Review Reports - Quality of Life in Patients with Treatment-Resistant Depression Treated with Intranasal Esketamine as Add-On Therapy to an SSRI—A Prospective, Comparative Study in Real-World Clinical Practice

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript addresses a clinically relevant topic and focuses on patient-reported quality-of-life outcomes. Prospective real-world design is valuable and complements existing clinical trial evidence. Use of both WHOQOL-BREF and SF-36, alongside MADRS and HDRS, provides a comprehensive assessment. The manuscript appropriately acknowledges the non-randomized design and its limitations.

Statistical tests are not clearly specified; the statistical analysis section needs greater detail.
The repeated measurements over 12 months require an appropriate longitudinal/repeated-measures analysis.
Effect sizes and 95% confidence intervals should be reported in addition to p-values.
Multiple comparisons across numerous QoL domains/subscales require consideration of multiplicity.
A sample-size/power calculation is not reported.
Potential confounding by indication and selection bias should be addressed more thoroughly because treatment allocation was non-randomized.
The definition and documentation of treatment-resistant depression should be described more explicitly.
Missing data, dropouts, and retention at the 12-month assessment should be reported clearly.
Safety/tolerability and adverse events associated with esketamine are not adequately presented.
Please correct terminology inconsistencies, particularly “ECC” vs. “ESK” in Tables 5–6 and “Sanity” in Table 3.

Author Response

The manuscript addresses a clinically relevant topic and focuses on patient-reported quality-of-life outcomes. The prospective real-world design is valuable and complements existing evidence from clinical trials. The use of both WHOQOL-BREF and SF-36, together with MADRS and HDRS, provides a comprehensive assessment. The manuscript appropriately acknowledges the non-randomized design and its limitations.

Comment 1:
The statistical tests are not clearly specified; the Statistical Analysis section requires more detail.

Response 1:
Thank you for this valuable comment. We agree with the Reviewer. The “Statistical Analysis” section has been expanded and clarified. In the revised version of the manuscript, we specified how quantitative and qualitative variables were presented and identified the tests used for between-group comparisons. For quantitative variables with an approximately normal distribution, the independent-samples Student’s t-test was used, whereas the Mann–Whitney U test was applied when normality assumptions were not met. Qualitative variables were compared using the chi-square test or Fisher’s exact test, depending on expected cell counts. We also clarified that all tests were two-sided and that the level of statistical significance was set at p < 0.05.

Revised manuscript excerpt – Section 4.6 “Statistical Analysis”:

Quantitative variables were presented as mean ± standard deviation (SD), and variables with distributions deviating from normality were presented as median and interquartile range (IQR). Qualitative variables were presented as counts and percentages. Distributional assumptions for continuous variables were assessed before inferential statistical testing. Between-group comparisons at baseline were performed using the independent-samples Student’s t-test for approximately normally distributed continuous variables and the Mann–Whitney U test when distributional assumptions were not met; qualitative variables were compared using the chi-square test or Fisher’s exact test, as appropriate. All tests were two-sided, and statistical significance was set at p < 0.05.

Comment 2:

Repeated measurements over 12 months require an appropriate longitudinal/repeated-measures analysis.

Thank you for this important comment. We agree with the Reviewer. In the revised version of the manuscript, the analysis of changes over time was supplemented with an approach accounting for repeated measurements within the same participants. The longitudinal analysis included the effects of time, treatment group, and the group × time interaction, allowing us to assess whether the pattern of change differed between the esketamine-treated group and the group receiving SSRI treatment. The corresponding description was added to the “Statistical Analysis” section.

Revised manuscript excerpt – Section 4.6 “Statistical Analysis”:

Because outcomes were assessed repeatedly in the same participants at T0, weeks 4, 8, and 12, and T_end, longitudinal changes in MADRS, HDRS, and quality-of-life scores were analyzed using repeated-measures analysis of variance (RM-ANOVA), with time treated as the within-subject factor and treatment group (ESK vs SSRI) as the between-subject factor. The group × time interaction was used to assess whether the pattern of change differed between treatment strategies. When the assumption of sphericity was violated, the Greenhouse–Geisser correction was applied.

Comment 3:
Effect sizes and 95% confidence intervals should be reported in addition to p-values.

Response 3:
Thank you for this important comment. In accordance with the Reviewer’s suggestion, the presentation of the results has been expanded to include effect-size measures and 95% confidence intervals alongside p-values. The appropriate description was added to the “Statistical Analysis” section, and the reporting of results was supplemented to allow assessment not only of statistical significance but also of the magnitude and precision of the observed differences.

Revised manuscript excerpt – Section 4.6 and Table 2:

For the main quality-of-life outcomes, the absolute change from T0 to T_end and the between-group difference in change were reported. In addition to p-values, descriptive between-group differences at T_end were presented together with 95% confidence intervals and Cohen’s d effect sizes, calculated from endpoint means and pooled standard deviations. WHOQOL-BREF: 8.2 (95% CI: 5.2–11.2); d = 1.41. SF-36: 8.3 (95% CI: 5.0–11.6); d = 1.31.

Comment 4:

Multiple comparisons across numerous quality-of-life domains/subscales require adjustment for multiplicity.

Response 4:
Thank you for this important comment. We agree that analyzing multiple quality-of-life domains and subscales increases the risk of a Type I error. Therefore, in the revised analysis, Holm correction was applied for multiple comparisons within the WHOQOL-BREF domains and SF-36 subscales. Both nominal p-values and adjusted p-values are presented in the tables. The corresponding description was also added to the “Statistical Analysis” section.

