Review Reports

Response to Reviewer 1 Comments

1. Summary

Thank you very much for taking the time to review our manuscript and provide feedback. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files.

2. Questions for General Evaluation

Reviewer’s Evaluation

Response and Revisions

Does the introduction provide sufficient background and include all relevant references?

Yes/Can be improved/Must be improved/Not applicable

We have added additional background around socioeconomic status (SES) and how SES has previously been shown to impact EQ-5D health utilities.

We have added additional references to the manuscript to improve context/provide justification for methods; including the reference suggested by this reviewer. Newly incorporated references appear as Reference #s 16, 17, 23, 25, 26, 30, 40.

Is the research design appropriate?

Yes/Can be improved/Must be improved/Not applicable

Are the methods adequately described?

Yes/Can be improved/Must be improved/Not applicable

We have provided additional context in the sections on Statistical Analysis and Software to be more explicit.

Are the results clearly presented?

Yes/Can be improved/Must be improved/Not applicable

Some additional discussion has been added ie. To discuss previous work from Iran looking at EQ-5D-3L and EQ-5D-5L and impact on ceiling effects.

Are the conclusions supported by the results?

Yes/Can be improved/Must be improved/Not applicable

Are all figures and tables clear and well presented?

Yes/Can be improved/Must be improved/Not applicable

3. Point-by-point response to Comments and Suggestions for Authors

Comments 1: In general, Qol should be health related quality of life

Response 1:  Thank you for flagging this. We have updated/clarified wording in the “Simple Summary” and throughout to confirm proper use.

Comments 2: 1-You should include the following reference to justify psychometric properties of the instrument in cancer patients. Psychometric properties of the EQ-5D-5L compared with EQ-5D-3L in cancer patients in Iran

Response 2:  Thank you for this suggestion, we have reviewed this reference and agree it is an important inclusion. It has been built into the manuscript as Reference # 40, and added to the discussion section (page 11, line 40 onwards), which now reads:

“In a sample of patients with cancer from Iran, the EQ-5D-5L demonstrated lower ceiling effects compared with the EQ-5D-3L.  Ceiling effects were observed for both EQ-5D-3L and EQ-5D-5L (12.07% and 9.44%, respectively), both lower than the acceptable limit of 15% for health status questionnaires [40]”.

Comments 3: 2-I suggest that you provide additional information on dependent and independent variables in OLS

Response 3: We recognize that providing as much transparency as possible is crucial, and thank you for this suggestion. Accordingly, you will note that additional information on independent variables in the in OLS are presented in Table 4. In this table, the model parameters of the two main effects models, with and without female and male cancers are presented.  In our Appendices Tables S3 and S4 outline model parameters of additional models that incorporate age (categorical) / birth sex and birth sex / age (categorical) interactions.

To explain the rationale of how our independent variables relate to our dependent variable, please read the paragraph (page 2, line 22 onwards):

“Socioeconomic status (SES), one’s access to material and social resources, aligns with the indicators used for socioeconomic position (SEP), in the World Health Organization’s social determinants of health conceptual framework [16]. These indicators (income, education, occupation), along with stratifiers such as social class, gender, and race /ethnicity, influence one’s exposure to advantage and disadvantage [16]. Social position determines health through intermediate factors, including material, psychosocial, behavioural, and biological factors [16], which impact HRQoL [17].”

Comments 4: 3-You can provide the values of the best and worst health states for Canadian value set.

Response 4: This is a good suggestion – we have added those values to the Methods section (page 3, line 42 onwards), which now reads:

“The best and worst health states for this Canadian TTO value set ranged from 1.000 for 11111 (best health) to a predicted mean (SE) -0.340 (0.013) for 33333 (worst health) [30]. ”

Response to Reviewer 2 Comments

1. Summary

Thank you very much for taking the time to review our manuscript and provide feedback. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files.

2. Questions for General Evaluation

Reviewer’s Evaluation

Response and Revisions

Does the introduction provide sufficient background and include all relevant references?

Yes/Can be improved/Must be improved/Not applicable

We have added additional background around socioeconomic status (SES) and how SES has previously been shown to impact EQ-5D health utilities.

See detailed responses to comments 1-6 below which focus on the introduction section.

Also/more broadly, we have added additional references to the manuscript to improve context/provide justifications where appropriate. New references appear as Reference #s 16, 17, 23, 25, 26, 30, 40.

Is the research design appropriate?

Yes/Can be improved/Must be improved/Not applicable

See additions/expansions to methods section overall for additional context including more explicit reference to our prior publication related to this work that has additional details on patient accrual (Tsui et al. 2025). Additional details have been provided to clarify statistical analysis.

Are the methods adequately described?

Yes/Can be improved/Must be improved/Not applicable

We have provided additional context in the sections on Statistical Analysis and Software to be more explicit.

See detailed comments 7-9 below for improvements to Methods section.

Are the results clearly presented?

Yes/Can be improved/Must be improved/Not applicable

Are the conclusions supported by the results?

Yes/Can be improved/Must be improved/Not applicable

Some additional discussion has been added ie. To discuss previous work from Iran looking at EQ-5D-3L and EQ-5D-5L and impact on ceiling effects.

