Discharge Readmission Analysis and Management in Sepsis (DReAMS-2)

Discharge Readmission Analysis and Management in Sepsis (DReAMS-2): Augmenting Readmission Risk Models With Smartphone PPG Signals and Deep Learning, an Adaptive Platform

This is an adaptive platform. This study is being done to collect information that will help us identify trends in patients with sepsis and other health conditions being readmitted into hospitals within 30 days of being discharged. This information will be used to create a computer tool that will help predict a patient's risk of being readmitted into the hospital after being discharged.

Participants will allow the study team to follow their health after they are discharged by taking their temperature once a day and placing their index finger over their smartphone camera when prompted by a text message. Participants will receive the text messages twice a day. When the participant receives the text message, they will click on the link and follow the instructions. Instructions include how to long to keep your finger on your phone camera and how to report your daily temperature. Additional questions will also be asked. After 30 days, the text messages will stop, and participation will be complete.

Patients admitted into the hospital for sepsis and other health conditions (see inclusion criteria) and planning to be discharged within 72 hours will be approached for this study. Once patient agrees to participate in this study, the following will occur:

• The study team member will guide the participant to open the study smartphone app on their phone.
• The app will provide a tutorial on how to take a measurement, after which the participant will complete their first measurement. The Duke research team member will assist the participant to ensure it is taken correctly. The measurement involves taking a video of the participant's fingertip using the app by placing their finger on the smartphone camera and holding their hand flat on a table. The video frames are immediately processed in real-time to extract the blood flow signal. The app automatically uploads the blood flow signal to the Measure Labs cloud backend.
• Once the patient has completed their first measurement, they will be considered enrolled in the study.
• Study team will measure participant's vital signs (heart rate, respiratory rate, blood pressure, temperature, and oxygen saturation) after they have been enrolled into the study.
• A digital thermometer will be sent home with the participants on day of discharge. The app will provide an input for the participant to enter their daily temperature.

Throughout the study, the participant will be expected to place their fingertip on their smartphone camera twice a day and take their temperature once a day. Participants will receive two text messages with a link to enter their temperature, scan their fingertip with their camera, and answer some questions about their current health status. Participation will occur from day of consent to 30 days after patient was enrolled. During the final week of the study, the patient will be asked if they would like to continue receiving text messages for an additional 60 days. If they decline or don't respond, after day 30, participant will no longer receive text messages.

Inclusion Criteria:

• Age 18 years or older

• Admitted to the hospital for 48 hours or greater

• Admitted with an index diagnoses of sepsis. We defined sepsis based on the 2018 Centers for Disease Control and Prevention's Adult Sepsis Event (ASE).

  [1] A subject must have an order for blood cultures, at least 4 qualifying antimicrobial days AND Evidence of at least 1 end organ dysfunction (new vasopressors, initiation of mechanical ventilation, doubling of serum creatinine (or GFR reduction by 50%), total bilirubin ≥ 2.0 mg/dl, an increase by 100% from baseline, platelet count < 100 cells/ul, and lactate ≥ 2.0.

• The adaptive design will allow the expansion into additional index hospitalization diagnosis (for example, heart failure exacerbation, sickle cell crises, COPD exacerbation, hip fractures, pneumonia, post surgical patients, diabetes mellitus)

Exclusion Criteria:

• Imminent death or anticipated death during index admission (to include comfort care or hospice)
• Patients due to be transferred to another acute care facility
• Participants not proficient with written and spoken English.
• Participants who do not have a smartphone and valid mobile number to receive text messages.
• Participants leaving the hospital against medical advice.
• Participants with essential tremor.
• Participants deemed non-compliant or not suitable for the study due to cognitive impairment or serious mental health conditions as determined by the Principal Investigator.
• Participants unable to independently navigate and operate smartphone applications.
• Participants with diminished decision-making capability.
• Participants likely to not be available to complete all protocol-required procedures (e.g., Clinical Outcome Assessments) to the best of the subject and investigator's knowledge.
• Participants with history or evidence of any other clinically significant disorder, condition or disease (with the exception of those outlined above) that, in the opinion of the investigator or study team, if consulted, would pose a risk to subject safety or interfere with the study evaluation, procedures or completion
• Vulnerable Subjects, this study will not enroll children < 18 years of age, pregnant women, or prisoners.