Revised manuscript excerpt – Section 4.6 “Statistical Analysis”:

In secondary analyses involving multiple WHOQOL-BREF domains and SF-36 subscales, the issue of multiple comparisons was addressed using the Holm procedure; both nominal and adjusted p-values were reported.

Comment 5:

No sample size/power calculation was provided.

Response 5:

Thank you for this comment. In the revised version of the manuscript, information on the sample size calculation was added. Assuming a moderate effect size (Cohen’s d = 0.5), a two-sided significance level of α = 0.05, and 80% power, the required sample size was estimated at approximately 64 participants per group. The final study sample consisted of 30 patients in each group because of recruitment feasibility. This limitation was also taken into account in the interpretation of the results and in the study limitations section.

Revised manuscript excerpt – Section 4.2 “Study Participants”:

Sample size calculation: An approximate a priori calculation for the comparison of two independent groups assumed a moderate between-group effect size (Cohen’s d = 0.5), a two-sided alpha level of 0.05, and 80% statistical power. This indicated a target sample size of approximately 64 participants in each group. Owing to recruitment feasibility in real-world clinical practice, the final cohort included 30 participants in each group (60 participants in total). The study therefore had limited power to detect small effects and to conduct extensive subgroup analyses, which should be considered when interpreting the results.

Comment 6:
Potential confounding by indication and selection bias should be addressed in greater detail because treatment allocation was not randomized.

Response 6:
Thank you for this important comment. We agree that the lack of randomization increases the risk of confounding by indication and selection bias. In the revised version of the manuscript, we expanded the description of the study design and the limitations section, noting that the choice of treatment strategy may have been related to the prior course of illness, symptom severity, previous treatment response, comorbidity, concomitant treatment, and treatment availability. We also emphasized that, despite the similarity of the groups in terms of measured baseline characteristics, the influence of unmeasured confounding factors cannot be excluded. Therefore, the findings should be interpreted as observational associations rather than evidence of a causal relationship.

Revised manuscript excerpt – Section 4.1 “Study Design”:

The study was a prospective, observational, comparative cohort study conducted under real-world clinical practice conditions at two specialized psychiatric centers in Poland. Treatment allocation was not randomized. The treatment strategy was selected by the treating physician based on clinical assessment, previous treatment history, clinical severity, medical indications, tolerability, comorbidities, concomitant treatment, patient preference, and treatment availability. Because these factors may have influenced both treatment selection and outcomes, confounding by indication and selection bias cannot be excluded. Accordingly, the study was designed to compare outcomes associated with treatment strategies used in routine practice rather than to estimate a randomized causal effect.

Comment 7:
The definition and documentation of treatment-resistant depression should be described in greater detail.

Response 7:
Thank you for this comment. In the revised version of the manuscript, the definition of treatment-resistant depression and the description of how it was documented were expanded. TRD was defined as the absence of a satisfactory clinical response to at least two consecutive adequate antidepressant treatment trials during the current depressive episode. We clarified that an adequate trial involved treatment at a therapeutic dose for an appropriate duration, generally at least 6 weeks, with confirmed adherence. The diagnosis of TRD was verified on the basis of medical records, treatment history, doses used, treatment duration, and assessment of the clinical response.

Revised manuscript excerpt – Section 4.2 “Study Participants”:

For the purposes of the study, treatment-resistant depression was operationally defined as a current major depressive episode with an insufficient clinical response despite at least two consecutive adequate antidepressant treatment trials during the current episode. An adequate trial required treatment at a therapeutic dose for an appropriate duration, generally at least 6 weeks for each trial, with adherence documented in the medical records. TRD status was verified on the basis of available medical documentation and treatment history, including the antidepressant used, dose, treatment duration, documented adherence, and clinical response to previous treatment.

Comment 8:
Missing data, dropouts, and retention during the 12-month assessment should be clearly reported.

Response 8:
Thank you for this comment. In the revised version of the manuscript, the description of participant retention and missing data was expanded. We specified the number of participants enrolled in the study, the number remaining in follow-up at 12 months, and the retention rates in both groups. We also clarified how missing data were handled and considered the potential bias resulting from loss to follow-up when interpreting the results.

Revised manuscript excerpt – Section 2.6 “Retention and Missing Data”:

A total of 60 participants were enrolled in the study (30 in each group). At the 12-month assessment, data were available for 29/30 participants in the ESK group (96.7% retention) and 28/30 participants in the SSRI group (93.3% retention). Thus, one participant in the ESK group and two participants in the SSRI group did not contribute data to the final assessment. The main analyses were based on available observations at each time point. Given the small amount of missing follow-up data, missing outcome values were not imputed in the primary analyses. The possibility of bias related to loss to follow-up was considered when interpreting the long-term results.

Comment 9:

Safety/tolerability and adverse events associated with esketamine are not adequately presented.

Response 9:
Thank you for this important comment. In the revised version of the manuscript, the section on treatment safety and tolerability was expanded. The frequencies of the most commonly observed adverse events in both groups are presented, including headache, dizziness, nausea, somnolence, anxiety, and transient increases in blood pressure. We also clarified that no serious adverse events were recorded and provided information on treatment completion and monitoring of patients receiving esketamine.

Revised manuscript excerpt – Section 2.5 “Safety and Tolerability”:

Both treatment strategies were generally well tolerated. In the ESK group, the most frequently documented adverse events were mild or moderate headache (6/30; 20%), transient disorientation or dizziness (5/30; 17%), nausea or vomiting (4/30; 13%), somnolence or fatigue (3/30; 10%), anxiety or inner restlessness (2/30; 7%), and a transient increase in blood pressure (2/30; 7%). In the SSRI group, headache (3/30; 10%), nausea or vomiting (2/30; 7%), somnolence or fatigue (5/30; 17%), and anxiety or inner restlessness (4/30; 13%) were observed, respectively. No serious adverse events were recorded in either group. The planned treatment and 12-month assessment were completed by 29/30 patients in the ESK group and 28/30 patients in the SSRI group. Esketamine was administered under medical supervision, with post-dose observation and blood pressure monitoring. Safety data were descriptive and did not constitute the primary endpoint of the study.

