1. Introduction and Overview. 2. Detecting Influential Observations and Outliers. 3. Detecting and Assessing Collinearity. 4. Applications and Remedies. 5. Research Issues and Directions for Extensions. Bibliography. Author Index. Subject Index.

Regression Diagnostics: Identifying Influential Data and Sources of Collinearity

Scope of the report This report is concerned with the evaluation of measures broadly designed to measure quality of life (QoL) in children and adolescents, either by self-report or proxy raters. Four research questions were identified: (1) To what extent are adult measures used in the evaluation of healthcare interventions in children? (2) How appropriate are adult measures for use with children? (3) To what extent do child self-reports correspond with assessments made by parents and carers? (4) How feasible and reliable are proxy measures of QoL in different disease contexts? Objectives (1) To review the state of the art with regard to measurement of QoL for children. (2) To make recommendations regarding the value of currently available measures for different purposes. (3) To identify further research needs. Method Electronic databases were searched for the period 1980-July 1999 for articles relating to measures of QoL, health status or well-being in children (under 18 years) with chronic disease. Handsearching of relevant journals and cross-referencing with reference lists in identified articles was also carried out. Key workers in the field were contacted for additional information, and the Internet was searched for relevant websites. Results Forty-three measures were identified (19 generic and 24 disease-specific). Sixteen measures allowed for completion by children and parent/caregiver; seven only allowed for completion by a proxy, and the remainder (n = 17) allowed only for child completion. The measures were described as QoL (n = 30), health status, (n = 2), perception of illness (n = 1), life satisfaction (n = 1) and quality of well-being (n = 1). RESULTS - TO WHAT EXTENT ARE ADULT MEASURES USED IN THE EVALUATION OF HEALTHCARE INTERVENTIONS IN CHILDREN?: Three studies were identified where adult measures were used with very few changes made for children. In 11 studies involving nine separate measures of QoL, adult measures were used as a model for work with children. RESULTS - HOW APPROPRIATE ARE ADULT MEASURES FOR USE WITH CHILDREN?: Adult measures may fail to tap the specific aspects of QoL that are important to the child. Measures based on adult work impose considerable response burden for children, in terms of length, reading skills and response scale. Wording and format of adult measures may need to be modified to account for children's cognitive and language skills. More basic research is needed to determine the level of response burden that children of different ages can manage. Assessments of difficulty (e.g. reading age) need to be routinely included with information about new measures. RESULTS - TO WHAT EXTENT DO CHILD SELF-REPORTS CORRESPOND WITH ASSESSMENTS MADE BY PARENTS AND CARERS?: Fourteen studies were identified in which concor-dance between child and parent was investigated, often as part of the development of a new measure. There was some evidence for greater concordance between child and parent for physical functioning compared with social and emotional domains, but greater heterogeneity in the latter measures may contribute to inconsistent results. There was no simple relationship between concordance and moderating variables such as age, gender and illness, but this conclusion was addressed only very rarely. RESULTS - HOW FEASIBLE AND RELIABLE ARE PROXY MEASURES OF QOL IN DIFFERENT DISEASE CONTEXTS?: Only five papers fulfilled the review criteria. Evaluation is difficult because authors fail to justify their choice of measures, and do not report critical information such as completion rates or missing data. Use of existing measures can potentially eliminate the time and expense required to develop a comprehensive measure of QoL, but a full battery of standardised tests may be expensive in terms of time for administration and scoring. In addition, battery measures tend to be lengthy and therefore demanding for sick patients. They are not recommended for work with children. RECOMMENDATIONS FOR RESEARCH - MINIMUM CRITERIA FOR NEW MEASURES: A set of procedures needs to be established for the development of new measures. These need to draw on the experience gained in development of child and adult measures to date. Basic research to enhance understanding of how children interpret questions in QoL measures is recommended. We need to understand the differences in meaning of items between children and adults, and between children of different ages. Some attempt to develop measures for children of 6 years or more have been reported, and these should be further developed. (ABSTRACT TRUNCATED)

