Survival rates after a diagnosis of childhood cancer have increased dramatically over the past four decades (30). This increase has resulted in an estimated 366,000 survivors of childhood cancer living in the United States (14). An expanding body of literature demonstrates that cancer treatment, which may consist of some combination of surgery, chemotherapy, or radiotherapy, can have long-term and damaging effects on growing children (2). Chronic health conditions are prevalent in over 70% of survivors 30 yr from cancer diagnosis and can include subsequent neoplasms, cardiopulmonary dysfunction, metabolic abnormalities, neuroendocrine disorders, neurocognitive disability, neurological or sensory impairment, and musculoskeletal disability (24).
Previous literature suggests that these late effects make cancer survivors at least five times more likely to have functional impairments and twice as likely to have activity limitations than siblings (22). The compound effects of treatment-related impairments and inactivity during and postcancer treatment contribute to muscle atrophy, cardiorespiratory deterioration, bone loss, and diminished physical performance abilities (34). These impairments and limitations have the potential to negatively affect survivors’ abilities not only for leisure time physical activities but also for social recreation that requires a certain degree of community mobility (23). At the extreme, significant loss of physical performance may even interfere with simple tasks required for daily living, like bathing, dressing, and preparing meals.
Prevalence estimates for physical performance limitations among cancer survivors range from 9.5% to 19.6% (21,33). This variation is likely because different methods of assessment may affect the accuracy of physical disability estimates (16,36). Several studies suggest that questionnaires, either self-reported or interviewer administered, tend to underestimate physical disability when compared with clinical evaluation (13,32). These discrepancies make documenting the burden of physical disability difficult. It is important to be able to accurately identify childhood cancer survivors (CCS) with clinically ascertained physical disability as these are the individuals who are most likely to benefit from intervention to remediate functional loss. With this in mind, the primary aim of this investigation was to evaluate the accuracy of self-reported physical performance limitations in CCS.
MATERIALS AND METHODS
Participants were members of the St. Jude Lifetime (SJLIFE) cohort study, a study of adult survivors of pediatric cancer treated at St. Jude Children’s Research Hospital (SJCRH). The primary aim of SJLIFE is to evaluate health outcomes among CCS as they age. Participants had a previous diagnosis of a childhood malignancy treated at SJCRH, were 18 yr or older, at least 10 yr from diagnosis, and were willing to return to SJCRH for evaluation. These analyses include survivors who completed an initial medical follow-up visit and functional assessment between November 2007 and April 2012. All procedures were approved by the SJCRH institutional review board. Written informed consent was obtained for each study participant before testing.
Among the 4263 potentially eligible members of the SJLIFE cohort, 4129 had been invited to participate as of April 30, 2012. In the first 63 blocks, 3034 patients were eligible for our study. Nonparticipants included 678 who actively or passively declined participation; 60 who were lost to follow-up and 270 who agreed to participate, but who had not yet been scheduled for their visit; 162 who agreed to complete a survey, but not to return for a medical evaluation; and 40 who completed a medical evaluation, but not a functional assessment (Fig. 1). Our analysis includes 1778 participants, and 58.6% of those were eligible.
Demographic and cancer treatment data were obtained from medical records by trained abstractors. All abstractions were reviewed and approved by a physician. Height and weight were measured without shoes using a wall-mounted stadiometer and an electronic scale, respectively. Body mass index (BMI) was calculated by dividing weight in kilograms by height in meters squared.
Self-reported physical performance limitation
As part of their SJLIFE evaluation, participants completed a battery of health questionnaires, one of which includes the 10-item physical functioning subscale of the Medical Outcomes Study Short Form 36 (SF-36) (3,38). The SF-36 is a widely used generic health-related quality of life questionnaire that has been tested in multiple populations, including cancer survivors (29,39). It is valid (r = 0.40) and reliable (Cronbach’s alpha = 0.82–0.90) (26,29,37,39) and takes 5–10 min to complete. Raw scores were summed for each 10 items on the physical functioning subscale and converted into T-scores with a population mean of 50 and a standard deviation of 10 (37,39). As in the previous analysis, we classified individuals with T-scores of ≤37 on the physical function subscale as having self-reported physical performance limitations (28). This corresponds to the lowest 10th percentile of the general population (3,37,39).
