This PubMed record (PMID 41641589; DOI 10.1097/CM9.0000000000003845) describes an article titled “Serum potassium as a predictor of 28-day and in-hospital mortality in patients with sepsis: A U-shaped association,” authored by Guang Zhao and colleagues. The work is published in Chinese Medical Journal (English) with citation details 2026 Aug 20;139(16):2502-2504 and an Epub date of 2026 Feb 5. Author affiliations include the Department of Emergency Medicine and the Department of Endocrinology at the Affiliated Kunshan Hospital of Jiangsu University, Kunshan, China. A Wolters Kluwer link is provided from the PubMed entry for full-text access.
The article title and bibliographic metadata are present in the PubMed entry and provide the primary descriptors necessary for citation and retrieval.
The PubMed entry contains no abstract and therefore does not report the study’s methods, sample size, inclusion or exclusion criteria, potassium measurement timing, definitions of low or high serum potassium, statistical analyses, effect estimates, or confidence intervals. The absence of an abstract in the PubMed record means that readers cannot evaluate the study design or results from the PubMed page alone.
Key identifiers (PMID, DOI, journal, authors, affiliations) are available and should be used to obtain the full text for complete study details. The PubMed page also lists similar articles on potassium and mortality in sepsis and critical illness, which may provide additional context while the full text is retrieved.
The article title states a U-shaped association between serum potassium and both 28-day mortality and in-hospital mortality among patients with sepsis. In epidemiologic terminology, a U-shaped association typically indicates that mortality risk is higher at both lower and higher levels of the exposure (here, serum potassium) and that an intermediate range is associated with lower risk.
From the PubMed record alone it is not possible to determine: the serum potassium thresholds defining the nadir and the increased-risk tails; whether the association was adjusted for confounders (age, comorbidities, renal function, illness severity); whether the analysis was observational or used a specific database; or the magnitude and statistical significance of the association. These details are not reported on the PubMed page and must be confirmed in the full text.
The title’s claim — that serum potassium predicts both 28-day and in-hospital mortality in sepsis with a U-shaped pattern — is clinically relevant because potassium disturbances are common in critically ill septic patients and are modifiable. However, clinicians should not change practice or infer clinical thresholds from the PubMed entry alone.
Before applying findings to patient care, readers should obtain the full article to review the following critical elements: study population and setting, timing and method of potassium measurement, definitions of hypokalemia and hyperkalemia, outcome definitions (28-day and in-hospital mortality), statistical models and covariates, sensitivity analyses, and limitations discussed by the authors.
The PubMed entry links to related and similar articles examining potassium levels and mortality in sepsis and critical illness, including analyses of the MIMIC databases and multicenter cohort studies. Those related studies may offer complementary data on potassium thresholds, adjustments for illness severity, and subgroup findings (for example, elderly patients or those with diabetes or chronic kidney disease). Clinicians and researchers may consult these related publications while seeking the full text of the Zhao et al. paper.
To evaluate the study’s methods and results, retrieve the full text via the Wolters Kluwer link provided on the PubMed page or through institutional access using the DOI 10.1097/CM9.0000000000003845. When appraising the article, focus on study design (retrospective vs prospective), potassium measurement protocol, covariate adjustment, robustness checks, and external validity to local patient populations.
Until the full text is reviewed, specific numeric findings, thresholds, statistical measures, and authors’ recommendations are not available from the PubMed entry and were not reported in the source record.