|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2006.07.7230 on December 17 2007 © 2008 American Society of Clinical Oncology. Quality of Life and Survival in the 2 Years After Surgery for Non–Small-Cell Lung Cancer
From the Centre for Health Economics Research and Evaluation (CHERE), University of Technology; Medical Oncology Unit, Sydney Cancer Centre; MBF Health Insurance (formerly CHERE); Department of PET and Nuclear Medicine, and Cardiothoracic Surgical Unit, Royal Prince Alfred Hospital, Sydney, Australia Corresponding author: Patricia Kenny, MPH, Centre for Health Economics Research and Evaluation, University of Technology, Sydney, PO Box 123, Broadway, Sydney, New South Wales 2007, Australia; e-mail: patsy.kenny{at}chere.uts.edu.au
Purpose Although surgery for early-stage non–small-cell lung cancer (NSCLC) is known to have a substantial impact on health-related quality of life (HRQOL), there are few published studies about HRQOL in the longer term. This article examines HRQOL and survival in the 2 years after surgery. Patients and Methods Patients with clinical stage I or II NSCLC (n = 173) completed HRQOL questionnaires before surgery, at discharge, 1 month after surgery, and then every 4 months for 2 years. HRQOL was measured with a generic cancer questionnaire (European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire [EORTC-QLQ] C30) and a lung cancer–specific questionnaire (EORTC QLQ-LC13). Data were analyzed to examine the impact of surgery and any subsequent therapy, and to describe the trajectories of those who remained disease free at 2 years and those with recurrent cancer diagnosed during follow-up. Results Disease recurred within 2 years for 36% of patients and 2-year survival was 65%. Surgery substantially reduced HRQOL across all dimensions except emotional functioning. HRQOL improved in the 2 years after surgery for patients without disease recurrence, although approximately half continued to experience symptoms and functional limitations. For those with recurrence within 2 years, there was some early postoperative recovery in HRQOL, with subsequent deterioration across most dimensions. Conclusion Surgery had a substantial impact on HRQOL, and although many disease-free survivors experienced recovery, some lived with long-term HRQOL impairment. HRQOL generally worsened with disease recurrence. The study results are important for informed decision making and ongoing supportive care for patients with operable NSCLC.
Surgical resection, although considered to offer the best survival outcomes for early-stage non–small-cell lung cancer (NSCLC),1 has the potential to affect health-related quality of life (HRQOL) separately from or in addition to disease effects (for example, through post-thoracotomy pain and activity limitation due to reduced lung capacity). Five-year survival rates have ranged from 55% to 72% for stage I disease and from 29% to 51% for stage II disease.2 There is also a small risk of an early postoperative death before the patient leaves the hospital; studies have reported hospital mortality of 3%3 and 5.8%.4 Although clinical research focuses on how to better identify the patients most likely to benefit from surgery in terms of survival, patients and surgeons also need information about the potential HRQOL outcomes. A number of cohort studies have described a substantial decrease in physical functioning and an increase in symptoms after surgery,3-6 with a return to preoperative HRQOL levels taking between 6 and 9 months,3,5 or 9 and 12 months.6 These studies were limited in the range of HRQOL dimensions assessed5 and sample size,6 and none reported follow-up beyond 1 year. Information describing the short- and long-term HRQOL consequences of lung cancer treatment is essential to enable clinicians to inform patients about the potential consequences of the treatment alternatives available to them and for the planning of ongoing supportive care for this patient group. This study aimed to examine the short- and long-term impact of surgery on HRQOL in patients with clinical stage I or II NSCLC. HRQOL and survival during the first 2 years are presented.
Recruitment Patients treated with surgery for clinical stage I or II NSCLC (n = 173) were recruited from the practices of six participating thoracic surgeons in Sydney, Australia, between April 1999 and December 2000. They were part of a randomized study of the role of positron emission tomography (PET) in preoperative assessment, which found no significant difference in management between the intervention and control groups.7 A secondary study objective was to describe the health outcomes for patients with operable NSCLC. Clinical staging was with chest x-ray and computed tomography scan of the thorax, upper abdomen, and brain, but not mediastinoscopy (as per the trial protocol).7 Of 183 PET study recruits, 173 underwent surgery and had a postoperative diagnosis of lung cancer (Appendix Fig A1, online only), making them suitable for the current study. The study was approved by the relevant institutional ethics committees and participants provided written consent.