Comment 10:
Please correct terminology inconsistencies, particularly “ECC” vs “ESK” in Tables 5–6 and “Rozum” in Table 3.

Response 10:
Thank you for drawing attention to these inconsistencies. The terminology has been standardized throughout the manuscript. The abbreviation “ECC” in Tables 5–6 was replaced with the correct abbreviation “ESK”, which is now used consistently throughout the text for the esketamine-treated group. In addition, the incorrect term “Rozum” in Table 3 was corrected and replaced with the appropriate name corresponding to the analyzed domain. The tables and text were also reviewed again for terminological consistency.

Revised manuscript excerpts following terminology correction – Tables 3, 5, and 6:

Table 3: “Mental Health”. Table 5: “ESK (n=30)”. Table 6: “ESK (n=30)”. The explanatory notes consistently use the wording: “ESK – esketamine”.

Reviewer 2 Report

Comments and Suggestions for Authors

The Methods state the ESK group received esketamine "as an add-on therapy to optimized antidepressant treatment," whereas the SSRI group received "standard treatment with SSRIs." The manuscript never clarifies whether the SSRI group also had their antidepressant dose optimized, or whether they simply continued on a previous SSRI regimen. If the ESK group's background SSRI was actively optimized while the SSRI group's was not, the comparison does not isolate the effect of esketamine, it compares (optimized SSRI + esketamine) versus (non-optimized SSRI).
ESK administration details are thoroughly explained at first in terms of the 84 mg dose, twice a week induction followed by weekly dosing, however the paper lacks the real experience of treatment during the 12-month observation period - frequency of treatment, treatment failures, discontinuation of treatment, dosage adjustments, adherence to the treatment, and use of antidepressants.
Section 4.6 only says that comparisons were made using “appropriate tests based on the type of data and distribution of the variables.” No details about the specific tests are provided.
The ESK group improved by +23.3 points on the WHOQOL-BREF overall score (from 52.1 to 75.4). The WHOQOL-BREF domains are scored 0–100, so this represents a ~45% relative improvement from baseline. In TRANSFORM-2—which the authors themselves cite as supporting evidence—the EQ-5D-5L health state index improved by only 0.310 (ESK + AD) vs. 0.235 (AD + PBO) at Day 28, on a 0–1 utility scale. The SSRI group's improvement of +15.5 points on WHOQOL-BREF is also far larger than what would be expected in treatment-resistant depression with SSRI continuation alone. In STAR*D, TRD patients had SF-12 mental component scores of 30 (vs. 45.7 for non-TRD) at baseline, with minimal improvement across treatment levels. The authors must explain how their observed effect sizes can be reconciled with the existing evidence base, or acknowledge that the effect sizes are implausibly large and may reflect measurement error, unblinding, or regression to the mean.
ESK has a well-characterized safety profile: dissociation occurs in approximately 27% of patients, sedation in 9.3%, somnolence in 18.2%, and transient blood pressure increases of 7–9 mmHg systolic and 4–6 mmHg diastolic are common. The manuscript reports zero safety outcomes, no adverse events, no dissociation, no blood pressure changes, no cognitive assessments. For a clinical study of esketamine in a real-world population, this is a major omission.

Author Response

In response to the Reviewer’s comments, we provide below our explanations and excerpts of the changes introduced into the manuscript.

1. Reviewer comment: “The methodology states that the ESK group received esketamine ‘as adjunctive therapy to optimized antidepressant treatment’, whereas the SSRI group received ‘standard treatment with SSRI medications’. The text does not explain whether the antidepressant dose was also optimized in the SSRI group or whether the existing treatment regimen with these drugs was simply continued. If the background SSRI dose in the ESK group was actively optimized whereas the SSRI group was not, this comparison does not isolate the effect of esketamine itself, but rather compares an ‘optimized SSRI + esketamine’ regimen with a ‘non-optimized SSRI’ regimen.”

Authors’ response: Thank you for this very important comment. We agree that the original wording ‘optimized antidepressant treatment’ could have suggested that the degree of SSRI optimization was known and comparable between the two groups. However, the available dataset did not contain a standardized variable comparable between groups that described the degree of SSRI dose optimization; therefore, we cannot demonstrate that the intensity of treatment modification was identical in both arms. To avoid implying information that was not documented in the study, we removed the term ‘optimized’ from the description of the ESK group. Oral SSRI treatment remained at the discretion of the treating physician as part of routine clinical practice. We also explicitly stated that the study compares two treatment strategies used in real-world clinical practice rather than the isolated pharmacological effect of esketamine. The lack of standardized data on the degree of treatment optimization in both groups has also been added to the study limitations.

Excerpts introduced into the manuscript:

Section 4.1. Study Design

Antidepressant pharmacotherapy remained at the discretion of the treating physician as part of routine clinical practice. The study did not collect a standardized measure, comparable between groups, of the degree of SSRI dose optimization. In the ESK group, esketamine was added to ongoing SSRI-based treatment, whereas the comparison group received SSRI-based treatment without esketamine. The comparison therefore concerned two treatment strategies used in clinical practice rather than the isolated pharmacological effect of esketamine.

Section 4.3. Esketamine and SSRI Treatment

Oral SSRI dosing remained at the discretion of the treating physician and could be modified according to clinical indications; the detailed degree of dose optimization was not recorded as a standardized research variable comparable between groups.