Quality-of-life measures in chronic diseases of childhood

AIMS—To identify currently available generic and disease specific measures of quality of life (QoL) for work with children; and make recommendations about the future development and application of QoL measures.
METHODS—Systematic searches were conducted to identify measures of QoL. Primary research papers were coded by the authors on the basis of predefined inclusion and exclusion criteria.
RESULTS—Of the 137 papers included in the review, 43 involved the development of a new measure. These included 19 generic and 24 disease specific measures. Almost half the measures were developed in the USA. Measures were identified which were appropriate for children across a broad age range, and included provision for completion by different respondents (child only, parent only, or both). There were no clear distinctions between measures of QoL, health, or functional status.
CONCLUSIONS—We have identified a small number of measures which fulfil basic requirements and could be used to assess QoL in clinical trials or following interventions. However, there remain a number of problems in measuring QoL in children. These include limited availability of disease specific measures; discrepancies between child and parent ratings; limited availability of measures for self completion by children; lack of precision regarding the content of domains of QoL; and the cultural appropriateness of measures developed elsewhere for children in the UK.

https://adc.bmj.com/content/archdischild/84/3/205.full.pdf

A review of measures of quality of life for children with chronic illness

STUDY DESIGN Outcome study to determine response distribution, internal consistency, reproducibility, and concurrent validity of the Scoliosis Research Society-22 (SRS-22) health-related quality-of-life (HRQL) questionnaire. OBJECTIVES Further refinement of an HRQL questionnaire specific for idiopathic scoliosis. SUMMARY OF BACKGROUND DATA Previous experience with the original and modified SRS HRQL questionnaires suggested a need for further refinement and more complete validation. METHODS The SRS-22 and Short Form 36 (SF-36) HRQL questionnaires were mailed to 83 previously surveyed postoperative idiopathic scoliosis patients. RESULTS Fifty-eight (70%) patients returned the first set of questionnaires. Their average age at surgery was 14.6 years, and their average follow-up interval since surgery was 10.8 years. Fifty-one (88%) of the 58 returned the second set of questionnaires an average of 28 days later. The psychometric attributes of the instruments were comparable: score distribution, SRS-22 56.9% ceiling and 1.7% floor, SF-36 79.3% ceiling and 1.7% floor; internal consistency (Cronbach alpha), SRS-22 0.92 to 0.75, SF-36 0.91 to 0.36; and reproducibility (intraclass correlation coefficient), SRS-22 0.96 to 0.85, SF-36 0.92 to 0.61. Concurrent validity, determined by Pearson Correlation Coefficients between SRS-22 and SF-36 domains, was 0.70 or greater ( < 0.0001) for 17 relevant comparisons. CONCLUSION The SRS-22 HRQL questionnaire is reliable with internal consistency and reproducibility comparable to SF-36. In addition, it demonstrated concurrent validity when compared to SF-36. It is shorter and more focused on the health issues related to idiopathic scoliosis than SF-36.

The reliability and concurrent validity of the scoliosis research society-22 patient questionnaire for idiopathic scoliosis.

http://public-files.prbb.org/publicacions/12060e10-c44a-012b-a7a8-000c293b26d5.pdf

Health-Related Quality of Life Measurement in Children and Adolescents: A Systematic Review of Generic and Disease-Specific Instruments

Study design A group of 102 brace-treated adolescents, aged 10-19 years with spine deformities participated in a cross-sectional study. Objectives To determine the effect of various types of orthoses on self-perceived health status. Summary of background data Spinal orthosis is an effective means of controlling progressive scoliosis, but bracing has shown a considerable impact on several aspects of adolescent functioning. Methods Skeletally immature patients with spine deformities (75% with idiopathic scoliosis) who visited consecutively for routine biannual follow-up evaluations of orthotic treatment were studied. Twenty-five patients used the Milwaukee brace, 30 the Boston brace, 13 the thoracolumbosacral orthosis (TLSO), and 34 the Charleston bending orthosis. Patients completed the Quality of Life Profile for Spine Deformities (QLPSD), a specific instrument that measures quality of life in five areas labeled psychosocial functioning, sleep disturbances, back pain, body image, and back flexibility. Higher QLPSD scores mean a high level of impairment of quality of life. Results Milwaukee brace-treated patients scored significantly higher than Boston brace-treated and TLSO-braced patients and patients with Charleston bending orthosis in the overall QLPSD score (mean +/- SD, 53.60 +/- 13.26 vs. 45.65 +/- 12.81 and 42.79 +/- 12.99, respectively) and in back flexibility and psychosocial functioning. Other quality-of-life-related variables selected in multivariate analysis were the Risser sign, clinical diagnosis, duration of brace treatment, and degrees of correction. Conclusion In cases of different orthoses of proven similar effectiveness in controlling the scoliotic curves, the use of bracing with the lowest impact on the quality of life should be recommended.

