The Quality of Life Questionnaire Core 30 (QLQ-C30) and Functional Assessment of Cancer-General (FACT-G) differ in responsiveness, relative efficiency, and therefore required sample size

doi:10.1016/j.jclinepi.2013.02.019

Journal of Clinical Epidemiology

Volume 67, Issue 1, January 2014, Pages 100-107

https://doi.org/10.1016/j.jclinepi.2013.02.019 Get rights and content

Abstract

Objectives

Quality of Life Questionnaire Core 30 (QLQ-C30) and Functional Assessment of Cancer Therapy–General (FACT-G) are widely used cancer-specific health-related quality of life (HRQOL) questionnaires. We aimed to compare their responsiveness with clinically important effects and statistical efficiency to detect such effects.

Study Design and Setting

Secondary analysis of QLQ-C30 and FACT-G data from a randomized controlled trial of Medical Qigong (n = 162 heterogeneous cancer patients). Difference in responsiveness (DR) and relative efficiency (RE) were calculated for five domains.

Results

FACT-G total score was more efficient than QLQ-C30 global scale for detecting change within the intervention arm [RE = 0.31 (0.083, 0.69)] and comparing change between trial arms [RE = 0.17 (0.009, 0.58)]. In the social domain, the QLQ-C30 scale was more responsive [DR = 0.28 (0.024, 0.54)] and more efficient within arm only [RE = 5.25 (1.21, 232.26)]. In the physical, functional/role, and emotional domains, neither questionnaire was more responsive or efficient.

Conclusion

FACT-G would require about one-third the sample of QLQ-C30 to detect a given change in overall HRQOL, whereas in the social domain, it would require five times the sample size. FACT-G won advantage in overall HRQOL by reduced “noise” (smaller standard deviation achieved by summing across 27 items), whereas QLQ-C30 won advantage in the social domain via a larger “signal” (achieved through well-targeted item content).

Introduction

Various criteria come into play when choosing among candidate health-related quality of life (HRQOL) questionnaires [1]. When choosing an HRQOL questionnaire to evaluate the effectiveness of an intervention, responsiveness to a clinically important effect is a key criterion [2]. The more responsive the measure, the smaller the sample size required to detect a given effect [3] or the greater the power for a fixed sample size [4]. Little information is currently available on the relative responsiveness of different HRQOL measures to guide researchers in their choice of such measures.

When selecting an HRQOL questionnaire for a cancer clinical trial, the European Organisation for the Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire Core 30 (QLQ-C30) [5] and Functional Assessment of Cancer Therapy–General (FACT-G) [6] are often obvious candidates [7]. There is a large body of evidence supporting the validity and utility of each across a wide range of clinical research contexts, and both are available in many languages. Luckett et al. [7] provided a thorough comparison of the two measures, highlighting important differences in scale structure, social domains, and tone that may inform choice for any particular study. Luckett et al. concluded that available psychometric evidence does not recommend one questionnaire over the other. Their review highlights the limitations of available evidence about responsiveness, in particular that there are no head-to-head comparisons. The value of a head-to-head comparison is that it allows estimation of both the difference in responsiveness (DR) [4] and the relative statistical efficiency [3] of two measures. The latter is particularly useful as it estimates the factor by which sample size may be reduced if the more responsive measure was used.

Our aim was to compare the responsiveness, statistical efficiency, and power of comparable scales from the EORTC QLQ-C30 and the FACT-G and to quantify the sample size implications.

Section snippets

Data set

We conducted secondary analysis of data from a randomized trial of Medical Qigong (breathing and movement exercises). Details of the trial methods and results are reported elsewhere [8]. In summary, the primary hypothesis of the trial was that patients randomized to Medical Qigong (intervention) would experience significant improvements in HRQOL compared with patients randomized to usual medical care (control). The Medical Qigong program ran for 10 weeks with two group-based supervised

Descriptive statistics of the HRQOL scales

Descriptive statistics for baseline data are shown in Table 2; results for the postintervention data were generally similar. In the physical domain, the FACT-G scale had higher interitem correlation, more items, and hence higher Cronbach α. In contrast, in the emotional domain, the FACT-G scale had lower interitem correlation, and despite more items, lower Cronbach α. In the remaining domains, the effect of lower interitem correlation in the FACT-G was countered by more items, such that

Discussion

We detected differences in responsiveness and statistical efficiency between FACT-G and QLQ-C30 for two of five pairs of scales. The FACT-G total score was more efficient than the QLQ-C30 global scale as a measure of overall HRQOL, for both change within the intervention arm (paired t-test) and for comparing change between trial arms (two-sample t-test). The estimate of RE for the latter case was 0.17, meaning that a randomized trial that used the QLQ-C30 to assess global HRQOL would require a

Acknowledgments

Professor M.T.K. and Dr M.L.B. are supported by the Australian Government through Cancer Australia. Professor P.B. holds an NHMRC Senior Principal Research Fellowship.

