Минимальная клиническая разница EQ-5D-5L, SF-6Dv2, и QLU-C10D в гематологовых халатных пациентах Китая.

Background

Hematologic malignancies, such as lymphoma and multiple myeloma, have steadily increased in incidence and now rank among the top ten cancers globally [ 1 ]. They are difficult to treat and often relapse, resulting in significant disease and economic burden. For example, the treatment costs for hematologic malignancies rank third among all cancer-related diseases in China [ 2 ]. Lymphoma, stemming from the lymphatic system, is the most common hematologic malignancy worldwide [ 3 ]. Key symptoms include painless lymphadenopathy, night sweats, fever, and weight loss [ 3 ]. Lymphomas are primarily classified as non-Hodgkin lymphoma (NHL) and Hodgkin lymphoma (HL), with the former being more prevalent [ 4 ], and diffuse large B-cell lymphoma (DLBCL) as its most common subtype [ 5 ]. In China, the estimated incidence of NHL) was 4.29 per 100,000 in 2016, with an increasing trend observed from 2006 to 2016 [ 6 , 7 ]. Multiple myeloma, originating from plasma cells in the bone marrow, presents with bone pain, anemia, and weight loss [ 8 ]. It is the second most common hematologic malignancy, accounting for approximately 10% of all hematologic tumors and 1% of all cancers in China, with an annual incidence ranging from 3.3 to 4.8 per 100,000 worldwide and 10 to 25 cases per million in China [ 9 – 11 ]. Given the chronic and often relapsing nature of these diseases, health-related quality of life (HRQoL) has become a critical consideration in both clinical management and research, guiding treatment decisions and patient-centered care strategies [ 12 ].

HRQoL information can also be translated into health utility score (HUS) for use in health technology assessments [ 13 ] if it is measured by utility instruments. Utility instrument can be broadly categorized into two types: generic and disease-specific scales. Generic scales, such as EQ-5D [ 14 ] and SF-6D [ 15 ], can be applicable to various populations and settings, and are widely used in cancer patients as well. EQ-5D includes the original 3-level (3L) and the new 5-level (5L) versions, with the latter being preferred in China due to its improved sensitivity [ 16 – 18 ]. Similarly, SF-6Dv2 was adopted for its refined structure and updated utility weights. However, they often lack specific dimensions targeting certain diseases or symptoms, limiting their sensitivity to HRQoL differences or changes. To address this limitation, cancer-specific utility tools like the EORTC Quality of Life Utility Measure-Core 10 Dimensions (QLU-C10D) were developed, to better reflect the unique health challenges faced by cancer patients [ 19 ].

Although HRQoL and HUS have become important endpoints in clinical trials, understanding the significance of these scores remains challenging for patients, clinicians, and researchers [ 20 ]. On the one hand, the units used in HRQoL tools are often unfamiliar to them, making it difficult in practice to grasp the meaning of the scales and the scores they generate. Unlike familiar physiological indicators, such as blood pressure or lung capacity, the clinical significance of changes in the scores is not immediately apparent, complicating their interpretation in clinical practice [ 21 , 22 ]. On the other hand, relying solely on statistical tests to analyze changes in the scores often reflects only statistical significance without directly conveying the clinical relevance of those changes [ 23 , 24 ].

To address this issue, the minimum clinically important difference (MCID) was developed, acting as a critical metric for assessing the clinical significance of changes in HRQoL scores [ 21 ]. The concept of MCID was first proposed by Jaeschke et al. [ 21 ], who defined it as “the smallest change in a treatment outcome that a patient would identify as important, and which would mandate a change in the patient’s management”. By determining the magnitude of score changes that represent meaningful score changes from the patient’s perspective, MCID helps translate HRQoL measurements into tools with greater practical relevance in clinical practice [ 25 ].

The methods for calculating MCID include anchor-based approach, distribution-based approach [ 26 ], and instrument-defined approach [ 27 ]. Each of them has its own characteristics and applications. According to anchor-based method, HRQoL score changes are linked to patient-reported outcomes or clinical outcomes, which are adopted as selecting external reference points [ 28 ]. Common statistical techniques in this category include categorical differentiation, receiver operating characteristic (ROC) analysis, and regression-based anchoring. Given that assessing HRQoL in cancer patients often relies on patient self-reports and clinical assessments, the anchor-based method has been widely adopted in oncology research [ 21 , 29 ]. Alternatively, the distribution-based method uses statistical calculations, such as standard deviation, standard error of measurement and Reliable Change Index, to determine the significance of HRQoL score changes [ 30 ]. While it may not fully capture the minimum importance from a patient or clinical perspective [ 31 ], it is particularly useful in large-sample studies. The instrument-defined method estimates MCID by calculating the mean absolute difference between health status levels within a specific classification system, making it ideal for scenarios that require precise detection of subtle health status changes [ 27 ].

In cancer patients, MCID studies have predominantly conducted on solid tumors such as glioma, breast, and colorectal cancers [ 32 – 35 ]. The studies have employed several utility instruments, with EQ-5D (both EQ-5D-3L and EQ-5D-5L) being the most commonly used [ 32 – 35 ]. Although several studies [ 36 – 39 ] have assessed MCID in patients with hematologic malignancies, none of them have included utility-based instruments in this population. In terms of the utility instruments, another widely used tool SF-6D is currently less used in estimating MCID in cancer patients, though it is often employed in other chronic diseases [ 40 ]. The disease-specific utility instrument QLU-C10D has not yet been adopted in previous studies.

Hence, the present study aims to fill this gap by estimating and comparing the MCID for EQ-5D-5L, SF-6Dv2 and QLU-C10D in Chinese patients with lymphoma and multiple myeloma employing anchor-based, distribution-based, and instrument-defined methods.

