Assessing Quality of Life in Research and Clinical Practice

September 3, 2002

There is a growing recognition in oncology of the importance of maintaining or improving patients’ quality of life (QOL) throughout the disease course. With this goal in mind, many clinical trials in oncology now seek to evaluate QOL end points.

ABSTRACT: There is a growing recognition in oncology of the importance of maintaining or improving patients’ quality of life (QOL) throughout the disease course. With this goal in mind, many clinical trials in oncology now seek to evaluate QOL end points. In using QOL measures as research tools, investigators need to consider which instrument is best suited to addressing the issues under study, how often and when to administer the instrument, and how to deal with data that may be missing due to toxicity, morbidity, or mortality. Findings from QOL research can inform clinical care by providing information about the likely impact of disease and its treatment on functioning and well-being, identifying common problems, and developing effective interventions to deal with these problems. The routine assessment of QOL may also have clinical uses at the individual patient level. These uses include fostering patient-provider communication, identifying frequently overlooked problems, prioritizing problems, and evaluating the impact of palliative and rehabilitative efforts. Although several barriers exist to routine assessment of quality of life in clinical practice, several strategies can be used to successfully overcome these barriers. [ONCOLOGY 16(Suppl 10):133-139, 2002]

In recent years, there hasbeen agrowing acceptance of the viewthat the goals of cancer treatment should include concerns about quality of life(QOL) as well as length of life. Patients with cancer experience a variety ofsymptoms due to their disease and its treatment, such as pain, fatigue, andnausea, that can have a significant negative impact on their well-being andfunctioning. The development of multidimensional self-report QOL instruments hasallowed investigators to measure the adverse impact of disease and its treatmenton well-being and functioning and evaluate the efficacy of interventionsdesigned to prevent or treat these adverse effects. Findings from QOL researchsuggest that routine use of QOL instruments as part of clinical practice has thepotential to improve the quality of care that patients receive as well as theirhealth status. However, in addition to its many benefits, there are also manychallenges to assessing quality of life in research and clinical practice.

Measurement of QOLEnd Points

Two features characterize most forms of QOL assessmentcurrently used in oncology. First, it is generally recognized that quality oflife is a multidimensional construct and is best measured using instruments thattap multiple domains of functioning and well-being.[1-3] Consistent with thisview, most QOL instruments measure physical, social, and emotional aspects offunctioning, as well as common symptoms of cancer and its treatment. Second,there is general agreement that quality of life is a subjective phenomenon andthat patients are the best judges of their own quality of life.[1,2] Indeed,studies have shown that considerable disparities exist between concurrentratings of quality of life made by patients and their physicians.[4,5]Accordingly, assessment of quality of life in oncology trials is typicallyperformed using patient self-report questionnaires.

Two of the most widely used multidimensional QOL instrumentsin oncology are the General Version of the Functional Assessment of CancerTherapy (FACT-G) [6] and the European Organization for Research and Treatment ofCancer Quality of Life Questionnaire (EORTC-QLQC30).[7] (For a comprehensivelist of these and other QOL scales discussed in this article, see the Appendixat the end of this supplement.)

The FACT-G (version 4) is a 27-item measure. For each item,respondents indicate on a 5-point rating scale (0 = not at all; 4 = very much)how true each statement (for example, "I have a lack of energy") hasbeen for them during the past 7 days. The FACT-G yields a total score foroverall quality of life as well as subscale scores for physical well-being,social/family well-being, emotional well-being, and functional well-being.