Clinical assessment of physical performance limitations
Two clinical measures were used to categorize physical performance limitations. During a comprehensive functional assessment, study participants completed the Physical Performance Test (PPT) and the 6-Minute Walk Test (6MW). The seven-item PPT, originally described by Reuben and Siu (27), is an assessment of the time it takes to complete each of a series of tasks typically performed during activities of daily living. It has been used in geriatric patient populations to identify mild to moderate frailty and to predict risk for falls (5). Scores on the PPT are inversely correlated with degree of disability, loss of independence, and early mortality (5,8). Scores on the PPT range from 0 to 28. Patients were observed and timed as they 1) wrote a brief sentence, 2) simulated eating, 3) lifted a book and put it on a shelf, 4) put on and removed a jacket, 5) picked up a penny, 6) walked 50 ft, and 7) turned around in place. Participants with PPT scores ≤17 were classified as having a clinically identified physical performance limitation. A cut point of <17 corresponds to being unlikely to function in the community independently (12). The 6MW test is a general measure of physical fitness. Researchers have used the test to evaluate cardiorespiratory fitness in specific populations including those with respiratory disease (6), cystic fibrosis (11), and cancer (17). Healthy reference populations have been evaluated and provide normative data for comparison (9,15). Participants were asked to walk indoors on a level surface for 6 min. They were instructed to walk as quickly as possible without running; standardized encouragement was provided each minute. Heart rate was monitored continuously with a polar heart rate (RS100, Lake Success, NY) and recorded along with a rating of perceived exertion (Borg scale) before beginning the test, at 2-min intervals throughout the test, and after a 2-min recovery period. The total distance walked was recorded in meters. Participants who walked distances ≤300 m were classified as having a clinically identified physical performance limitation. The cut point of <300 m corresponds to an aerobic capacity equivalent to moderate housework (i.e., sweeping floors or carrying groceries) (7,31,40).
Descriptive statistics were used to characterize the study population and the distribution of scores on the SF-36 physical function subscale, the PPT, and the 6MW test. Demographic and treatment variables were compared between participants and nonparticipants with two-sample t-tests, nonparametric equivalents, or chi-squared statistics (or Fisher’s exact tests) as appropriate. Statistical diagnostic tests (sensitivity or the proportion of positives correctly identified, specificity or the proportion of negatives correctly identified, accuracy or the proportion of positives and negatives correctly identified, and Cohen’s kappa coefficient or interrater agreement) (1,35) were used to evaluate the diagnostic accuracy of, and agreement between, self-reported physical performance limitations when each clinical assessment of physical performance was used as the “gold” standard. Logistic regression models were used to identify survivors who were most likely to report a limitation when one was not clinically apparent (“false positives”) after adjusting for sex, age, diagnosis, obesity status, and time since diagnosis. SAS version 9.2 (SAS Institute, Cary, NC) was used for all analyses.
The demographic and treatment characteristics of the 1778 participants are shown in Table 1. Slightly more than half of the survivors were female (50.8%), and leukemia comprised the most common childhood malignancy (45.3%). The median age at diagnosis was 6.8 yr (range = 0–24.8) and the median time since diagnosis was 24.9 yr (10.9–48.2). Overweight and obesity were present among 29.0% and 36.3% of survivors, respectively. Participants did not differ from nonparticipants by primary diagnosis or age at diagnosis but were more likely to be female (P < 0.01).