2. Reviewer comment: “Details of esketamine administration are discussed comprehensively with respect to the 84 mg dose and the regimen consisting of an induction phase (twice weekly), followed by once-weekly administration; however, information is lacking on the actual course of treatment during the 12-month follow-up period—including treatment frequency, treatment failures, discontinuations, dose modifications, treatment adherence, and the use of antidepressant medications.”

Authors’ response: Thank you for this comment. We expanded the description of the treatment course and clearly distinguished the planned regimen from the data actually available on treatment exposure during the 12-month follow-up. At the endpoint, data were available for 29/30 patients in the ESK group (96.7%) and 28/30 in the SSRI group (93.3%). However, the detailed visit-by-visit number of esketamine administrations, short treatment interruptions, dose modifications, as well as dose changes and adherence to oral antidepressants were not prospectively collected as standardized research variables. We do not wish to reconstruct these parameters retrospectively in a way that could introduce bias. Therefore, we clarified in the manuscript that the described regimen represented the treatment framework, while subsequent administration frequency was individualized according to clinical response and tolerability. We also added a separate paragraph on treatment exposure and identified the lack of detailed adherence and treatment-modification data as a study limitation.

Excerpts introduced into the manuscript:

Section 2.7. Treatment Exposure

The dosing regimen described above defined the treatment framework; however, after the induction phase, the frequency of subsequent esketamine administrations was individualized according to clinical response and tolerability. Final 12-month data were available for 29/30 participants in the ESK group and 28/30 participants in the SSRI group. The detailed number of administrations, short treatment interruptions, esketamine dose modifications, and changes in dosing and adherence to oral antidepressants were not prospectively recorded as standardized research variables; therefore, they were not reported as quantitative measures of exposure.

3. Reviewer comment: “Section 4.6 merely states that comparisons were made using ‘appropriate tests selected according to the type of data and distribution of variables’, without providing details on the specific tests applied.”

Authors’ response: Thank you for this comment. Section 4.6 has been fully clarified. Quantitative variables were presented as mean ± SD or median and IQR depending on distribution, and categorical variables as counts and percentages. Between-group comparisons were performed using the independent-samples Student’s t-test or the Mann–Whitney U test, whereas categorical data were analyzed using the chi-square test or Fisher’s exact test. Longitudinal changes were analyzed using RM-ANOVA with time as the within-subject factor, group as the between-subject factor, and the group × time interaction; when sphericity was violated, the Greenhouse–Geisser correction was applied. For the main outcomes, 95% CIs and Cohen’s d were also reported, and Holm’s procedure was used for multiple comparisons across WHOQOL-BREF domains and SF-36 subscales. All tests were two-sided, with p < 0.05 considered statistically significant.

Excerpts introduced into the manuscript:

Section 4.6. Statistical Analysis

Quantitative variables were presented as mean ± standard deviation (SD), and variables with distributions deviating from normality as median and interquartile range (IQR). Between-group comparisons were performed using the independent-samples Student’s t-test or the Mann–Whitney U test; categorical variables were compared using the chi-square test or Fisher’s exact test. Longitudinal changes were analyzed using RM-ANOVA with evaluation of the group × time interaction. When the sphericity assumption was violated, the Greenhouse–Geisser correction was applied. For the main outcomes, 95% confidence intervals and Cohen’s d were reported, and Holm’s procedure was used for multiple comparisons. All tests were two-sided, and p < 0.05 was considered statistically significant.

4. Reviewer comment: “In the ESK group, the overall WHOQOL-BREF score improved by 23.3 points (from 52.1 to 75.4). The scoring range for individual WHOQOL-BREF domains is 0–100, which corresponds to a relative improvement of approximately 45% compared with baseline. By comparison, in the TRANSFORM-2 study—cited by the authors themselves as evidence supporting efficacy—the EQ-5D-5L health-status index (utility scale 0–1) improved at day 28 by 0.310 in the ESK + antidepressant (AD) group and by 0.235 in the AD + placebo (PBO) group. Likewise, the 15.5-point improvement in WHOQOL-BREF in the SSRI group is considerably larger than would be expected in treatment-resistant depression managed solely by continuing SSRI therapy. In the STAR*D study, patients with treatment-resistant depression (TRD) had a baseline mental component score of 30 on the SF-12 (compared with 45.7 points in patients without TRD), with only minimal improvement observed during subsequent treatment steps. The authors must explain how the effect magnitudes observed in their study can be reconciled with the existing scientific evidence, or acknowledge that these values are implausibly high and may result from measurement error, unblinded treatment allocation, or regression to the mean.”

Authors’ response: Thank you for this detailed and important comment. We agree that directly comparing the numerical change in WHOQOL-BREF with EQ-5D-5L or SF-12 results is not methodologically appropriate. These instruments differ in construction, scoring range, and psychometric properties, and the studies being compared also differ in assessment time point (12 months in our study versus 28 days in TRANSFORM-2) and design. The relative change calculated from WHOQOL-BREF points should not be interpreted as a standardized effect size or as a measure directly comparable with the EQ-5D utility index. At the same time, we acknowledge the Reviewer’s concern that the magnitude of improvement observed in both groups is large and, in a non-randomized, unblinded observational study, may partly reflect regression to the mean, the natural variability of the depressive episode, intensity of clinical care, concomitant treatment modifications, patient expectations, participant selection, and retention. We therefore revised the interpretative language to be explicitly associative, added a separate paragraph to the Discussion, and expanded the limitations section. The numerical values were retained as observed descriptive data, but they are no longer presented as a direct, isolated effect of esketamine or as a percentage improvement comparable with results from other scales.

