Abstract
Introduction
We appreciate the reader’s interest in our study 1 and their valuable observations regarding its methodology, analysis, and clinical relevance. Below, we respond to the points raised in the letter.
Methodological Considerations
Sample Size and Statistical Power
The letter highlights the lack of an explicit power analysis. However, it is important to highlight that the study was conducted as a census of all deaths that occurred in 2023 at Hospital III Daniel Alcides Carrión. By including the entire population of interest (n = 239), the risk of sampling error inherent in convenience sampling studies is minimized. Furthermore, the application of Poisson regression models with robust variance allowed for the identification of statistically significant associations, which supports the robustness of the analysis performed. 2
Definition of “Terminal Patients”
The absence of an explicit definition of “terminal patient” overlooks the fact that, in the context of retrospective studies, the inclusion of patients who died during hospitalization and had an advanced disease already constitutes an objective criterion for inferring a terminal condition. 3
Low Rate of Patients with Previous Palliative Care
The finding that only 2.1% of patients had received palliative care before hospitalization reflects, rather than a bias, the reality of clinical practice in the studied setting. Far from indicating an error in data collection, this finding underscores a critical issue: the limited integration of palliative care in certain hospital environments. This, in turn, justifies the relevance of the study in promoting improvements in end-of-life care.
Considerations in the Presentation of Results
Potential Biases
Although the letter states that the study did not sufficiently address sources of bias and the handling of incomplete data, the authors do acknowledge in the limitations section that the retrospective nature and single-center design may limit the generalizability of the results. The exclusion of cases with incomplete records is a common practice to ensure data quality, and the omission of a detailed report on these cases does not compromise the integrity or interpretation of the findings.
External Validity
The criticism regarding the limited generalizability due to the single-center design is, in fact, an aspect that the authors already acknowledge as an inherent limitation of the study. However, it is essential to consider that the study’s main contribution lies in providing data from an underexplored context (Peru) and demonstrating that the observed prescribing patterns are consistent with findings from other countries. This suggests that, despite the geographic limitation, the study provides valuable evidence for the global discussion on the inappropriate use of antibiotics in palliative care. 4
Clinical Implications and Recommendations
Regarding the suggestion to expand the discussion on local guidelines, antimicrobial resistance, and economic aspects, it is important to note that the primary objective of the study was to describe prescribing patterns at the end of life. While a broader discussion on these topics would enrich the debate, its absence does not diminish the relevance of the findings, which in themselves highlight a critical issue and pave the way for future research aimed at improving clinical practice and implementing antimicrobial stewardship programs.