Data Collection The European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire C30 [EORTC QLQ-C30] (version 3), a generic cancer instrument validated in patients with lung cancer,8 and the EORTC QLQ-LC13, a validated lung cancer–specific instrument,9 were used to measure HRQOL. Questionnaires were self-completed at recruitment, hospital admission, and discharge (9 days after surgery on average), 1 and 4 months after surgery, and then every 4 months until 2 years after surgery. Additional assessments were completed at the beginning and end of adjuvant therapy, and participants diagnosed with recurrent disease were asked to complete monthly assessments for as long as possible. Preoperative and discharge assessments were completed at the clinic or hospital, and the remaining assessments were completed by postal survey. Surveys were completed in the participant's preferred language, and sociodemographic characteristics were collected at the recruitment interview. Survival, disease recurrence, and other clinical information was collected from individual hospital medical records and treating doctors. Regular telephone contact with patients (or next of kin) facilitated information tracking. Survival data were censored at 2 years or the date of last contact if lost to follow-up. Disease recurrence was defined as the diagnosis of local or distant recurrence within 2 years of surgery by the treating doctors in their clinical practice.
Analysis The EORTC QLQ-C30 and QLQ-LC13 items were summarized into scales as per the manual.10 The QLQ-C30 has nine multi-item scales (overall quality of life [QOL]; physical, role, emotional, cognitive, and social functioning; and pain, nausea and vomiting, and fatigue) and six single items (lack of appetite, constipation, diarrhea, dyspnea, insomnia and financial difficulties). The QLQ-LC13 has one multi-item scale (dyspnea) and nine single items (pain in arm/shoulder, chest, and other parts; cough; hemoptysis; dysphagia; peripheral neuropathy; alopecia; and sore mouth). Multi-item scales were the mean score of the relevant items transformed to a score between 0 and 100. A higher score represents better HRQOL for the overall QOL and functioning scales, and worse HRQOL (more symptoms) for the symptom scales. The single items have four responses: not at all, a little, quite a bit, and very much. The average of each participant's preoperative assessments (recruitment and hospital admission) was used as baseline HRQOL. The multi-item scales were analyzed in two ways: mixed models of the HRQOL score over time (see statistical methods), and individual change in HRQOL relative to preoperative levels was calculated to illustrate the proportion of patients with impaired or improved HRQOL after surgery. Given that a 10-point change in each of the EORTC multi-item scales is generally considered to be a moderate or meaningful change,11-13 the proportion of surviving respondents scoring 10 or more points better or worse than their preoperative scores is reported for each time-point. For the single items, which have only four levels, the change from preoperative levels was assessed as deteriorated (a higher symptom level than before surgery), no change (the symptom was present and the level had not changed), improved (the symptom was present and the level had decreased), resolved (the symptom was present before surgery but not at the postoperative time point), or not present (the symptom was not present before surgery or at the postoperative time point). Two single items (pain in other body parts and dyspnea) were omitted because multi-item scales also measured these symptoms.
Statistical Methods Each model included the HRQOL scale (dependent variable), an intercept, and independent variables representing time after surgery. The models for the no-recurrence group included two indicator variables for the first two time points (hospital discharge and 1 month after surgery), a 4-month intercept adjustment variable (indicator coded 0 for observations before 4 months and 1 for all other observations), and one continuous variable for time in months after 4 months and up to 24 months after surgery or death or loss to follow-up (all observations up to and including 4 months were coded 0). Given that recurrence was diagnosed before 4 months in 20% of the recurrence group, the models for this group included only one indicator variable (hospital discharge), a 1-month intercept adjustment variable (indicator coded 0 for observations before 1 month and 1 for all other observations), and one continuous variable for time in months after 1 month and up to 24 months or death or loss to follow-up (all observations up to and including 1 month were coded 0). Thus the model intercept should be interpreted as average baseline HRQOL, the estimates for each indicator should be interpreted as the average difference from baseline at that time point, the intercept adjustment should be interpreted as the average difference from baseline over the period it represents, and the estimates for the continuous variables should be interpreted as representing the average change per month from the adjusted intercept over the period it represents (4 months to the end of follow-up for those without recurrence; 1 month to the end of follow-up for the recurrence group). In addition, two random effects were included: a random patient effect representing individual preoperative HRQOL (intercept) and a random patient by time effect representing individual linear change per month (slope). These two random effects are reported as standard deviations (square root of the estimated variance parameter), so that the degree of variation among individuals (random effects) can be interpreted relative to the mean of all individuals (fixed effects). Clinical and sociodemographic covariates were tested in these models and retained if statistically significant. The type of surgery (pneumonectomy or less than pneumonectomy) was tested as an interaction with the intercept adjustment variable and found to be statistically significant for three HRQOL scales (role and social functioning and dyspnea) in the recurrence group and was retained in these models. An additional eight patient-level effects (PET study randomization, pathologic stage, age, sex, education, country of birth, private health insurance, and years of smoking) and three time-varying treatment effects (adjuvant radiotherapy, palliative radiotherapy, and chemotherapy) were tested but were not significant in any models and therefore were not retained. The significance of fixed effects was tested with the t test using the containment method to compute denominator df.15 Because of the number of hypotheses tests, Hochberg's modified Bonferroni procedure was used so that the overall type I error rate would not exceed 0.05.18 This was applied within each recurrence group, to all tests of fixed effects for all domains. Thus, the procedure was applied once to 43 tests in the recurrence group analysis and once to 50 tests in the no-recurrence group analysis.