The mean scores on the SF-36 physical function scale and the PPT and the mean distance walked in 6 min are shown in the supplementary material (see Table, SDC 1 http://links.lww.com/MSS/A342, reported and measured physical function scores) by sex, diagnosis group, age at diagnosis, age at evaluation, time since diagnosis and BMI. Males walked farther than females on the 6MW test but scored similarly on the SF-36 and PPT. Individuals with either a bone or CNS tumor had the lowest scores on the physical function subscale of the SF-36 and PPT and walked the shortest distances during the 6MW test. Age at diagnosis was associated with scores on the SF-36 physical function subscale. Older study participants scored lower on the SF-36 physical function subscale, PPT, and walked shorter distances during the 6MW test. A BMI of ≥40 kg·m−2 was associated with a lower score on the physical function subscale of the SF-36. A BMI of ≤18.5 or ≥35 kg·m−2 was associated with lower PPT scores. Similarly, a BMI of ≥35 kg·m−2 was also associated with shorter walking distances during the 6MW test.
The percentages of individuals classified with physical performance limitations according to the selected cut points for each measure are shown in Table 2. The percentage of individuals who self-reported physical performance limitations was highest among survivors treated for CNS tumors (22.0%), bone tumors (18.8%), or lymphoma (18.1%). Survivors older than 50 yr (34.5%), who had survived longer than 40 yr (32.3%) or who had BMI values <18.5 kg·m−2 (20.3%) or >40 kg·m−2 (25.2%), were also most likely to report physical performance limitations. Physical performance limitations assessed with the PPT were most prevalent among CNS tumor survivors (12.8%), in individuals who had survived ≥40 yr from diagnosis (11.3%), and in individuals with BMI <18.5 kg·m−2 (10.9%). Physical performance limitations, assessed with the 6MW test, were most prevalent in bone and CNS tumor survivors (15.6% and 9.2%), survivors older than 50 yr (13.8%), and among survivors whose BMI was <18.5 kg·m−2 (12.5%) or >40 kg·m−2 (11.2%).
Sensitivity, specificity, and accuracy
The overall sensitivities and specificities of self-reported physical performance limitations when compared with physical performance limitations measured and classified according to the two clinical assessments were 0.59 and 0.89 for the 6MW and 0.69 and 0.87 for the PPT, respectively (Table 3). When self-reported physical function was compared with 6MW or PPT, 13% of participants in this cohort were misclassified (Table 4). The positive predictive values for the 6MW (26%) and the PPT (14%) indicate that using the physical function subscale of the SF-36 with a cut point of 37 overestimates the prevalence of physical performance limitations when the clinical measures are considered the “gold standard.” The strength of the kappa coefficients in overall comparisons showed only fair or slight agreement between the SF-36 physical function subscale and the 6MW test (0.30, 95% confidence interval [CI] = 0.23–0.36) or PPT (0.19, 95% CI = 0.13–0.25).
The sensitivity, specificity, accuracy, and percent agreement of the SF-36 physical function subscale measure varied by diagnostic group and the outcome standard used. When the 6MW test was used as the comparison standard, sensitivity and specificity ranged from 0.50 and 0.84 among lymphoma survivors to 0.85 and 0.84 in CNS survivors. Accuracies ranged from 0.82 in lymphoma survivors to 0.88 in leukemia survivors. When the PPT was used as the comparison standard, sensitivity and specificity ranged from 0.67 and 0.82 in bone cancer survivors to 0.80 and 0.90 in leukemia survivors. Accuracy ranged from 0.80 in bone cancer to 0.90 in leukemia survivors. Kappa values were better for agreement between the SF-36 physical function subscale measure and the 6MW test than they were for agreement between the SF-36 physical subscale measure and the PPT.
Characteristics of those who are misclassified by self-report
When the 6MW test was used as the comparison standard for the SF-36 physical function subscale, 13.0% of the participants were misclassified. Nearly all (10.5%) of the misclassifications were individuals whose SF-36 physical function subscale score indicated that they had a physical performance limitation but whose distance walked in 6 min did not indicate a limitation (Table 4). Survivors of lymphoma (15.0%) and CNS tumor (14.2%), survivors >50 yr (25.9%), those with >40 yr of survivorship (24.2), and those with a BMI >40 kg·m−2 (16.1%) had the highest rates of false-positive diagnosis when the 6MW considered the gold standard.