References (19)

T. Luckett et al.
Choosing patient-reported outcome measures for cancer clinical research—practical principles and an algorithm to assist non-specialist researchers
Eur J Cancer
(2010)
K.N. Lohr et al.
Evaluating quality-of-life and health status instruments: development of scientific review criteria
Clin Ther
(1996)
M.R. Tuley et al.
Estimating and testing an index of responsiveness and the relationship of the index to power
J Clin Epidemiol
(1991)
T. Luckett et al.
Choosing between the EORTC QLQ-C30 and FACT-G for measuring health-related quality of life in cancer clinical research: issues, evidence and recommendations
Ann Oncol
(2011)
B. Oh et al.
Impact of medical Qigong on quality of life, fatigue, mood and inflammation in cancer patients: a randomized controlled trial
Ann Oncol
(2010)
M.T. King et al.
Meta-analysis provides evidence-based effect sizes for a cancer-specific quality-of-life questionnaire, the FACT-G
J Clin Epidemiol
(2010)
D. Revicki et al.
Recommended methods for determining responsiveness and minimally important differences for patient-reported outcomes
J Clin Epidemiol
(2008)
M.H. Liang et al.
Comparative measurement efficiency and sensitivity of five health status instruments for arthritis research
Arthritis Rheum
(1985)
N.K. Aaronson 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
(1993)

There are more references available in the full text version of this article.

Cited by (35)

