Validation of Single-Item Linear Analog Scale Assessment ofQuality of Life in Neuro-Oncology Patients

Dona E.C. Locke, PhD, Paul A. Decker, MS, Jeff A. Sloan, PhD, Paul D. Brown, MD, James F.Malec, PhD, Matthew M. Clark, PhD, Teresa A. Rummans, MD, Karla V. Ballman, PhD, Paul L.Schaefer, MD, and Jan C. Buckner, MDDepartment of Psychiatry and Psychology (D.E.C.L.), Mayo Clinic, Scottsdale, Arizona; Division ofBiostatistics (P.A.D., J.A.S., K.V.B.), Department of Radiation Oncology (P.D.B.), Department ofPsychiatry and Psychology (J.F.M., M.M.C., T.A.R.), and Division of Medical Oncology (J.C.B),Mayo Clinic, Rochester, Minnesota; and Toledo Community Hospital (P.L.S.), Toledo, Ohio, USA

AbstractPatient quality-of-life (QOL) assessment requires balancing the detail provided by multi-itemassessments with the reduced burden of single-item assessments. In this project, we investigated thepsychometric properties of single-item linear analog scale assessments (LASAs) for patients withnewly diagnosed high-grade gliomas. Measures included QOL LASAs (overall, physical, emotional,spiritual, intellectual), Symptom Distress Scale (SDS), Profile of Mood States (POMS; overall,confusion, fatigue), and Functional Assessment for Cancer Therapy-Brain (FACT-Br; overall, brain,physical, emotional). Associations of LASA measures with SDS, POMS, and FACT-Br domains andwith Eastern Cooperative Oncology Group performance score (PS) and Mini-Mental StateExamination (MMSE) were assessed. Repeated measures ANOVA models compared the changeover time of LASAs and SDS, POMS, and FACT-Br. Two hundred five patients completed theassessments across three time points. To allow comparison across measures, all scores wereconverted to a scale of 0–100 with higher scores indicating better QOL. LASA mean scores rangedfrom 60–78; SDS, POMS, and FACT-Br ranged from 62–81. FACT-Br physical (P<0.001) andPOMS fatigue subscale (P=0.005) decreased over time, as did LASA physical (P=0.08). LASA scaleswere strongly associated with corresponding scales on SDS, POMS, and FACT-Br (0.44<rho<0.65;P<0.001). LASA was negatively associated with PS and positively with MMSE, with associationssimilar in magnitude to the other QOL and psychosocial measures. The data suggest that the single-item LASA scales are valid for assessing QOL of cancer patients and are an appropriate alternativewhen a shorter instrument is warranted.

KeywordsQuality of life; measurement; LASA; validation; neuro-oncology

© 2007 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved.Address reprint requests to: Dona E.C. Locke, PhD, Division of Psychology, Mayo Clinic Arizona, 13400 East Shea Boulevard,Scottsdale, AZ 85259, USA, locke.dona@mayo.edu.Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customerswe are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resultingproof before it is published in its final citable form. Please note that during the production process errors may be discovered which couldaffect the content, and all legal disclaimers that apply to the journal pertain.The authors have no conflict of interest or financial involvement with this manuscript.These data were presented as a poster at the 42nd Annual American Society of Clinical Oncology Conference, Atlanta, GA, June 2–6,2006.

NIH Public AccessAuthor ManuscriptJ Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

Published in final edited form as:J Pain Symptom Manage. 2007 December ; 34(6): 628–638. doi:10.1016/j.jpainsymman.2007.01.016.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

IntroductionQuality-of-life issues for cancer patients and survivors have received much recent empiricalinvestigation.1 In neuro-oncology, the description and measurement of quality of life (QOL)continues to evolve. QOL is defined in broad terms and thought to encompass several differentaspects of life, including emotional functioning, physical functioning, cognitive functioning,social functioning, and spiritual well-being.2 There have been a large number of multi-iteminstruments developed to capture the cancer patient’s self-report of QOL,3–5 such as theFunctional Assessment of Cancer Therapy (FACT-G) 6 or the European Organization forResearch and Treatment of Cancer (EORTC).7 Many studies have shown that several of thesemultiple-item inventories with resulting summary scores are helpful in measuring QOL inadults with many different types of cancers.8 Multi-item assessments can provide great detailabout and individual’s QOL and have high internal consistency.9

However, multiple-item instruments can be lengthy and, therefore, time consuming for patientsto complete, and this time burden may result in poor completion rates, especially in brain tumorpatients.10 In addition, cancer patients experience many symptoms related to their cancer andits treatment, such as nausea and fatigue. Thus, patient burden is an important consideration.If proven to be psychometrically sound and to provide helpful information for the practitioner,simpler measurements such as single-item scales could help reduce patient burden and yieldimportant QOL information. The rational for single-item assessments has been maderepeatedly across a broad spectrum of applications.11–14

Sloan and colleagues9,15 have proposed measuring QOL with single items or with briefmultiple-item scales. Their basic tenet is that the research question involved will determinewhether a single-item assessment will suffice or if a more detailed assessment will be required.In their review of published literature, they conclude that both single-item and multi-item indexscores provide clinically important information and are statistically robust. Sloan andcolleagues9 suggest that single items may be sufficient when only a global impression of QOLis needed, when screening is desired to determine if a patient needs more in-depth assessment,or in a Phase II study attempting to assess whether a treatment has any impact on QOL.

