Table 3.
Analyses of quality-of-life outcomes after four treatment cycles.
| Variable | Control group (n=30) | JK-5G group (n=30) | Difference between JK-5G and Control group (95% CI) | P a | Effect Size c |
|---|---|---|---|---|---|
| Scores on quality-of-life measures | |||||
| FACT-L scale | 115.63 ± 11.30 | 122.57 ± 4.52 | 6.93(2.43 – 11.43) | 0.003 | 0.805 |
| Lung-cancer subscale | 33.60 ± 1.22 | 33.70 ± 1.71 | 0.10(-0.67 – 0.87) | 0.795 | 0.067 |
| Trial Outcome Index | 74.43 ± 8.15 | 78.87 ± 3.21 | 4.43(1.19 –7.67) | 0.007 | 0.715 |
| EORCT QLQ-C30 | |||||
| Functional Scales | |||||
| Physical Functioning | 85.11 ± 21.06 | 92.45 ± 6.00 | 7.47(-0.63 – 15.57) | 0.071 | 0.484 |
| Role Functioning | 82.78 ± 25.70 | 100 ± 0 | 17.23(7.61 – 26.86) | 0.001 | 0.945 |
| Emotional Functioning | 90.00 ± 12.26 | 96.95 ± 4.08 | 7.00(2.25 – 11.75) | 0.005 | 0.773 |
| Cognitive Functioning | 97.22 ± 6.32 | 98.89 ± 4.23 | 1.70(-1.14 – 4.54) | 0.235 | 0.301 |
| Social Functioning | 82.78 ± 24.95 | 97.78 ± 8.46 | 15.10(5.32 – 24.88) | 0.003 | 0.809 |
| Global Health | 56.67 ± 14.91 | 52.78 ± 29.06 | -3.97(-16.05 – 8.11) | 0.517 | 0.171 |
| Symptom Scales | |||||
| Fatigue | 9.26 ± 21.06 | 0 | -9.33(-17.26 – -1.41) | 0.019 | 0.622 |
| Nausea and Vomiting | 0 | 0 | 0 | 1 | 0 |
| Pain | 13.89 ± 20.57 | 1.67 ± 5.09 | -12.13(-19.99 – -4.27) | 0.003 | 0.811 |
| Dyspnoea | 3.33 ± 10.17 | 4.44 ± 11.52 | 1.10(-4.46 – 6.66) | 0.694 | 0.102 |
| Insomnia | 16.67 ± 25.90 | 2.22 ± 8.46 | -14.47(-24.58 – -4.35) | 0.005 | 0.75 |
| Appetite Loss | 11.11 ± 25.27 | 0 | -11.17(-20.65 – -1.68) | 0.019 | 0.622 |
| Constipation | 0.00 | 0.00 | 0 | 1 | 0 |
| Diarrhoea | 0.00 | 0.00 | 0 | 1 | 0 |
| Financial Difficulties | 27.78 ± 12.63 | 12.22 ± 16.34 | -15.40(-22.88 – -7.92) | <0.01 | 1.065 |
Data are presented as means ± SD. CI, confidence interval; FACT-L, Functional Assessment of Cancer Therapy-Lung; EORCT QLQ-C30, European Organization for Reasearch and Treatment of Cancer Quality-of-Life Questionnaire Core 3.
p-values calculated with the independent-samples Student’s t-tests for continuous variables.
Cohen’s d statistic was used to determine effect size, a measure of the difference between two means divided by the standard deviation (in this case, the mean in the JK5G group minus the mean in the control group) divided by a standard deviation for the pooled data. As described in conventional classification, effects bigger than 0.20 are small, those bigger than 0.50 are moderate, and those bigger than 0.80 are large.