The EORTC QLQ-C30 is a 30-item measure. For each item,respondents indicate the rating that best applies to them. Seven items are ratedyes or no for an unspecified time frame (eg, "Do you have any troubletaking a long walk?"); 21 items are rated on a 4-point scale (1 = not atall; 4 = very much) for the past week (eg, "Were you tired?"); and 2items are rated on the 7-point scale (1 = very poor; 7 = excellent) for the pastweek (eg, "How would you rate your overall quality of life?"). TheEORTC QLQ-C30 yields scores for five functional scales (physical, rolecognitive, social, and emotional), three symptom scales (nausea, pain, andfatigue), and a global health and QOL scale. The measure also yields single-itemratings of additional symptoms commonly reported by cancer patients (dyspnea,appetite loss, sleep disturbance, constipation, and diarrhea) as well as theperceived financial impact of disease and its treatment.

Both the FACT-G and the EORTC QLQ-C30 have been shown to haveadequate validity and reliability and to be able to distinguish patientsaccording to their performance status.[6,7] A number of disease- specificmodules (eg, breast, lung, and prostate) have been developed to supplement eachof these core measures. These modules assess additional symptoms and QOL issuesthat are relatively specific to certain forms of cancer.

Linear Analogue Self-Assessment (LASA) scales are also widelyused in QOL research in oncology.[8] A LASA scale consists of a 100-millimeterline with descriptors at each end. Respondents mark their current statussomewhere along the line, and then the distance in millimeters from the lowerend point (0 point) is measured to obtain their scores. LASA scales have beendeveloped to measure a variety of symptoms (eg, pain) and aspects of functioning(eg, physical activity), as well as overall quality of life.

These measures are popular, in part, because they arerelatively easy and quick to administer. Moreover, there is evidence to suggestthat many LASA scales compare favorably with more established QOL measures interms of both validity and ability to detect changes over time.[9] Although theuse of LASA scales is appealing, caution is advised. Investigators need todetermine whether the specific set of LASA scales to be administered has beenvalidated for its intended use. In the absence of existing validity data, LASAscales should be used in combination with the more established FACT-G andEORTC-QLQ-C30 measures. (For a discussion of the use of these and otherinstruments in the assessment of anemia and anemia thearpy, see Dr. David Cella’sarticle in this supplement.)

Methodologic Issues in the Evaluation of QOL End Points

Perhaps the most important methodologic issue to consider inevaluating QOL end points in an oncology clinical trial is the selection ofappropriate outcome measures. In most instances, the use of a well-validatedmultidimensional self-report QOL instrument (eg, FACT-G, EORTC QLQ-C30) willmeet this requirement. Depending on the nature of the trial, it may be necessaryto supplement these core measures with additional measures that provide moreinformation about those symptoms that are most relevant to the patientpopulation under study. For example, the lung subscale for Functional Assessmentof Cancer Therapy (FACT-L) [1] includes several items assessing respiratorydifficulties. Likewise, in trials where relief of pain is a primary goal, it maybe useful to collect additional information about the subjective experience ofpain using a LASA scale or a measure such as the Brief Pain Inventory.[10]

Number and Timing of Assessments

A second important issue to consider is the number and timingof QOL assessments. The desire to collect self-reported information atrelatively brief intervals in order to increase the likelihood of detectingchanges over time must be weighed against concerns about the burden to patientsand the financial cost of conducting frequent assessments. Osoba[11] hasproposed a set of guidelines that may be useful in determining the timing of QOLassessments in oncology clinical trials. A baseline QOL assessment carried outbefore the initiation of treatment can be considered necessary for two reasons:

First,in randomized trials, the baseline assessment will indicate whether there arepreexisting differences in quality of life between patients in the varioustreatment arms; if present, these differences would need to be adjusted forstatistically in order to accurately determine treatment effects.

Second,the baseline assessment conducted prior to intervention provides an essentialpoint of reference for identifying changes over time that may be attributable tothe treatment under investigation.

In most instances, one or more on-treatment assessments arealso necessary. As noted by Osoba,[11] the frequency and timing of theseassessments will depend on the research question(s) being asked. If, forexample, the goal is to determine whether chemotherapy improves quality of lifein patients experiencing disease-related symptoms (eg, pain), on-treatmentassessments should be conducted just before the start of subsequent chemotherapycycles to reduce the likelihood that results will reflect short-term treatmentside effects. In instances where multiple chemotherapy cycles are beingadministered, the nature of the research question being asked and the financialcosts of data collection will determine whether on-treatment assessments areconducted after each cycle or at less frequent intervals.