Sample Characteristics The sample was predominantly Australian born males aged between 42 and 82 years, and only 3% had never smoked. The majority (93%) were assessed as clinical stage I before surgery, but after surgery only half were stage I (Table 1). Within the 2 years of follow-up, 36% were diagnosed with recurrence; of these, 55% received palliative radiotherapy and 24% received palliative chemotherapy. Among those with recurrent disease, the mean time from surgery to diagnosis of recurrence was 10.7 months (standard deviation, 7.1 months).
Survival Postoperative mortality in the hospital was 3%. The survival rate was 78.7% at 1 year and 65.2% at 2 years, whereas the disease-free survival rate was 70.1% at 1 year and 51.9% at 2 years (Fig 1). The 2-year survival rate for the recurrence group was 36.3% (median survival after surgery, 18 months) and 81.4% for the no-recurrence group (Fig 1).
HRQOL Completion The HRQOL analysis excluded 10 participants with no postoperative HRQOL data. The remaining 163 participants completed questionnaires in English (93%), Chinese (3%), Greek (2%), Italian (1%), and Portuguese (1%). The number providing data at each time point declined during follow-up, due largely to death rather than missing data (Appendix Table A1, online only). Forty-three percent completed all HRQOL assessments and an additional 13% completed all assessments until death. The attrition rate was greater for the recurrence group. Of 62 participants with recurrence, 95% completed HRQOL preoperatively (50% at 1 year and 27% at 2 years); 61% recorded observations after the diagnosis of recurrence (35% of this group's data). Five participants died in the hospital, and of the remaining 106 without recurrence, 96% completed HRQOL preoperatively (76% at 1 year and 74% at 2 years). The completion rate as a percent of surviving participants ranged from 95% preoperatively to 71% at 2 years for those with recurrence, and 96% preoperatively to 86% at 2 years for those without recurrence.
HRQOL Average preoperative functioning and overall QOL scores, represented by the model intercepts, ranged from 57.1 to 82.3 for the recurrence group (Table 2) and from 68.0 to 89.3 for the no-recurrence group (Table 3). After surgery, both groups showed substantial deterioration (18.1 to 55.2 points) in all domains except emotional functioning, where the decline was smaller (5.3 to 6.4 points; Tables 2 and 3). Although there was considerable improvement 1 month after surgery, the recurrence group remained below preoperative levels for physical and role functioning, and those having pneumonectomy had substantially worse role and social functioning scores, which persisted to the end of follow-up. After 1 month, overall QOL as well as physical and cognitive functioning deteriorated significantly in the recurrence group. The average decline was 0.6 to 0.8 points per month until the end of follow-up, with standard deviations of a similar size, indicating considerable variation among individuals for the rate of change (Table 2).