Methods

Study design and participants

This study utilized longitudinal data of both lymphoma and multiple myeloma patients, treated at the Hematology Department of a tertiary hospital in Shanghai, China. Baseline interviews took place between September and December 2022, with follow-up interviews conducted between February and June 2023. The baseline eligibility criteria required: (1) a clinical diagnosis of either lymphoma or multiple myeloma; (2) 18 years or older; and (3) cognitive ability sufficient to comprehend the survey and communicate effectively in Chinese. Eligible patients were recruited upon hospital admission, provided informed consent, and were interviewed by trained staff. Patients diagnosed with DLBCL, the most prevalent subtype of lymphoma, as well as all multiple myeloma patients, were invited for follow-up. The time interval between the baseline and follow-up interviews was set between 3 and 6 months, depending on the treatment regimen.

Participants were randomly assigned to one of two versions of the questionnaire at baseline, which contained identical questions in different sequences: EQ-5D-5L (or SF-6Dv2), EORTC QLQ-C30, SF-6Dv2 (or EQ-5D-5L), followed by sociodemographic questions such as age, place of residence, education, and occupation. The follow-up questionnaire was similar with baseline questionnaire except for the absence of sociodemographic section. Ultimately, 200 lymphoma and 139 multiple myeloma patients were included at baseline, with 78 DLBCL and 89 multiple myeloma patients completing the follow-up.

This study adhered to the principles outlined in the Declaration of Helsinki, involved human participants, and received ethical approval from the Ethics Committee of Zhongshan Hospital, Fudan University.

HRQoL instrument

EQ-5D-5L

The EQ-5D-5L evaluates the respondent’s health status on the day of survey. It consists of two parts: the health state descriptive system and the EQ-VAS. The descriptive system includes five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression [ 41 ]. Each dimension offers a five-level response scale: no problems, slight problems, moderate problems, severe problems, and extreme problems. In total, it measures 3,125 health states. In this study, we used the utility value set developed by Luo et al. [ 42 ], which is based on the general Chinese population, to calculate the EQ-5D-5L HUS. The range of EQ-5D-5L HUS values is from − 0.391 to 1.0, where 1 indicates “full health” and 0 represents “death”; and negative HUS indicates a health state worse than death.

SF-6Dv2

The SF-6Dv2, originating from the second version of the SF-36, assesses participants’ HRQoL over the previous four weeks [ 43 ]. Its descriptive system includes six dimensions: physical functioning, role limitation, social functioning, pain, mental health, and vitality [ 44 ]. While most dimensions have five response levels, the pain dimension has six levels, ranging from no impact to extreme discomfort. This allows the measurement of 18,500 unique health states. This study utilized the utility value system developed by Wu et al., based on the general Chinese mainland population, to calculate the SF-6Dv2 HUS [ 45 ]. The range of SF-6Dv2 HUS spans from − 0.277 (the worst health state) to 1.0 (the best health state).

EORTC QLU-C10D

The QLU-C10D is derived from the EORTC QLQ-C30 by selecting and combining four functional domains (physical, role, social, and emotional) and six symptoms (pain, fatigue, sleep, appetite, nausea, and bowel problems), resulting in a total of 10 items [ 19 ]. The QLU-C10D inherits the one-week recall period of the original QLQ-C30 instrument. Each C10D item has four levels: “Not at all,” (Level 1) “A little,” (Level 2) “Quite a bit,” (Level 3) and “Very much” (Level 4). In our study, the QLU-C10D HUS was computed using the Chinese value set, with utility values ranging from 0.083 (the worst state) to 1.0 (the best state) [ 46 ].

EORTC QLQ-C30

The EORTC QLQ-C30 assessed patients’ HRQoL over the past week [ 47 ]. It consists of 30 items divided into 15 domains: five functional domains (physical, role, cognitive, emotional, and social functioning), three symptom domains (fatigue, pain, nausea, and vomiting), one overall health status/global quality of life domain, and six individual items (domains). Responses for the overall health status/global quality of life item range from 1 to 7, and for the functional dimension items, from 1 to 4, with higher scores indicating better overall quality of life and functional status. Other items are scored on a 4-point scale ranging from 1 (“not at all”) to 4 (“very much”), with higher scores indicating more severe symptoms [ 48 ]. The QLQ-C30 total score is calculated as the mean of the sum of the scores of 13 dimensions (excluding financial impact and global quality of life), ranging from 0 to 100, with higher scores indicating a better HRQoL [ 49 ].

Clinical information

Baseline clinical data of the hematologic cancer patients were extracted from electronic medical records and laboratory tests. The former encompassed chronic conditions, diagnosis timing, treatment initiation, treatment protocols, and tumor staging; the latter included routine assessments of blood, liver, and kidney function, as well as imaging studies such as Positron Emission Tomography and Computed Tomography, along with pathological findings from biopsy samples. Lymphoma staging was categorized into four stages (I, II, III, and IV) based on the extent of lymphoma lesion [ 50 ]; multiple myeloma staging was classified into three levels (I, II, and III) according to the revised International Staging System [ 51 ]. Tumor relapse was assessed according to the treatment regimen [ 51 – 53 ].

For lymphoma patients, clinical changes were categorized as “worsened,” “stable,” or “improved” based on positron emission tomography-computed tomography results, following the Response Evaluation Criteria in Solid Tumors 1.1 (RECIST 1.1) [ 53 ].“Improvement” includes patients with complete response (CR) or partial response (PR), defined as the disappearance of all target lesions or a reduction in the total diameter of target lesions by at least 30%. “Deterioration” refers to an increase of at least 20% in the sum of the diameters of target lesions or the appearance of new lesions. “Stable” indicates a condition that does not meet the criteria for either improvement or deterioration.