In multivariable models, CNS survivors were 2.6 times (95% CI = 1.5–4.6), and lymphoma survivors were 1.5 times (95% CI = 1.0–2.4) more likely than leukemia survivors to report a physical performance limitation when their 6MW distance did not indicate a limitation (Table 5).
In multivariable models, age and body size were also associated with reporting a physical performance limitation when one was not present. For each 1 yr increase in age, the odds ratio of reporting a limitation when one was not present was 1.05 (95% CI = 1.01–1.08) in multivariable models. In multivariable models, normal weight individuals were less likely to report a physical performance limitation when one was not present compared with obese individuals (odds ratio [OR] = 0.7, 95% CI = 0.5–0.9).
When the PPT was used as the comparison standard for the SF-36 physical function subscale, 13.0% of the participants were misclassified. Nearly all (12.2%) of the misclassifications were individuals whose SF-36 physical function subscale score classified them with a physical performance limitation but whose score on the PPT did not indicate a limitation. Survivors of lymphoma (17.1%) and bone cancer (17.2%), survivors >50 yr (31.0%), those with >40 yr of survivorship (25.8), and those with a BMI >40 kg·m−2 (21.7%) had the highest rates of false-positive diagnosis when the PPT considered the gold standard.
In multivariable models, CNS tumor, bone tumor, and lymphoma survivors were 2.5 times (95% CI = 1.4–4.4), 1.9 (95% CI = 1.1–3.3) and 1.7 times (95% CI = 1.1–2.6) more likely, respectively, than leukemia survivors to report a physical performance limitation when their PPT score did not indicate a limitation. In multivariable models, age and body size were also associated with reporting a physical performance limitation when one was not present. For each 1 yr increase in age, the odds of reporting a limitation when one was not present was 1.05 times (95% CI = 1.02–1.08) in multivariable models. In multivariable models, normal weight individuals were less likely to report a physical performance limitation when their PPT score did not indicate a limitation than obese individuals (OR = 0.6, 95% CI = 0.5–0.8).
Self-reported physical function has been used widely to indicate the abilities of CCS to successfully navigate their homes, schools, work places, and communities for everyday living, social interaction, and recreation. Our study indicates that most CCS accurately report their physical performance limitations. Those who incorrectly report physical performance limitations report a problem when one is not clinically apparent. Those with CNS or bone tumors, those who are older, and those who are either over or underweight are most likely to misclassify their physical performance status.
To our knowledge, no study has previously evaluated the accuracy of self-reported physical performance, measured with the physical function subscale of the SF-36, in a cohort of CCS. Our findings are similar to those reported in elderly cohorts from the United States and Spain (10,18,19). Kelly-Hayes et al. (18) reported that 11% of an elderly cohort in the United States were misclassified when reported task limitation was used to assess poor performance and compared with the clinician’s observation of limitation on the same activities. As in our study, among those who were misclassified, the majority reported a performance limitation when one was not present according to the clinical measure (18). Ferrer et al. (10) reported agreement between self-reported physical limitations on an interview based survey and performance on a 4-m walk test in an elderly Spanish cohort that mirror those seen in our cohort.
Conversely, our findings are in contrast to research among older adults that have evaluated the influence of data collection methods with physical performance limitations as the outcome of interest. These studies consistently report that surveys, self-administered or interviewer administered, underestimate the prevalence of physical performance limitations when compared with clinical evaluations (13,16,32,36). This difference may be because CCS, several decades younger than members of these elderly cohorts, have different perceptions of impaired physical function and focused more on aerobic capacity and mobility and less on activities necessary for simple daily living. Accordingly, in our cohort, the 6MW had better agreement with self-reported physical function than did the PPT.
