This is the first study to assess the measurement characteristics of two generic and two cancer specific PBMs in the CHL patient group. Overall, general measurements (eg EQ-5D-5 L and SF-6DV2) showed better measurement performance than two cancer specific measurements (ie, QLU-C10D and Fact-8D). The EQ-5D-5 L showed excellent convergence and known group effectiveness for SF-6DV2, while the SF-6DV2 showed a gentle ceiling and bottom effect than EQ-5D-5 L. However, it should be noted that more than half of our samples are composed of CHL patients currently being treated. As a result, their quality of life measured by cancer measurement was low when it was evaluated as a general scale. This inconsistency occurs because it can be more accurate because cancer-specific measurements can be more sensitive to the domain that can specifically affect the quality of life of cancer patients.
We showed that the average utility score of the SF-6D, QLU-C10D and Fact-8D was lower than the average utility score of EQ-5D-5 L, and their ceiling effect (range, 9.4-65.4%) was weaker than EQ-5D-5 L (range, 40.8-91.6%). Previous research provided evidence mixed between cancer patients. For example, Gamper et al. Patients in the UK, Australia and Italy showed similar results when comparing QLU-C10D with EQ-5D-3 L (21). Nahvijou et al. SF-6D showed that the utility score was higher than the EQ-5D-5 L in breast cancer patients (22). Kim et al. The QLU-C10D reported that the utility score was produced slightly higher than the EQ-5D-3 L (22). PAN et al. We compared QLU-C1-D and EQ-5D-5 L among 243 cancer patients in China and found the mixed results when comparing the QLU-C10D and EQ-5D-5 L utility scores. EQ-5D-5 L and Fact-8D were limited. Herdman et al. EQ-5D-5 L is a multi-center RCT study, including 250 cancer patients, is far or better than Fact-8D (24). There may be some explanations for these results. For example, EQ-5D-5 L has less dimensions than SF-6D, QLU-C10D and Fact-8D, which can reduce usefulness. Another reason is that the utility of two cancer specific PBMs was developed based on individual selection experiments (DCEs) including duration, and most EQ-5D-5 L utility was estimated based on the combination of time-off (TTO) and DCE technology. The difference in evaluation methods can lead to a systematic difference in utility. For example, XIE et al. Using the period method, the health utility score derived from the DCE has been found that the utility score is less likely than the derived score using TTO and DCE (25).
The correlation between two cancer-specific PBMs (QLU-C10D and FACT-8D) is found to be stronger than the cancer specific PBM and general PBM. This is not surprising because two cancer measurements have been developed to evaluate the health of cancer patients, and their description system contains the most important aspects of HRQOL for this group. For example, the correlation coefficient of the nausea between the QLU-C10D and Fact-8D was the highest among all assumed pairs. However, we have shown that the connection between the utility scores of these two measures is weak (ICC = 0.31). One possible explanation for this discovery is that there is no China value for Fact-8D, and the Australian preference weight used to calculate the utility score may not reflect the preference of the Chinese population. We also used the Australian value set to estimate the utility score, and we have further evaluated the correlation between the Fact-8D and the three different measurements. We have an increase in ICC but not enough. Currently, there is a lack of quantitative evidence of direct comparison between QLU-C10D and Fact-8D. Previous studies show that the pit-8D pit was quite lower than the QLU-C10D. This suggests that the PITS status described in the Fact-8D is more serious compared to the QLU-C10D, and potentially low correlation between the two measurements (26). Another study that evaluated the content effectiveness of five PBMs in cancer patients showed that the Fact-8D had the best content in the relevant aspects of all evaluated measurements, including QLU-C10D 27. In addition, we discovered a strong connection between the EQ-5D and SF-6D utility scores, but the SF-6D dimension showed a stronger correlation with the similar dimensions of two cancer-specific PBMs than the EQ-5D-5 L. More Strong Correlation. QLU-C10D. Previous research shows that the EQ-5D-5 L has a good content in terms of understanding (27), but it provides qualitative evidence that there is a lack of evidence of SF-6D’s performance compared to cancer-specific PBMs, which requires additional evaluation.