For multiple myeloma patients, the changes were also classified into three categories according to the treatment guidelines, which consider changes in M protein levels and bone marrow plasma cell proportions before and after treatment [ 51 ]. CR is defined as M protein becoming negative and bone marrow plasma cells < 5%. PR is indicated by a > 50% reduction in bone marrow plasma cells or M protein. Progressive disease (PD) is characterized by a > 25% increase in bone marrow plasma cells or a > 25% rise in 24-hour urine M protein. Stable disease (SD) falls between PD and PR. CR and PR are classified as “improvement,” SD as “stable,” and PD as “deterioration”.

Statistical analysis

Descriptive statistics

Descriptive statistics were utilized to analyze the patients’ characteristics, the EQ-5D-5L, SF-6Dv2 and QLU-C10D HUSs, the QLQ-C30 total score, and the changes in the scores of the these instruments. Continuous variables were presented as means and SD, while categorical variables were expressed as numbers and percentages (%). To compare key demographic and clinical characteristics between follow-up participants and the baseline cohort, independent-samples t-tests and chi-square tests were applied.

Measurement of minimal clinically important difference

Anchor-based method

The anchor-based method relies on external criteria to determine the minimum but meaningful change scores of participants, using a Spearman correlation coefficient of ≥ 0.3 as the threshold for evaluating the anchor’s validity [ 28 ].

This study employed two anchors (i.e., the QLQ-C30 total score and clinical indicators). For the former, we defined minimal change using half of the SD of the anchor scores at baseline as the lower boundary and twice this value as the upper boundary [ 54 , 55 ]. The participants were then divided into three categories based on their anchor change scores: no change (< 0.5 baseline SD), minimal change (≥ 0.5 baseline SD and ≤ 1 baseline SD), and significant change (> 1 baseline SD). The MCID for improvement and deterioration was calculated by comparing the mean change scores between the relevant groups [ 56 ].

For the latter, both the two kinds of patients were categorized into three groups (i.e., improvement, stable, and deterioration) according to clinical indicators described above [ 51 , 53 ]. The MCID for improvement, stable, and deterioration was derived by calculating the mean differences in EQ-5D-5L, SF-6Dv2 and QLU-C10D HUSs between baseline and follow-up within the respective groups.

Distribution-based method

Previous studies suggest that half the SD of baseline scores and one standard error of measurement (SEM) can approximate the MCID [ 55 , 57 , 58 ]. Consequently, we calculated the MCID as half the SD of the baseline HUSs of the three utility instruments. And the SEM was computed using the formula: SEM = SD (58) , in which r represents the test-retest reliability or Cronbach’s α coefficient. The Cronbach’s α coefficients for EQ-5D-5L, SF-6Dv2 and QLU-C10D were determined at both baseline and follow-up in this study. The SEMs for both baseline and follow-up scores were calculated, and the mean SEM was used to estimate the MCID.

Instrument-defined method

The method calculates the MCID by examining the mean differences in HUSs between baseline health status and single-level transitions to other health states [ 59 ]. MCID is categorized based on the direction of these transitions: improvement to a better health state and deterioration to a worse health state [ 54 ]. To prevent overestimation, the following transitions were excluded [ 33 , 54 ]. First, the transition between level 3 (moderate problems) and level 4 (severe problems) across all dimensions of EQ-5D-5L, as it is at least 1.4 times larger than that of other adjacent levels according to the Chinese EQ-5D-5L value set [ 33 , 60 ]. Second, the transition from level 3 to level 4 in physical functioning, role functioning, and bowel issues of QLU-C10D, as it is at least 1.38 times greater than transitions between other adjacent levels. When calculating MCID for improvements, baseline health states in which all dimensions were rated as “1” (indicating the best possible health state) were excluded, as further improvement was not possible. In addition, the method was not applied to SF-6Dv2 due to the limited sample sizes of lymphoma patients ( N = 8) and multiple myeloma patients ( N = 7) meeting the criteria for a single-level transition of SF-6Dv2.

Statistical analyses were performed using SPSS software (IBM SPSS Statistics 27.0). All statistical tests were two-tailed, with a significance level set at 0.05.

Results

Characteristics of study participants

At baseline, the study included 200 lymphoma patients and 139 multiple myeloma patients (Table 1 ). Among the lymphoma patients, the majority were male (58.0%), aged ≤ 60 years (53.0%), and did not have chronic diseases (54.0%). Most patients were diagnosed within the past 6 months (58.0%), had tumors staged III-IV (83.5%), had their last treatment within the past 30 days (83.5%), and had no recurrence (74.5%). A total of 78 patients were followed up, resulting in a follow-up rate of 56.5%. Compared to their baseline health status, 14 patients had deteriorated, 24 remained stable, and 40 had improved.

In terms of the multiple myeloma patients, the majority were male (53.2%), aged > 60 years (68.3%), and retired (66.2%). Most patients were diagnosed more than 6 months ago (69.8%), had tumors staged I-II (65.5%), had their last treatment within the past 30 days (61.8%), and had no recurrence (53.2%). A total of 89 multiple myeloma patients were followed up, resulting in a follow-up rate of 64.0%. Compared to their baseline health status, 14 patients had deteriorated, 44 remained stable, and 30 had improved. Independent-samples t-tests and chi-square tests indicated no significant differences in key demographic and clinical characteristics between follow-up participants and the baseline cohort (eTable 1 ).

Table 2 detailed the distributions of EQ-5D-5L, SF-6Dv2, QLU-C10D and QLQ-C30 scores at the two time points. Figure 1 showed the health utility score distributions at both time points.