Are all figures and tables clear and well presented?

Yes/Can be improved/Must be improved/Not applicable

3. Point-by-point response to Comments and Suggestions for Authors

Comment 1: In simple summary, the section devoted to Implications states that "... can be impacted ...". However, I recommend avoiding this term because it implies causality and an experimental approach, neither of which corresponds to the nature of the presented study.

Response 1: Thank you for this suggestion. To avoid/reduce this implication, we have adjusted wording in the Simple Summary (page 1, line 21) and Introduction (page 2, line 53) to “…may be associated with…” instead of “…can be impacted…”

Comment 2 Introduction

In addition to the objective of the current study, it would be helpful to list the specific hypotheses tested by the study at the end of this section.

Response 2 We appreciate this comment/suggestion, and have added an explicit hypothesis statement at the end of our introduction which reads:

“We hypothesized that patient characteristics of age, sex, education, marital status, employment status, income, primary cancer site, and ethnicity will impact EQ-5D-3L health utilities.”

In the Methods section, we also mention our rationale for assigning our reference categories:

“Reference categories were assigned based on either the largest sample size or the group thought to be most privileged, based on literature [27,28].”

Comment 3 Furthermore, consider explaining the relationship between perceived quality of life and socioeconomic status. The current version of the text seems somewhat superficial and unspecific in this regard.

Response 3 We understand the importance of being as clear and specific as possible and thank you for this suggestion. Accordingly, we had added some further exploration in the Introduction. Our section on socioeconomic status now reads (page 2, lines 22 onwards):

“Socioeconomic status (SES), one’s access to material and social resources, aligns with the indicators used for socioeconomic position (SEP), in the World Health Organization’s social determinants of health conceptual framework [16]. These indicators (income, education, occupation), along with stratifiers such as social class, gender, and race /ethnicity, influence one’s exposure to advantage and disadvantage [16]. Social position determines health through intermediate factors, including material, psychosocial, behavioural, and biological factors [16], which impact HRQoL [17].

SES is associated with EQ-5D health utilities in oncology [18-20]. In Canada, a large survey of cancer survivors (n = 1,759) found higher EQ-5D-3L health utilities among those with university/college education compared with no university / college education p<0.001, and lower health utilities among individuals not married compared with married or common law (p = 0.001). The same study found a large difference in mean (SE) health utilities by cancer site, from 0.76 (0.03) in pancreatic cancer to 0.90 (0.05) in chronic lymphocytic leukemia [18]. In China, among a sample of patients with colorectal cancer, lower household income was associated with lower EQ-5D-5L health utility with a mean (SD) 0.505 (0.419) for <¥20K to 0.759 (0.315) for >¥80K.These examples illustrate how factors such as marital status, income, education level, and cancer site have been associated with EQ-5D-3L or 5L responses in Canada and China [18,19]. ” 

Comment 4  As part of the introductory definition of the issue, please specify your arguments for choosing the EQ-5D-3L. Which other health-related quality of life (HRQoL) indicators did you consider, and why did you use the EQ-5D-3L? I assume the reported frequency of use was not the only argument

Response 4 Thank you for this suggestion. We added an explicit comment near the end of the introduction to explain our choice, and added a supporting reference (page 2, lines 46 onwards):

“We used the EQ-5D-3L instrument instead of the more recent EQ-5D-5L version as the 3L version has more historic use than the 5L version as a clinical outcome assessment in health technology assessment, regulatory reviews, and systematic literature reviews [23]”

Comment 5 The introductory section should also include a more detailed summary of the current knowledge regarding the relationship between quality of life (QoL) and socioeconomic status (SES).

Response 5  Thank you for this suggestion; we have provided a response above where it was first flagged to improve discussion of QoL and SES.

Comment 6 Lastly, it would be appropriate to explain the concept of socioeconomic status in more detail. There is no consensus among relevant sources on the conceptual definition, and different approaches can be found at the level of the indicators used. Furthermore, your study effectively narrows the concept of SES down to income alone, which does not capture the full breadth of SES.

Response 6 Thank you for this thoughtful observation.  We added a paragraph to explain the concept of socioeconomic status, and how it relates to the World Health Organization’s conceptual framework on the social determinants of health (page 2, lines 22 onwards) which reads:

“Socioeconomic status (SES), one’s access to material and social resources, aligns with the indicators used for socioeconomic position (SEP), in the World Health Organization’s social determinants of health conceptual framework [16]. These indicators (income, education, occupation), along with stratifiers such as social class, gender, and race /ethnicity, influence one’s exposure to advantage and disadvantage [16]. Social position determines health through intermediate factors, including material, psychosocial, behavioural, and biological factors [16], which impact HRQoL [17].”

We would like to clarify that we tested multiple key factors in our models, and found that income was the only significant SES indicator associated with EQ-5D-3L health utilities (see Table 4).

Comment 7 Materials and Methods

2.1

I recommend adding an explicit mention of the data collection method used. In this context, it would be appropriate to describe the specific data collection procedures (e.g., data collection procedure, possible use of health or administrative records for respondent recruitment). Lastly, it would be useful to include information on the response rate.