Impact of the type of brace on the quality of life of Adolescents with Spine Deformities.

Study design The development and construction of a specific instrument for measuring quality of life in adolescents with spine deformities was investigated. Objectives To assess the validity and reliability of the Quality of Life Profile for Spine Deformities. Summary of background data An 88-item questionnaire was self-administered to 174 patients ranging in age from 10 to 20 years with spine deformities. Items were rated on a five-point Likert scale. Higher scores means high level of impairment in quality of life. Age, gender, menarche or voice change, salient symptoms in the medical record, ordinary parameters on physical examination, and measurements on standard anteroposterior and lateral radiographs were recorded. The retest was done 10 days after the initial administration in a subsample of 35 patients. Methods The test-retest reliability was analyzed calculating the intraclass correlation coefficient. Internal consistency was measured with the Cronbach's alpha method. Factor analysis was used to obtain a reduced number of variables. Construct validity was assessed using the principal components model of factor analysis based on the correlation matrix and using the varimax computer algorithm for orthogonal rotation. Discriminant validity was assessed using the Kruskal-Wallis test. Results The Quality of Life Profile for Spine Deformities contained 21 items and five factors in conceptual terms labeled psychosocial functioning, sleep disturbances, back pain, body image, and back flexibility. The overall questionnaire score showed an internal consistency of 0.88 and a test-retest correlation of 0.91. Patients with structural curves showed significantly higher scores in all dimensions of the Quality of Life Profile for Spine Deformities except for the subscale of body image than patients with postural curves. When patients were grouped according to the symptom of back pain, those with backache had a significantly higher quality of life overalls score and scores in the dimensions of sleep disturbances and pain. Brace-treated patients showed statistically significant differences in the quality of life overall score and scores in the dimensions of psychosocial functioning and back flexibility. Conclusions The instrument developed for measuring quality of life in patients with spine deformities during the period of bone growth has validity, internal consistency, and high test-retest reliability. The conceptualization of quality of life of the Quality of Life Profile for Spine Deformity includes psychosocial dimensions and pain and function.

Construction and validation of a specific quality of life instrument for adolescents with spine deformities

Study Design. A cross-sectional multicenter study was performed to validate the Spanish version of the Scoliosis Research Society-22 (SRS-22) Patient Questionnaire. Objectives. To determine the construct validity and convergent validity of the instrument. Methods. The Spanish version of the SRS-22 was given to 175 patients with scoliosis (mean age, 19 years old and 86% women). A subgroup of 31 patients also received the Quality of Life for Spine Deformities Profile (QLSDP). Construct validity was studied by factor analysis. Discriminant validity was assessed analyzing the relation between SRS-22 scores and the variables of deformity severity. Convergent validity as related to the QLSDP was studied with a multitrait-multimethod matrix analysis. Results. Factor analysis offered a solution of four factors coherent with the dimensions of the original instrument. SRS-22 scores were worse in older patients (r = -0.34); patients using analgesics demonstrated lower scores (P < 0.001). Patients treated with a brace had a poorer self-image and were less satisfied with their treatment (P < 0.001) than the other treatment groups. Angular improvement of the curves was associated with better self-image scores (r = 0.34). The SRS-22 and QLSDP demonstrated high correlation coefficients in the convergent validity tests (r = 0.84). Conclusion. The Spanish version of the SRS-22 is valid. It has a factorial structure similar to that of the original questionnaire. Moreover, it relates to known severity characteristics of the disease, distinguishes among scoliosis patient groups, and shows concordant values with another valid instrument for measuring self-perceived health.

Validity of the Spanish version of the Scoliosis Research Society-22 (SRS-22) Patient Questionnaire.