A secondary analysis of those who self-reported a physical performance limitation showed that those most likely to report a limitation were also most likely to be misclassified by the SF-36 physical function subscale. For example, survivors of lymphoma, CNS tumors, and bone tumors were more likely than leukemia survivors to self-report a limitation according to the SF-36 physical function subscale, and also more likely to report a limitation when one was not detectable by either the 6MW test or PPT criteria. In addition, rates of disability by self-report among older study participants or those who were under weight or obese were higher. These survivors were more likely to report a limitation when one was not clinically detectable. These findings complicate assessment of physical function since those who appear most at risk for physical performance limitations are also most likely to be misclassified when self-report is used to capture these outcomes.
Measurement of physical performance limitations is difficult, and previous work suggests that a global approach to function is needed to correctly classify physical function status (4,6). When evaluating physical function among CCS, it is also important to consider that they may have emotional and cognitive outcomes that influence both their abilities to report and their perceptions of their physical abilities (20). The physical function subscale of the SF-36 is a component of a larger generic health-related quality of life instrument, and reporting on this measure certainly is influenced by cognitive and emotional constructs. The PPT (a tool designed to assess dimensions of physical function common in everyday life) and 6MW test (a tool that evaluates aerobic capacity and mobility) are more direct measures of immediate performance and are less likely than a self-report measure to be influenced by cognitive abilities and emotional overlay. The inclusion of tools that directly evaluate physical function may be a useful addition to traditional self-reported measures.
The findings of this study should be considered in the context of potential limitations. First, not all of the individuals eligible for our study participated. It is possible that those who chose not to or who were unable to participate had more or fewer physical performance limitations than those who were able to participate. Because we could not assess physical performance limitations in nonparticipants, we have no way to evaluate the magnitude or direction of this bias. Second, the participants in our cohort were more likely to be female than the nonparticipants. Because our results did not differ by sex, this is unlikely to have affected our findings (25). In addition, the instruments and cut points we selected, although validated and widely used in cancer survivor and other populations, (12,15,26,31,40), are not the only measures of physical performance available. Other self-report measures may have better concordance with the 6MW test or the PPT. Finally, cell sizes were small for the comparisons between the SF-36 and the PPT, and when data were stratified by diagnosis and sex. This increased the variability of our estimates and made it difficult to draw conclusions about measured versus self-reported physical performance limitations in specific subgroups.
Nevertheless, our analysis provides preliminary information that indicates that self-report, although not perfect, accurately identifies individuals without physical performance limitations. In contrast, self-report is less reliable for identifying individuals with physical performance limitations with only a modest level of sensitivity. Interpretation of self-report data regarding physical performance should take into account the potential for misclassification. Our results show that using self-report misclassified some survivors of childhood cancer as having a physical performance limitation when one is not detected with clinical performance measures, which likely inflates overall prevalence estimates of this outcome in the CCS population. Our results also identify survivors whose self-report data may be less optimistic about their performance than is their actual physical performance when evaluated with a clinical tool.
This work was supported in part by the Cancer Center Support (CORE; grant no. CA 21765), the National Cancer Institute (grant no. 5R25CA023940), the National Cancer Institute, and the American Lebanese Syrian Associated Charities (ALSAC).
The authors have no conflicts of interest.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Altman DG, Bland JM. Statistics notes: diagnostic tests 1: sensitivity and specificity. BMJ
. 1994; 308 (6943): 1552.
2. Alvarez JA, Scully RE, Miller TL, et al. Long-term effects of treatments for childhood cancers. Curr Opin Pediatr
. 2007; 19 (1): 23–31.
3. Bohannon RW, DePasquale L. Physical Functioning Scale of the Short-Form (SF) 36: internal consistency and validity with older adults. J Geriatr Phys Ther
. 2010; 33 (1): 16–8.
4. Bowling A. Mode of questionnaire administration can have serious effects on data quality. J Public Health (Oxf)
. 2005; 27 (3): 281–91.
