Original Paper Using Reference Data on Quality of LifeÐthe Importance of Adjusting for Age and Gender, Exempli®ed by the EORTC QLQ-C30 (+3) M.J. Hjermstad, 1 P.M. Fayers, 2 K. Bjordal 3 and S. Kaasa 2,4 1 The Norwegian Cancer Society and Department of Medical Oncology and Radiotherapy, The Norwegian Radium Hospital, Montebello, N-0310,Oslo; 2 Unit for Applied Clinical Research, Faculty of Medicine, University of Science and Technology, Trondheim; 3 Department of Oncology, Trondheim University Hospital; and 4 Palliative Medicine Unit, Department of Oncology, Trondheim University Hospital, Trondheim, Norway Interpretation of health related quality of life (HRQOL) results in cancer patients is facilitated by knowledge of the levels of HRQOL in the general population. However, direct comparisons can be misleading unless age and gender are considered. We demonstrate the derivation of age- and gender- speci®c `expected' values from population reference values by means of simple calculations. This survey included 3000 randomly selected Norwegians above 18 years of age who received the European Organization for Research and Treatment of Cancer Core Quality of Life Questionnaire (EORTC QLQ-C30 (+ 3)) by mail. 1965 responses from 2,892 eligible persons (68%) were received. The popu- lation was divided into six disease groups based on self-reported health problems. The observed mean scale scores of the diVerent groups deviated greatly from those obtained in the general population. The score for physical function, for example, was 72 for cancer patients and ranged from 73.3 to 82.5 in other disease groups, as opposed to 89.9 in the general population and 98.9 in those with no health problems. The range for one of the quality of life (QOL) scales was 57.7 to 84.7 compared with 73.7 in the general population. Expected mean scores for the patient groups were computed from the refer- ence values, based on the concept of equivalence of age and gender. The diVerences between the observed mean scores and the reference values were strongly mediated by this method. The expected scores for physical function then ranged from 83.3 to 93.1 and from 70.3 to 75 for the QOL scale. The impact of age and gender on the reference data from the EORTC QLQ-C30 (+ 3) obtained in a general population shows that these variables must be considered when interpreting data on HRQOL for cancer patients. The demonstration of how to generate mean values which are adjusted according to the age and gender distribution of a population should increase the usefulness of this questionnaire among clinicians. # 1998 Elsevier Science Ltd. All rights reserved. Key words: health related quality of life, reference data, clinical importance, EORTC QLQ-C30 Eur J Cancer, Vol. 34, No. 9, pp. 1381±1389, 1998 INTRODUCTION Cancer-specific questionnaires, like the European Organi- zation for Research and Treatment of Cancer Core Quality of Life Questionnaire (EORTC QLQ-C30) [1], the Rotterdam Symptom Checklist (RSCL) [2], the Functional Living Index Cancer (FLIC) [3], the Functional Assessment of Cancer Therapy Scale (FACT) [4] and the Cancer Rehabilitation Evaluation System (CARES) [5], have been developed for assessment of health related quality of life (HRQOL) in vari- ous cancer populations and for comparisons across diagnoses and stages of disease. Reference data from the general population on generic HRQOL questionnaires like the Sickness Impact Pro®le (SIP) [6] and the SF-36 in particular [7], have been derived European Journal of Cancer, Vol. 34, No. 9, pp. 1381±1389, 1998 # 1998 Elsevier Science Ltd. All rights reserved Pergamon Printed in Great Britain PII: S0959-8049(98)00136-1 0959-8049/98 $19.00+0.00 1381 Correspondence to M.J. Hjermstad. Received 17 Oct. 1997; revised 13 Feb. 1998; accepted 25 Mar. 1998.