In addition, four devices aim to measure the dimensions of HRQOL, but some configuration may be different. For example, the “sleep” item of QLQ-C30 is negatively expressed, “Did you have a hard time sleeping?” In fact, G is suggested positively, “I’m sleeping well.” This change in phrases can potentially affect the way the participants interpret and respond to the question, which affects the overall evaluation of HRQOL. A tool for measuring similar structures may be a potential problem that can calculate other results. This inconsistency may be due to the subtle difference of the configuration itself or the change of precision and methodology of the measurement device (28). We need additional exploration of the content effectiveness of this population.
Overall, the four PBMs we evaluated showed excellent discriminatory abilities in differentiating HRQOL between risk groups, which represents the effectiveness of known groups. Fact-8D was performed more effectively than the other three PBMs when differentiated by the cancer stages reported by previous studies (24). The reason for the inconsistency may be because the Australian preference weight weight is used to estimate the fact -8D utility score. This method may not fully capture the preference of the Chinese population on the HRQOL of the cancer. In addition, Australia’s fact -8D evaluation studies have shown that the same coefficients are assigned to other severity levels in certain dimensions (eg fatigue, sleep, work, support and worry). This can potentially limit the ability to distinguish various levels of health problems (18). The F-State Tics confirmed that general PBMs are more sensitive than cancer-specific PBMs in discriminating patients. We found that the utility score of the general PBM is associated with daily life functions such as self-management, auxiliary tools and caregivers, but the utility score of QLU-C10D is more related to treatment. This discovery contradicts the previous study of Gamper et al. (21), a more comprehensive explanatory system found that QLU-C10D provides greater advantages compared to the general PBM when distinguishing a clinically known group. But gamper et al. More than 80% of the participants used an old sample that saw the cancer stage of 0 or i. In contrast, PAN et al. The EQ-5D-5 l utility used a sample that generated a F-state tic value higher than that QLU-C10D utility (23). In addition, the change in the EQ-5D-5 l utility score was less than the other three variations. Perhaps it is because it has a minor effect on the disease on the entire score, including only five dimensions (23). In addition, according to our known group validation analysis, the EQ-5D-5 L showed the biggest difference in the average utility score between the lower groups in three comparisons. For example, the average utility score of the EQ-5D-5 L among patients using auxiliary tools is 0.92, while among those who never used it, the difference of 0.6 is 0.32. In contrast, the average difference in utility scores for SF-6D, QLU-C10D and Fact-8D was 0.29, 0.3 and 0.19, respectively. This suggests that the utility gain of cost-utility analysis is more likely to be more likely for EQ-5D-5 L than other measurements.
The discovery that the general preference -based measurement shows better measurement characteristics than each cancer measurement in CHL has a significant effect. It suggests that general measures can provide reliable and valid evaluations of the quality of life that is not worse than certain measures in this group of patients. This enables more consistent comparison and evaluation by promoting a wide range of applications across various HL types. In the case of medical service providers and policymakers, these reliability can help the resource allocation and decision -making process, so that the intervention is evaluated as standardized standards. Moreover, in the context of clinical trials and research, the excellent measurement characteristics of the general measurement suggest that the priority for evaluating the patient’s results can be simplified without sacrificing accuracy. Cancer-specific measurements are designed to capture the unique aspects of cancer patients, but in this study suggests that the effect of general tools is sensitive to detecting changes in quality of life. This balance between specificity and generalization can improve the efficiency of clinical practice and research, which can ultimately contribute to CHL’s patient -centered treatment and simplified health economy evaluation.
limits
When interpreting the findings, some limitations should be taken into account. First, the sample was recruited from the volunteer pool through the internal network of the patient tissue. These volunteers may be healthier than average CHL survivors and frequent Internet and social media users. This can potentially introduce a selection bias. Second, online surveys are commonly used in these types of studies, but the Internet -based format may not be fully guaranteed by data quality. CHL’s survivors may not have completely participated in long investigations because they are poor in physical and mental health, which can affect the reliability of our discovery. Finally, considering that we have collected data in cooperation with patient tissues, medical information such as companionship and clinical symptoms was not collected from the patient’s record, which can affect our findings, especially the feasibility evaluation of known groups.
