Letter regarding: »Arterial pH and short-term mortality in adult non-traumatic acute patients«

Regarding the publication on »Arterial pH and short-term mortality in adult non-traumatic acute patients« [1].

Studies on the relationship between arterial pH and mortality provide useful information for risk categorisation in the acute period. However, various statistical and methodological concerns must be addressed in order to improve the findings and ensure accurate conclusions. A critical concern is the potential impact of confounding circumstances. Although confounding factors were controlled in this investigation, it is unclear which variables were included in the analysis and whether all relevant confounding factors were appropriately addressed. For example, underlying comorbidities, acute illness severity, and therapy measures can all have a major impact on pH levels and mortality. Details about these modifications are insufficient, therefore a more detailed discussion of the statistical models utilised and the precise confounding factors addressed is required to contextualise the results.

Another problem is that this study is based on hazard ratios (HRs) for mortality associated with various pH values. Although HRs are a valuable comparison, this study does not investigate whether the link between pH and mortality is linear or if thresholds or nonlinear interactions should be considered. The high hazard ratio for severe acidosis (pH < 7.20) shows a substantial correlation, but it's important to assess the dose-response relationship between pH and death. Furthermore, failure to account for time-dependent elements or the precise timing of pH shift (e.g., within the first few hours of acute care consultation) may impair the accuracy of the findings obtained, particularly in terms of early and late mortality outcomes.

Future research could use a more sophisticated approach, focussing on patient-specific characteristics such pre-existing medical disorders and medications, as well as investigating the impact of different pH categories in predicting short- and long-term outcomes. Furthermore, prospective studies that constantly monitor pH during patients' therapy may provide more dynamic information about the link between pH and mortality. Future research should look into interventions that can rectify acidosis and/or alkalosis early in the illness course and see whether they have a direct impact on mortality risk. In conclusion, while this study provides helpful information, future research should focus on developing statistical models to include new confounding factors and accounting for more precise characteristics such as time and patient heterogeneity.