There has been significant interest in developing and validating simple, single-item measuresof aspects of QOL for use in appropriate situations, as described by Sloan and colleagues.9,15 Linear analog self-assessment scales have been utilized as an assessment format in manymultiple-item questionnaires.3,16–20 Essentially, as described by Cella,3 linear analog self-assessment scales are a type of assessment scale in which 100 mm lines are used. Anchoringdescriptors are placed at each end of the line and patients mark their current state on thatsymptom somewhere on that line. The score is the measured number of millimeters from thezero end of the scale to the patient’s indication of where they fall.

The development effort culminated in the Linear Analog Scale assessment (LASA), whichconsists of five single items asking respondents to rate, on zero to ten scales, their perceivedlevel of functioning (see Appendix A);21,22 this LASA approach has since been utilized inmultiple trials.18,23,24 As mentioned above, the definition of QOL and its domains has evolvedover time, but at the time of the initial LASA development and data collection, QOL wasconceptualized as a multidimensional construct with four domains (physical, functional,emotional, and social). We utilized that definition,2 as well as the experience of ourmultidisciplinary team of treating professionals, including oncologists, physiatrists,psychiatrists, psychologists, physical therapists, and chaplains, to choose the domains includedin the LASA. With this screening measure, patients are empowered to communicate with theirphysicians about their QOL overall and across four major life domains that are most distressingfrom their point of view. If necessary, these symptoms can then be evaluated in more detail.

Locke et al. Page 2

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

While the LASA has been widely used in clinical studies, its psychometric properties have notbeen explored extensively. The goal of this paper is to present reliability and validity data forthe LASA utilizing data obtained in neuro-oncology patients.

Patients and MethodsSample

The participants included in this analysis are patients from the Glioma QOL study, which wasa prospective companion protocol for three Institutional Review Board–approved NorthCentral Cancer Treatment Group (NCCTG) high-grade glioma treatment protocols. Thesetreatment protocols and patient treatment and medical history have been previously describedin detail23,24 and will only be described here briefly. The prospective treatment trials involveda total of 220 newly diagnosed high-grade glioma patients treated with chemotherapy andradiotherapy. Patients were deemed evaluable if they were alive without tumor progression.

MeasuresComplete details of the study measures are included in the reports by Brown andcolleagues23,24 and will not be repeated in detail here. Rather, we will provide brief summariesof the measures and their intended assessment constructs.

In these protocols, the LASA, along with one additional QOL measure and two psychosocialmeasures (see below), was completed by the patient (after receiving instructions on how tocomplete the forms) within 72 hours of study entry (baseline), at two months, and at fourmonths. At baseline, at two months, and at four months, patients also underwent assessmentwith the Folstein Mini-Mental State Examination (MMSE).25 The Eastern CooperativeOncology Group (ECOG) performance score (PS), a simple, widely administered measure ofphysical function, was also graded by the health care provider at baseline, at two months, andat four months. Collectively, patients completed the QOL measures, psychosocial measures,and the MMSE in approximately 30 minutes.

Linear Analog Scale Assessment (LASA) includes five simple items, each of which targets aspecific domain of quality of life (see Appendix A).21 Specific domains include physical well-being (i.e., fatigue, activity level), emotional well-being (i.e., depression, anxiety, stress),spiritual well-being (i.e., sense of meaning, relationship with God), and intellectual well-being(i.e., ability to think clearly, concentrate). An item for overall QOL is also included. The Likertscales run from 0 (as bad as it can be) to 10 (as good as it can be). Thus, higher ratings suggesthigher QOL.

Functional Assessment of Cancer Therapy-Brain (FACT-Br)6,26 is a self-administered 46-itemmeasure, with 27 general items to assess physical, social/family, emotional, and functionalwell-being. The brain tumor specific version includes an additional subscale of 19 brain tumorspecific QOL items. The scores on individual items run from 0 (not at all) to 4 (very much).Items are totaled to produce the following subscales, along with an overall QOL score: physicalwell-being, social/family well-being, emotional well-being, functional well-being, and braintumor specific scale. For validation purposes, we included the overall QOL score, the physicalwell-being subscale, the emotional well-being subscales, and the brain tumor specificsymptoms subscale, as those are the scales that have corresponding items on the LASA.Overall, higher ratings suggest higher QOL.

A recent study suggests that the FACT-G has a large effect size in differentiating patients withvarious ECOG performance scores.4 Short-term test-retest reliability is strong (0.92).6 TheFACT-G Overall score also has been shown to be appropriately correlated with other measuresof global QOL.4,6

Locke et al. Page 3

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Profile of Mood States-Short Form (POMS-SF)27 includes a self-administered list of 30adjectives that the patient rates on a scale from 0 (not at all) to 4 (extremely). Subscales includetension/anxiety, vigor/activity, fatigue/inertia, depression/dejection, confusion/bewilderment,and anger/hostility. For LASA validation purposes, we utilized the overall score, the fatigue/inertia subscale, and the confusion/bewilderment subscale. Higher scores indicate less distressor fewer symptoms of each scale.

Validation information27 shows that the POMS-SF has an acceptable internal consistency forthe total score (0.91). The POMS-SF total score and subscales correlate appropriately withother psychosocial measures providing both convergent and discriminant validity.27

Symptom Distress Scale (SDS) is a self-administered 13-item questionnaire developed tomeasure physical concerns in cancer patients.28 This measure is frequently utilized in oncologytrials.29,30

Statistical MethodsThe LASA, along with the FACT-Br, POMS, and SDS, was completed by subjects at baselineand at two and four months after enrollment per the dictates of the larger treatment protocols.The raw summative scores from each QOL or psychosocial measure were transformed foranalysis. For a given raw score, the transformed score was a value between 0 and 100 inclusive;this represented the relative position of the raw score along the theoretical dimension for thescore.22,31,32 In all cases, 0 reflected the worst QOL and 100, the best. By using the transformedscores, we were able to compare scores designed to evaluate the same construct as well ascompare relative scores designed for different QOL domains.