STUDY DESIGN Validation of the transcultural adaptation of a questionnaire for measuring health-related quality of life. OBJECTIVES To translate and culturally adapt the SRS-22 questionnaire to Spanish. To determine the metric qualities (internal consistency and test-retest reproducibility) of this questionnaire. SUMMARY OF BACKGROUND DATA The SRS-22 Patient Questionnaire has proven to be a valid instrument for clinical assessment of patients with idiopathic scoliosis. The widespread use of the SRS-22 in non-English-speaking countries requires its transcultural adaptation. METHODS Transcultural adaptation of the SRS-22 was carried out according to the International Quality of Life Assessment Project guidelines and included two translations and two back-translations of the material. A committee of experts decided on the final version. The questionnaire was administered to 175 individuals (152 women and 23 men) with idiopathic scoliosis. The mean age of the participants at the time they received the questionnaire was 18.9 years, thoracic curve magnitude was 28.8 degrees, and lumbar curve magnitude was 28.1 degrees. At this time, 85 patients had been treated surgically, 45 had been treated with orthesis, and 45 were under observation. A subgroup of 30 patients completed the questionnaire a second time 1 week later. Internal consistency was determined with Cronbach's alpha coefficient and test-retest reliability with the intraclass correlation coefficient. RESULTS The overall alpha coefficient of the questionnaire was 0.89. Coefficients for individual domains were as follows: function/activity, 0.67; pain, 0.81; mental health, 0.83; self-image, 0.73; and satisfaction, 0.78. The questionnaire as a whole had an intraclass correlation coefficient of 0.96. Intraclass correlation coefficients for individual domains were as follows: pain, 0.93; function, 0.82; self-image, 0.94; mental health, 0.94; and satisfaction, 0.98. CONCLUSIONS The Spanish version of the SRS-22 Patient Questionnaire demonstrated adequate internal consistency for the majority of domains and excellent reproducibility. These results suggest that the process of adaptation has produced an instrument that is apparently equivalent to the original and suitable for clinical research.

The Spanish version of the SRS-22 patient questionnaire for idiopathic scoliosis: transcultural adaptation and reliability analysis.

The Walter Reed Visual Assessment Scale (WRVAS) was designed to allow idiopathic scoliosis patients to describe their perception of their deformity. In a previous stduy, the scale has shown good correlation with magnitude of the curve The study included 70 patients (60 women and 10 men), mean age 19.4 years (range 12–40), with idiopathic scoliosis. Each patient filled out the WRVAS and the SRS-22 questionnaire. Thoracic and lumbar curve angles were determined in standing X-rays and the largest was named Cobbmax. WRVAS internal consistency was assessed with Cronbach's alpha. Correlation coefficients were calculated between Cobbmax and the various WRVAS questions, and Cobbmax and the SRS-22 scales. The correlation between the WRVAS and SRS-22 was also determined Mean magnitudes were thoracic curve, 36.6° and lumbar curve, 33.2°; average Cobbmax was 37.9°. The mean total WRVAS score was 15.6. Mean scores for the various SRS-22 scales were function 4.6, pain 4.3, self-image 3.7, mental health 4.2, and total score 84.1. Internal consistency for the WRVAS was excellent (Cronbach's alpha, 0.9), and there were no signs of collinearity among the seven questions (tolerance range 0.2–0.5). All the items on the WRVAS correlated significantly with Cobbmax (correlation coefficients, 0.4 to 0.7). The correlation between the total WRVAS and total SRS-22 score was -0.54 (P = .0001) and between WRVAS total score and SRS-22 image domain score was -0.57 (p = 0.0001) The WRVAS showed excellent internal consistency and absence of collinearity. There was a highly significant correlation between the results of the test and the magnitude of the deformity. The WRVAS correlated significantly with the SRS-22 image scale. The WRVAS is a valid instrument to assess scoliosis patients perception of their deformity

/pdf/validity-of-the-walter-reed-visual-assessment-scale-to-478a8gnr79.pdf

Jose M Climent

Papers

Impact of the type of brace on the quality of life of Adolescents with Spine Deformities.

Construction and validation of a specific quality of life instrument for adolescents with spine deformities

Validity of the Spanish version of the Scoliosis Research Society-22 (SRS-22) Patient Questionnaire.

The Spanish version of the SRS-22 patient questionnaire for idiopathic scoliosis: transcultural adaptation and reliability analysis.

Validity of the Walter Reed Visual Assessment Scale to measure subjective perception of spine deformity in patients with idiopathic scoliosis