Finally, there is the issue of off-treatment assessments—thoseconducted following the completion or cessation of treatment. Once again, thenature of the research question and issues of cost will be the primary factorsdetermining the number and timing of these assessments. In studies of patientswith advanced disease (and a poor prognosis for survival), it may be bothdesirable and feasible to follow patients until disease progression occurs oreven until death.[11] Data collected during the off-treatment period wouldindicate if and for how long any of the observed on-treatment benefits toquality of life may have persisted.

Handling Missing Data

A third important methodologic issue to consider is thehandling of missing data. This issue is of particular relevance to studies ofquality of life end points. As Moinpour[12] has noted, "In the very settingwhere quality of life questions are most compelling, they are the most difficultto evaluate because the missing data mechanism is often dependent on the veryoutcome being assessed—the health status and quality of life of thepatient." That is, patients who are experiencing negative health outcomes,such as treatment toxicity or progressive disease, are also most likely to havemissing QOL data. Under these circumstances, analyses based only on available (nonmissing)data may lead to erroneous conclusions. For example, if QOL data are missing ona consistent basis due to treatment toxicity, the analysis of only nonmissingdata is likely to lead to an overestimate of the actual QOL benefits of theagent under study.

At present, there is no consensus on the optimal method fordealing with nonrandom missing QOL data in clinical trials. As a generalstrategy, Fairclough and colleagues[13] suggest that two questions be consideredin attempting to evaluate the impact of missing data. First, why are the datamissing? If data are missing for reasons related to treatment toxicity ordisease progression, then the missing data mechanism is "nonignorable"and statistical models appropriate for this situation should be explored.Second, how sensitive are the study results to different assumptions about themissing data mechanism? In the absence of a consensus on the "best"approach, sensitivity analyses are recommended to examine the effects of severaldifferent methods of handling missing data. Readers interested in learning aboutthese methods may wish to consult a special issue of Statistics in Medicine (volume17, numbers 5-7, 1998) devoted specifically to the topic of missing QOL data inoncology clinical trials.

Clinical Uses of QOL Findings

Although QOL assessment in oncology research is a relativelynew phenomenon, a number of clinical uses of QOL findings can be identified.First and foremost, QOL findings have permitted clinicians to understand thetypical impact of specific forms of cancer treatment on patients’ functioningand well-being. This type of information can be used to communicate accurateexpectations about likely limitations in functioning to patients about to begintreatment. For example, patients about to undergo blood and marrowtransplantation can be advised as to the usual nature and duration oflimitations in physical functioning or expected changes in emotional well-being.At the same time, this information can be used to identify the types of servicesmost patients are likely to require to restore functioning or well-being. Forexample, QOL data have suggested that certain forms of cancer treatment canresult in persistent fatigue that limits a patient’s ability to work or engagein social activities.[14] These findings have prompted efforts to developinterventions to prevent or reduce fatigue in the posttreatment period.[15]

Predicting Treatment Response

A second clinical use involves the potential for quality oflife information to predict treatment response and survival time over and abovestandard clinical indicators such as disease stage and performance status. Inother words, QOL data collected at the start of treatment may be helpful inidentifying patients likely to have a poorer response to therapy or shortersurvival. Along these lines, Coates and colleagues[16] demonstrated thepredictive value of QOL information in a clinical trial that comparedintermittent vs continuous chemotherapy in women with metastatic breast cancer.QOL scores at the start of treatment, and subsequent changes in those scores,were found to predict survival duration independent of treatment group,performance status, and treatment response.