The following techniques were used to provide information supportive of the validity of theLASAs. Because of the ordinal nature of the data, Spearman’s correlation coefficients wereused to measure the associations of each LASA with the corresponding scales on the FACT-Br, POMS, and SDS. Repeated measures analysis of variance models were fit for each LASAand the corresponding scales on the FACT-Br, POMS, and SDS to determine whether thechange in LASA over time was similar to that of the other QOL and psychosocial scales. Forthese models, the scale was the dependent variable and measurement time point was theindependent variable. These models were fit with an autoregressive correlation (AR1)structure. Finally, the association of the LASA and each QOL or psychosocial scale with ECOGperformance score and MMSE score was assessed using Spearman’s correlation coefficient.In all cases, P-values ≤0.05 were considered statistically significant. Correlations <0.3 wereconsidered weak, correlations from 0.3–0.5 were considered moderate, and correlations >0.5were considered strong.33

ResultsOverall, there were 205 participants enrolled in the three clinical trials who completed theLASA at a minimum of one of the three time points: baseline, two months, and four months.One hundred ninety-four subjects (88% of evaluable patients) completed the LASA at baseline,127 (71% of evaluable patients) completed the LASA at the two-month follow-up, and 82 (66%of evaluable patients) completed the LASA at the four-month follow-up. There were nosignificant differences in completion rates among QOL and psychosocial measures across thethree studies (P > 0.10). Baseline demographics, tumor information, and treatment informationfor the 205 participants are presented in Table 1. The median age was 56 years and the medianbaseline MMSE score was 29. The vast majority of participants were ECOG performance score0 or 1 (89%). The median overall survival was 10.2 months and the median progression freesurvival was 4.7 months

Locke et al. Page 4

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

A typical reliability calculation when validating new questionnaires is internal consistency orCronbach’s alpha. However, this coefficient is typically provided when the individual itemsare summed together to provide a total score on the measure. This is not the purpose of theLASA, as each item is intended to stand alone. However, Cronbach’s alpha is a measure ofhow much an individual’s response to one item is related to his/her responses on all the otheritems. As the LASAs are theorized to measure different facets of QOL, one might theorize thatthe LASAs could be related to one another, though perhaps not perfectly so. Thus, we presentCronbach’s alpha values for each time point despite the fact that the five items are not summedtogether to form a composite score. Cronbach's alpha at time 0 was 0.83; at month two, it was0.88; and at month four, it was 0.88. We do not provide test-retest reliability, as QOL couldbe expected to be variable across time. However, we present below the change in each QOLscale over time.

Table 2 summarizes the mean, standard deviation, median, and range on the LASA, SDS,POMS, and FACT-Br across the three time points. There was no significant change across timefor the overall score or emotional, spiritual, and intellectual scores for the LASA, the SDS, theoverall and confusion scores for the POMS, and the overall score, emotional score, and brainscore of the FACT-Br. The FACT-Br physical score (P<0.001) and the POMS-fatigue score(P=0.005) significantly decreased over time (P<0.001), while the LASA physical scoredisplayed a decreasing trend over time but did not achieve statistical significance (P=0.083).

To supplement the above repeated measures analysis, the change from baseline to the two-month follow-up was calculated for each scale. All correlation coefficients were significantlydifferent than 0 (P<0.001 in each case except where noted). The associations of the changefrom baseline for each scale with the LASA change from baseline were as follows: ΔLASAoverall with ΔSDS overall (r=0.41), ΔPOMS overall (r=0.43), and ΔFACT-Br overall (r=0.41);ΔLASA emotional with ΔFACT-Br emotional (r=0.23, P =0.018) and ΔPOMS overall(r=0.36); ΔLASA intellectual with ΔFACT-Br brain (r=0.26, P=0.006) and ΔPOMS confusion(r=0.22, P=0.026); and ΔLASA physical with ΔFACT-Br physical (r=0.42) and ΔPOMSfatigue (r=0.48). Additionally, the change from two months to four months was calculated foreach scale and the associations were as follows: ΔLASA overall with ΔSDS overall (r=0.37,P=0.001), ΔPOMS overall (r=0.52), and ΔFACT-Br overall (r=0.42); ΔLASA emotional withΔFACT-Br emotional (r=0.32, P=0.006) and ΔPOMS overall (r=0.32, P=0.009); ΔLASAintellectual with ΔFACT-Br brain (r=0.30, P=0.011) and ΔPOMS confusion (r=0.58); andΔLASA physical with ΔFACT-Br physical (r=0.45) and ΔPOMS fatigue (r=0.50).

The association of the LASA scales with each corresponding scale on the SDS, POMS, andFACT-Br at baseline, two months, and four months is summarized in Table 3. LASA scaleswere significantly associated with the corresponding scales on the SDS, POMS, and FACT-Br (P<0.001 in each case). All correlations were moderate to strong (r=0.50–0.65) except forthe correlation between the LASA emotional scale and the FACT-Br emotional scale at cycle0 (r=0.44), which was moderate in strength.