For the no-recurrence group, there was substantial improvement 1 month after surgery in overall QOL and functioning, but most dimensions remained well below the preoperative levels. There was additional improvement between 1 and 4 months after surgery for this group. However, the average scores at 4 months remained significantly worse than the preoperative scores for physical, role, and social functioning, which persisted to the end of follow-up; however, this difference was less than 10 points for physical and social functioning (Table 3). Emotional functioning in this group showed a small but significant improvement 4 months after surgery. There were no additional significant changes after 4 months in the overall QOL and functioning scores in this group (Table 3). A simple description of how to interpret the model parameters in Tables 2 and 3 is provided in the Appendix (online only), with worked examples and graphs of the predicted HRQOL for each clinical subgroup at each time point, for two illustrative domains. The average change identified in the models was reflected in the numbers with a moderate to large change relative to their preoperative scores at each time point. At discharge from the hospital, the majority reported worse overall QOL and physical, role, social, and cognitive functioning. Between hospital discharge and 4 months after surgery, the proportion with worse HRQOL decreased and the proportion reporting HRQOL within 10 points of preoperative levels increased. Although changes between 4 months and 2 years in the proportion of the no-recurrence group with worse HRQOL were small, 52% of this group had worse physical functioning 2 years after surgery (Fig 2) and 47% had worse role functioning (Appendix Fig A2, online only). Between 4 months and 2 years after surgery, the proportion of the recurrence group with worse HRQOL increased so that the majority had worse HRQOL on all dimensions except emotional functioning (Fig 2; Appendix Fig A2). The majority of both groups reported stable or improved emotional functioning.
Symptoms The average preoperative scores (model intercept) for the four multi-item symptom scales (dyspnea, fatigue, pain, and nausea and vomiting) ranged from 7.5 to 29.3 for the recurrence group (Table 2) and from 2.9 to 19.9 for the no-recurrence group (Table 3). After surgery, both groups showed a substantial increase in these symptoms (11.8 to 46.8 points) followed by improvement. For the recurrence group, pain and fatigue remained below preoperative levels 1 month after surgery, as was dyspnea for those in this group who had pneumonectomy. After 1 month, fatigue, dyspnea, and nausea and vomiting worsened in the recurrence group; the level increased at 0.6 to 0.9 points per month until the end of follow-up on average (Table 2). One month after surgery, all four symptoms remained significantly worse than at baseline for the no-recurrence group, although this difference was small for nausea and vomiting (Table 3). Four months after surgery, the pain, fatigue, and dyspnea scores in the no-recurrence group remained significantly worse than at baseline, but the difference was less than 10 points (Table 3). After 4 months, with the exception of pain, there was no significant change in symptoms in the no-recurrence group. Pain scores decreased (improved) at 0.5 points per month until the end of follow-up on average (Table 3). The most prevalent preoperative symptoms were cough (recurrence group, 83%; no-recurrence group, 82%), dyspnea (recurrence group, 78%; no-recurrence group, 71%), fatigue (recurrence group, 76%; no-recurrence group, 74%), and pain (recurrence group, 69%; no-recurrence group, 48%). At discharge from the hospital, the majority reported worse pain, fatigue, and dyspnea than preoperatively. Between discharge and 4 months, this decreased and the proportion returning to preoperative levels increased. However, 53% of the no-recurrence group still had worse dyspnea 2 years after surgery (Fig 3) and 40% had worse fatigue (Fig 2). Between 4 months and 2 years after surgery, the proportion of the recurrence group with worse symptoms increased for all multi-item symptom scales (Figs 2 and 3). Although cough improved for some during the 2 years after surgery, it remained unchanged for many (64% of the no-recurrence group and 94% of the recurrence group reported a cough at 2 years; Fig 3). Compared with preoperatively, approximately half (recurrence group, 54%; no-recurrence group, 48%) reported worse chest pain at discharge, and this proportion declined during 2 years for the no-recurrence group but not the recurrence group. However, 2 years after surgery, 31% to 33% of the no-recurrence group reported the same or worse chest pain and arm or shoulder pain compared with before (Fig 3). See Appendix Figures A3 and A4 (online only) for the remaining single items.