Identifying Effective Palliative Care Interventions

Yet a third clinical use is to identify palliative careinterventions that are effective in improving quality of life beyondpretreatment values. One of the most notable examples of this potential use ofQOL data comes from a clinical trial that compared mitoxantrone (Novantrone)plus prednisone to prednisone alone in men with hormone-refractory prostatecancer.[17] Previous single-arm studies [18-21] had yielded preliminary evidencethat prednisone, mitoxantrone, and mitoxantrone plus prednisone each producedsome palliative benefit and had limited toxicity. Based on these findings,Tannock and colleagues designed a randomized trial to test the hypothesis thattreatment with mitoxantrone plus prednisone would provide better palliation thanprednisone alone. Results indicated that, although survival did not differsignificantly between the two treatment arms, the primary criterion of apalliative response (ie, pain relief) was met by significantly more patients whoreceived mitoxantrone plus prednisone than prednisone alone and that theduration of the palliative response also was longer in the combined therapygroup.[17]

Additional analyses of quality of life data from this trialshowed that, after 12 weeks of treatment, patients receiving prednisone alonedid not exhibit any significant improvements relative to their baseline scores.In contrast, patients that received combined therapy exhibited significantimprovements in physical, role, emotional, and social functioning and in globalquality of life.[22] The results of this trial illustrate how inclusion of QOLend points in oncology clinical trials has the potential to improve palliationof patients with advanced cancer.

Applying QOL Results at the Individual Patient Level

The clinical uses of QOL findings described above representways in which the results of research studies incorporating QOL measures canimprove the clinical care of patients in general. Can these findings also beapplied to improve the care of an individual patient? At least four potentialbenefits of administering a QOL measure to an individual patient can beidentified.

One obvious benefit is the potential for improvingpatient-provider communication. QOL measures can provide clear and unambiguousinformation about a patient’s current state of well-being and functioning. Inaddition, information obtained from patient self-report measures is notdependent upon the ability or willingness of a patient or provider to initiate adiscussion of QOL concerns.

The potential also exists to identify frequently overlookedproblems. By systematically prompting patients to respond to a standardized itemset, QOL measures probe for problems that providers may be reluctant to inquireabout and patients may be reluctant to complain about, such as problems withsexual functioning.

Administration of a QOL measure may also help in prioritizingproblems. Since patients often present with multiple symptoms,[23] providers maybe challenged in deciding where to focus their efforts in improving quality oflife. By attaching numerical ratings to perceptions of symptom severity andproblems in functioning, QOL measures can help both patients and providersidentify those issues that are of more immediate concern.

Last but not least, QOL measures may help clinicians evaluatethe impact of palliative care efforts. Direct comparison of scores obtained byadministering the same QOL measures on multiple occasions permits adetermination of both the direction and magnitude of changes. This informationcan be quite helpful in deciding whether or not a patient is benefiting from aparticular therapy or intervention.

Barriers to AssessingQuality of Life in Routine Clinical Practice

Given the many potential benefits of routine clinicalassessment of quality of life, why is the practice not more widespread? Threebroad categories of barriers to routine assessment of quality of life inclinical practice can be identified. These are provider inexperience withquality of life assessment, methodologic concerns about reliability andvalidity, and logistic barriers that inhibit feasibility of clinicalimplementation and integration. Nine specific barriers falling under these threebroad categories are summarized in Table 1 and discussed briefly below.

Provider InexperienceWith QOL Assessment

Lack of provider experience conducting formal QOL assessmentsis one of the most important barriers to routine use in clinicalpractice.[24,25] Because providers are the information gatekeepers in thehealth-care setting, their support in the use of QOL data to inform and guideclinical practice is essential.

Presently, there is a large discrepancy between the number ofproviders who report that QOL assessments are useful and those who actually usean instrument to assess quality of life.[24] Surveys of providers reveal thatthe knowledge base about assessment is limited but that with greater experienceand exposure to available instruments, providers may be willing to use them on aroutine basis.[26] Understanding the role that provider inexperience with QOLmeasures plays and determining ways to increase health-care professionals’exposure to QOL measures are critical for any attempt at integrating these datainto routine clinical practice.