The association of the QOL and psychosocial scales with the ECOG performance score at eachtime point is displayed in Table 4. At baseline, the associations between the ECOG performancescore and the LASA scales were similar to associations between the other measures and theECOG performance score (i.e., weak negative correlations). The correlations for each scalewith the ECOG performance score were small and were as follows: LASA overall (r= −0.25),SDS (r= −0.17), POMS overall (r= −0.24), and FACT-Br overall (r= −0.29). The associationbetween the physical scale on the LASA and the ECOG performance score was moderate (r=−0.36) and similar to the correlation of the physical scale on the FACT-Br (r=−0.32) and thePOMS-fatigue (r=−0.31) with the ECOG performance score at baseline. However, theassociation of the ECOG performance score with the LASA emotional scale (r= −0.20) was

Locke et al. Page 5

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

larger in magnitude than the FACT-Br emotional scale (r= −0.07). The association of the LASAintellectual and the ECOG performance score was weak (r=−0.16) as was the correlation ofthe POMS-confusion with ECOG performance score (r=−0.13); however, the association ofthe FACT-Br brain with ECOG performance score was larger in magnitude (r=−0.33). Similarfindings were observed at two months and four months, with the exception that at two monthsPOMS-confusion (r=−0.08) and FACT-Br emotional (r=−0.10) and at four months the LASAemotional (r= −0.0004), SDS overall (r= −0.05), POMS confusion (r=−0.05), and FACT-Bremotional (r= −0.04) had no significant association with the ECOG performance score.

Finally, the correlation of each QOL and psychosocial scale with the MMSE score at baselinewas investigated (Table 5). Each scale was positively, but weakly, associated with the MMSEscore. The associations of the LASA scales with the MMSE were similar in magnitude to theassociation of the other QOL and psychosocial scales with the MMSE. All correlationcoefficients were significantly different than 0 (P<0.014 in each case), with the exception ofthe FACT-Br emotional scale (P=0.094), POMS-fatigue (P=0.088), and LASA spiritual scale(P=0.111).

DiscussionOur aim was to provide initial psychometric information for a five-item self-report measure ofQOL in cancer patients. While there are several existing well validated and reliable, but muchlonger, measures of QOL in cancer patients, there has been a scientific call9 in support of thedevelopment of single-item assessment measures, such as the LASA. A validated, brief QOLmeasure, such as the LASA, would be advantageous because it would reduce patient burden,both in clinical situations and in clinical trials, and have more clinical utility for the busypractitioner. The results of this study provide initial evidence suggesting that the LASA maybe as valid as lengthier measures of QOL in neuro-oncology patients.

Our foremost finding was that the summary statistics suggest that the LASA items haveadequate variability within them to be clinically meaningful. That is, patients score along theentire spectrum of possible scores on the scales, avoiding restriction of range, ceiling or flooreffects, and providing information that is meaningful and can be different across individualpatients and across time. These results are consistent with findings from other studies.9,32,34,35

Second, we found that the LASA items correlated strongly with similar scales on other, well-validated, multi-item scales that are frequently used in cancer patients, thereby supporting theconcurrent validity and construct validity of the LASA. Specifically, the Overall QOL LASAitem correlates strongly with the overall score on the FACT-Br, the Overall SDS score, andthe Overall POMS score. The LASA physical item correlates strongly with the FACT-Brphysical scale and with the Fatigue subscale of the POMS. The LASA emotional item correlatesstrongly with the FACT-Br emotional scale and the overall POMS. The LASA intellectualitem correlates strongly with the FACT-Br brain subscale and the confusion subscale of thePOMS. In terms of sensitivity to change, we also found that the LASA items change over timein a manner similar to the longer QOL and psychosocial scales.

Third, we found that the LASA items were negatively correlated with ECOG performancestatus and were positively correlated with MMSE scores at baseline. The direction of theserelationships is as expected and supports the construct validity of the LASA items. Themagnitude of all these relationships is weak, which is also expected given the constructdifference and, therefore, provides divergent validity for the items. For example, theperformance status rating is determined by the provider and is heavily influenced by thepatient’s physical functioning. The LASA is rated by the patient and is considered a broader

Locke et al. Page 6

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

assessment of overall QOL. In addition, the correlations between the ECOG performance statusand MMSE and the other QOL and psychosocial scales are similarly weak.

This study does have several limitations. We only utilized neuro-oncology patients, so theseresults may not apply to other cancer populations. However, given that the measures used forvalidation purposes in this study are also used broadly with patients with many types of cancer,we might expect that the LASA would be similarly useful across populations. Also, though noformal validation trial of the LASA with other cancer patients has been published, the LASAhas been used in other trials, with concurrent validation data published alongside the individualtrial results.36–44 Further work also has been done using the LASA across a wide spectrum ofpatient populations45–49 and healthy populations.50 Secondly, we assessed QOL at three timepoints across four months, which is appropriate for patients, such as high-grade glioma patients,with a relatively short median survival. Perhaps longer follow-up, such as twelve to thirty-sixmonths is needed to fully validate these scales in patients with other types of cancer with amuch longer median survival or for whom cure is expected.

In summary, the LASA is a psychometrically-validated brief measure of QOL. Despite itsbrevity, it provides an overall estimate and covers four major subcomponents of QOL (physical,emotional, spiritual, and intellectual). It therefore gives a global, comprehensive view of apatient’s QOL. We do not propose that the LASA items should replace longer, well-validatedmeasures of QOL. When a more detailed assessment of QOL is desired or QOL is consideredthe primary outcome, either clinically or for research purposes, a longer measure may remaina clinician’s or researcher’s best choice. However, our results do suggest that the LASA is apsychometrically-validated brief measure of QOL that may be useful as a screening measure,when only a broad view of QOL is desired or when time constraints require a very briefassessment.

