SEARCH

SEARCH BY CITATION

Keywords:

  • patient-reported outcomes;
  • oncology;
  • health-related quality of life;
  • reference values

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES
  10. Supporting Information

BACKGROUND

Health-related quality of life (HRQOL) measures are commonly used in oncology research. Interest in their use for monitoring or screening is increasing. The Functional Assessment of Cancer Therapy (FACT) is one of the most widely used HRQOL instruments. Consequently, oncology researchers and practitioners have an increasing need for reference values for the Functional Assessment of Cancer Therapy–General (FACT-G) and its 7-item rapid version, the Functional Assessment of Cancer Therapy–General 7 (FACT-G7), to compare FACT scores across specific subgroups of patients in research trials and practice. The objectives of this study are to provide 1) reference values from a sample of the general US adult population and a sample of adults diagnosed with cancer and 2) cutoff scores for quality of life.

METHODS

A sample of the general US population (N = 1075) and a sample of patients with cancer from 12 studies (N = 5065) were analyzed. Cutoff scores were established using distribution- and anchor-based methods. Mean values for the cancer sample were analyzed by performance status, cancer type, and disease status. Also, t tests and established criteria for meaningful differences were used to compare values.

RESULTS

FACT-G and FACT-G7 scores in the general US population sample and cancer sample were generally comparable. Among the sample of patients with cancer, FACT-G and FACT-G7 scores worsened with declining performance status and increasing disease status.

CONCLUSIONS

These data will aid interpretation of the magnitude and meaning of FACT scores, and allow for comparisons of scores across studies. Cancer 2014;120:2902–2909. © 2014 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES
  10. Supporting Information

As cancer treatments continue to extend the survival of patients, researchers and oncologists increasingly focus on integrating patient-reported outcomes in clinical trials and in oncology practice.[1, 2] The National Cancer Institute now recommends that clinical trial endpoints include patient-reported outcomes (PROs).[3, 4] Health-related quality of life (HRQOL) is an inclusive PRO incorporating symptoms and functioning in physical, psychological, and social domains. HRQOL assessments supplement more traditional clinical endpoints such as survival and tumor response,[5] and measuring HRQOL has been associated with improved communication regarding symptom burden, salutary outcomes such as reduced symptom distress, and better symptom management.[6-8]

The Functional Assessment of Chronic Illness Therapy (FACIT; http://www.facit.org), includes the widely-used Functional Assessment of Cancer Therapy–General, commonly referred to as the FACT-G.[9] Recently, a rapid, 7-item version of the FACT-G was published to provide clinicians and researchers with a brief version of the FACT-G.[10] This rapid version, labeled the FACT-G7, is a good option for estimating the HRQOL (ie, FACT-G score) of patients in clinical practice or research investigations, with or without additional disease- or treatment-specific assessments.

Despite the wide use of the FACT-G, interpretation is difficult if scores cannot be anchored to meaningful reference values. Availability of reference values would also facilitate calculation of standardized effect sizes, which can be useful when comparing effect sizes across heterogeneous studies.[11] A previous report provided normative data for the FACT-G from a sample of 1075 adults in the general US population and a sample of 2236 adult individuals diagnosed with cancer.[12] That report had insufficient data to provide FACT scores by performance status, cancer type, or disease status and did not provide guidance regarding clinically useful cutoff scores. This information can prove useful when interpreting research results or clinical change scores across specific subgroups of oncology patients, because since then, considerably more data have been collected, enabling further stratification of reference scores by these key variables. Furthermore, reference value data from a large sample of adults with cancer has not been published for the newer, rapid 7-item FACT-G.[10] Reference values also provide meaningful information for researchers interested in comparative effectiveness research.

The objectives of this study were: 1) to provide reference value data for the FACT-G and FACT-G7 measures across 2 reference groups (adults diagnosed with cancer and the general US adult population), and 2) To provide cutoff values for poor HRQOL to guide in the interpretation of FACT scores. These reference values include information by performance status categories, cancer types, and disease status.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES
  10. Supporting Information

US Adult General Sample

The US adult general population data were collected by Knowledge Networks (Menlo Park, Calif), a survey research firm. The data were drawn from a random sample of 1400 people, age 18 years and older, from more than 100,000 individuals who were members of an Internet-based survey panel. The panel was a demographically representative sample of the general US adult population compared to the 2010 US Census. African American and Hispanic individuals were oversampled. The participants were similar in sex, age, race/ethnicity, and education to the general US adult population. Details of the Knowledge Networks approach, as compared to the more traditional random digit dialing approach to sampling, are available.[13, 14] Panel members already had a demographic and illness history on file.