Minimal clinically important difference

Anchor-based analysis

The changes in EQ-5D-5L and SF-6Dv2 scores showed moderate correlations with QLQ-C30 total score, with Spearman correlation coefficients being 0.666 and 0.573 in lymphoma, and 0.545 and 0.613 in multiple myeloma, respectively. The correlation between QLU-C10D score and QLQ-C30 total score was stronger with the coefficients of 0.779 and 0.772 in the two kinds of patients. The QLQ-C30-based MCID values for the EQ-5D-5L, SF-6Dv2, and QLU-C10D were as follows (Table 3 ). Bolded values indicated the key MCID thresholds, which were primarily referenced for clinical interpretation as they provided the most meaningful differentiation in HRQoL changes. For lymphoma patients, the MCID values for the improvement group were 0.043 (EQ-5D-5L), 0.095 (SF-6Dv2) and 0.035 (QLU-C10D), respectively. In the deterioration group, the MCID values were − 0.058, -0.055 and − 0.076. For multiple myeloma patients, the improvement group had an EQ-5D-5L MCID of 0.057, SF-6Dv2 MCID of 0.116 and QLU-C10D of 0.054, while the deterioration group had MCID values of -0.086, -0.121, and − 0.074.

According to the clinical indicators, the MCID values for lymphoma patients were 0.057 (EQ-5D-5L), 0.095 (SF-6Dv2), and 0.026 (QLU-C10D) for the improvement group; 0.023, 0.058, and 0.011 for the stable group; and − 0.151, -0.084, and − 0.067 for the deterioration group, respectively. In multiple myeloma patients, the improvement group had an EQ-5D-5L MCID of 0.116, SF-6Dv2 MCID of 0.057, and QLU-C10D of 0.068; the stable group showed 0.029, 0.085, and 0.004; while the deterioration group recorded MCID values of -0.084, -0.047, and − 0.068 (Table 4 ).

Distribution-based analysis

The MCID estimates, calculated using half the baseline SD and the SEM value (Table 5 ), were as follows. For lymphoma patients, EQ-5D-5L showed values of 0.084 (0.5 SD) and 0.087 (1 SEM), SF-6Dv2 yielded 0.076 and 0.084, and QLU-C10D recorded 0.065 and 0.038. For multiple myeloma patients, the MCID values were 0.111 (0.5 SD) and 0.082 (1 SEM) for EQ-5D-5L, 0.086 and 0.096 for SF-6Dv2, and 0.062 and 0.049 for QLU-C10D.

Instrument-defined analysis

Table 6 provided the estimated MCID values for EQ-5D-5L and QLU-C10D across different health state transitions in lymphoma and multiple myeloma patients. For lymphoma, transitions to an improved health state yielded the values of 0.058 for EQ-5D-5L and 0.023 for QLU-C10D. In contrast, transitions to a worse state resulted in the values of -0.064 and − 0.043, respectively. Among multiple myeloma patients, improvements corresponded to MCID values of 0.055 for EQ-5D-5L and 0.013 for QLU-C10D, while deteriorations were associated with MCID values of -0.053 and − 0.028.

Mean minimal clinically important difference value according to the three approaches

Table 7 presented various MCID averages for lymphoma and multiple myeloma patients. Averaging results from three calculation methods, the mean MCID values for EQ-5D-5L, SF-6Dv2, and QLU-C10D were 0.070, 0.078, 0.044 for lymphoma patients, and 0.075, 0.087, 0.044 for multiple myeloma patients. The average MCID values from two anchors were 0.064, 0.077, 0.046 (lymphoma patients) and 0.074, 0.091, 0.056 (multiple myeloma patients). The mean distribution-based MCID values, derived from 0.5 SD and SEM, were higher, with values of 0.086 (EQ-5D-5L), 0.080 (SF-6Dv2), 0.052 (QLU-C10D) for lymphoma patients and 0.097, 0.091, 0.056 for multiple myeloma patients. The lowest mean MCID values were obtained from the instrument-defined estimates. Based on the three single-level transitions, the values were 0.061 (EQ-5D-5L), 0.033 (QLU-C10D) for lymphoma patients and 0.054, 0.021 for multiple myeloma patients, respectively.

Discussion

To our knowledge, this is the first study to derive MCID values for hematologic malignancies by simultaneously incorporating both generic (EQ-5D-5L, SF-6Dv2) and cancer-specific (QLU-C10D) utility instruments (Table 8 ). Moreover, all HUSs calculated were based on respective China-specific utility value sets. By employing multiple methods to calculate MCID, this study provided a comprehensive reference for future research. Also, the use of a multi-dimensional, disease-specific instrument QLQ-C30 as an anchor captured HRQoL changes across various dimensions. As a result, we have provided more accurate tools for interpreting HRQoL changes, thereby improving clinical management and treatment adjustments for hematologic cancer patients in China. While this study estimated MCID values for lymphoma and multiple myeloma, the lymphoma cohort consisted solely of DLBCL patients who completed follow-up, making the findings particularly relevant to this subtype.