AcknowledgmentsThis study was conducted as a collaborative trial of the North Central Cancer Treatment Group and Mayo Clinic andwas supported in part by Public Health Service grants CA 25224, CA 37404, CA 15083, CA 35415, and the LinseBock Foundation.

References1. Sloan J, Frost M, Berzon R, et al. The clinical significance of quality of life assessments in oncology:

a summary for clinicians. Support Care Cancer 2006;14(10):988–998. [PubMed: 16794811]2. Rummans T, Bostwick M, Clark M. Maintaining quality of life at the end of life. Mayo Clin Proc

2000;75:1305–1310. [PubMed: 11126840]3. Cella D. Methods and problems in measuring quality of life. Support Care Cancer 1995;3:11–22.

[PubMed: 7697298]4. Cheung Y, Goh C, Thumboo J, et al. Variability and sample size requirements of quality of life

measures: a randomized study of three major questionnaires. J Clin Oncol 2005;23:4936–4944.[PubMed: 16051946]

5. Osoba D. A taxonomy of the uses of health-related quality-of-life instruments in cancer care and theclinical meaningfulness of the results. Med Care 2002;40(6 Suppl):III31–III38. [PubMed: 12064756]

6. Cella D, Tulsky D, Gray G, et al. The functional assessment of cancer therapy (FACT) scale:development and validation of the general measure. J Clin Oncol 1993;11:570–579. [PubMed:8445433]

7. de Haes J, Curran D, Young T, et al. Quality of life evaluation in oncological clinical trials -the EORTCmodel.The EORTC Quality of Life Study Group. Eur J Cancer 2000;36:821–825. [PubMed:10785585]

Locke et al. Page 7

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

8. Coates A, Porzsolt F, Osoba D. Quality of life in oncology practice: prognostic value of EORTC QLQ-C30 scores in patients with advanced malignancy. Eur J Cancer 1997;33:1025–1030. [PubMed:9376182]

9. Sloan J, Aaronson N, Cappelleri J, et al. Clinical significance consensus meeting group.Assessing theclinical significance of single items relative to summated scores. Mayo Clin Proc 2002;77:479–487.[PubMed: 12004998]

10. Roa W, Brasher P, Bauman G, et al. Abbreviated course of radiation therapy in older patients withglioblastoma multiforme: a prospective randomized clinical trial. J Clin Oncol 2004;22:1583–1588.[PubMed: 15051755]

11. Bernhard J, Sullivan M, Hurny C, et al. Clinical relevance of single item quality of life indicators incancer clinical trials. Br J Cancer 2001;84:1156–1165. [PubMed: 11336464]

12. Vickers A, McCarney R. Use of a single global assessment to reduce missing data in a clinical trialwith follow-up at one year. Control Clin Trials 2003;24:731–735. [PubMed: 14662279]

13. Abdel-Khalek A. Single-virus multi-item scales in measuring death anxiety. Death Stud 1998;22:763–772. [PubMed: 10346701]

14. Sutherland H, Lockwood G, Cunningham A. A simple, rapid method for assessing psychologicaldistress in cancer patients: evidence of validity for linear analog scales. J Psychosoc Oncol1989;7:31–43.

15. Sloan J, Cella D, Frost M, et al. Assessing clinical significance in measuring oncology patient qualityof life: introduction to the symposium, content overview, and definition of terms. Mayo Clin Proc2002;77:367–370. [PubMed: 11936934]

16. Brown P, Clark M, Atherton P, et al. Will improvement in quality of life (QOL) impact fatigue inpatients receiving radiation therapy for advanced cancer? Am J Clin Oncol 2006;29:52–58. [PubMed:16462503]

17. McCormack H, Horne D, Sheather S. Clinical applications of visual analogue scales: a critical review.Psychol Med 1988;18:1007–1019. [PubMed: 3078045]

18. Rummans T, Clark M, Sloan J, et al. Impacting quality of life for patients with advanced cancer witha structured multidisciplinary intervention: a randomized controlled trial. J Clin Oncol 2006;24:635–642. [PubMed: 16446335]

19. Wewers M, Lowe N. A critical review of visual analogue scales in the measurement of clinicalphenomena. Res Nurs Health 1990;13:227–236. [PubMed: 2197679]

20. Bruera E, Kuehn N, Miller M, et al. The Edmonton Symptom Assessment System (ESAS): a simplemethod for the assessment of palliative care patients. J Palliat Care 1991;7:6–9. [PubMed: 1714502]

21. Bretscher M, Rummans T, Sloan J, et al. Quality of life in hospice patients.A pilot study.Psychosomatics 1999;40:309–313. [PubMed: 10402876]

22. Sloan J, Loprinzi C, Kuross S, et al. Randomized comparison of four tools measuring overall qualityof life in patients with advanced cancer. J Clin Oncol 1998;16:3662–3673. [PubMed: 9817289]

23. Brown P, Maurer M, Rummans T, et al. A prospective study of quality of life in adults with newlydiagnosed high-grade gliomas: the impact of the extent of resection on quality of life and survival.Neurosurgery 2005;57:495–503. [PubMed: 16145528]

24. Brown P, Ballman K, Rummans T, et al. Prospective study of quality of life in adults with newlydiagnosed high-grade gliomas. J Neurooncol 2006;76:283–291. [PubMed: 16163448]

25. Folstein M, Folstein S, McHugh P. “Mini-mental state.” A practical method for grading the cognitivestate of patients for the clinician. J Psychiatr Res 1975;12:189–198. [PubMed: 1202204]

26. Weitzner M, Meyers C, Gelke C, et al. The functional assessment of cancer therapy (FACT) scale.Development of a brain subscale and revalidation of the general version (FACT-G) in patients withprimary brain tumors. Cancer 1995;75:1151–1161. [PubMed: 7850714]