5. Brown M, Sinacore DR, Binder EF, Kohrt WM. Physical and performance measures for the identification of mild to moderate frailty. J Gerontol A Biol Sci Med Sci
. 2000; 55 (6): M350–5.
6. Butland RJ, Pang J, Gross ER, Woodcock AA, Geddes DM. Two-, six-, and 12-minute walking tests in respiratory disease. BMJ
. 1982; 284 (6329): 1607–8.
7. Cote CG, Casanova C, Marín JM, et al. Validation and comparison of reference equations for the 6-min walk distance test. European Respir J
. 2008; 31 (3): 571–8.
8. Delbaere K, Van den Noortgate N, Bourgois J, Vanderstraeten G, Tine W, Cambier D. The Physical Performance Test as a predictor of frequent fallers: a prospective community-based cohort study. Clin Rehabil
. 2006; 20 (1): 83–90.
9. Enright PL, Sherrill DL. Reference equations for the six-minute walk in healthy adults. Am J Respir Crit Care Med
. 1998; 158 (5 pt 1): 1384–7.
10. Ferrer M, Lamarca R, Orfila F, Alonso J. Comparison of performance-based and self-rated functional capacity in Spanish elderly. Am J Epidemiol
. 1999; 149 (3): 228–35.
11. Gulmans VA, van Veldhoven NH, de Meer K, Helders PJ. The six-minute walking test in children with cystic fibrosis: reliability and validity. Pediatr Pulmonol
. 1996; 22 (2): 85–9.
12. Hayes KW, Johnson ME. Measures of adult general performance tests: the Berg Balance Scale, Dynamic Gait Index (DGI), Gait Velocity, Physical Performance Test (PPT), Timed Chair Stand Test, Timed Up and Go, and Tinetti Performance-Oriented Mobility Assessment (POMA). Arthritis Care Res (Hoboken)
. 2003; 49 (S5): S28–42.
13. Hebert R, Raiche M, Gueye NR. Survey disability questionnaire does not generate valid accurate data compared to clinical assessment on an older population. Arch Gerontol Geriatr
. 2012; 54 (2): e57–62.
14. Howlader N, Noone AM, Krapcho M, Neyman N, Aminou R, Altekruse SF, Kosary CL, Ruhl J, Tatalovich Z, Cho H, Mariotto A, Eisner MP, Lewis DR, Chen HS, Feuer EJ, Cronin KA editors. SEER Cancer Statistics Review, 1975–2009 (Vintage 2009 Populations)
15. Jay SJ. Reference equations for the six-minute walk in healthy adults. Am J Respir Crit Care Med
. 2000; 161 (4 pt 1): 1396.
16. Jette AM. How measurement techniques influence estimates of disability in older populations. Soc Sci Med
. 1994; 38 (7): 937–42.
17. Kasymjanova G, Correa JA, Kreisman H, et al. Prognostic value of the six-minute walk in advanced non-small cell lung cancer. J Thorac Oncol
. 2009; 4 (5): 602–7.
18. Kelly-Hayes M, Jette AM, Wolf PA, D’Agostino RB, Odell PM. Functional limitations and disability among elders in the Framingham Study. Am J Public Health
. 1992; 82 (6): 841–5.
19. Merrill SS, Seeman TE, Kasl SV, Berkman LF. Gender differences in the comparison of self-reported disability and performance measures. J Gerontol A Biol Sci Med Sci
. 1997; 52 (1): M19–26.
20. Ness KK, Gurney JG, Zeltzer LK, et al. The impact of limitations in physical, executive, and emotional function on health-related quality of life among adult survivors of childhood cancer: a report from the Childhood Cancer Survivor Study. Arch Phys Med Rehabil
. 2008; 89 (1): 128–36.
21. Ness KK, Hudson MM, Ginsberg JP, et al. Physical performance limitations in the Childhood Cancer Survivor Study cohort. J Clin Oncol
. 2009; 27 (14): 2382–9.
