Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review this article.

This manuscript offers an important contribution to the emerging discussion on the use of AI in the provision of care for older adults and in palliative care. By drawing on Actor-Network-Theory, the authors propose three central themes: (1) the tensions between rules-based logic and patient-centered approaches; (2) the ethical discomfort raised by the use of AI in contexts involving emotional care; and (3) the structural inequalities that create barriers to its implementation in resource-limited settings. The manuscript is well written, clearly structured, and engaging to read.

While the paper makes a valuable contribution, I would like to suggest several areas where clarification or further elaboration would strengthen the work:

At the outset, the manuscript emphasizes the importance of integrating palliative and geriatric care. However, this theme is not revisited in the remainder of the text. A clearer alignment between the introduction and subsequent text would enhance coherence.

The rationale for selecting the field site requires further justification. Beyond its location in a region with an aging population and limited resources, why was this particular hospital chosen? Were other hospitals in the region considered, and if so, why was this one prioritized?

The authors note that care in this hospital combines Traditional Mongolian Medicine, Traditional Chinese Medicine, Western biomedicine, and geriatrics. It would be useful to explain more concretely how this integration is enacted in practice, particularly in relation to the utilisation of AI.

Details on recruitment methods are lacking. How were healthcare professionals invited to participate—through posters, email communication, or presentations on care units? How long did the interviews last?

In Table 1, the column “Preferred AI Tools” does not mention Deepseek, despite the claim that some providers use it. In addition, the entry “be cautious about AI” is ambiguous and would benefit from clarification.

For future research, a comparative dimension would be valuable. For instance, perspectives on AI use could be explored in a hospital formally integrated into a palliative care program, or in an urban setting, to better capture the dynamics at play across different contexts.

In sum, this manuscript presents original insights and addresses an important and timely topic. With the clarifications and elaborations suggested above, the paper will make an even stronger contribution to the field.

Author Response

Comments 1: At the outset, the manuscript emphasizes the importance of integrating palliative and geriatric care. However, this theme is not revisited in the remainder of the text. A clearer alignment between the introduction and subsequent text would enhance coherence.

 

Response 1:  Thank you for pointing out the importance of integrating palliative and geriatric care in light of AI implementation. We agree it is important to contextualise our article within this unique setting throughout. Therefore, we have made comprehensive changes across the manuscript, including in the Introduction (pp. 2, line:42-68; 72-89), the Research Questions (pp. 3, line 126-131), Methodology (pp. 3, line 136), Results (pp. 6, line 226-232, pp. 7, line 281-287 and pp. 9 line 363-364; 371-372), and Discussion (pp. 11, line 466-467; pp. 12, line 532-533 and pp. 13, line 548-552). In so doing, we aim to provide a clearer account of how AI is engaged differently in general geriatric versus specialised palliative care, emphasise the relational and ethical dimensions of care, and highlight the limited but targeted role AI can play in supporting complex clinical practices.

 

 

Comments 2: The rationale for selecting the field site requires further justification. Beyond its location in a region with an aging population and limited resources, why was this particular hospital chosen? Were other hospitals in the region considered, and if so, why was this one prioritized?

Response 2: We have now added further details explaining why the field site was selected, considering both its marginalised status and its representativeness in terms of integrated geriatric and palliative care (section 2.1, pp. 4, line 149-156).

 

Comments 3: The authors note that care in this hospital combines Traditional Mongolian Medicine, Traditional Chinese Medicine, Western biomedicine, and geriatrics. It would be useful to explain more concretely how this integration is enacted in practice, particularly in relation to the utilisation of AI. 

 

Response 3: Thank you for pointing out the unique setting of our study site, which integrates multiple medical systems. While these diverse practices were intertwined within integrated geriatric and palliative care, our data captured little of this multidisciplinary context in relation to AI engagements. Instead, the data focused more on general attitudes towards AI and its impact on clinical care, as well as how ethical and structural barriers may shape such engagements. Therefore, we have removed the information about this context at the end of Section 2.1. However, we acknowledge that this unique context has the potential to yield further insights into the application of AI in diverse clinical settings, and we have added this point as a limitation in Section 4.1 (pp. 13, line 594-597, pp. 14, line 598-602).