Qualitative analysis of gynecologic oncology patients’ experience with treatment holidays
2022, Gynecologic Oncology Reports
Patients with advanced or recurrent gynecologic malignancies occasionally take breaks from systemic treatment colloquially referred to as “treatment holidays” or “chemotherapy holidays.” There are no data from the patient perspective that help describe this experience.
Patients with recurrent or advanced primary gynecologic malignancies who had decided to enter a treatment holiday were recruited and interviewed. A treatment holiday was defined as a planned temporary break or delay in treatment for a patient with recurrent or advanced primary gynecologic malignancy for reasons other than pursuit of hospice or best supportive care, research protocol violation or unacceptable toxicity. Interviews were audiotaped, transcribed and then analyzed using an inductive thematic analysis.
Of 6 total patients identified for participation, 5 completed interviews with ages ranging from 57 to 80 years. Two participants returned to their previous treatment regimen after their holiday therapy, two switched therapies, and one remained on an extended break from systemic treatment. Treatment holidays were experienced as a break from the physical and psychological routine of being a cancer patient, but also brought about feelings of a lack of structure, uncertainty, and led to a confrontation with mortality issues. Overall, participants had favorable experiences which were initiated by their providers in whom they had a deep sense of trust.
Patients experience treatment holidays as a positive and valuable break from the physical and psychosocial routine of cancer treatment and illness. These experiences produce distinct emotional needs that clinicians should address to best support patients electing treatment holidays.
Rice bran arabinoxylan compound and quality of life of cancer patients (RBAC-QoL): Study protocol for a randomized pilot feasibility trial
2020, Contemporary Clinical Trials Communications
Citation Excerpt :
QLQ-C30 is recommended over FACT-G when social activities, financial impact, and symptom scales are outcome measures of interest in a clinical trial [48]. However, the global QoL scale of QLQ-C30 is less responsive than the FACT-G total score and thus requires more trial participants to detect changes in overall QoL [49]. With the symptom scales being one of the primary outcomes of interest in this research, QLQ-C30 is the QoL instrument of choice.
Rice bran arabinoxylan compound (RBAC) is a nutraceutical for enhancing a depleted immune system during and after cancer treatment. This pilot feasibility trial aims to evaluate the effects of RBAC on cancer patients' quality of life during active treatment, compared to placebo, using a validated questionnaire. Other outcome measures include changes in inflammatory and nutritional status, cytokine profile, and gut microbiota.
The study will recruit 50 participants from a regional cancer center in Australia. Patients aged 18–70, diagnosed with solid organ cancers stage II and above, and currently undergoing active systemic therapies, are eligible. Random allocation of participants into two groups is stratified based on metastatic status and treatment type. The dosage is either 3 g/day of RBAC or placebo in identical packaging. The participants, study coordinator, and treating oncologists are blinded to the interventions. Data collections are at baseline and at four follow-up sessions, which are six weeks apart (24 weeks). Statistical analysis will involve a protected p-value with multiple dependent values and analyzed by ANOVA with repeated measures on the occasion of testing and with both a full Bonferroni or Sidak corrections applied to protect against Type I errors. Any observed significance warrants further analysis with pairwise comparisons. Analysis of covariance will also be performed to assess any influence of the demographic data, cancer diagnosis, as well as changes in physical activity, dietary habits, and complementary medicine usage. Comparisons of gut microbiota will be based on the analysis of the fecal microbiome using 16S ribosomal ribonucleic acid amplicon sequencing. The proposed research timeline is from October 2018 to May 2022.
ANZCTR. Reg No: ACTRN12619000562178p.
Impact of lorlatinib on patient-reported outcomes in patients with advanced ALK-positive or ROS1-positive non-small cell lung cancer
2020, Lung Cancer
Citation Excerpt :
The EORTC QLQ-C30 [19] consists of 30 questions that are incorporated into five functional scales (physical, role, cognitive, emotional and social), a global QoL scale, three symptom scales, and six single symptom items. It is one of the most widely used measures of cancer-specific QoL [20–22]. The EORTC QLQ-LC13 consists of one multi-item scale and nine single items that assess specific symptoms, side effects and pain medication use of lung cancer patients.
To evaluate patient-reported outcomes (PROs) from a phase 1/2 study (NCT01970865) in patients with anaplastic lymphoma kinase (ALK)- or ROS1-positive advanced non-small cell lung cancer (NSCLC) treated with lorlatinib 100 mg once daily.
PRO measures, including global quality of life (QoL), functioning domains and symptoms, were assessed by the European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire Core 30 (QLQ-C30) and the 13-item Lung Cancer (QLQ-LC13) module. Mean changes of absolute scores from baseline were assessed. Percentages of patients showing improvement, stability or worsening on each scale were reported, with a change of ≥10 points considered clinically meaningful (CM).
255 patients completed baseline and ≥1 post-baseline PRO assessment. Most patients had CM improvement (42.4 %) or stable (38.0 %) scores for global QoL. Functioning domains with the greatest proportion of patients with improved scores were role (37.6 %) and emotional (36.9 %); only one domain had more patients showing worsening than improving function (cognitive [24.3 % vs 22.4 %]). Most patients showed improved or stable scores for disease-related symptoms. No QLQ-C30 symptom domains had more patients worsening than improving. Symptoms on the QLQ-C30 scale with the greatest proportion of patients with improved scores were fatigue (49.4 %) and insomnia (46.3 %). Four QLQ-LC13 domains had more patients worsening than improving (two most affected were peripheral neuropathy [37.3 % vs 13.7 %] and alopecia [19.2 % vs 13.3 %]). Symptoms on the QLQ-LC13 scale with the greatest proportion of patients with improved scores were cough (42.7 %) and pain in other parts (32.9 %).
Lorlatinib treatment showed CM improvement from baseline in global QOL that was maintained over time. Additionally, there were improvements in physical, emotional, social, and role functioning. Improvements were shown in appetite loss and key symptoms such as pain, dyspnea, cough and fatigue; a worsening in peripheral neuropathy was noted.
Validation of the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 Summary Score in Patients With Hematologic Malignancies
2019, Value in Health
Citation Excerpt :
In that study, the QLQ-C30 Summary Score showed the second largest group difference, with only the PF scale being more discriminative. The good performance of the global QOL scale observed for our disease-specific comparisons in AML and MDS patients differs from that seen in previous studies in patients with solid tumors,7,32,33 which questioned the sensitivity of the global QOL scale. Although this finding warrants further research, we believe that, overall, our results support the strategy of employing the QLQ-C30 Summary Score as a study endpoint in cases where previous knowledge does not allow for domain-specific hypotheses or when the sample being studied includes mixed hematologic populations.