22. Ness KK, Mertens AC, Hudson MM, et al. Limitations on physical performance and daily activities among long-term survivors of childhood cancer. Ann Intern Med
. 2005; 143 (9): 639–47.
23. Ness KK, Wall MM, Oakes JM, Robison LL, Gurney JG. Physical performance limitations and participation restrictions among cancer survivors: a population-based study. Ann Epidemiol
. 2006; 16 (3): 197–205.
24. Oeffinger KC, Mertens AC, Sklar CA, et al. Chronic health conditions in adult survivors of childhood cancer. N Engl J Med
. 2006; 355 (15): 1572–82.
25. Ojha RP, Oancea SC, Ness KK, et al. Assessment of potential bias from non-participation in a dynamic clinical cohort of long-term childhood cancer survivors: results from the St. Jude Lifetime cohort study. Pediatr Blood Cancer
. 2012; 60 (5): 856–64.
26. Pinar R. Reliability and construct validity of the SF-36 in Turkish cancer patients. Qual Life Res
. 2005; 14 (1): 259–64.
27. Reuben DB, Siu AL. An objective measure of physical function of elderly outpatients. The Physical Performance Test. J Am Geriatr Soc
. 1990; 38 (10): 1105–12.
28. Reulen RC, Winter DL, Lancashire ER, et al. Health-status of adult survivors of childhood cancer: a large-scale population-based study from the British Childhood Cancer Survivor Study. Int J Cancer
. 2007; 121 (3): 633–40.
29. Reulen RC, Zeegers MP, Jenkinson C, et al. The use of the SF-36 questionnaire in adult survivors of childhood cancer: evaluation of data quality, score reliability, and scaling assumptions. Health Qual Life Outcomes
. 2006; 4: 77.
30. Ries L, Melbert D, Krapcho M, et al. SEER Cancer Statistics Review, 1975–2005
. Bethesda, MD: U.S. National Institutes of Health, National Cancer Institute; 2008.
31. Rostagno C, Olivo G, Comeglio M, et al. Prognostic value of 6-minute walk corridor test in patients with mild to moderate heart failure: comparison with other methods of functional evaluation. Eur J Heart Fail
. 2003; 5 (3): 247–52.
32. Rothwell PM, McDowell Z, Wong CK, Dorman PJ. Doctors and patients don’t agree: cross sectional study of patients’ and doctors’ perceptions and assessments of disability in multiple sclerosis. BMJ
. 1997; 314 (7094): 1580.
33. Rueegg CS, Michel G, Wengenroth L, von der Weid NX, Bergstraesser E, Kuehni CE. Physical performance limitations in adolescent and adult survivors of childhood cancer and their siblings. PLoS One
. 2012; 7 (10): e47944.
34. Schmitz KH, Holtzman J, Courneya KS, Masse LC, Duval S, Kane R. Controlled physical activity trials in cancer survivors: a systematic review and meta-analysis. In: Cancer Epidemiol Biomarkers Prev
. 2005; 14 (7): 1588–95.
35. Siegel S, Castellan NJ. Nonparametric Statistics for the Behavioral Sciences
. New York: McGraw-Hill; 1988. p. 311.
36. Walsh EG, Khatutsky G. Mode of administration effects on disability measures in a sample of frail beneficiaries. Gerontologist
. 2007; 47 (6): 838–44.
37. Ware JE. SF-36 Physical and Mental Health Summary Scales : A User’s Manual
. Boston: The Health Institute, New England Medical Center; 1994. p. 188.
38. Ware JE. SF-36 health survey : manual and interpretation guide. In: Lincoln RI, editor. QualityMetric
. Boston (MA): Health Assessment Lab; 2000. p. 238.
39. Ware JEJ. SF-36 Health Survey Update. Spine
. 2000; 25 (24): 3130–9.
40. Zugck C, Krüger C, Dürr S, et al. Is the 6-Minute Walk Test a reliable substitute for peak oxygen uptake in patients with dilated cardiomyopathy? Eur Heart J
. 2000; 21 (7): 540–9.