Adult Cancer Patient Sample

The current study merged and analyzed data from 12 different studies in which adult participants diagnosed with cancer completed FACT items. The 12 studies employed a range of eligibility criteria, some convenience samples and some with tighter criteria. They were drawn from available data sets in our possession, pooled to provide a wide range of people with various cancer diagnoses. All studies were approved by institutional review board. See Supporting Table 1 for a description of the studies (see online supporting information).

Measures

The FACT-G is a 27-item instrument containing four subscales: Physical Well-Being (PWB; 7 items), Functional Well-Being (FWB; 7 items), Social Well-Being (SWB; 7 items), and Emotional Well-Being (EWB; 6 items) on a 5-point Likert-type scale ranging from 0 (not at all) to 4 (very much) with a recall period of the past 7 days. Responses are summed to create a total FACT-G score and individual subscale scores, with higher scores reflecting better HRQOL. The highest possible score is 28 for the PWB, SWB, and FWB subscales, 24 for the EWB subscale, and 108 for the FACT-G total score. The FACT-G has demonstrated good validity and reliability in cancer and general US population samples.[12]

The FACT-G7 is a rapid, 7-item version of FACT-G that contains questions from the PWB, FWB, and EWB subscales. Scores range from 0 to 28. The FACT-G7 has demonstrated good validity and reliability in cancer and general US population samples.[10]

Two of the 12 samples comprising the adult cancer patient sample completed the 27-item FACT-G Version 3 and for consistency were scored using rules applied to Version 4 (see www.facit.org for scoring rules and information on versions 3 and 4). Respondents from the US general population sample did not complete 6 of the 27 FACT-G items, because these items addressed issues specific to illness or treatment and were not considered appropriate to administer to a general US adult population sample. To ensure that the range of scores for this 21-item version of the FACT-G would be equivalent to scores for the complete 27-item instrument, the 4 subscales scores (PWB, SWB, EWB, FWB) were prorated.

Statistical Analyses

Scores on the FACT-G (including subscales) and FACT-G7 were analyzed by performance status, cancer type, and disease status. Performance status was assessed by the Eastern Cooperative Oncology Group Performance Status Rating (ECOG PSR). The ECOG PSR is a widely used scale to classify patient functional status, ranging from 0 (“normal activity without symptoms”) to 4 (“unable to get out of bed”).[15] After examining the entire sample, we elected to analyze the data by the 4 most prevalent cancer types: breast, lung, colorectal, and prostate.[16] These 4 cancer types account for more than 800,000 new cancer diagnoses per year, and 48% of all cancer cases, in the United States.[17] In addition, we analyzed the data by disease status, dividing respondents into those with localized disease, those with regional spread, and those with metastatic disease. Disease status was determined by medical chart review or patient self-report, depending on the specific study procedures.

Statistical Differences

T tests were used to determine statistically significant mean score differences between patients with cancer and the US general population and between clinical subgroups within the cancer sample. Multiple comparisons were adjusted for using a Tukey's test. For all statistical tests, a significance level of P < .05 was used.

Meaningfulness of Differences

We relied on previously established meaningful differences. Because established estimates were not available for the FACT-G7, we used distribution- and anchor-based methods to derive meaningful differences for this measure. Previous research has suggested 2- to 3-point differences on the FACT-G subscale scores and 4- to 7-point differences on the FACT-G total score are associated with meaningful differences in clinical and subjective anchors.[18-20] With regard to the FACT-G7, data from Yanez et al[10] are consistent with setting 2 to 3 points as a meaningful difference. The range of observed 0.33 standard deviation and 1.0 standard error of the mean in the current data was 1.83 to 2.57 for the FACT-G7. This has been suggested as a likely magnitude associated with meaningful difference and change, and was closest to a combined ECOG PSR of 2/3/4.[20, 21] As a result, we set 2 to 3 points on the FACT-G7 as a difference score likely to be important for group change.