We investigated the validity of the recently developed European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) summary score in patients with hematologic malignancies. Specifically, we evaluated the adequacy of a single-factor measurement model for the QLQ-C30, and its known-groups validity and responsiveness to change over time.
We used confirmatory factor analysis to test the single-factor model of the QLQ-C30, using baseline QLQ-C30 data (N = 2134). The QLQ-C30 summary score was compared to the original QLQ-C30 scales using general (age, sex, Eastern Cooperative Oncology Group performance status, comorbidity) and disease-specific (red blood cell transfusion dependency) groups. Repeated measurements allowed us to investigate responsiveness to change in a subgroup of patients with acute myeloid leukemia.
The single-factor model of the QLQ-C30 exhibited adequate fit in patients with hematologic malignancies. Known-group comparisons generally supported the construct validity of the summary score when using more general grouping variables (sociodemographics, broad clinical parameters). Nevertheless, when groups were formed on the basis of disease-specific variables (eg, transfusion dependency), the summary score performed less well the some of the original, separate scales of the QLQ-C30.
Our findings provide support for the validity of the single-factor model of the EORTC QLQ-C30 in patients with hematologic malignancies. Specifically, the results suggest that the summary score can be used as an endpoint in this population when symptom- or other health domain-specific hypotheses are not available.
Development of the EORTC QLQ-CAX24, A Questionnaire for Cancer Patients With Cachexia
2017, Journal of Pain and Symptom Management
Citation Excerpt :
Although there may be some modification to the QLQ-CAX24 after Phase 4, through a reduction in length and adjustment to the scale structure, items are unlikely to change, so it is now available for use in clinical trials. Whether clinicians and researchers should choose to use the QLQ-CAX24 or the FAACT, the only other instrument for the assessment of HRQOL in patients with cancer cachexia will partly be determined by which of the core questionnaires from the EORTC and Functional Assessment of Chronic Illness Therapy measurement systems is most applicable to their requirements.16,27 An additional consideration is that the FAACT is a single scale whereas the QLQ-CAX24 comprises scales allowing more precise hypothesis testing.
Cachexia is commonly found in cancer patients and has profound consequences; yet there is only one questionnaire that examines the patient's perspective.
To report a rigorously developed module for patient self-reported impact of cancer cachexia.
Module development followed published guidelines. Patients from across the cancer cachexia trajectory were included. In Phase 1, health-related quality of life (HRQOL) issues were generated from a literature review and interviews with patients in four countries. The issues were revised based on patient and health care professional (HCP) input. In Phase 2, questionnaire items were formulated and translated into the languages required for Phase 3, the pilot phase, in which patients from eight countries scored the relevance and importance of each item, and provided qualitative feedback.
A total of 39 patients and 12 HCPs took part in Phase 1. The literature review produced 68 HRQOL issues, with 22 new issues arising from the patient interviews. After patient and HCP input, 44 issues were formulated into questionnaire items in Phase 2. One hundred ten patients took part in Phase 3. One item was reworded, and 20 items were deleted as a consequence of patient feedback.
The QLQ-CAX24 is a cancer cachexia-specific questionnaire, comprising 24 items, for HRQOL assessment in clinical trials and practice. It contains five multi-item scales (food aversion, eating and weight-loss worry, eating difficulties, loss of control, and physical decline) and four single items.
Replication and validation of higher order models demonstrated that a summary score for the EORTC QLQ-C30 is robust
2016, Journal of Clinical Epidemiology
Citation Excerpt :
However, in many studies, it may be difficult to prespecify which QLQ-C30 scales are of most interest. In such cases, investigators frequently rely on the two-item scale assessing overall quality of life [5–8]. The disadvantage of this very brief two-item overall quality of life scale is that it may have less measurement precision than is desired for detecting group differences over time.
To further evaluate the higher order measurement structure of the European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire Core 30 (QLQ-C30), with the aim of generating a summary score.
Using pretreatment QLQ-C30 data (N = 3,282), we conducted confirmatory factor analyses to test seven previously evaluated higher order models. We compared the summary score(s) derived from the best performing higher order model with the original QLQ-C30 scale scores, using tumor stage, performance status, and change over time (N = 244) as grouping variables.
Although all models showed acceptable fit, we continued in the interest of parsimony with known-groups validity and responsiveness analyses using a summary score derived from the single higher order factor model. The validity and responsiveness of this QLQ-C30 summary score was equal to, and in many cases superior to the original, underlying QLQ-C30 scale scores.
Our results provide empirical support for a measurement model for the QLQ-C30 yielding a single summary score. The availability of this summary score can avoid problems with potential type I errors that arise because of multiple testing when making comparisons based on the 15 outcomes generated by this questionnaire and may reduce sample size requirements for health-related quality of life studies using the QLQ-C30 questionnaire when an overall summary score is a relevant primary outcome.

View all citing articles on Scopus

: Conflict of interest: The authors have no conflicts of interest to disclose. There are no financial arrangements between any author and the custodians of the FACT-G or the QLQ-C30; no other products or competing products play any role in this manuscript.

View full text

Original ArticleThe Quality of Life Questionnaire Core 30 (QLQ-C30) and Functional Assessment of Cancer-General (FACT-G) differ in responsiveness, relative efficiency, and therefore required sample size

Abstract

Objectives

Study Design and Setting

Results

Conclusion

Introduction

Section snippets

Data set

Descriptive statistics of the HRQOL scales

Discussion

Acknowledgments

Eur J Cancer

Clin Ther

J Clin Epidemiol

Ann Oncol

Ann Oncol

J Clin Epidemiol

J Clin Epidemiol

Comparative measurement efficiency and sensitivity of five health status instruments for arthritis research

Arthritis Rheum

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

Original Article
The Quality of Life Questionnaire Core 30 (QLQ-C30) and Functional Assessment of Cancer-General (FACT-G) differ in responsiveness, relative efficiency, and therefore required sample size