The mean EQ-5D-5L MCID values were 0.064 (lymphoma) and 0.074 (multiple myeloma) according to anchor-based method; and were 0.086 and 0.097 according to distribution-based method. Both of the two kinds of values were substantially higher than those in cervical intraepithelial neoplasia (anchor-based: 0.041, distribution-based: 0.025) [ 33 ] and in Australian cancer patients (anchor-based: 0.03, distribution-based: 0.07) [ 35 ]. Compared to solid tumor patients, those with hematologic cancers often experience more complex symptoms like fatigue, infections, and anemia [ 61 , 62 ]. These symptoms show smaller variations, requiring larger improvements to be perceived as meaningful, leading to higher MCID values. Also, according to Hu et al. [ 33 ], the cervical intraepithelial neoplasia patients exhibited relatively mild disease and higher baseline HRQoL, limiting the space for improvement, which leads to smaller clinically significant changes and lower MCID values. Bourke et al. [ 35 ] also noted that higher baseline HRQoL in their cohort likely contributed to lower MCID values. Such findings aligned with the systematic review conducted by Cheng et al. [ 63 ], which demonstrated that the EQ-5D MCID values generally decrease as baseline scores increase. This phenomenon can be explained by the “ceiling effect”, where in the patients with better baseline health status are closer to the upper limit of the health scale, making them more responsive to subtle health improvements [ 63 ]. However, the MCID value derived from the instrument-defined method was 0.046 for both lymphoma and multiple myeloma patients, slightly lower than the value of 0.064 for cervical intraepithelial neoplasia patients [ 33 ] using the same method. This lower MCID value might reflect differences in the health status transitions between patient groups. Also, the differences in the social and cultural backgrounds of patients, accessibility of medical resources, and variations in disease management across studies may also influence the measurement of MCID values. In contrast, Pickard et al. [ 32 ] and Sagberg et al. [ 34 ] reported anchor-based EQ-5D-3L MCID values for solid tumor patients (UK: 0.11, US: 0.08) and glioma patients (0.15), which were higher than ours (lymphoma: 0.06; multiple myeloma:0.07). The primary reason may lie in the choice of HRQoL instrument. We adopted the EQ-5D-5L with its finer gradations, which is better equipped to detect smaller changes in health status more precisely.

According to the three approaches, the mean MCID values for lymphoma patients were lower than those for multiple myeloma patients (Table 7 ). This difference may be associated with the higher baseline HRQoL scores observed in lymphoma patients compared to those with multiple myeloma (Table 2 ). Similarly, this phenomenon can also be explained by the “ceiling effect“ [ 63 ] described above. In our sample, 24.6% of lymphoma patients and 15.7% of multiple myeloma patients reported the maximum EQ-5D-5L utility score at baseline, indicating a moderate ceiling effect. A scatterplot of baseline utility scores versus follow up utility scores (eFigure 1 ) further illustrates that higher baseline values were associated with limited improvement, supporting the presence of ceiling effects.

The choice of calculation methods significantly impacted MCID estimations. Anchor-based method relying on patient feedback or clinical changes, resulted in smaller MCID values compared to that derived from distribution-based method (Table 8 ). This is because the method could capture more subtle variations as it focuses on individual-level health changes; while distribution-based method reflects overall sample change. Emphasizing population-level change, the distribution-based method could better assess substantial treatment effects in larger groups but is less sensitive to individual-level variations. The instrument-defined method yielded the smallest MCID value (Table 7 ), indicating a strong ability to detect subtle changes in health status. The highest sensitivity stems from its focus on single-level health state transitions, allowing precise capture of minor fluctuations. However, the method requires sufficient transition observations to ensure stability in estimations. Indeed, the approach cannot be adopted to SF-6Dv2 due to the small number of patients meeting the requirement in the analysis.

The study used two anchoring standards: QLQ-C30 total score and clinical indicators. When anchored to the QLQ-C30 total score, the EQ-5D-5L MCID values were consistently lower for both lymphoma and multiple myeloma patients, and the SF-6Dv2 MCID was lower for lymphoma patients. When anchored to clinical indicators, the QLU-C10D MCID values were consistently lower for both lymphoma and multiple myeloma patients, and the SF-6Dv2 MCID was lower for multiple myeloma patients (Table 8 ). These findings demonstrated the differences in MCID sensitivity between the subjective anchoring and the objective anchoring across different disease types and scales, offering valuable clinical insights. Subjective anchor methods (e.g., the QLQ-C30 total score) rely on patients’ self-assessment, measuring aspects like quality of life, emotional state, and physical function. They capture individualized feedback and are effective at detecting subtle health changes. In contrast, objective anchor methods (e.g., clinical indicators) depend on healthcare professionals’ diagnoses and measurements, reflecting physiological or functional changes such as tumor shrinkage, disease progression, or recovery. Therefore, when selecting an MCID calculation method, it is important to select and apply subjective or objective anchoring according to the disease characteristics. To facilitate clinical applicability, it is important to consider the method by which MCID values are derived. In general, anchor-based MCIDs—particularly those linked to clinical indicators such as physician assessments or patient-reported improvement—are more reflective of real-world treatment response. In our study, anchor-based estimates were consistently lower and more consistent across instruments, suggesting greater sensitivity to subtle changes. Therefore, when applying MCID values in clinical practice for patients with hematologic malignancies, we recommend prioritizing anchor-based MCIDs derived from objective clinical indicators to guide treatment evaluation and decision-making.

In this study, QLU-C10D consistently demonstrated the lowest MCID values irrespective of the estimation approaches, reflecting its superior sensitivity in identifying clinically meaningful health changes [ 64 – 66 ]. This may be attributed to its cancer-specific domains (e.g., fatigue and nausea) and relatively narrow utility range (0.083–1.0), both of which enhance its ability to detect MCID. As a result, cancer-specific interventions (e.g., targeted therapy, immunotherapy, or palliative care) would benefit from QLU-C10D’s sensitivity to subtle symptom changes (e.g., a utility gain ≥ 0.044). Nevertheless, its cancer specific nature limits its use in cross-disease comparisons and population-level assessments. In contrast, generic instruments such as EQ-5D-5L and SF-6Dv2 offer broader applicability in the two aspects. Also, for interventions with broad health impacts (e.g., chemotherapy, stem cell transplantation), they are better suited to capture multidimensional improvements in physical and overall health, ensuring comprehensive evaluation of clinical effectiveness and supporting health-related decision-making. Among them, EQ-5D-5L is preferred due to its lower MCID thresholds, greater responsiveness, and wider usage; while SF-6Dv2 may serve as a complementary tool in assessing functional and psychological dimensions.

