The invention relates to the technical field of smart
medical treatment, in particular to an inpatient sign data alarm method and
system, and the method comprises the steps: obtaining multi-source
sign detection data of an inpatient, and carrying out the clustering
decomposition, and obtaining the sub-
feature detection data of the multi-source
feature detection data; collecting real-time illness state parameters of the inpatient and personal attributes of the inpatient, and endowing weighted values to the sub-
feature detection data according to the real-time illness state parameters and the personal attributes of the inpatient; based on the
weight value of the sub-feature detection data and in combination with the standard upper and
lower limit thresholds of the sub-feature parameters, the early warning priority of the inpatient is evaluated; and when the multi-source physical
sign detection data of the plurality of
hospitalized patients are abnormal,
nursing personnel are reminded to nurse the
hospitalized patients according to the early warning priorities of the
hospitalized patients. According to the method, personalized upper and
lower limit setting is adopted for the scenes, facing the
nursing sequence, of multiple patients in the monitoring period, early warning priority configuration is carried out adaptively according to the current conditions of the patients, the early warning monitoring process is optimized, and efficiency is improved.