Excerpts introduced into the manuscript:

Discussion – Interpretation of the Magnitude of Change

Direct comparison of the numerical magnitude of changes in WHOQOL-BREF, EQ-5D-5L, and SF-12 is not justified because these instruments differ in construction, scoring range, and psychometric properties, and the studies considered involve different assessment time points. The large absolute changes observed in both groups should be interpreted cautiously. In a non-randomized and unblinded observational study, they may have been influenced by regression to the mean, the natural variability of the depressive episode, intensity of clinical contact, concomitant treatment modifications, patient expectations, and selective retention. Therefore, the observed changes were not interpreted as the isolated effect of esketamine or as values directly comparable with results obtained using other quality-of-life scales.

5. Reviewer comment: “The safety profile of esketamine (ESK) is well established: dissociative symptoms occur in approximately 27% of patients, sedation in 9.3%, and somnolence in 18.2%; transient increases in blood pressure are also frequently observed (7–9 mmHg for systolic and 4–6 mmHg for diastolic blood pressure). The present study reports no safety data: no adverse events, dissociative symptoms, or blood pressure changes were recorded, and cognitive function was not assessed. For a clinical study of esketamine involving a real-world patient population, this represents a significant omission.”

Authors’ response: Thank you for this important comment. We supplemented the manuscript with a separate section on safety and tolerability and expanded the monitoring description in Section 4.3. In the ESK group, the most frequently documented adverse events were headache (20%), transient disorientation or dizziness (17%), nausea or vomiting (13%), somnolence or fatigue (10%), anxiety or restlessness (7%), and transient elevation of blood pressure (7%); no serious adverse events were recorded. Blood pressure was monitored before and after esketamine administration. At the same time, we emphasize that dissociative symptoms were not assessed using a dedicated standardized scale, and cognitive function was not a prespecified endpoint. Therefore, we do not report the frequency of dissociation or cognitive changes as quantitative outcomes, and we do not equate transient disorientation or dizziness with dissociation. This limitation has now been clearly stated in the manuscript.

Excerpts introduced into the manuscript:

Section 2.5. Safety and Tolerability

In the ESK group, the most frequently documented adverse events were headache (6/30; 20%), transient disorientation or dizziness (5/30; 17%), nausea or vomiting (4/30; 13%), somnolence or fatigue (3/30; 10%), anxiety or inner restlessness (2/30; 7%), and transient elevation of blood pressure (2/30; 7%). No serious adverse events were recorded in either group.

Section 4.3. Safety Monitoring

Dissociative symptoms were not assessed using a dedicated standardized scale, and cognitive function was not a prespecified endpoint. Therefore, the frequency of dissociation and cognitive changes was not reported as quantitative study outcomes; transient disorientation or dizziness was not equated with dissociation.

Reviewer 3 Report

Comments and Suggestions for Authors

Abstract

There is a confusing statement: "real-world clinical trial." However, it is more accurately a prospective, observational study, not a clinical trial.

The authors mentioned add-on therapy to optimized antidepressant therapy in the Esketamine group. This is confusing, as the SSRI group also received SSRI monotherapy. Clarifying what "optimized" means or how "standard treatment" is defined would improve transparency

The conclusion looks overstated: "Treatment with intranasal 31 esketamine as add-on therapy was associated with greater improvement in quality of life." The study is observational, and the authors may not be able to conclude causality unless the study design and data are robust.

Introduction

As stated earlier, the authors mentioned add-on therapy to optimized antidepressant therapy in the Esketamine group. This is confusing, as the SSRI group also received SSRI monotherapy. Clarifying what "optimized" means or how "standard treatment" is defined would improve transparency.

Provide guidelines as a reference for the statement "patients with treatment-resistant depression treated with intranasal esketamine as add-on therapy and in patients receiving standard SSRI-based antidepressant treatment"

All abbreviations should be written in full at their first appearance (e.g., NMDA, WHOQOL-BREF).

At the end of the introduction, the primary objective and hypothesis seem redundant; streamline this for clarity.

Also, clearly state the study's primary and/or secondary objectives (if any).

Methods

Provide a reference for the STROBE guidelines, and include a filled form as supplementary material.

The statement "Randomization was not used"may confuse readers, as this was an observational study. In this context, why randomization

The term "optimized" antidepressant therapy in the esketamine group might not reach the general audience, which needs more clarity with the reference. More detail in protocols on SSRI (e.g., preference, dose, and any adjustments) and the exact protocol for "optimized" antidepressant therapy in the esketamine group would aid replication. Provide a standardized procedure for quality-of-life assessments.

Neither the sample size nor the power analysis was reported, and the limitations state that "the sample size was moderate, comprising 60 patients, 30 in each group." On what basis do the authors consider the sample size moderate?. The lack of a prior sample size determination is a potential limitation.

The statistical analysis section lacks information on the analyses performed, their justification, and the database (with version) used. More details would improve clarity for readers, particularly for replication.

Results

Table 2: What do you mean by general index? Specify whether this is PCS, MCS, or a documented composite, and define how it was calculated.

Table 3: Social relations should be "Social relationship"

Table 4: Labels do not adhere to the domains of the SF-36. E.g., there is no such "role limitations" domain in the SF-36. This may confuse readers.

Tables 2-4: Mention the statistical analysis used in the footnote.

Table 5–6 group label: "ECC" should be ESK; also standardize "SSRIs" to SSRI (singular) to match text/footnotes.

Table labeling should be consistent: "4th week" vs. "Week 12" vs. T0/T_end

Discussion

Address potential limitations in more detail, as the current discussion is not adequate, as follows. Address unmeasured or residual confounding and offer concrete ways future studies could address these (e.g., propensity scoring, matched designs). How the lack of randomization and potential selection bias may influence the interpretation of group differences.