Our study has some limitations. The sample was drawn from a single hospital, limiting its generalizability to the broader patients. Also, MCID can vary depending on study context, interventions, and patient characteristics, meaning these results may not apply to other settings. Due to the limited sample sizes within different age subgroups(e.g., 40 and 38 in lymphoma), we were unable to estimate age-specific MCID values, which may have overlooked potential age-related differences in meaningful health changes. Additionally, we did not account for specific chemotherapy regimens, which may have influenced patients’ HRQoL trajectories and potentially affected the estimated MCID values. Furthermore, only two time points were measured, precluding the of MCID differences across distinct phases such as active treatment, maintenance, or long-term survivorship. Future studies should include patients from more diverse regions and backgrounds to further validate our findings.

In conclusion, this study provides MCID estimates for EQ-5D-5L, SF-6Dv2, and QLU-C10D in Chinese patients with lymphoma and multiple myeloma based on several approaches.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Acknowledgements

We are grateful to all the patients who participated in this study.

Abbreviations

Complete response

Diffuse large B-cell lymphoma

EuroQol five-dimensional five-level

Hodgkin lymphoma

Health-related quality of life

Health utility score

Minimum clinically important difference

Non-Hodgkin lymphoma

Progressive disease

Partial response

Quality of Life Utility-Core 10 Dimensions

Stable disease

Standard error of measurement

Author contributions

All authors contributed to the study’s conception and design. Material preparation, data collection, and analysis were performed by F.L., J.L., and A.Z. The first draft of the manuscript was written by F.L., and all authors commented on and revised earlier versions of the manuscript. J.W. and N.L. provided critical review and editing of the manuscript. Supervision and oversight were provided by P.L. and P.W. All authors read and approved the final manuscript.

Funding

This study was funded by the National Natural Science Foundation of China (Project No. 72274037) awarded to P. Wang.

Data availability

The datasets generated and analyzed during the current study are not publicly available due to patient confidentiality but are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study was approved by the Ethics Committee of Zhongshan Hospital, Fudan University.

Consent for publication

All authors have reviewed the final manuscript and consented to its publication.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Fangyu Li, Jing Li, Aixue Zhang these authors contributed equally to this work.

Contributor Information

Peng Liu, Email: liu.peng@zs-hospital.sh.cn.

Pei Wang, Email: wang_p@fudan.edu.cn.

References

1. Gongqi W, Xiaoyan X, Yan H. Health-related quality of life in patients with tuberculosis who completed treatment. Nurs J Chin People’s Liberation Army. 2018;35(24):29–33.

2. Dong H, Tang Z, Yan X, Jiang Q. Analysis on malignant tumor treatment cost in Anhui Province——Based on the SHA 2011. Health Econ Res. 2021;38(02):58–61.

3. Binder AF, Brody JD. Non-Hodgkin Lymphoma. Mount Sinai Expert Guides: Oncology. 2019:342– 53.

4. Matasar MJ, Zelenetz AD. Overview of lymphoma diagnosis and management. Radiol Clin North Am. 2008;46(2):175–98.

5. Mai Y, Yu JJ, Bartholdy B, Xu-Monette ZY, Knapp EE, Yuan F, et al. An oxidative stress-based mechanism of doxorubicin cytotoxicity suggests new therapeutic strategies in ABC-DLBCL. Blood. J Am Soc Hematol. 2016;128(24):2797–807.

6. Cao W, Chen HD, Yu YW, Li N, Chen WQ. Changing profiles of cancer burden worldwide and in China: a secondary analysis of the global cancer statistics 2020. Chin Med J (Engl). 2021;134(7):783–91.

7. Liu W, Liu J, Song Y, Zeng X, Wang X, Mi L, et al. Burden of lymphoma in China, 2006–2016: an analysis of the global burden of disease study 2016. J Hematol Oncol. 2019;12:1–9.

8. Jie ZQZJZYMLNGWQL. Analysis of psychological elasticity and its influencing factors in patients with multiple myeloma. Harbin Med J. 2023;43(02):89–92.

9. Peng L. Practice guidelines for diagnosis and treatment of multiple myeloma in Zhongshan hospital,fudan University(v1.2019). Chin J Clin Med. 2018;25(05):855–9.

10. Li X, Liu J, Chen M, Gu J, Huang B, Zheng D, et al. Health-related quality of life of patients with multiple myeloma: A real-world study in China. Cancer Med. 2020;9(21):7896–913.

11. Kazandjian D. Multiple myeloma epidemiology and survival: A unique malignancy. Semin Oncol. 2016;43(6):676–81.

12. Xu RH, Wong EL-y, Jin J, Huang H, Dong D. Health-related quality of life measured using EQ-5D in patients with lymphomas. Support Care Cancer. 2021;29:2549–60.

13. NICE U. Guide to the methods of technology appraisal. National Institute for Health and Clinical Excellence (NICE) London, UK. 2013.

14. Brooks R, Group E. EuroQol: the current state of play. Health Policy. 1996;37(1):53–72.

15. Brazier J, Roberts J, Deverill M. The Estimation of a preference-based measure of health from the SF-36. J Health Econ. 2002;21(2):271–92.