An implication of provider inexperience with QOL assessmentis the potential for limited patient-provider communication. Although somephysicians have expressed concerns that QOL data will be used as a substitutefor direct communication rather than as an adjunct to care,[27] the opposite appears to be true. Recentstudies have reported that patients and staff find QOL assessments helpful foridentifying patient concerns and improving communication.[28,29]

Methodologic Concerns About Reliability and Validity

A second category of barriers relates to methodologicproblems and several specific concerns have been identified within thiscategory. Historically, physicians have been skeptical about the sensitivity andspecificity of QOL instruments.[24,30] Inspite of significant efforts to define quality of life and improve measures,doubts about the methodologic development of instruments persist.

At present, clinicians doubt the ability of availableinstruments to measure subtle individual changes. This is of particular concernfor cancer patients who may present from one week to the next with slightclinical changes that represent significant QOL changes according to thepatients themselves and/or family members. To date, there are not enough dataavailable for most instruments to determine the meaning of change scores withrespect to their clinical significance. The more the instruments are used,however, the greater the amount of information available to determine clinicallymeaningful change scores over time.

A final methodologic concern is the limited ability tocompare scores across different instruments. Technology such as computeradaptive testing (CAT) can address this and other methodologic concerns.[31,32]

Feasibility and Logistical Problems

A third category of barriers pertains to feasibility andlogistic problems associated with data collection and recording. These barriersapply to both providers and patients. Assessments need to be understandable,user-friendly, and short.[25] Ifnot, health-care providers are less likely to use the assessments.[25]Additionally, results must be ready in "real-time," at the visit whenthe data are gathered. Concern about the timeliness of results was the mostcommon problem cited by oncologists in a study of QOL assessment with advancednon-small-cell lung cancer.[33]

Oncologists also stated that more frequent assessments werenecessary to help guide treatment planning and decision-making. Routineassessment (at each visit) of quality of life will require a major commitment ofresources from both clinical and administrative areas.

Scoring and interpretation of data for clinical usefulnessconstitutes another barrier. Scores must be clinically meaningful to bothproviders and patients. Results must be presented in a format that is easy toread, provides useful information, and facilitates direct discussion abouttopics such as treatment options and general and specific aspects of quality oflife.[25,29]

Factors Critical to Success in Assessing Quality of Life in Clinical Practice

Several factors are likely to be critical to the success ofefforts to increase the use of QOL assessment in routine clinical practice. Aswith barriers, these can be grouped into three broad categories of nine specificfactors. The categories are an acceptable set of core measures, clinicalrelevance and ease of use, and "buy-in" from providers and patients.Each of the three categories and their associated factors are summarized in Table 2 and discussed briefly below.

An Acceptable Set ofCore Measures

A core set of QOL measures needs to be accepted and endorsedby health-care providers. Having a variety of standardized measurements fromwhich clinicians can choose may increase the use of routine assessment. This setwould include an array of tools with adequate validity and reliability that aregood estimates of longer instruments. The establishment of a core set ofmeasures would allow clinicians to tailor the use of assessments to meet theirneeds and those of their patients.

Broader use of more modern item response theory (IRT) modelsin the development of instruments should be considered. IRT measurement modelsmay allow ways to compare scores across instruments as well as estimate howpatients would score on a given instrument based on their responses to anotherinstrument.[34] Computer adaptive tests are one way these types of measurementscould be developed and used while preserving other critical factors. Thispractice could result in more precise assessments of an individual’s ability,while simultaneously decreasing the number of questions that need to beadministered.

Ease of Use

Ease of use is one of the most important factors necessaryfor assessing quality of life as part of routine clinical practice.Computer-based testing (CBT) is one way that more frequent assessments can beconducted with minimal burden on patients and providers. CBT eliminates the needfor a test administrator, as needed for traditional paper and pencil formats.Rather, the patient administers the instrument to him/her self. This givespatients a sense of control that is particularly important for cancer patientswho often feel that the cancer diagnosis severely limits their control in life.