Response 7  Thank you for this suggestion. This information was reported previously in Tsui, et al. Patient Experiences Regarding Feasibility of Implementing Real-World EQ-5D Collection at an Oncology Centre in Ontario, Canada. Curr Oncol 2025, 32, 308. We have added an explicit note to the Methods section 2.1 to specify/reference this earlier work (page 3, line 10).

Comment 8 2.2 Please comment on the adequacy of the sample size. If applicable, provide links to relevant sources that support your arguments.

Response 8 Thank you for this opportunity to justify our sample sizes. We recognize the importance of sample size in avoiding Type II errors, and for this reason rely on coefficient estimates and their 95% confidence intervals to make inferences about the associations between SES and HRQoL, rather than relying on p-values alone[1].  We note that for several aspects of SES (e.g. ethnicity, sex) we did not have significant p-values but had confidence intervals that included important differences in HRQoL. We have been careful to interpret the results accordingly.

We have added a summarized justification in the Methods Study sources and measures section (page 3, lines 22 onwards) that reads:

“These sample sizes allowed the estimation of model parameters and their 95% confidence intervals (CIs) with reasonable precision, according to linear regression best practices [25], and the accuracy in parameter estimation principle [26].”

 We also added a sentence to our limitations to acknowledge the wider estimated CIs in the model that includes the birth sex variable (page 11, lines 46 onwards).

“Our model that includes the birth sex variable (28 df, n = 111) has wider 95% CI estimates, even though it satisfies model parsimony and clinical relevance.” 

Comment 9  2.3. Provide a brief explanation for choosing a nonparametric test (Spearman).

Response 9 Thank you for asking for an explanation. As mentioned in the methods section (page 3, lines 35 onwards),

“Spearman rank correlations were explored between numerical income categories (<$29,999 to > $150,000) and each dimension. Positive coefficients indicate that lower income is associated with higher problem severity.”

Given that income categories are ordinal (Table 1), and the distribution of participant’s income categories within each dimension level are not normally distributed (Figure 1), we chose to use the Spearman rank correlation.

Comment 10  Would you be able to reiterate the significance of your study and its overall contribution to the field?

Response 10  Thank you for this thoughtful suggestion.  We added a sentence to the Conclusion section (page 12, lines 9 onwards):

“This study provides the first Canadian real-world estimates that quantify the effect of income on EQ-5D-3L health utilities in oncology.”  

Comment 11: In conclusion, the manuscript may contribute to the field, and with the above improvements, its impact could be further enhanced. Thank you for your commitment to advancing our understanding in this area. I look forward to seeing the new version of your paper.

Response 11: Thank you for the broad overall support, and again for the detailed and thoughtful feedback on our work. We believe our edits have indeed improved the manuscript significantly, and look forward to your further review.

Response to Reviewer 3 Comments

1. Summary

Thank you very much for taking the time to review our manuscript and provide feedback. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files.

2. Questions for General Evaluation

Reviewer’s Evaluation

Response and Revisions

Does the introduction provide sufficient background and include all relevant references?

Yes/Can be improved/Must be improved/Not applicable

We have added additional background around socioeconomic status (SES) and how SES has previously been shown to impact EQ-5D health utilities.

We have added additional references to the manuscript to improve context/provide justification for methods; including the reference suggested by this reviewer. Newly incorporated references appear as Reference #s 16, 17, 23, 25, 26, 30, 40.

Is the research design appropriate?

Yes/Can be improved/Must be improved/Not applicable

Are the methods adequately described?

Yes/Can be improved/Must be improved/Not applicable

We have provided additional context in the sections on Statistical Analysis and Software to be more explicit.

Are the results clearly presented?

Yes/Can be improved/Must be improved/Not applicable

Are the conclusions supported by the results?

Yes/Can be improved/Must be improved/Not applicable

Some additional discussion has been added ie. To discuss previous work from Iran looking at EQ-5D-3L and EQ-5D-5L and impact on ceiling effects.

Are all figures and tables clear and well presented?

Yes/Can be improved/Must be improved/Not applicable

3. Point-by-point response to Comments and Suggestions for Authors

Comment 1: For Table 1, present (n = 170) and not with capital N

Response 1: Thank you for catching this – we have updated appropriately (page 4, line 9).

Comment 2: Explain the rationale of the two models “Table 4. Two main effects models with and without female and male cancers” in the statistical analyses section.

Response 2: We appreciate this suggestion. Accordingly we have added an explanation on the rationale of the two models in the Methods, Statistical analysis section (page 3, lines 18 onwards).

“Models with and without the birth sex variable were tested to assess the effect of sex-specific cancers. First, we included data with all cancers, and excluded sex as a variable (n = 170). To incorporate sex as a variable, we excluded individuals with gynecological, breast, and prostate cancers in our multivariable analysis (n = 111).” 

Comment 3: State the outcome variable in title (Table 4)

Response 3: Thank you for suggesting this, we have updated the title of Table 4 (page 9, lines 1 and 2) to:

“EQ-5D-3L health utility as the outcome predicted by socioeconomic status: two main-effects models with and without sex-specific cancers.”