Address what constitutes clinically meaningful change on the WHOQOL-BREF and SF-36, not just statistical significance.

Add how clinicians might implement quality-of-life monitoring or esketamine treatment, referencing real-world practice challenges.

Author Response

ABSTRACT

Reviewer comment: “There is a misleading statement: ‘real-world clinical trial’. More precisely, this is a prospective observational study, not a clinical trial.”

Authors’ response: Thank you for this comment. We agree that the term “real-world clinical trial” was imprecise. Throughout the manuscript, it has been replaced with “a prospective, observational, comparative cohort study conducted in a real-world clinical practice setting.” This wording clearly defines the study design and does not suggest an interventional clinical trial.

Text added to the manuscript:

Abstract – study design

A prospective, observational, comparative cohort study was conducted in a real-world clinical practice setting.

Reviewer comment: “The authors mentioned adjunctive therapy to optimized antidepressant treatment in the esketamine group. This is misleading because the SSRI group also received SSRI monotherapy. Clarifying what ‘optimized’ means or how ‘standard treatment’ is defined would improve clarity.”

Authors’ response: Thank you. We removed the term “optimized antidepressant treatment” because it could imply a separate protocol-driven dose-escalation algorithm used only in the ESK group. No such algorithm was imposed in this study. In both groups, SSRI treatment was managed by the treating physician and reflected routine clinical practice; the difference was the addition of esketamine in the ESK group. We also replaced the term “standard treatment” with the more precise wording “SSRI-based treatment without esketamine.”

Text added to the manuscript:

Abstract – treatment groups

The study included 60 adults with TRD: 30 received intranasal esketamine as add-on therapy to ongoing SSRI-based treatment, and 30 received SSRI-based treatment without esketamine. The study protocol did not impose a separate SSRI dose “optimization” algorithm; oral antidepressant treatment was managed according to routine clinical practice.

Reviewer comment: “The conclusion appears overstated: ‘Intranasal esketamine treatment 31 as adjunctive therapy was associated with greater improvement in quality of life.’ The study is observational, and the authors may not be able to draw causal conclusions unless the study design and data are robust.”

Authors’ response: We agree. The conclusion has been toned down and is now strictly associative. We explicitly state that treatment assignment was not controlled by the investigators and that the non-randomized design and the possibility of residual confounding do not allow the observed differences to be interpreted as an isolated causal effect of esketamine.

Text added to the manuscript:

Abstract – conclusion

In this observational cohort, a treatment strategy including intranasal esketamine was associated with greater improvement in several HRQoL measures than SSRI-based treatment without esketamine. Because treatment assignment was not controlled by the investigators and residual confounding cannot be excluded, the findings should not be interpreted as evidence of an isolated causal effect of esketamine.

INTRODUCTION

Reviewer comment: “As mentioned previously, the authors referred to adjunctive therapy to optimized antidepressant treatment in the esketamine group. This is misleading because the SSRI group also received SSRI monotherapy. Clarifying what ‘optimized’ means or how ‘standard treatment’ is defined would improve clarity.”

Authors’ response: Thank you. Consistent with the changes made in the Abstract, we removed the term “optimized” and described the compared strategies as SSRI-based treatment with the addition of intranasal esketamine and SSRI-based treatment without esketamine. We also added information that the choice of SSRI and any dose modifications were clinical decisions made by the treating physician rather than elements of the study protocol.

Text added to the manuscript:

Introduction – description of treatment strategies

The primary aim of the study was to compare 12-month changes in HRQoL between patients receiving a treatment strategy including intranasal ESK and patients treated with SSRIs without esketamine.

Reviewer comment: “Provide guidelines as a reference point for the statement ‘patients with treatment-resistant depression treated with intranasal esketamine as adjunctive therapy and patients receiving standard SSRI-based antidepressant treatment’.”

Authors’ response: Thank you. We added the current CANMAT 2023 guidelines (published in 2024) as a reference point. These guidelines include intranasal esketamine as an add-on treatment option for patients with an inadequate response to at least two adequate antidepressant trials. The new reference was added to the bibliography as reference 22.

Text added to the manuscript:

Introduction – guidelines

Current Canadian Network for Mood and Anxiety Treatments (CANMAT) guidelines include intranasal esketamine as an add-on treatment option for patients with an inadequate response to at least two adequate antidepressant trials (22).

Reviewer comment: “All abbreviations should be written out in full at first occurrence (e.g., NMDA, WHOQOL-BREF).”

Authors’ response: Thank you. We reviewed the Abstract and the main text and expanded abbreviations at first occurrence, including TRD, SSRI, ESK, HRQoL, NMDA, WHOQOL-BREF, SF-36, MADRS, and HDRS. Abbreviations are used consistently thereafter.

Text added to the manuscript:

Introduction – abbreviations

World Health Organization Quality of Life–BREF (WHOQOL-BREF) ... 36-Item Short Form Health Survey (SF-36) ... N-methyl-D-aspartate (NMDA) receptor.

Reviewer comment: “At the end of the Introduction, the primary aim and hypothesis appear redundant; they should be simplified for clarity.”

Authors’ response: We agree. The final paragraph of the Introduction was shortened and the repeated statement of the aim was removed. It was replaced with one concise paragraph containing the primary aim, the secondary aim, and the hypothesis.

Text added to the manuscript:

Introduction – simplified aims paragraph

The primary aim of the study was to compare 12-month changes in HRQoL ... The secondary aim was to describe changes in the severity of depressive symptoms ... We hypothesized that the strategy including ESK would be associated with greater improvement in HRQoL.