This study of survival and HRQOL after surgery for NSCLC found a 2-year survival rate of 65% and long-term HRQOL consequences for many of the survivors. The mixed models showed substantial initial deterioration in the physical dimensions and overall QOL, with subsequent improvement between hospital discharge and 4 months after surgery provided lung cancer did not recur. For those not diagnosed with recurrent disease, there was little change between 4 months and 2 years after surgery in all but the pain scale, which improved during this period. Those diagnosed with recurrent disease deteriorated across all dimensions of HRQOL except emotional functioning. Approximately half of those who were alive and apparently disease free 2 years after surgery continued to experience increased dyspnea and fatigue, and a decline in physical and role functioning relative to their preoperative levels. The mixed models showed that although these dimensions remained below preoperative levels, the difference was only clinically significant for role functioning. This long-term impairment in disease-free survivors was not accounted for by either the extent of surgery or advanced pathologic stage (associated with recurrence but not independently associated with HRQOL). We did not collect the data to examine associations with pulmonary function. Emotional functioning was the dimension least impaired by surgery and quickly returned to preoperative levels. Overall QOL returned to preoperative levels for most of those without recurrence, despite persisting physical impairment; these results provide additional support for the importance of emotional functioning to the overall QOL assessment.19 Unlike studies reporting recovery of preoperative HRQOL levels within six to twelve months,3,5,6 we found that even disease-free survivors continued to report functional limitations and symptoms 2 years after surgery. Our sample was older than that of Zieren et al,6 which may contribute to the different findings, but was similar in age to that of Dales et al5 and Win et al.3 The latter's longitudinal analysis excluded respondents with missing data who may have had worse HRQOL. Some methodologic limitations should be considered when interpreting these results. First, we defined two patient groups by recurrence status at 2 years, analyzing their HRQOL separately to describe patterns in these clinically distinct groups. Although separate analysis is appropriate, it would be inappropriate to make statistical comparisons across these groups, which were defined by an outcome. Second, the regression analysis assumes missing data to be missing at random,20 when it is likely that much of the missing data is related to the respondent's health status. The likely consequence for our results is that potential bias in the recurrence and no-recurrence group analyses differs in both direction and extent. It is generally accepted that those with advanced cancer do not complete surveys when they become too ill,20 leaving only those in better health to provide data. Hence, our results may overestimate HRQOL in the group with recurrence, and thus underestimate the impact of disease progression. In contrast, respondents without recurrence may not complete surveys because they prefer to put the cancer experience behind them. Consequently, our results for this group may underestimate HRQOL and therefore the extent of recovery from surgery. However, the extent of this bias is limited by the low missing data rate in this group (2-year response, 86%). These results can provide useful information for clinicians preparing patients for lung cancer surgery. Although few (if any) would decline surgery, informed decision making requires a full understanding of the potential outcomes (including HRQOL). Information about the continuing HRQOL effects can also contribute to ensuring that appropriate ongoing management and support are provided. Our results indicate that many patients live with persisting cough, dyspnea, and fatigue, and the associated functional limitations. If clinicians understand the range of potential long-term outcomes, they will be alert to patients with ongoing rehabilitation and supportive care needs, ensuring referral to the appropriate services.
The author(s) indicated no potential conflicts of interest.
Conception and design: Patricia M. Kenny, Madeleine T. King, Rosalie C. Viney, Michael J. Boyer Provision of study materials or patients: Brian C. McCaughan Collection and assembly of data: Christine A. Pollicino, Jocelyn M. McLean Data analysis and interpretation: Patricia M. Kenny, Madeleine T. King, Michael J. Boyer Manuscript writing: Patricia M. Kenny, Madeleine T. King, Rosalie C. Viney, Michael J. Boyer, Christine A. Pollicino, Jocelyn M. McLean, Michael J. Fulham, Brian C. McCaughan Final approval of manuscript: Patricia M. Kenny, Madeleine T. King, Rosalie C. Viney, Michael J. Boyer, Christine A. Pollicino, Jocelyn M. McLean, Michael J. Fulham, Brian C. McCaughan
Interpretation of mixed models. This Appendix text explains how to interpret the model parameters presented in Tables 2 and 3. In these models, mean health-related quality of life (HRQOL) is represented as the sum of the intercept and a series of coefficients (fixed effects) multiplied by the value of the relevant effect variable. Thus the mean predicted HRQOL at any particular time point can be derived easily. For example the discharge effect variable is an indicator variable coded 1 for a discharge observation and 0 for all other observations. Therefore, mean HRQOL at discharge would be the sum of the intercept and the discharge coefficient (ie, discharge coefficient multiplied by 1 and all the other coefficients multiplied by 0). The predicted mean HRQOL scores at all time points are illustrated in Appendix Figures A5 and A6 for physical and role functioning.
The standard deviations for the intercept and change per month after 4 months (no-recurrence group) or 1 month (recurrence group) represent the extent of individual variation around the mean (fixed) effects.
Worked example: Patients who experienced a recurrence during the study. The recurrence group (Table 2) models included an indicator variable at hospital discharge, an intercept adjustment at 1 month (indicator coded 0 for observations before 1 month and 1 for all other observations at or after 1 month), and one continuous variable (time after 1 month), so that
Worked example: Patients who did not experience a recurrence during the study. The no-recurrence group (Table 3) models included two indicator variables (discharge and 1 month), a 4-month intercept adjustment (indicator coded 0 for observations before 4 months and 1 for all other observations at or after 4 months), and one continuous variable (time after 4 months), so that
We thank surgeons Paul Bannon, Matthew Bayfield, Bruce French, Nick Hendel, and Michael Wilson, and research nurses Vera Cvetanovksi and Sandra Wojcinski. We also thank the participating patients and staff at Royal Prince Alfred, Concord, Strathfield Private, Liverpool, and St Vincent's Hospitals.