Cutoff Scores

To establish individual and group cutoff scores for low HRQOL, we examined the distribution of the scores. We mapped distribution-based values onto clinically meaningful anchors such as the mean values associated with ECOG PSR categories and compared to the distribution of scores. This allowed us to make clinically meaningful comparisons between cutoff scores on the FACT scales.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES
  10. Supporting Information

Reference Values

Table 1 presents demographic and clinical characteristics for the sample of patients with cancer and the US general population sample. Table 2 compares the total score and subscales of the FACT-G and total score on the FACT-G7 for the adult cancer patient sample and US general adult sample. With the exception of the SWB subscale, which was worse among the general adult US sample than the sample of patients with cancer, scores on the FACT-G and FACT-G7 were comparable across the 2 samples. Internal consistency (ie, Cronbach's alphas) for the FACT-G and FACT-G7 were acceptable (eg, > .70).[22] Given the expected age difference between the general population and cancer samples, we calculated the effect size for these differences in age on FACT means among the cancer sample. These effect sizes were small (ie, generally below 0.2 and occasionally in the 0.2-0.4 range) and therefore we elected not to pursue additional age-related analyses.

Table 1. Adult Cancer Patient Sample and Adult US General Sample Demographic and Clinical Characteristics
CharacteristicAdult Cancer Patient Sample N = 5065Adult US General Sample N = 1075
N or Mean% or SDN or Mean% or SD
  1. Abbreviations: ECOG PSR, Eastern Cooperative Oncology Group performance status rating; N/A, not applicable; SD, standard deviation.

Demographics    
Male238347.053149.4
Age, y (SD)56.813.345.916.6
Race/ethnicity    
Hispanic109421.6
White (including Hispanic)81675.9
Non-Hispanic white271053.5
Non-Hispanic black98719.511010.2
Other961.9918.2
Not collected/missing1783.5585.4
Education    
Less than high school129625.712411.5
High school212641.973368.3
College or graduate degree128925.421119.6
Not collected3547.070.7
ECOG PSR    
Normal activity, without symptoms173234.2
Some symptoms, but do not require bed rest during waking day153930.4
Require bed rest for <50% of waking day113522.4
Require bed rest for >50% of waking day3907.7
Unable to get out of bed250.5
Not collected/missing2444.8
Cancer Diagnosis  N/AN/A
Breast cancer126324.9  
Lung cancer64112.7  
Prostate cancer4869.6  
Colorectal cancer61112.1  
Leukemia1112.2  
Head and neck cancer58211.5  
Kidney cancer1893.7  
Other cancer118223.3  
Current Stage  N/AN/A
No evidence of disease79915.7  
Local58711.6  
Regional61612.2  
Distal81916.2  
N/A2655.2  
Not collected189837.5  
Missing811.6  
Table 2. Adult Cancer Patient Sample and US General Adult Sample
 FWBPWBEWBSWBaFACT-GFACT-G7b
  1. a

    Due to the observed meaningful differences between the cancer and the US general adult sample on the SWB scale, we removed the 2 cancer-specific items for the cancer FACT-G and rescored the subscale using the same scoring methods used for the general population sample. The new mean of 20 remained greater than the mean on the general population mean.

  2. b

    The range for meaningful differences for the FACT-G is 4-7 points. The range for meaningful differences for both the FACT-G subscales and the FACT-G7 is 2-3 points.

  3. Abbreviations: EWB, Emotional Well Being; FACT-G, Functional Assessment of Cancer Therapy–General; FACT-G7, Functional Assessment of Cancer Therapy–General 7-item; FWB, Functional Well Being; PWB, Physical Well Being; SWB, Social Well Being.

Adult Cancer Patient Sample
N491849314918493049124937
Mean18.221.018.122.079.319.1
Standard deviation6.66.04.65.317.05.5
Cronbach's alpha0.870.860.770.780.910.78
US General Adult Sample
N107510751075107510751075
Mean18.522.719.919.180.120.1
Standard deviation6.85.44.86.818.15.4
Cronbach's alpha0.870.840.800.800.910.81

Table 3 contains scores by ECOG PSR category for the sample of patients with cancer. Results indicate that scores on the FACT-G and FACT-G7 decrease (eg, worsen) with declining ECOG PSR category. The FACT-G and FACT-G7 significantly differentiated between ECOG PSR categories (P values < .05), with small-to-medium effect sizes. The score differences between all adjacent categories for the FACT-G total score, PWB, FWB and FACT-G7 exceeded the respective meaningful difference thresholds.

