ABSTRACT
Acute diarrhea is associated with a reduced absorption of both vitamin K antagonists (VKA) and vitamin K itself. To date, the net effect on the coagulation status of subjects with VKA remains elusive. We performed a systematic retrospective single-center analysis using an electronic data extraction approach to identify subjects with plasmatic anticoagulation (either VKA or direct oral anticoagulant (DOAC)) and diarrhea in a German University Hospital over a period of eight years. Acute diarrhea and complete documentation of coagulation status on admission were defined as inclusion criteria, anticoagulation other than VKA/DOAC and obvious inadherence as exclusion criteria. Subjects with VKA/DOAC admitted for hypertension served as control group. Data extraction yielded 356 subjects with gastrointestinal diagnoses and 198 hypertensive subjects, 55 and 83 of whom fulfilled all in- and exclusion criteria. INR values of subjects with VKA were significantly higher in subjects with diarrhea than in hypertensive controls (4.3 ± 3.7 vs. 2.3 ± 0.7, p < 0.001). The distribution of subjects having INR values lower, higher or within the target range differed significantly among groups with a substantially higher prevalence of overanticoagulation in the diarrhea group (46.4% vs. 14.3%, p < 0.001). In a multinomial logistic regression model, acute diarrhea was significantly associated with overanticoagulation (odds ratio 7.2, 95% confidence interval 2.163-23.921; p < 0.001), whereas age, sex, creatinine, and indication of anticoagulation were not (p > 0.05 each). Acute diarrhea is associated with a highly increased risk for overanticoagulation in patients with VKA. Thus, gastroenteritis necessitates a close monitoring of INR in order to identify subjects needing a temporary pause of VKA therapy.