Supported by a National Health and Medical Research Council Project Grant No. 991255. Presented in part at the Australian Health Outcomes Conference, September 15-16, 2004, Canberra, Australia, and the annual conference of the International Society for Quality of Life Research, October 16-19, 2004, Hong Kong, China. Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Manser R, Wright G, Hart D, et al: Surgery for early stage non-small cell lung cancer. Cochrane Database Syst Rev 1:CD004699, 2005[Medline] 2. Nesbitt JC, Putnam JB Jr, Walsh GL, et al: Survival in early-stage non-small cell lung cancer. Ann Thorac Surg 60:466-472, 1995 3. Win T, Sharples L, Wells FC, et al: Effect of lung cancer surgery on quality of life. Thorax 60:234-238, 2005 4. Handy JR Jr, Asaph JW, Skokan L, et al: What happens to patients undergoing lung cancer surgery? Outcomes and quality of life before and after surgery. Chest 122:21-30, 2002[CrossRef][Medline] 5. Dales RE, Belanger R, Shamji FM, et al: Quality-of-life following thoracotomy for lung cancer. J Clin Epidemiol 47:1443-1449, 1994[CrossRef][Medline] 6. Zieren HU, Muller JM, Hamberger U, et al: Quality of life after surgical therapy of bronchogenic carcinoma. Eur J Cardiothorac Surg 10:233-237, 1996[Abstract] 7. Viney RC, Boyer MJ, King MT, et al: Randomized controlled trial of the role of positron emission tomography in the management of stage I and II non-small-cell lung cancer. J Clin Oncol 22:2357-2362, 2004 8. Aaronson NK, Ahmedzai S, Bergman B, et al: The European Organization for Research and Treatment of Cancer QLQ-C30: A quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 85:365-376, 1993 9. Bergman B, Aaronson NK, Ahmedzai S, et al: The EORTC QLQ-LC13: A modular supplement to the EORTC core quality of life questionnaire (QLQ-C30) for use in lung cancer clinical trials. Eur J Cancer 30:635-642, 1994[CrossRef] 10. Fayers P, Aaronson N, Bjordal K, et al: EORTC QLQ-C30 Scoring Manual (ed 3). Brussels, Belgium, European Organisation for Research and Treatment of Cancer, 2001 11. King MT: The interpretation of scores from the EORTC quality of life questionnaire QLQ-C30. Qual Life Res 5:555-567, 1996[CrossRef][Medline] 12. Osoba D, Bezjak A, Brundage M, et al: Analysis and interpretation of health-related quality-of-life data from clinical trials: Basic approach of The National Cancer Institute of Canada Clinical Trials Group. Eur J Cancer 41:280-287, 2005[CrossRef][Medline] 13. Osoba D, Rodrigues G, Myles J, et al: Interpreting the significance of changes in health-related quality-of-life scores. J Clin Oncol 16:139-144, 1998 14. Brown H, Prescott R: Applied Mixed Models in Medicine. Chichester, United Kingdom, Wiley, 1999 15. SAS Institute Inc: SAS/STAT 9.1 User's Guide. Cary, NC: SAS Institute Inc, 2004 16. Fairclough DL: Design and Analysis of Quality of Life Studies in Clinical Trials. Boca Raton, FL, Chapman & Hall/CRC, 2002, pp 54-57; 66-67 17. Singer JD, Willett JB: Applied Longitudinal Data Analysis: Modelling Change and Event Occurrence. New York, NY, Oxford University Press, 2003, pp 190-208 18. Hochberg Y: A sharper Bonferroni procedure for multiple tests of significance. Biometrika 75:800-802, 1988 19. Sarna L, Padilla G, Holmes C, et al: Quality of life of long-term survivors of non-small-cell lung cancer. J Clin Oncol 20:2920-2929, 2002 20. Fairclough DL: Design and Analysis of Quality of Life Studies in Clinical Trials. Boca Raton, FL, Chapman & Hall/CRC, 2002, pp 69-72 Submitted June 19, 2006; accepted October 12, 2007.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2008 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