Table 3. Adult Cancer Patient Sample Known Group Differences by ECOG PSR (Entire Sample, N = 5065)
ECOG PSR CategoriesaNMeanSDAdjacent Category DifferenceP ValuebEffect Sizec
  1. a

    This sample had 244 observations missing for ECOG variable.

  2. b

    The Tukey multiple comparison adjustment was used for P values.

  3. c

    Adjacent category mean difference divided by overall standard deviation for the scale or subscale (Table 2).

  4. d

    The last 2 categories of ECOG PSR were collapsed due to smaller cell sizes for these categories.

  5. Abbreviations: ECOG PSR, Eastern Cooperative Oncology Group performance status rating; EWB, Emotional Well Being; FACT-G, Functional Assessment of Cancer Therapy–General; FACT-G7, Functional Assessment of Cancer Therapy–General 7-item; FWB, Functional Well Being; PWB, Physical Well Being; SD, standard deviation; SWB, Social Well Being.

FACT-G      
Normal activity, without symptoms169387.814.29.0<.0010.53
Some symptoms, but do not require bed rest during waking day150678.815.27.7<.0010.45
Require bed rest for <50% of waking day108971.115.49.8<.0010.58
Require bed rest for >50% of waking day/unable to get out of bedd38461.315.2   
FWB      
Normal activity, without symptoms169121.55.83.6<.0010.55
Some symptoms, but do not require bed rest during waking day150517.96.02.8<.0010.42
Require bed rest for <50% of waking day109515.15.93.0<.0010.45
Require bed rest for >50% of waking day/unable to get out of bed38712.15.5   
PWB      
Normal activity, without symptoms169824.34.23.3<.0010.55
Some symptoms, but do not require bed rest during waking day151021.05.13.3<.0010.55
Require bed rest for <50% of waking day109717.75.64.4<.0010.73
Require bed rest for >50% of waking day/unable to get out of bed38513.35.8   
EWB      
Normal activity, without symptoms169419.44.11.3<.0010.28
Some symptoms, but do not require bed rest during waking day150318.14.51.1<.0010.24
Require bed rest for <50% of waking day109417.04.71.2<.0010.26
Require bed rest for >50% of waking day/unable to get out of bed38715.85.0   
SWB      
Normal activity, without symptoms170022.75.30.9<.0010.17
Some symptoms, but do not require bed rest during waking day151021.85.30.60.010.11
Require bed rest for <50% of waking day109421.25.21.10.0040.21
Require bed rest for >50% of waking day/unable to get out of bed38520.15.5   
FACT-G7      
Normal activity, without symptoms170022.04.43.2<.0010.58
Some symptoms, but do not require bed rest during waking day151118.84.82.5<.0010.45
Require bed rest for <50% of waking day109616.35.02.6<.0010.47
Require bed rest for >50% of waking day/unable to get out of bed38813.75.0   

Supporting Tables 2a through 2d contain scores by ECOG PSR category and cancer type for the FACT-G, FACT-G7, and FWB, PWB, SWB, EWB subscales. Most score differences between adjacent ECOG PSR categories were significantly different for the FACT-G, FACT-G7, PWB, and FWB subscales (P values < .05). Score differences were associated with small-to-large effect sizes, and for the majority of comparisons exceeded their respective meaningful difference thresholds. Relative to other subscales, the EWB and SWB subscales had less variability by ECOG PSR category.

Table 4 contains FACT-G and FACT-G7 scores by disease status and type. Results indicate that the FACT-G and FACT-G7 scores decreased with increasing disease status. The FACT-G and FACT-G7 significantly differentiated between most disease status categories (P < .05), with small to large effect sizes. The score difference for the “no evidence of disease” and “local disease” categories exceeded the meaningful difference threshold for the FACT-G. Scores for regional and metastatic spread were generally similar.