27. Baker F, Denniston M, Zabora J, et al. A POMS short form for cancer patients: psychometric andstructural evaluation. Psychooncology 2002;11:273–281. [PubMed: 12203741]

28. McCorkle R, Benoliel J. Symptom distress, current concerns, and mood disturbance after diagnosisof life-threatening disease. Soc Sci Med 1983;17:431–438. [PubMed: 6867788]

29. Degner L, Sloan J. Symptom distress in newly diagnosed ambulatory cancer patients and as a predictorof survival in lung cancer. J Pain Symptom Manage 1995;10:423–431. [PubMed: 7561224]

Locke et al. Page 8

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

30. Goldberg R, Sargent D, Morton R, et al. A randomized controlled trial of fluorouracil plus leucovorin,irinotecan, and oxaliplatin combinations in patients with previously untreated metastatic colorectalcancer. J Clin Oncol 2004;22:23–30. [PubMed: 14665611]

31. Osoba D, Rodrigues G, Myles J, et al. Interpreting the significance of changes in health-related quality-of-life scores. J Clin Oncol 1998;16:139–144. [PubMed: 9440735]

32. Sloan J, Dueck A. Issues for statisticians in conducting analyses and translating results for quality oflife end points in clinical trials. J Biopharm Stat 2004;14:73–96. [PubMed: 15027501]

33. Cohen, J. Statistical power analysis for the behavioral sciences. Vol. 2nd ed. Hillsdale, NJ: LawrenceErlbaum Associates; 1988.

34. Sloan J, Loprinzi C, Laurine J, et al. A simple stratification factor prognostic for survival in advancedcancer: the good/bad/uncertain index. J Clin Oncol 2001;19:3539–3546. [PubMed: 11481361]

35. Buchanan D, O'Mara A, Kelaghan J, et al. Quality-of-life assessment in the symptom managementtrials of the National Cancer Institute-supported Community Clinical Oncology Program. J ClinOncol 2005;23:591–598. [PubMed: 15659506]

36. Loprinzi CL, Levitt R, Barton DL, et al. Evaluation of shark cartilage in patients with advanced cancer:An NCCTG trial. Cancer 2005;104:176–182. [PubMed: 15912493]

37. Giordano K, Jatoi A, Stella P, et al. Docetaxel and capecitabine in patients with metastaticadenocarcinoma of the stomach and gastroesophageal junction: a phase II study from the NorthCentral Cancer Treatment Group. Ann Oncol 2006;17:652–656. [PubMed: 16497828]

38. Crawford J, Cella D, Cleeland C, et al. Relationship between changes in hemoglobin level and qualityof life during chemotherapy in anemic cancer patients receiving epoetin alfa therapy. Cancer2002;95:888–895. [PubMed: 12209734]

39. Shasha D, George M, Harrison L. Once-weekly dosing of epoetin-alpha increases hemoglobin andimproves quality of life in anemic cancer patients receiving radiation therapy either concomitantlyor sequentially with chemotherapy. Cancer 2003;98:1072–1079. [PubMed: 12942577]

40. Herr H. Quality of life in prostate cancer patients. CA Cancer J Clin 1997;47:207–217. [PubMed:9242169]

41. Witzig T, Silberstein P, Loprinzi C, et al. Phase III, randomized, double-blind study of epoetin alfacompared with placebo in anemic patients receiving chemotherapy. J Clin Oncol 2005;23:2606–2617. [PubMed: 15452187]

42. Hudis C, Vogel C, Gralow J, et al. Weekly epoetin alfa during adjuvant chemotherapy for breastcancer: effect on hemoglobin levels and quality of life. Clin Breast Cancer 2005;6:132–142.[PubMed: 16001991]

43. Crawford J, Demetri G, Gabrilove J, et al. Clinical benefits of epoetin alfa therapy in patients withlung cancer. Clin Lung Cancer 2002;3:180–190. [PubMed: 14662041]

44. Glaspy J, Degos L, Dicato M, et al. Comparable efficacy of epoetin alfa for anemic cancer patientsreceiving platinum- and nonplatinum-based chemotherapy: a retrospective subanalysis of two large,community-based trials. Oncologist 2002;7:126–135. [PubMed: 11961196]

45. Gill P, Kaur J, Rummans T, et al. The hospice patient's primary caregiver. What is their quality oflife? J Psychosom Res 2003;55:445–451. [PubMed: 14581099]

46. Saag M, Bowers P, Leitz G, et al. Once-weekly epoetin alfa improves quality of life and increaseshemoglobin in anemic HIV+ patients. AIDS Res Hum Retroviruses 2004;20:1037–1045. [PubMed:15587490]

47. Provenzano R, Garcia-Mayol L, Suchinda P, et al. Once-weekly epoetin alfa for treating the anemiaof chronic kidney disease. Clin Nephrol 2004;61:392–405. [PubMed: 15224803]

48. Afdhal N, Dieterich D, Pockros P, et al. Epoetin alfa maintains ribavirin dose in HCV-infectedpatients: a prospective, double-blind, randomized controlled study. Gastroenterology2004;126:1302–1311. [PubMed: 15131791]

49. Loprinzi C, Sloan J, Perez E, et al. Phase III evaluation of fluoxetine for treatment of hot flashes. JClin Oncol 2002;20:1578–1583. [PubMed: 11896107]

50. Sloan J, Huschka M, Atherton P, et al. What role do cancer patients want to play in treatment decisionmaking: a pooled-analysis. [abstract]. J Clin Oncol 2006;24:8521.

