More Concern About Transfusion Requirement When Evaluating Quality of Life in Anemic Patients

Article Tools SPECIAL DEPARTMENTS Article Tools OPTIONS & TOOLS Export Citation Track Citation Add To Favorites Rights & Permissions COMPANION ARTICLES No companion articles ARTICLE CITATION DOI: 10.1200/JCO.2002.20.14.3182 Journal of Clinical Oncology - published online before print November 1, 2016 PMID: 12118036 More Concern About Transfusion Requirement When Evaluating Quality of Life in Anemic Patients Esther N. OlivaxEsther N. OlivaSearch for articles by this author , Arianna D’AngeloxArianna D’AngeloSearch for articles by this author , Bruno MartinoxBruno MartinoSearch for articles by this author , Francesco NobilexFrancesco NobileSearch for articles by this author , Borislav D. DimitrovxBorislav D. DimitrovSearch for articles by this author , Annalisa PernaxAnnalisa PernaSearch for articles by this author Janice L. GabrilovexJanice L. GabriloveSearch for articles by this author , Brenda SarokhanxBrenda SarokhanSearch for articles by this author , Pierre CremieuxxPierre CremieuxSearch for articles by this author Show More Azienda Ospedaliera “Bianchi-Melacrino-Morelli”, Reggio Calabria, ItalyClinical Research Centre for Rare Diseases “Aldo e Cele Daccò”, Mario Negri Institute for Pharmacological Research, Ranica (BG), Italy, Medical University*, Plovdiv, BulgariaMount Sinai Medical Center, New York, NYOrtho Biotech Products, L.P., Raritan, NJAnalysis Group/Economics, Cambridge, MA https://doi.org/10.1200/JCO.2002.20.14.3182 First Page Full Text PDF Figures and Tables © 2002 by American Society of Clinical OncologyjcoJ Clin OncolJournal of Clinical OncologyJCO0732-183X1527-7755American Society of Clinical OncologyResponse15072002In Reply:Drs Oliva et al have raised two relevant issues that pertain to our community-based study of epoetin alfa in anemic cancer patients, providing further insight into the interpretation of the published results.1 First, they point out that establishing the internal validity of quality-of-life (QOL) questionnaires is important in determining the reliability of the QOL analyses. Indeed, while such consistency results were not reported in the original article because of space constraints, we had performed a Cronbach’s alpha calculation for the 20-item anemia subscale of the Functional Assessment of Cancer Therapy-Anemia (FACT-An) scale. The FACT-An questions were grouped into two domains: 13 fatigue-related items and seven items not pertaining to fatigue. Overall and fatigue-specific Cronbach’s alpha values were very high (> 0.9), as shown in Table 1. This finding suggests that the FACT-An was reliable and provided stable and consistent responses across individual items within the questionnaire. Furthermore, the correlation between the linear analog scale assessment (LASA) QOL instrument and the FACT-An subscale was high (r = .66, P < .05). Table 1. Internal Consistency of FACT-An AnalysisTable 1. Internal Consistency of FACT-An AnalysisFACT-An (20 items)Fatigue (13 items)Nonfatigue (7 items)Cronbach’s alpha at baseline0.9200.9290.627Cronbach’s alpha at month 20.9330.9400.661Cronbach’s alpha at completion0.9400.9470.685Oliva et al also suggested that the reported relationship between hemoglobin (Hb) and QOL may have stemmed from a change in transfusion use. In other words, it is possible that patients with higher Hb values reported increased QOL simply because they received fewer, if any, transfusions. To examine the merit of this assertion, we conducted two analyses, reported in Tables 2 and 3. In the first multivariate analysis, we compared the relationship between Hb and overall QOL scores from the LASA for clinic visits during which no transfusion was received with visits during which at least one unit of blood was transfused. Patients’ overall QOL scores during each visit were regressed with age, sex, and Hb as covariates. If transfusion was the mechanism through which Hb affected QOL, we should have observed (1) a significant and large coefficient on the Hb variable among patients who received transfusions and (2) a small and insignificant coefficient among patients who did not receive transfusions. However, the coefficients on the Hb variable were similar in size and not statistically significantly different from each other. Similar results were observed for the energy and activity QOL parameters (data not shown). Table 2. Comparative Panel Regression AnalysisTable 2. Comparative Panel Regression AnalysisFactorPatient Visits With No TransfusionPatient Visits With TransfusionEstimateSDPEstimateSDPNOTE. A t test showed that the regression coefficients associated with Hb were not statistically significantly different between the two models (P = .7188).Intercept31.463.24< .000124.505.27< .0001Time0.430.06< .00010.540.10< .0001Age−0.110.03.0002−0.110.06.0884Female0.840.83.31460.351.56.8227Hb2.400.25< .00012.560.38< .0001 Table 3. Multivariate Linear Regression Model on the Relationship Between Transfusion, Hb, and Overall QOLTable 3. Multivariate Linear Regression Model on the Relationship Between Transfusion, Hb, and Overall QOLFactorEstimateSDPIntercept32.193.04< .0001Time0.470.05< .0001Age−0.110.03< .0001Female0.680.75.36Hb2.310.24< .0001Transfusion−8.494.00.03Transfusion*Hb0.460.38.22A second analysis based on multivariate analysis of all patients is reported in Table 3 and confirmed the first results. In this analysis, both transfusions and the interaction of transfusions with Hb were included as additional covariates. If QOL is more sensitive to changes in Hb among patients with low Hb who receive transfusions, then the interaction term should be highly significant (thereby indicating that the relationship between QOL and Hb is greatest among patients with low Hb who received transfusions). However, the interaction variable between Hb and transfusion use is small and insignificant, thereby confirming that the effect of Hb on QOL was the same regardless of whether patients received transfusions.Taken together, these results indicate that the direct relationship between Hb and QOL described in our article was present regardless of RBC transfusion.REFERENCE1. Gabrilove JL, Cleeland CS, Livingston RB, et al: Clinical evaluation of once-weekly dosing of epoetin alfa in chemotherapy patients: Improvements in hemoglobin and quality of life are similar to three-times-weekly dosing. J Clin Oncol 19:: 2875,2001-2882, Link, Google Scholar