Table 4. Adult Cancer Patient Sample Statistics by Disease Status and Type
Disease StatusNMeanSDAdjacent Category DifferenceP ValueaEffect Sizeb
  1. a

    The Tukey multiple comparison adjustment was used for P values.

  2. b

    Adjacent category mean difference divided by overall standard deviation for the scale or subscale (Full sample: FACT-G = 16.6, FACT-G7 = 5.6; breast cancer patients: FACT-G = 17.1, FACT-G7 = 5.3; lung cancer patients: FACT-G = 17.0, FACT-G7 = 5.8; prostate cancer patients: FACT-G = 14.9, FACT-G7 = 4.9; colorectal cancer patients: FACT-G = 15.9, FACT-G7 = 5.1).

  3. Abbreviations: FACT-G, Functional Assessment of Cancer Therapy–General; FACT-G7, Functional Assessment of Cancer Therapy–General 7-item; SD, standard deviation.

Full Sample
FACT-G      
No evidence of disease79086.7515.026.10<.0010.37
Local disease57280.6515.922.78.020.17
Regional spread59677.8716.921.17.550.07
Metastatic76476.7016.65
FACT-G7      
No evidence of disease79221.554.931.95<.0010.35
Local disease57719.605.400.73.090.13
Regional spread59918.875.631.04.0030.18
Metastatic77017.835.61
Breast Cancer Patients
FACT-G      
No evidence of disease37688.014.25.6<.0010.32
Local disease18682.416.22.8.300.16
Regional spread15979.616.42.3.500.13
Metastatic18781.914.3
FACT-G7      
No evidence of disease37622.04.51.7<.0010.32
Local disease18720.35.21.2.090.23
Regional spread16019.15.50.3.890.06
Metastatic18819.45.0
Lung Cancer Patients      
FACT-G      
No evidence of disease6982.615.54.1.420.24
Local disease9078.515.31.5.920.09
Regional spread13877.016.74.5.080.27
Metastatic18772.517.7
FACT-G7      
No evidence of disease6920.55.51.4.420.24
Local disease9219.15.90.5.920.08
Regional spread13818.65.51.9.010.33
Metastatic18916.75.6
Prostate Cancer Patients      
FACT-G      
No evidence of disease7290.911.55.8.300.39
Local disease2785.114.92.8.940.19
Regional spread1587.913.010.4.070.70
Metastatic7577.517.8
FACT-G7      
No evidence of disease7223.34.02.9.050.60
Local disease2820.44.81.9.630.39
Regional spread1522.34.34.9.0041.00
Metastatic7517.46.0
Colorectal Cancer Patients      
FACT-G      
No evidence of disease9086.314.76.7.020.42
Local disease9079.616.10.1.990.01
Regional spread11379.516.02.6.530.17
Metastatic15176.916.0
FACT-G7      
No evidence of disease9021.44.62.3.020.45
Local disease9019.15.20.6.840.12
Regional spread11319.75.21.5.070.29
Metastatic15318.25.2

Cutoff Scores

Supporting Table 3 contains the cutoff scores for low HRQOL for the FACT-G, its subscales, the FACT-G7, and distribution-based and anchor-based information. Based on the distribution of the means, 1 and 0.5 standard deviations below the mean were selected as cutoff scores for low HRQOL for individuals and groups, respectively. At the individual level, cutoff scores for low HRQOL were associated with the combined ECOG PSR categories 3/4. At the group level, cutoff scores for low HRQOL were associated with an ECOG PSR category of 2. To validate these cutoff scores, we used the chi-square test to determine whether there was an association between categorical data. More specifically, we sought to determine whether low HRQOL (HRQOL was dichotomized at the suggested cutoff value indicated in Supporting Table 3) was associated with poor ECOG PSR (dichotomized at the suggested category indicated in Supporting Table 3). Findings indicated that low HRQOL was associated with poor ECOG PSR (all P values < .001).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES
  10. Supporting Information

Clinicians and researchers can reference these data to inform interpretation of the magnitude and meaning of their own research findings and allow for comparisons of their patients to the general population and cancer-specific scores reported in this study. More specifically, these data are useful when comparing FACT scores to the general population to help understand the realistic ceiling of HRQOL scores one can expect, given that the general population does not report perfect HRQOL scores. In clinical settings, this will allow oncology providers to refer patients scoring in the low range of QOL to appropriate resources in supportive oncology such as social work, psychology, and psychiatry.