16. Janssen MF, Bonsel GJ, Luo N, Is. EQ-5D-5L better than EQ-5D-3L? A Head-to-Head comparison of descriptive systems and value sets from seven countries. PharmacoEconomics. 2018;36(6):675–97.

17. Mulhern B, Feng Y, Shah K, Janssen MF, Herdman M, van Hout B, et al. Comparing the UK EQ-5D-3L and english EQ-5D-5L value sets. PharmacoEconomics. 2018;36(6):699–713.

18. Ma YXA. Study on reliability and validity of Chinese version of EQ-5D-5L. Shanghai Med Pharm J. 2013;34(09):40–3.

19. King M, Costa D, Aaronson N, Brazier J, Cella D, Fayers P, et al. QLU-C10D: a health state classification system for a multi-attribute utility measure based on the EORTC QLQ-C30. Qual Life Res. 2016;25:625–36.

20. Yost KJ, Eton DT. Combining distribution- and anchor-based approaches to determine minimally important differences: the FACIT experience. Eval Health Prof. 2005;28(2):172–91.

21. Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials. 1989;10(4):407–15.

22. Sloan J, Symonds T, Vargas-Chanes D, Fridley B. Practical guidelines for assessing the clinical significance of Health-Related quality of life changes within clinical trials. Drug Inform J. 2003;37(1):23–31.

23. Copay AG, Subach BR, Glassman SD, Polly DW Jr., Schuler TC. Understanding the minimum clinically important difference: a review of concepts and methods. Spine J. 2007;7(5):541–6.

24. Fayers PM, Machin D. Quality of life: the assessment, analysis and interpretation of patient-reported outcomes. Wiley; 2013.

25. Tsiplova K, Pullenayegum E, Cooke T, Xie F. EQ-5D-derived health utilities and minimally important differences for chronic health conditions: 2011 Commonwealth fund survey of sicker adults in Canada. Qual Life Res. 2016;25(12):3009–16.

26. Crosby RD, Kolotkin RL, Williams GR. Defining clinically meaningful change in health-related quality of life. J Clin Epidemiol. 2003;56(5):395–407.

27. McClure NS, Sayah FA, Xie F, Luo N, Johnson JA. Instrument-Defined estimates of the minimally important difference for EQ-5D-5L index scores. Value Health. 2017;20(4):644–50.

28. Revicki D, Hays RD, Cella D, Sloan J. Recommended methods for determining responsiveness and minimally important differences for patient-reported outcomes. J Clin Epidemiol. 2008;61(2):102–9.

29. King MT. A point of minimal important difference (MID): a critique of terminology and methods. Expert Rev PharmacoEcon Outcomes Res. 2011;11(2):171–84.

30. Guyatt GH, Osoba D, Wu AW, Wyrwich KW, Norman GR. Methods to explain the clinical significance of health status measures. Mayo Clin Proc. 2002;77(4):371– 83.

31. de Vet HC, Terwee CB, Ostelo RW, Beckerman H, Knol DL, Bouter LM. Minimal changes in health status questionnaires: distinction between minimally detectable change and minimally important change. Health Qual Life Outcomes. 2006;4:54.

32. Pickard AS, Neary MP, Cella D. Estimation of minimally important differences in EQ-5D utility and VAS scores in cancer. Health Qual Life Outcomes. 2007;5:70.

33. Hu X, Jing M, Zhang M, Yang P, Yan X. Responsiveness and minimal clinically important difference of the EQ-5D-5L in cervical intraepithelial neoplasia: a longitudinal study. Health Qual Life Outcomes. 2020;18(1):324.

34. Sagberg LM, Jakola AS, Solheim O. Quality of life assessed with EQ-5D in patients undergoing glioma surgery: what is the responsiveness and minimal clinically important difference? Qual Life Res. 2014;23(5):1427–34.

35. Bourke S, Bennett B, Oluboyede Y, Li T, Longworth L, O’Sullivan SB, et al. Estimating the minimally important difference for the EQ-5D-5L and EORTC QLQ-C30 in cancer. Health Qual Life Outcomes. 2024;22(1):81.

36. Goswami P, Oliva EN, Ionova T, Salek S. Responsiveness and the minimal clinically important difference for HM-PRO in patients with hematological malignancies. Blood. 2018;132(Supplement 1):2294.

37. Breeze J, Mehta J, Tremblay G, Shacham S, Shah J, Li L et al. Pcn324 minimal clinically important difference (Mcid) for the functional assessment of Cancer Therapy - Multiple myeloma (Fact-Mm): two approaches. Value Health. 2020;23.

38. Quinten C, Kenis C, Decoster L, Debruyne PR, De Groof I, Focan C, et al. Determining clinically important differences in health-related quality of life in older patients with cancer undergoing chemotherapy or surgery. Qual Life Res. 2019;28(3):663–76.

39. Murao MKT, Hamada R, Miyasaka J, Matsushita M, Otagaki A, Kajimoto T, Arai Y, Kanda J, Nankaku M, Ikeguchi R, Takaori-Kondo A, Matsuda S. Minimal important difference of the 6-minute walk test after allogenic hematopoietic stem cell transplantation. Disabil Rehabil. 2024;46(15):3449–56.

40. Walters SJ, Brazier JE. Comparison of the minimally important difference for two health state utility measures: EQ-5D and SF-6D. Qual Life Res. 2005;14(6):1523–32.

41. Herdman M, Gudex C, Lloyd A, Janssen M, Kind P, Parkin D, et al. Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Qual Life Res. 2011;20(10):1727–36.

42. Luo N, Li M, Liu GG, Lloyd A, de Charro F, Herdman M. Developing the Chinese version of the new 5-level EQ-5D descriptive system: the response scaling approach. Qual Life Res. 2013;22(4):885–90.