CBT also decreases the burden on clinic staff, whileproviding immediate "real-time" feedback. Information from assessmentscan be displayed in graphic reports as visual aids that help guide discussionsabout treatment options and care planning. Patients have reported that thesetypes of discussions improve communication between themselves and providers.[28]Additionally, discussions help patients feel understood both physically andemotionally.[28] These positive feelings experienced by patients can impact"buy-in" and encourage clinicians to use QOL assessments on a regularbasis.

‘Buy-in’ From Providersand Patients

The more providers are educated about quality of life and thebenefits to clinical practice, the greater the likelihood that they will usethese data. This is supported by results from a study by Bezjak andcolleagues,[25] whoreported that the majority (82%) of oncologists surveyed felt that theirknowledge about quality of life was limited. These same oncologists said theybelieved in the benefit of QOL assessment for patient care and that they wouldlikely increase their use of QOL information in the future.

This "buy-in" from key staff is critical to asuccessful program.[24,25,27,28,34-39] If an introduction to QOL assessment wasincorporated into general medical education, it might eventually become aroutine component of patient care.[35] Educationneeds to focus on information about quality of life conceptually, interpretationof specific measures, and uses of QOL data in clinical practice. Organizationalrequirements for inclusion of QOL end points in clinical trials, as have beenmandated by the National Cancer Institute of Canada,[40] may also increaseexposure, knowledge, and use of QOL assessments by clinicians.

"Buy-in" from patients, including the desire forand willingness to undergo routine assessments, is equally important. Supportcan be attained by presenting QOL assessments as a standard part of patient careto highlight unstated patient concerns or issues typically only indirectlyassessed by health-care professionals. Providers can explain that assessments atregular intervals are a way to facilitate patient-provider communication. Detmarand Aaronson[28] reportedthat most patients felt that after multiple assessments, physicians were moreinformed and aware of the physical and psychosocial impact of treatment on theirdaily life activities.

Summary

Patients with cancer suffer from a variety of symptoms thatcan adversely affect their well-being and functioning. In recent years, therehas been a growing recognition that maintaining or improving the quality of lifefor cancer patients is an important treatment goal. Accordingly, manyclinical oncology trials now seek to evaluate QOL end points.

There is general agreement that quality of life is a complexconstruct best measured by using instruments which assess multiple aspects offunctioning and well-being as viewed from the patients’ perspective. Reliableand valid multidimensional self-report QOL measures have been developed that aresensitive to changes in patients’ health status. In seeking to use thesemeasures as research tools, there are a number of important methodologic issuesto consider. These issues include selecting instruments that measure QOLoutcomes pertinent to the patient population under study, determining theoptimal number and timing of QOL assessments to include in the study, andresolving the handling of data missing due to toxicity, morbidity, andmortality.

There have been several clinical uses of findings from QOLresearch. These include the ability to provide clinicians and patients withaccurate expectations about the likely impact of treatments on well-being andfunctioning, the ability to identify common problems that will need to beaddressed, and the ability to identify therapies and interventions effective inaddressing these problems. In addition, findings suggest that QOL data mayimprove clinicians’ ability to predict treatment response and survival time incertain contexts.

The clinical uses of QOL measures may extend beyond theapplication of research findings. Routine assessment of quality of life as partof clinical practice has the potential to improve communication between patientsand providers, identify frequently overlooked problems, prioritize problems, andevaluate the impact of therapeutic efforts at the individual patient level.There are, however, a number of barriers to the routine use of QOL measures inclinical practice. Provider inexperience with QOL assessment, unresolvedmethodologic issues in QOL assessment, and questions about the feasibility andlogistics of integrating QOL assessment into routine clinical care are among themore commonly identified barriers. These barriers are not insurmountable iffactors that can contribute to successful implementation and use of QOL data inroutine clinical practice are carefully considered.