To date, this study provides the most comprehensive information on the FACT-G and its subscales by performance status, disease status, and cancer type, and is the first to our knowledge to provide individual cutoff scores for low HRQOL for individuals. Furthermore, this is the first study we are aware of that provides reference values and cutoff scores for the FACT-G7. Similar to all cutoff scores, the ones provided here are subject to change with the accumulation of more data regarding their research and clinical utility.

Findings indicate that values on the FACT-G, FACT-G7, FWB, EWB, and PWB subscales are similar across the sample of patients with cancer and US general population sample. Within the adult cancer sample, scores on the FACT-G, FACT-G7, and all subscales worsened with declining ECOG PSR category. Furthermore, lower FACT-G and FACT-G7 scores were found with worsening disease status.

Mean cancer sample scores on the FACT and FWB, PWB, EWB, and SWB subscales were similar to those reported by Brucker et al.[12] In the current study and the study by Brucker et al,[12] cancer survivors reported better SWB relative to the general population. This may be attributable to the increased social support cancer patients receive during diagnosis and treatment[23] or could be attributable to posttraumatic growth (benefit finding) in the area of social support.[24]

Among the cancer sample, results indicate that patients reported better scores on PWB than on the FWB across all ECOG PSR categories. In addition, relative to patients with lung cancer, patients with prostate cancer reported better HRQOL scores across most ECOG PSR categories and across all disease statuses on the FACT-G, its subscales, and the FACT-G7. These differences may be attributed to more aggressive treatments typically received by patients with lung cancer, or some systemic adverse effect of lung cancer not attributable to simple performance status.

It is important to note that the current sample does not contain a representative number of minorities, and the mean age is younger than the average age of individuals diagnosed with cancer. In addition, cancer type was not weighted to match data from the Surveillance, Epidemiology, and End Results (SEER) Program, although many of the cancers are similar to SEER data and the most commonly diagnosed cancers were represented. The relative youth of our cancer sample likely translated into improved overall health and quality of life, bringing FACT reference values closer to those of the general population sample. It is also possible that the attempt to oversample African American and Hispanic participants in the general US population sample led to more parity between cancer and general population values, given the known health disparities in African American and Hispanic populations. Moreover, although participants in the US adult general population sample were drawn from an online panel similar to the US general population, these findings may not generalize to individuals who do not have Internet access, although effort was made in this regard by providing computers and Internet access to individuals who did not have this. Despite the study limitations, the large sample size, variation in performance status categories, representation of numerous cancer types, and varying stages of disease are strengths of this study.

Although direct comparisons are somewhat compromised by the fact that these are pooled, reanalyzed data from a range of different studies, these data provide the best set of currently available reference scores for the FACT-G and FACT-G7 in individuals with cancer and in the general US population. Future research should consider constructing a population-based sampling frame of patients with cancer and the general population, matched on key demographic variables, and administer the FACT-G to all participants to obtain fully comparable scores with regard to FACT-G differences that can be attributed to cancer diagnosis or stage. In addition, more research on optimal cutoff scores for a range of purposes is recommended.

In summary, HRQOL is an important outcome measure in cancer care, and the widespread use of the FACT-G and FACT-G7 underscores need for enhancing the interpretability of the scores generated from these assessments. This report provides reference values for the FACT-G and FACT-G7 by physical performance status, cancer type, and disease status and provides practical cutoff scores for the FACT-G and FACT-G7. The reference values and cutoff scores presented here may assist clinicians and researchers to interpret the magnitude and meaning of their results with the FACT measures and to compare scores across patient groups in clinical trials.

FUNDING SUPPORT

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES
  10. Supporting Information

Mr. Peipert and Ms. Beaumont report support from FACIT (www.facit.org). Research reported in this manuscript was supported in part by a National Cancer Institute Diversity Supplement awarded to Betina Yanez, Ph.D. under grant number R01 CA 157809.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
cncr28758-sup-0001-supptables.docx41KSupporting Information Tables

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.