 

 

Comments 4: Details on recruitment methods are lacking. How were healthcare professionals invited to participate—through posters, email communication, or presentations on care units? How long did the interviews last?

 

Response 4: We have added further details about our recruitment methods in section 2.2 (pp. 4, line 177-185).

 

 

Comments 5: In Table 1, the column ‘Preferred AI Tools’ does not mention Deepseek, despite the claim that some providers use it. In addition, the entry ‘be cautious about AI’ is ambiguous and would benefit from clarification.

 

Response 5: We have provided further clarification about ‘Preferred AI Tools’, particularly how practitioners perceive and engage with AI, including Doubao and DeepSeek, in the footnote on pp. 5. 

 

Comments 6: For future research, a comparative dimension would be valuable. For instance, perspectives on AI use could be explored in a hospital formally integrated into a palliative care program, or in an urban setting, to better capture the dynamics at play across different contexts.

 

Response 6: We strongly agree that a comparative lens would further enrich both the research design and findings to explore AI engagements in more established palliative care and urban contexts. Therefore, we have added this to the limitations to acknowledge that our current study captures only a single rural hospital context and that additional research is needed to understand how AI is integrated across diverse clinical settings (pp. 13, line 594-597, pp. 14, line 598-602).

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Colleagues,

Thank you for this manuscript. Below you can find the list of my suggestions to improve your manuscript:

1. Please consider Shortening the Introduction: Trim the lengthy introduction (pages 1-3) to focus on core research gaps and questions, condensing policy and background details.
2. Please Clarify AI applications: Specify targeted AI uses in geriatric/palliative care (e.g., symptom management or decision support) to align with participant references to tools like Doubao.
3. Include Interview Guide Table: Add a table or appendix with sample interview questions for transparency.
4. Summarize Findings in a Table: Present key themes, sub-themes, and quotes in a table to enhance readability.
5. Address Saturation: Note if/when thematic saturation was reached with the 14 participants.
6. Detail Selection Criteria: Explain purposive sampling rationale, including how age/experience variations were considered to mitigate potential biases.

Respectfully

Author Response

Comments 1: Please consider Shortening the Introduction: Trim the lengthy introduction (pages 1-3) to focus on core research gaps and questions, condensing policy and background details.

 

Response 1: We agree that the introduction could be more concise, with a stronger focus on emphasising the complexity of adopting AI in integrated geriatric and palliative care. Thank you for your comments; we have now streamlined the policy and background details in the introduction (pp. 2, pp. 3) to improve the rigorousness of the structure. 

 

Comments 2: Please Clarify AI applications: Specify targeted AI uses in geriatric/palliative care (e.g., symptom management or decision support) to align with participant references to tools like Doubao.

 

Response 2: We have provided further details on how our participants engaged with AI applications in practice at the beginning of Section “3. Results” (pp. 6, line 226-232), outlining different uses of AI and how these engagements are interconnected with the themes. In so doing, we aim to contextualise participants’ AI use within their everyday clinical work and highlight the diversity of AI adoption.

 

 

Comments 3: Include Interview Guide Table: Add a table or appendix with sample interview questions for transparency.

 

Response 3: Thank you for your suggestion. We have added the interview guide as “Appendix 1” to the article.

 

Comments 4: Summarize Findings in a Table: Present key themes, sub-themes, and quotes in a table to enhance readability.

 

Response 4: Thank you for your suggestion. We have added the interview guide as “Appendix 2” to the article.

 

Comments 5: Address Saturation: Note if/when thematic saturation was reached with the 14 participants.

Response 5: We have added further details about the saturation process in section 2.2 (pp. 4, line 173-176).

 

 

Comments 6: Detail Selection Criteria: Explain purposive sampling rationale, including how age/experience variations were considered to mitigate potential biases.

 

Response 6: We have added further details about our detail selection criteria in section 2.2 (pp. 4, line 180-185).