Locke et al. Page 9

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Appendix

Appendix A

Linear Analog Self Assessment

Name:_________________________ Date:_______________________Directions: Please circle the number (0–10) best reflecting your response to the following thatdescribes your feelings during the past week, including today.1. How would you rate your physical well being over the past week? This question refers to such things as fatigue, activity, etc. 0 1 2 3 4 5 6 7 8 9 10 As bad as As good as it can be it can be2. How would you rate your emotional well being over the past week? This question refers to such things as depression, anxiety, stress, etc. 0 1 2 3 4 5 6 7 8 9 10 As bad as As good as it can be it can be

3. How would you rate your spiritual well being over the past week? This question refers to such things as a sense of meaning and purpose, relationship with God, etc. 0 1 2 3 4 5 6 7 8 9 10 As bad as As good as it can be it can be4. How would you rate your intellectual well being over the past week? This question refers to such things as the ability to think clearly, to concentrate, to remember, etc. 0 1 2 3 4 5 6 7 8 9 10 As bad as As good as it can be it can be5. How would you rate your overall well being over the past week? 0 1 2 3 4 5 6 7 8 9 10 As bad as As good as it can be it can be

Locke et al. Page 10

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Locke et al. Page 11

Table 1Baseline Characteristics (n=205)

Characteristic Summary

Age, years

 Mean±SD 54.9±11.4

 median; min-max 56.0; 19.0–79.0

Gender, n(%)

 Female 74 (36)

 Male 131 (64)

ECOG Performance Score, n(%)

 0 88 (43)

 1 95 (46)

 2 22 (11)

Family History of Brain Tumor, n(%)

 Yes 19 (9)

 No 186 (91)

Extent Resection, n(%)

 Biopsy 35 (17)

 Subtotal resection 91 (44)

 Gross total resection 79 (39)

Anticonvulsants, n(%)

 Yes 149 (73)

 No 56 (27)

Corticosteroid Therapy, n(%)

 Yes 124 (60)

 No 81 (40)

Tumor location, n(%)

 Temporal 46 (22)

 Frontal 51 (25)

 Parietal 27 (13)

 Occipital 6 (3)

 Other 75 (37)

MMSE score

 Mean±SD 27.7±3.4

 median; min-max 29.0; 8.0–30.0

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Locke et al. Page 12Ta

ble

2Ps

ycho

soci

al a

nd Q

ualit

y of

life

Ass

essm

ents

at B

asel

ine,

Tw

o M

onth

s, an

d Fo

ur M

onth

s

Bas

elin

eT

wo

Mon

ths

Four

Mon

ths

nM

ean±

SDM

edia

n; r

ange

nM

ean±

SDM

edia

n; r

ange

nM

ean±

SDM

edia

n; r

ange

Pa

LASA

 O

vera

ll19

470

.0±2

0.9

126

69.5

±21.

782

69.1

±21.

60.

471

75.0

; 12.

5–10

0.0

75.0

; 0.0

–100

.075

.0; 2

5.0–

100.

0

 Ph

ysic

al19

465

.2±2

3.8

127

63.9

±23.

582

60.2

±22.

10.

083

75.0

; 0.0

–100

.075

.0; 0

.0–1

00.0

56.3

; 12.

5–10

0.0

 Em

otio

nal

194

68.5

±23.

012

766

.0±2

4.2

8266

.1±2

3.3

0.27

3

75.0

; 0.0

–100

.075

.0; 6

.3–1

00.0

75.0

; 25.

0–10

0.0

 Sp

iritu

al19

477

.8±2

1.6

126

77.0

±21.

282

78.9

±21.

40.

819

75.0

; 12.

5–10

0.0

75.0

; 12.

5–10

0.0

81.3

; 25.

0–10

0.0

 In

telle

ctua

l19

466

.1±2

4.0

127

68.9

±20.

482

66.8

±24.

90.

239

75.0

; 0.0

–100

.075

.0; 1

2.5–

100.

075

.0; 0

.0–1

00.0

SDS

 O

vera

ll18

975

.8±1

8.0

124

77.9

±14.

479

78.4

±13.

80.

476

79.2

; 8.3

–100

.079

.2; 2

9.2–

100.

083

.3; 3

7.5–

100.

0

POM

S

 O

vera

ll17

269

.1±1

5.7

115

68.8

±14.

077

69.0

±15.

40.

416

71.3

; 13.

3–97

.572

.5; 2

3.3–

90.8

72.5

; 20.

0–97

.5

 C

onfu

sion

182

77.3

±19.

012

280

.9±1

7.9

8079

.5±1

8.8

0.19

2

81.3

; 0–1

00.0

87.5

; 18.

8–10

0.0

84.4

; 18.

8–10

0.0

 Fa

tigue

184

68.0

±22.

712

262

.0±2

2.8

8162

.0±2

5.3

0.00

5

70.0

; 5.0

–100

.067

.5; 0

–100

.065

.0; 0

–100

.0

FAC

T-B

r

 O

vera

ll19

072

.8±1

2.6

117

72.1

±13.

978

71.9

±14.

40.

101

73.5

; 23.

9–98

.472

.5; 3

2.6–

99.5

73.7

; 33.

8–96

.2

 Ph

ysic

al19

281

.3±1

6.1

126

74.8

±20.

381

73.3

±19.

9<0

.001

85.7

; 7.1

–100

.078

.6; 0

.0–1

00.0

78.6

; 21.

4–10

0.0

 Em

otio

nal

193

72.4

±19.