Reviewer comment: “The primary and/or secondary objectives of the study (if any) should be clearly stated.”

Authors’ response: Thank you. The primary objective was defined as the comparison of 12-month changes in HRQoL across WHOQOL-BREF domains and SF-36 subscales between the two treatment strategies. The secondary objective was to describe longitudinal changes in MADRS/HDRS scores and interpret them alongside the course of HRQoL.

Text added to the manuscript:

Introduction – objectives

The primary aim of the study was to compare 12-month changes in HRQoL ... The secondary aim was to describe changes in depressive symptom severity measured with MADRS and HDRS and to interpret them in relation to the course of HRQoL.

METHODS

Reviewer comment: “Provide a reference to the STROBE guidelines and include the completed checklist as supplementary material.”

Authors’ response: Thank you. In Section 4.1, we added the full name of STROBE together with a citation to the existing reference (von Elm et al.). We also prepared a completed STROBE checklist as Supplementary Table S1.

Text added to the manuscript:

4.1. Study design

The study was reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (18). The completed STROBE checklist was included as supplementary material (Supplementary Table S1).

Reviewer comment: “The statement ‘Randomization was not used’ may be misleading to readers because this was an observational study. In this context, why randomization…”

Authors’ response: We agree. The phrase “Randomization was not used” has been removed. The revised wording explains that, in this observational study, treatment exposure was not assigned by the investigators; groups were defined on the basis of treatment selected in routine clinical practice.

Text added to the manuscript:

4.1. Study design

Treatment exposure was not assigned by the investigators; patients were classified according to the strategy selected by the treating physician in routine clinical practice.

Reviewer comment: “The term ‘optimized’ antidepressant therapy in the esketamine group may not be clear to a general readership and requires further clarification. More detailed SSRI protocols (e.g., preferred agents, dose, and possible modifications) and the exact protocol for ‘optimized’ antidepressant therapy in the esketamine group would facilitate replication. A standardized quality-of-life assessment procedure should be presented.”

Authors’ response: Thank you for this important comment regarding reproducibility. We clarified that the protocol did not impose a preference for any specific SSRI or a fixed titration algorithm; the choice of agent and dose modifications were clinical decisions made by the treating physician based on response and tolerability. Specific agents and doses at the individual-patient level were not prospectively standardized study variables; therefore, we did not reconstruct them post hoc and identified this as a limitation. The HRQoL assessment procedure was standardized by specifying the same instruments, instructions, scoring rules, and time points in both groups.

Text added to the manuscript:

4.3. Treatment

The study protocol did not impose a preference for any specific SSRI agent or a fixed dose-escalation algorithm. The choice of oral SSRI and any dose modifications remained at the discretion of the treating physician.

4.4. Assessment instruments

In both groups, the same versions of the instruments, instructions, scoring rules, and prespecified assessment time points were used.

Reviewer comment: “Neither sample size nor power analysis is reported, while the limitations state that ‘the sample size was moderate and included 60 patients, 30 in each group.’ On what basis do the authors consider the sample size moderate? The absence of a prespecified sample-size calculation is a potential limitation.”

Authors’ response: Thank you. We added a justification of the sample size: assuming Cohen’s d = 0.5, alpha = 0.05, and 80% power, the approximate target sample size was about 64 participants per group. Ultimately, 30 participants per group were recruited because of recruitment feasibility in a real-world setting. We removed the term “moderate sample size” and clearly stated that the study had limited power to detect small effects and to perform subgroup analyses.

Text added to the manuscript:

4.2. Participants – sample size

An approximate a priori calculation ... Cohen’s d = 0.5, alpha = 0.05, and 80% power ... approximately 64 participants in each group. ... the final cohort included 30 participants in each group.

Reviewer comment: “The statistical analysis section lacks information about the analyses performed, their rationale, and the software/database used (including version). More detail would improve transparency for readers, particularly with respect to replication.”

Authors’ response: Thank you. Section 4.6 was expanded to describe data presentation, selection of tests according to distribution, analysis of categorical variables, comparison of T0–T_end changes, and repeated-measures analysis. The significance criterion was also added. The name and version of the statistical software were not stated in the supplied manuscript; therefore, the file contains a clearly marked placeholder to be completed by the authors rather than introducing unverified information.

Text added to the manuscript:

4.6. Statistical analysis

Baseline comparisons ... Student’s t-test or Mann–Whitney U test; categorical variables ... chi-square or Fisher’s exact test. ... longitudinal course ... RM-ANOVA ... Statistical analyses were performed using [STATISTICAL SOFTWARE NAME AND VERSION — TO BE CONFIRMED BY THE AUTHORS].

RESULTS

Reviewer comment: “Table 2: What do you mean by general index? Specify whether this is PCS, MCS, or a documented composite index, and define how it was calculated.”

Authors’ response: We agree. The term “SF-36 general index” was ambiguous and could imply PCS, MCS, or a validated composite index. Because the manuscript did not document such a score according to the standard SF-36 scoring system, we removed the non-standard “general index” from Table 2, the Abstract, and the interpretation. SF-36 results are now reported only at the level of the standard subscales; PCS and MCS were not calculated.

Text added to the manuscript:

Results – SF-36

Neither the Physical Component Summary (PCS), Mental Component Summary (MCS), nor any other validated overall SF-36 index was calculated; therefore, the non-standard “general index” from the previous version was removed from the inferential interpretation.

Reviewer comment: “Table 3: Social relationships should be ‘Social relationships’.”

Authors’ response: Thank you. The label was standardized to the standard term “Social relationships.”

