Skip to main content

Table 2 Psychometric properties of the FACT/GOG-Ntx subscalee over time

From: Psychometric testing of the Functional Assessment of Cancer Therapy/Gynecologic Oncology Group—Neurotoxicity (FACT/GOG-Ntx) subscale in a longitudinal study of cancer patients treated with chemotherapy

Properties

Methods

Analytical methods and criteria

Results in this study

Results in prior studies

Floor/ceiling effect

Examining the proportions of participants with the lowest/highest scores at the scale level

Frequency endorsement (15% of endorsement with lowest/highest scores)

Floor effects (28.3–50.6%)

NA

Internal consistency

Cronbach’s method

Cronbach’s alpha statistics (alpha > 0.70)

Total score: 0.82–0.89

Sensoryd: 0.80–0.90

Motord: 0.70–0.79

Hearingd: 0.64–0.90

Dysfunctiond: 0.71–0.96

Total score only:

0.82–0.86a

0.62–0.90b

0.80–0.85c

Examining the corrected item-total correlations

Pearson product-moment correlation coefficient (r = 0.30–0.70)

Q1 (Numbness or tingling in hands): 0.47–0.71

Q2 (Numbness or tingling in feet): 0.57–0.73

Q3 (discomfort in hands): 0.59–0.73

Q4 (discomfort in feet): 0.66–0.76

Q5 (joint pain or muscle cramps): 0.44–0.64

Q6 (weak all over): 0.47–0.72

Q7 (trouble hearing): 0.30–0.59

Q8 (ringing or buzzing in ears): 0.26–0.51

Q9 (trouble buttoning buttons): 0.34–0.63

Q10 (trouble feeling the shape of small objects): 0.34–0.65

Q11 (trouble walking): 0.55–0.69

Q1: 0.37–0.69c

Q2: 0.37–0.81c

Q3: 0.30–0.73c

Q4: 0.35–0.78c

Q5: 0.15–0.44c

Q6: 0.39–0.58c

Q7: 0.19–0.27c

Q8: 0.13–0.38c

Q9: 0.21–0.46c

Q10: 0.24–0.48c

Q11: 0.35–0.52c

Convergent validity

Examining the correlation between the FACT/GOG-Ntx and the other instruments with similar construct

Spearman’s correlation coefficient (rs > 0.40)

EORTC QLQ-CIPN20e: 0.79–0.93 (p < 0.01)

NCI-CTCAE (sensory)e: 0.23–0.45 (p < 0.01)

NCI-CTCAE (motor)e: 0.15–0.50 (p < 0.01)

Light touch using 10-g Monofilamente: 0.23–0.47 (p < 0.01)

NA

Factorial validity

Examining the hypothesized factor structure of the scale

Confirmatory factor analysis (χ2/df < 3, CFI > 0.95, TLI > 0.95, RMSES < 0.08)

χ2/df = 2.26–8.50, CFI = 0.79–0.95, TLI = 0.63–0.91, RMSEA = 0.07–0.14; SRMR = 0.046–0.079,

NA

Sensitivity to change

Measuring changes in scores of the FACT/GOG-Ntx subscale over time

Generalized estimating equation (p < 0.05)

Wald chi-square = 113.6, p < 0.001

N/A

Effect size \( r=\frac{Z}{\sqrt{N}} \)

(r = 0.1, 0.3 and 0.5 as small-, moderate- and large effect size)

r = 0.17–0.37

r = 0.37–0.91a

Responsiveness

Examining the MCID of the FACT/GOG-Ntx subscale

0.3 SD and 0.5 SD of the baseline FACT/GOG-Ntx subscale score.

MCID = 1.38–3.68

N/A

  1. Note: FACT/GOG-Ntx subscale = Functional Assessment of Cancer Therapy/Gynecologic Oncology Group—Neurotoxicity subscale; EORTC QLQ-CIPN20 = European Organization for Research and Treatment of Cancer Quality of Life-Chemotherapy-Induced Peripheral Neuropathy Scale 20; NCI-CTCAE = National Cancer Institute -Common Terminology Criteria for Adverse Events; CFI = comparative fit index; TLI = Tucker-Lewis index; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual; MCID = minimal clinically important difference; SD = standard deviation
  2. aCella et al. (2003) [31]: 230 patients with advanced non-small cell lung carcinoma
  3. bCalhoun et al.(2003) [12]: 56 chemotherapy-naive ovarian patients
  4. cHuang et al. (2007) [13]: 134 advanced endometrial cancer patients
  5. dThe hypothesized four-factor structure of the FACT/GOG-Ntx subscale was proposed by Huang et al. (2007) and not confirmed in this study
  6. eA higher score indicates worse neurotoxicity