and MH), as implied by existing summary scoring systems. METHODS: A total of 1,714 heterogeneous cancer and HIV/AIDS patients were recruited from five institutions. Of these, 56% were women; 81% Caucasians; and about 10% were from each of the major cancer types and HIV/AIDS. RESULTS: Analyses of the SF-36 confirmed the two dimensions of health namely physical and mental. However, item fit statistics and residual factor analysis revealed that some items intended to represent the PH dimension fit better with the MH dimension. CONCLUSION: This paper demonstrated the value of Rasch residual factor analysis for understanding and enhancing interpretation of health.     

Sensitivity to change of health status measures in a randomized controlled trial: comparison of the COOP charts and the SF-36

This study compared the sensitivity to change of comparable dimensions of a multi-item multi-dimensional health status measure (the SF-36) with the equivalent single item domains on the Dartmouth COOP charts. One hundred and twenty nine patients were randomized to either day case laparoscopic surgery (n=60) or open inguinal hernia repair (n=69). Respondents completed the SF-36 and COOP charts at baseline (prior to surgery) and at follow up at 10 days and 6 weeks. Equivalent dimensions of physical functioning, mental health/emotional condition, social activities, pain and overall condition/general health on the two questionnaires were compared. Despite slightly different pictures of change provided by the physical functioning and overall condition/general health dimensions the general picture of change provided by the two instruments was similar. At 10 days, patients who underwent open surgery reported far greater levels of dysfunction than those who underwent laparoscopic surgery on both questionnaires. At 6 weeks the pain dimension of both questionnaires indicated a large improvement from baseline, whilst no other domain on either questionnaire for either group indicated such improvement. The general picture of change provided by the two measures was similar. The results suggest that both the SF-36 and the COOP charts may prove suitable for the assessment of health perception outcomes in surgical clinical trials. Differences on certain domains were caused in large measure by the nature of the questions posed. The study once again highlights the importance of checking item content to determine the suitability of any particular measure for a given study.     

The SF-36 scales are not accurately summarised by independent physical and mental component scores

Objectives The Short Form 36 Health Status Questionnaire (SF-36) has eight scales that can be condensed into two components: physical component summary (PCS) and mental component summary (MCS). This paper investigates: (1) the assumption that PCS and MCS are orthogonal, (2) the applicability of a single model to different condition-specific subgroups, and (3) a reduced five-scale model. Study design and setting We performed a secondary analysis of two large-scale data sets that utilised the SF-36: the Health Survey for England 1996 and the Welsh Health Survey 1998. We used confirmatory factor analysis to compare hypothetical orthogonal and oblique factor models, and exploratory factor analysis to derive data-driven models for condition-specific subgroups. Results Oblique models gave the best fit to the data and indicated a considerable correlation between PCS and MCS. The loadings of the eight scales on the two component summaries varied significantly by disease condition. The choice of model made an important difference to norm-referenced scores for large minorities, particularly patients with a mental illness or mental–physical comorbidity. Conclusions We recommend that users of the SF-36 adopt the oblique model for calculating PCS and MCS. An oblique five-scale model provides a more universal factor structure without loss of predictive power or reliability.     

A comparison of the discriminative and evaluative properties of the SF-36 and the SF-6D index

Purpose  

To examine whether the move from the multidimensional SF-36 patient-reported outcome measure to the single-index preference-based SF-6D entails a loss in discriminative and evaluative properties, the magnitude of that loss and whether it matters.     

The SF-36 component summary scales and the daytime diurnal cortisol profile

Purpose  

Higher daytime cortisol output has been associated with higher levels of perceived stress and worse mental and physical health outcomes. Hypothalamic–pituitary–adrenal (HPA) axis dysregulation, such as elevated secretion of daytime cortisol, occurs in many mental and physical illnesses. However, the nature of the association between functional health status and daytime cortisol production has not been established.     

Sample size and power estimation for studies with health related quality of life outcomes: a comparison of four methods using the SF-36

We describe and compare four different methods for estimating sample size and power, when the primary outcome of the study is a Health Related Quality of Life (HRQoL) measure. These methods are: 1. assuming a Normal distribution and comparing two means; 2. using a non-parametric method; 3. Whitehead's method based on the proportional odds model; 4. the bootstrap. We illustrate the various methods, using data from the SF-36. For simplicity this paper deals with studies designed to compare the effectiveness (or superiority) of a new treatment compared to a standard treatment at a single point in time. The results show that if the HRQoL outcome has a limited number of discrete values (< 7) and/or the expected proportion of cases at the boundaries is high (scoring 0 or 100), then we would recommend using Whitehead's method (Method 3). Alternatively, if the HRQoL outcome has a large number of distinct values and the proportion at the boundaries is low, then we would recommend using Method 1. If a pilot or historical dataset is readily available (to estimate the shape of the distribution) then bootstrap simulation (Method 4) based on this data will provide a more accurate and reliable sample size estimate than conventional methods (Methods 1, 2, or 3). In the absence of a reliable pilot set, bootstrapping is not appropriate and conventional methods of sample size estimation or simulation will need to be used. Fortunately, with the increasing use of HRQoL outcomes in research, historical datasets are becoming more readily available. Strictly speaking, our results and conclusions only apply to the SF-36 outcome measure. Further empirical work is required to see whether these results hold true for other HRQoL outcomes. However, the SF-36 has many features in common with other HRQoL outcomes: multi-dimensional, ordinal or discrete response categories with upper and lower bounds, and skewed distributions, so therefore, we believe these results and conclusions using the SF-36 will be appropriate for other HRQoL measures.     

A structural equation model for predictors of mental and physical health in women]

OBJECTIVE: To measure the extent to which stress, social support, and self-esteem are predictors of an individual's mental and physical health. Structural equations were integrated with previously-estimated partial models, which simplify the relationships among variables. METHODS: The study sample included 283 women with children. All of the participants resided in the municipality of General Escobedo, state of Nuevo León, Mexico. The surveys were carried out in the second semester of 2003, in the participants homes, using self-evaluation questionnaires to measure each of the variables included in the model. Each participant completed the questionnaire in one sitting. Results were analyzed with AMOS 5.0, employing the maximum likelihood method, often utilized in structural equation models. RESULTS: The results indicate an acceptable adjustment in the proposed model: (chi2/gl=3.03, goodness of fit (GFI)=0.894, adjusted goodness of fit (AGFI)=0.848, root mean square error of approximation (RMSEA)=0.08, incremental fit index (IFI)=0.910). Variances were 31.9% with regard to stress, 27.4% with regard to physical health, and 72.1% with regard to mental health. CONCLUSIONS: Social support and self-esteem are predictors of stress; age and stress are predictors of physical health; and stress, self-esteem, and physical health are predictors of mental health.