Text added to the manuscript:

Table 3

Social relationships

Reviewer comment: “Table 4: The labels do not correspond to the SF-36 domains. For example, SF-36 does not include a domain called ‘role limitations’. This may mislead readers.”

Authors’ response: We agree. The SF-36 subscale labels in Table 4 were standardized according to standard terminology, including “Role limitations due to physical health (RP),” “Emotional well-being / mental health (MH),” and “Bodily pain (BP).”

Text added to the manuscript:

Table 4

Role limitations due to physical health (RP); Mental health / emotional well-being (MH); Bodily pain (BP).

Reviewer comment: “Tables 2–4: The statistical analysis used should be specified in the footnote.”

Authors’ response: Thank you. The footnotes to Tables 2–4 were expanded to specify which tests correspond to the reported p-values and to mention the additional longitudinal RM-ANOVA analysis described in Section 4.6.

Text added to the manuscript:

Footnotes to Tables 2–4

P-values refer to between-group comparisons of T0–T_end changes using the independent-samples Student’s t-test or Mann–Whitney U test, depending on data distribution. The longitudinal course was additionally assessed using RM-ANOVA as described in Section 4.6.

Reviewer comment: “Group label in Tables 5–6: ‘ECC’ should be ESK; ‘SSRI’ should also be standardized to SSRI (singular) to match the text/footnotes.”

Authors’ response: Thank you. In Tables 5–6, “ECC” was corrected to “ESK,” and the comparison-group label was standardized to “SSRI (n=30).”

Text added to the manuscript:

Tables 5–6

ESK (n=30); SSRI (n=30).

Reviewer comment: “Table labels should be consistent: ‘Week 4’ compared with ‘Week 12’ and T0/T_end.”

Authors’ response: Thank you. The time points in Table 1 were standardized to the format: Baseline (T0), Week 4, Week 8, Week 12, Month 12 (T_end), with an analogous Polish version.

Text added to the manuscript:

Table 1

Baseline (T0); Week 4; Week 8; Week 12; Month 12 (T_end).

DISCUSSION

Reviewer comment: “Discuss the potential limitations in greater detail, as the current discussion is insufficient, as follows. Discuss unmeasured or residual confounding and propose specific ways to address it in future studies (e.g., propensity scores, adjusted models). How might the lack of randomization and potential selection bias affect the interpretation of between-group differences?”

Authors’ response: Thank you. The limitations section was substantially expanded. Confounding by indication, selection bias, and unmeasured confounders, as well as their impact on the interpretation of differences, were discussed. We also added specific approaches for future studies: prospective collection of treatment exposure and adherence data, propensity-score matching/weighting, inverse-probability-of-treatment weighting, adjusted mixed-effects models, sensitivity analyses, and—where feasible—blinded outcome assessment.

Text added to the manuscript:

Discussion – limitations

Future observational studies should ... use methods to reduce measured confounding, such as propensity-score matching or weighting, inverse-probability-of-treatment weighting, and adjusted longitudinal mixed-effects models.

Reviewer comment: “Define what constitutes a clinically meaningful change in WHOQOL-BREF and SF-36, rather than reporting statistical significance alone.”

Authors’ response: Thank you. A separate paragraph on clinical relevance was added. We noted that there is no single universal MCID threshold for WHOQOL-BREF in TRD; as an indicative reference, we cited a TRD study using a 10-point threshold for WHOQOL-BREF domains scored from 0 to 100. For SF-36, the commonly used range of approximately 3–5 points was indicated, with an explicit caveat that this depends on the population and domain. The interpretation was defined as exploratory because these thresholds were not prespecified in our cohort.

Text added to the manuscript:

Discussion – clinical relevance

In a TRD study using WHOQOL-BREF, a practical threshold of 10 points ... was used as an indicative clinically meaningful change (23). For SF-36 scales, differences of approximately 3–5 points are often considered potentially clinically meaningful ... (24).

Reviewer comment: “Add how clinicians can implement quality-of-life monitoring or esketamine treatment, with reference to practical challenges.”

Authors’ response: Thank you. The practical section of the Discussion was expanded. We proposed repeated HRQoL monitoring at key treatment time points in parallel with MADRS/HDRS and discussed real-world requirements for implementing esketamine: patient eligibility assessment, supervised administration, monitoring of blood pressure and adverse events, post-dose observation, staffing resources, appointment availability, and the burden of frequent visits.

Text added to the manuscript:

Discussion – clinical implementation

A practical approach includes assessment with WHOQOL-BREF or SF-36 before initiation of the treatment strategy, repetition of the assessment after key treatment stages, and periodic reassessment during maintenance treatment ... For esketamine, implementation requires appropriate patient qualification, supervised administration, monitoring of blood pressure and adverse events, post-dose observation, trained staff, and appointment availability (22).

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Author Response

We would like to sincerely thank the Reviewer for the careful evaluation of our manuscript and for the valuable and constructive comments. We appreciate the time and effort devoted to the review. The comments helped us improve the clarity, methodological transparency, and overall quality of the manuscript. We have addressed all points raised and revised the manuscript accordingly.

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for considering all comments.

Author Response

Reviewer 3 Report

Comments and Suggestions for Authors

I really appreciate the authors' effort in addressing the comments and making appropriate revisions. However, one minor correction remains:

Section 4.6 still states: Statistical analyses were performed using [STATISTICAL SOFTWARE NAME 501 AND VERSION – TO BE CONFIRMED BY THE AUTHORS]. Replace it with the exact software and version.

Author Response

Thank you for this comment. The requested information has been added to Section 4.6 of the revised manuscript. The following sentence is now included in the new version:

“Statistical analyses were performed using TIBCO Statistica 13.3 (TIBCO Software Inc., Palo Alto, CA, USA).”