511

975

.1±1

9.0

8077

.3±1

9.4

0.13

0

75.0

; 0.0

–100

.079

.2; 2

5.0–

100.

083

.3; 2

9.2–

100.

0

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Locke et al. Page 13

Bas

elin

eT

wo

Mon

ths

Four

Mon

ths

nM

ean±

SDM

edia

n; r

ange

nM

ean±

SDM

edia

n; r

ange

nM

ean±

SDM

edia

n; r

ange

Pa

 B

rain

184

69.2

±15.

512

470

.1±1

5.9

7969

.9±1

5.3

0.46

7

71.1

; 14.

5–98

.769

.7; 3

5.5–

98.7

69.4

; 21.

1–10

0.0

a Rep

eate

d m

easu

res a

naly

sis o

f var

ianc

e w

ith a

n au

tore

gres

sive

cor

rela

tion

stru

ctur

e (A

R1)

.

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Locke et al. Page 14Ta

ble

3C

orre

latio

n w

ith L

ASA

Sca

les

Cyc

le 0

Cyc

le 1

Cyc

le 2

nr

nr

Nr

SDS

 O

vera

ll18

90.

5312

30.

5679

0.57

POM

S

 O

vera

ll17

20.

6111

40.

6477

0.65

 W

ith L

ASA

em

otio

nal

172

0.60

115

0.58

770.

52

 C

onfu

sion

with

LA

SA in

telle

ctua

l18

20.

5812

20.

5380

0.62

 Fa

tigue

with

LA

SA p

hysi

cal

184

0.63

122

0.56

810.

60

FAC

T-B

r

 O

vera

ll19

00.

5811

60.

6378

0.56

 Ph

ysic

al19

20.

5912

60.

5381

0.54

 Em

otio

nal

193

0.44

119

0.53

800.

50

 B

rain

with

LA

SA in

telle

ctua

l19

10.

5412

40.

5479

0.50

r=Sp

earm

an c

orre

latio

n co

effic

ient

. All

corr

elat

ion

coef

ficie

nts w

ere

sign

ifica

ntly

diff

eren

t fro

m 0

(P<0

.001

in e

ach

case

).

The L

ASA

ove

rall

was

corr

elat

ed w

ith th

e SD

S, P

OM

S, an

d FA

CT-

Br o

vera

ll; th

e LA

SA p

hysi

cal w

as co

rrel

ated

with

the F

AC

T-B

r phy

sica

l and

the P

OM

S-fa

tigue

; the

LA

SA in

telle

ctua

l was

corr

elat

edw

ith th

e PO

MS-

conf

usio

n an

d th

e FA

CT-

Br b

rain

; and

the

LASA

em

otio

nal w

as c

orre

late

d w

ith th

e FA

CT-

Br e

mot

iona

l and

PO

MS-

over

all.

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Locke et al. Page 15Ta

ble

4C

orre

latio

n w

ith E

CO

G P

erfo

rman

ce S

core

Bas

elin

eT

wo

Mon

ths

Four

Mon

ths

nr

Pan

rPa

nr

Pa

LASA

 O

vera

ll19

4−0

.25

<0.0

0112

5−0

.21

0.02

081

−0.1

90.

082

 Ph

ysic

al19

4−0

.36

<0.0

0112

6−0

.25

0.00

581

−0.1

60.

148

 Em

otio

nal

194

−0.2

00.

006

126

−0.1

60.

073

81−0

.000

40.

997

 Sp

iritu

al19

4−0

.15

0.04

412

5−0

.20

0.02

781

−0.0

70.

559

 In

telle

ctua

l19

4−0

.16

0.02

912

6−0

.22

0.01

681

−0.0

80.

460

SDS

 O

vera

ll18

9−0

.17

0.02

212

3−0

.14

0.12

078

−0.0

50.

643

POM

S

 O

vera

ll17

2−0

.24

0.00

211

4−0

.28

0.00

276

−0.2

10.

066

 C

onfu

sion

182

−0.1

30.

073

121

−0.0

80.

378

79−0

.05

0.65

9

 Fa

tigue

184

−0.3

1<0

.001

121

−0.3

0<0

.001

80−0

.27

0.01

5

FAC

T-B

r

 O

vera

ll19

0−0

.29

<0.0

0111

6−0

.24

0.00

877

−0.2

10.

073

 Ph

ysic

al19

2−0

.32

<0.0

0112

5−0

.21

0.01

880

−0.1

10.

332

 Em

otio

nal

193

−0.0

70.

352

118

−0.1

00.

260

79−0

.04

0.74

7

 B

rain

191

−0.3

3<0

.001

123

−0.2

70.

003

78−0

.23

0.04

0

r=Sp

earm

an c

orre

latio

n co

effic

ient

.

a Com

parin

g co

rrel

atio

n co

effic

ient

to 0

.

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Locke et al. Page 16

Table 5Correlation with MMSE Score at Cycle 0

n r Pa

LASA

 Overall 177 0.22 0.003

 Physical 177 0.21 0.005

 Emotional 177 0.26 <0.001

 Spiritual 177 0.12 0.111

 Intellectual 177 0.29 <0.001

SDS

 Overall 172 0.19 0.014

POMS

 Overall 156 0.20 0.012

 Confusion 165 0.26 <0.001

 Fatigue 167 0.13 0.088

FACT

 Overall 173 0.34 <0.001

 Physical 175 0.19 0.013

 Emotional 176 0.13 0.094

 Brain 174 0.34 <0.001

r=Spearman correlation coefficient.

aComparing correlation coefficient to 0.

J Pain Symptom Manage. Author manuscript; available in PMC 2009 August 26.