Abstract
Acute kidney injury (AKI) is common among hospitalized patients, but laboratory-defined AKI may not be documented. We aimed to determine the frequency of undocumented AKI in internal medicine wards, compare clinical characteristics and in-hospital outcomes by documentation status, and explore associated factors. We conducted a retrospective single-center study of adult hospitalizations with AKI defined by serum creatinine criteria; urine-output criteria were unavailable. Because reliable pre-admission creatinine values were not consistently available, the lower of admission and discharge creatinine was used as reference and compared with maximal creatinine. Among 161 hospitalizations in 161 unique patients, 39 (24.2%) were undocumented. Undocumented cases had lower admission creatinine (2.1 vs 2.9 mg/dL; p p p = 0.016). Length of stay and in-hospital mortality did not differ significantly. In 149 hospitalizations with a demonstrable creatinine change, 33 (22.1%) were undocumented. Higher maximal creatinine remained associated with lower odds of underdocumentation (adjusted OR 0.67 per 1 mg/dL, 95% CI 0.50-0.89; p = 0.006). Approximately one quarter of serum creatinine-defined AKI cases were undocumented, particularly those with lower maximal creatinine. Prospective studies should assess whether structured AKI detection and documentation improve kidney-related care and outcomes.