Automated mobility assessment
Automatically evaluating a level of mobility of a patient using a machine-learning model and load sensors integrated into a patient support system. A machine-learning model implemented on a computing system analyzes the load sensor data to determine whether the patient has difficulty standing up. This automated analysis can replace existing mobility assessments that were previously performed by caregivers. The mobility assessment data generated by the computing system can be recorded in the patient's electronic medical record (EMR) and utilized as a factor in determining patient fall risk.
1 . A system for automatically evaluating a level of mobility of a patient, the system comprising:
a patient support system comprising at least one load sensor, wherein the patient support system is a bed and the at least one load sensor comprises at least two load beams embedded in the bed and positioned to measure patient movements while lying on the bed;
at least one processor in communication with the at least one load sensor; and
a memory encoding instructions which, when executed by the at least one processor, cause the at least one processor to:
record load sensor data from the patient support system for a length of time for each of a plurality of patients;
determine a level of mobility of each of the plurality of patients during the length of time based on observations of the patient movements;
train a machine-learning algorithm with the load sensor data and corresponding level of mobility of each patient to identify patterns of the load sensor data indicative of particular levels of mobility;
output a mobility model operable to determine a level of a mobility of a patient based on the load sensor data;
evaluate a falls risk level of the patient based on the level of the mobility of the patient, the level of the mobility of the patient including a length of time to exit the bed, wherein the length of time is determined from when the patient is in a position on the patient support structure to when the patient is in a standing position, and wherein the length of time is compared to a range of values;
record the falls risk level of the patient; and
automatically implement a safety measure on the bed based on the falls risk level by arming guard rails on the bed when the falls risk level exceeds a threshold.
2 . The system of claim 1 , further comprising a camera configured to capture images of the patient on the patient support system during the length of time, wherein observations of the patient movements are performed by reviewing the images captured by the camera.
3 . The system of claim 1 , wherein the length of time is at least 10 seconds.
4 . The system of claim 1 , wherein the machine-learning algorithm is a supervised machine-learning algorithm.
5 . The system of claim 1 , wherein if the falls risk level exceeds the threshold, an alert is issued to a hospital information system.
6 . The system of claim 1 , further comprising an infrared sensor configured to detect movement of the patient and activate a camera.
7 . The system of claim 1 , further comprising one or more location monitoring devices configured to track locations of caregivers.
8 . A method of automatically evaluating a level of mobility of a patient, the method comprising:
recording load data received from load sensors embedded in a patient support system, wherein the patient support system is a bed and the load sensors comprise at least two load beams embedded in the bed and positioned to measure patient movements while lying on the bed;
analyzing the load data at a computing system using a mobility model;
outputting a level of mobility for the patient, the level of mobility of the patient including a length of time to exit the bed, wherein the length of time is determined from when the patient is in a position on the patient support structure to when the patient is in a standing position, and wherein the length of time is compared to a range of values;
evaluating a falls risk level of the patient based on the level of the mobility of the patient;
recording the falls risk level of the patient; and
automatically implementing a safety measure on the bed based on the falls risk level by arming guard rails on the bed when the falls risk level exceeds a threshold.
9 . The method of claim 8 , wherein the mobility model is generated by:
observing movements of a plurality of patients for a length of time, each patient being on a patient support system;
determining mobility scores for each of the plurality of patients based on the observed movements;
recording load data from each patient support system for the length of time;
correlating observed movements and mobility scores with patterns of load data; and
training a machine-learning algorithm with the load data and corresponding mobility scores to identify patterns of load sensor data indicative of the patient movements corresponding to mobility scores.
10 . The method of claim 8 , further comprising analyzing the level of mobility in combination with additional patient data to determine the falls risk level for the patient.
11 . The method of claim 10 , wherein the additional patient data is automatically retrieved from the patient's EMR.
12 . The method of claim 10 , wherein the additional patient data is received at a computing device from inputs provided by a caregiver.
13 . The method of claim 8 , further comprising computing a mobility score based on the level of mobility of the patient, and recording the mobility score to an EMR of the patient.
14 . The method of claim 13 , wherein the mobility score is based on the Get Up and Go Test.
15 . The method of claim 10 , wherein the falls risk level is determined using the Hendrich II Fall Risk Model.
16 . The method of claim 10 , further comprising issuing an alert when the falls risk level is high.
17 . One or more non-transitory computer-readable media having computer-executable instructions embodied thereon that, when executed by one or more computing devices, cause the one or more computing devices to:
generate a mobility model, including to:
record, over a period of time with a video camera, video images of each of a plurality of patients on respective patient support systems, wherein each of the respective patient support systems is a bed including at least one load sensor having at least two load beams embedded therein and positioned to measure patient movements while lying on the bed;
record, over the period of time with the at least two load beams embedded in the patient support systems, load data for each of the plurality of patients;
analyze the load data and video images with a patient movement analyzer to determine patterns of the patient movements of each patient;
assign a mobility score to each pattern of movement based on a caregiver evaluating the video images corresponding to the pattern of movement with a mobility test; and
train a supervised machine-learning algorithm with the load sensor data and corresponding mobility scores to output the mobility model;
evaluate a level of mobility for an individual patient by recording load sensor data from the at least two load beams embedded in a patient support system, the patient support system including the bed on which the individual patient is resting; and
analyze the load sensor data with the mobility model, including to:
evaluate the individual patient's fall risk by analyzing the individual patient's mobility level with other factors of a fall risk assessment, wherein the individual patient's mobility includes a length of time to exit the bed before a successful exit of the bed, wherein the length of time is determined from when the patient is in a position on the patient support structure to when the patient is in a standing position, and wherein the length of time is compared to a range of values, and wherein the other factors are determined by one or more of receiving input from the caregiver, accessing information from the individual patient's EMR, and receiving data from a patient monitoring device;
record the individual patient's mobility score and fall risk in the individual patient's EMR; and
if the individual patient's mobility score or fall risk exceed a predetermined threshold, automatically implement a safety measure by arming guard rails on the bed and issue an alert to a hospital information system.