References:

1. Cella DF, Bonomi AE: Measuring quality of life: 1995update. Oncology 9(suppl 11):47-60,. 1995.

2. Osoba D: Lessons learned from measuring health-relatedquality of life in oncology. J Clin Oncol 12:608-616, 1994.

3. Moinpour CM: Measuring quality of life: An emergingscience. Semin Oncol 21(suppl 10):48-63, 1994.

4. Kahn SB, Houts PS, Harding SP: Quality of life andpatients with cancer: A comparative study of patient versus physicianperceptions and its implications for cancer education. J Cancer Educ 7:241-249,1992.

5. Stephens RJ, Hopwood P, Girling DJ, et al: Randomizedclinical trials with quality of life endpoints: Are doctors’ ratings ofpatients’ physical symptoms interchangeable with patients’ self-ratings? QualLife Res 6:225-236, 1997.

6. Cella DF, Tulsky DS, Gray G, et al: The FunctionalAssessment of Cancer Therapy (FACT) Scale: Development and validation of thegeneral measure. J Clin Oncol 3:570-579, 1993.

7. Aaronson NK, Ahmedzai S, Bergman B, et al: The EuropeanOrganization for Research and Treatment of Cancer QLQ-C30: A quality-of-lifeinstrument for use in international clinical trials in oncology. J NatlCancer Inst 85:365-376, 1993.

8. Cella DF: Measuring quality of life in palliative care. SemOncol 22:73-81, 1995.

9. Stockler MR, Osoba D, Goodwin P, et al: Responsiveness tochange in health-related quality of life in a randomized clinical trial: Acomparison of the Prostate Cancer Specific Quality of Life Instrument (PROSQOLI)with analogous scales from the EORTC QLQ-C30 and a trial specific module. JClin Epidemiol 51:137-145, 1998.

10. Daut RL, Cleeland CS, Flanery RC: Development of theWisconsin Brief Pain Questionnaire to assess pain in cancer and other diseases. Pain17:197-210, 1983.

11. Osoba D: Rationale for the timing of health-relatedquality-of-life (HQL) assessments in oncological palliative therapy. CancerTreat Rev 22(suppl A):69-73, 1996.

12. Moinpour CM: Do quality of life assessments make adifference in the evaluation of cancer treatments? Controlled Clin Trials18:311-317, 1997.

13. Fairclough DL, Peterson HF, Chang V: Why are missingquality of life data a problem in clinical trials of cancer therapy? Stat Med17:667-677, 1998.

14. Sadler IJ, Jacobsen PB: Progress in understanding fatigueassociated with breast cancer treatment. Cancer Investigation 19:723-731, 2001.

15. Atkinson A, Barsevick A, Cella D, et al: NCCN practiceguidelines for cancer-related fatigue. Oncology 14(11A):151-61, 2000.

16. Coates A, Gebski V, Signorini D, et al: Prognostic valueof quality-of-life scores during chemotherapy for advanced breast cancer. JClin Oncol 10:1833-1838, 1992.

17. Tannock IF, Osoba D, Stockler MR, et al: Chemotherapywith mitoxantrone plus prednisone or prednisone alone for symptomatichormone-resistant prostate cancer: A Canadian randomized trial with palliativeend points. J Clin Oncol 14:1756-1764, 1996.

18. Tannock I, Gospodarowicz M, Meakin W, et al: Treatment ofmetastatic prostatic cancer with low-dose prednisone: Evaluation of pain andquality of life as pragmatic indices of response. J Clin Oncol 7:590-597,1989.

19. Osborne CK, Drelichman A, Von Hoff DD, et al:Mitoxantrone: Modest activity in a phase II trial in advanced prostate cancer. CancerTreat Rep 67:1133-1135, 1983.