43. Zhang A, Mao Z, Wang Z, Wu J, Luo N, Wang P. Comparing measurement properties of EQ-5D and SF-6D in East and South-East Asian populations: a scoping review. Expert Rev Pharmacoecon Outcomes Res. 2023;23(5):449–68.

44. Brazier J, Usherwood T, Harper R, Thomas K. Deriving a preference-based single index from the UK SF-36 health survey. J Clin Epidemiol. 1998;51(11):1115–28.

45. Wu J, Xie S, He X, Chen G, Bai G, Feng D, et al. Valuation of SF-6Dv2 health States in China using time Trade-off and Discrete-Choice experiment with a duration dimension. PharmacoEconomics. 2021;39(5):521–35.

46. Cao Y, Xu J, Norman R, King MT, Kemmler G, Huang W et al. Chinese utility weights for the EORTC cancer-specific utility instrument QLU-C10D. Qual Life Res. 2024:1–15.

47. Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, Duez NJ, 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;85(5):365–76.

48. van Roij J, Kieffer JM, van de Poll-Franse L, Husson O, Raijmakers NJH, Gelissen J. Assessing measurement invariance in the EORTC QLQ-C30. Qual Life Res. 2022;31(3):889–901.

49. Pilz MJ, Gamper EM, Efficace F, Arraras JI, Nolte S, Liegl G, et al. EORTC QLQ-C30 general population normative data for Italy by sex, age and health condition: an analysis of 1,036 individuals. BMC Public Health. 2022;22(1):1040.

50. Carbone PP, Kaplan HS, Musshoff K, Smithers DW, Tubiana M. Report of the committee on Hodgkin’s disease staging classification. Cancer Res. 1971;31(11):1860–1.

51. Palumbo A, Avet-Loiseau H, Oliva S, Lokhorst HM, Goldschmidt H, Rosinol L, et al. Revised international staging system for multiple myeloma: A report from international myeloma working group. J Clin Oncol. 2015;33(26):2863–9.

52. Ren L, Ren Y, Yang Y, Wang W, Xu T, Wang Y, et al. Clinical profiles in multiple myeloma patients with extreme survivals: a study from a National medical center in China. Hematology. 2023;28(1):2277503.

53. Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST to PERCIST: evolving considerations for PET response criteria in solid tumors. J Nucl Med. 2009;50(Suppl 1):s122–50.

54. McClure NS, Sayah FA, Ohinmaa A, Johnson JA. Minimally important difference of the EQ-5D-5L index score in adults with type 2 diabetes. Value Health. 2018;21(9):1090–7.

55. Norman GR, Sloan JA, Wyrwich KW. Interpretation of changes in health-related quality of life: the remarkable universality of half a standard deviation. Med Care. 2003;41(5):582–92.

56. Cheung YT, Foo YL, Shwe M, Tan YP, Fan G, Yong WS, et al. Minimal clinically important difference (MCID) for the functional assessment of cancer therapy: cognitive function (FACT-Cog) in breast cancer patients. J Clin Epidemiol. 2014;67(7):811–20.

57. Wyrwich KW, Nienaber NA, Tierney WM, Wolinsky FD. Linking clinical relevance and statistical significance in evaluating intra-individual changes in health-related quality of life. Med Care. 1999;37(5):469–78.

58. Wyrwich KW, Tierney WM, Wolinsky FD. Further evidence supporting an SEM-based criterion for identifying meaningful intra-individual changes in health-related quality of life. J Clin Epidemiol. 1999;52(9):861–73.

59. Luo N, Johnson J, Coons SJ. Using instrument-defined health state transitions to estimate minimally important differences for four preference-based health-related quality of life instruments. Med Care. 2010;48(4):365–71.

60. Jin X, Liu GG, Gerstein HC, Levine MAH, Guan H, Li H, et al. Minimally important difference and predictors of change in quality of life in type 2 diabetes: A community-based survey in China. Diabetes Metab Res Rev. 2018;34(8):e3053.

61. Burstein R, Aviv A, Even-Zohar NG, Nachmias B, Haran A, Braun M, et al. Comparing end-of-life care of hematologic malignancy versus solid tumor patients in a tertiary care center. Eur J Haematol. 2023;111(4):528–35.

62. Hochman MJ, Yu Y, Wolf SP, Samsa GP, Kamal AH, LeBlanc TW. Comparing the palliative care needs of patients with hematologic and solid malignancies. J Pain Symptom Manage. 2018;55(1):82–8. e1.

63. Cheng LJ, Chen LA, Cheng JY, Herdman M, Luo N. Systematic review reveals that EQ-5D minimally important differences vary with treatment type and May decrease with increasing baseline score. J Clin Epidemiol. 2024;174:111487.

64. Cosci F, Christensen KS, Ceccatelli S, Patierno C, Carrozzino D. Patient health Questionnaire-9: A clinimetric analysis. Braz J Psychiatry. 2024;46:e20233449.

65. Cosci F, Carrozzino D, Patierno C, Romanazzo S, Berrocal C, Chiarugi A, et al. Mental pain questionnaire: clinimetric properties of a potential global Person-Centred outcome measure. Clin Psychol Psychother. 2024;31(6):e70022.

66. Carrozzino D, Patierno C, Pignolo C, Christensen KS. The concept of psychological distress and its assessment: A clinimetric analysis of the SCL-90-R. Int J Stress Manage. 2023;30(3):235–48.

Associated Data

Supplementary Materials

Supplementary Material 1

Data Availability Statement

The datasets generated and analyzed during the current study are not publicly available due to patient confidentiality but are available from the corresponding author on reasonable request.