20. Raghavan D, Coorey G, Rosen M, et al: Management ofhormone-resistant prostate cancer: An Australian trial. Sem Oncol 23(suppl 14):20-23, 1996.

21. Moore MJ, Osoba D, Murphy K: Use of palliative end pointsto evaluate the effects of mitoxantrone and low-dose prednisone in patients withhormonally resistant prostate cancer. J Clin Oncol 12:689-694, 1994.

22. Osoba D, Tannock IF, Ernst DS, et al: Health-relatedquality of life in men with metastatic prostate cancer treated with prednisonealone or mitoxantrone and prednisone. J Clin Oncol 17:1654-1663, 1999.

23. Portenoy RK, Thaler HT, Kornblith AB, et al: The MemorialSymptom Assessment Scale: An instrument for the evaluation of symptomprevalence, characteristics, and distress. Eur J Cancer 30A:1326-1336,1994.

24. Morris J, Perez D, McNoe B: The use of quality of lifedata in clinical practice. Qual Life Res 7:85-91, 1998.

25. Bezjak A, Ng P, Skeel R, et al: Oncologists’ use ofquality of life information: Results of a survey of Eastern Cooperative OncologyGroup Physicians. Qual Life Res 10:1-13, 2001.

26. Taylor K, Macdonald K, Bezjak A, et al: Developing ameasuring instrument to assess physician use of quality of life (QOL)information: The MD-QOL. Clin Investigator Med 17:B88, Abstract 518, 1994

27. Higginson I, Carr A: Using quality of life measures inthe clinical setting. Brit Med J 322:1297-1300, 2001.

28. Detmar S, Aaronson N: Quality of life assessment in dailyclinical oncology practice: A feasibility study. Eur J Cancer34:1181-1186, 1998.

29. Carlson L, Speca M, Hagen N, et al: Computerized qualityof life screening in a cancer pain clinic. J Palliative Care 17:46-52,2001.

30. Taylor K, MacDonald K, Bezjak A: Physicians’perspective on quality of life: An exploratory study of oncologists. QualLife Res 5:5-14, 1996.

31. Lohr K: Applications of health status assessment measuresin clinical practice. Overview of the third conference on advances in healthstatus assessment. Med Care 30:MS1-MS14, 1992.

32. Bezjak A, Ng P, Taylor K, et al: A preliminary survey ofoncologists’ perceptions of quality of life information. Psycho-Oncol6: 107-113, 1997.

33. Chang C-H, Cella D, Masters G, et al: Real-time clinicalapplication of quality of life assessment in lung cancer. Submitted.

34. Chang C, Cella D: Equating health-related quality of lifeinstruments in applied oncology settings. Physical Med and Rehab: States ArtRev 11; 397-406, 1997.

35. Tanaka T, Gotay C: Physicians’ and medical students’perspectives on patient’s quality of life. Acad Med 73:1003-1005, 1998.

36. Tsevat J, Weeks J, Guadagnoli E, et al: Usinghealth-related quality of life information: Clinical encounters, clinicaltrials, and health policy. J Gen Internal Med 9:576-582, 1994.

37. O’Boyle C, McGee H, Hickey A, et al: Individual qualityof life in patients undergoing hip replacement. Lancet 339:1088-1091,1992.

38. Testa M: Interpreting quality-of-life clinical trial datafor use in clinical practice of antihypertensive therapy. J Hypertension5(Suppl 1): S9-S13, 1997.

39. Greenfield S, Nelson E: Recent developments and futureissues in the use of health status assessment measures in clinical settings. MedCare 30:MS23-MS41, 1992.

40. Osoba D: The Quality of Life Committee of the ClinicalTrials Group of the National Cancer Institute of Canada: Organization andfunctions. Qual Life Res 1:211-218, 1992.

41. Davis K, Cella D: Assessing quality of life in oncology clinicalpractice: A review of barriers and critical success factors. J Clin OutcomesManage. In press.