Use of a Novel Toilet Seat to Passively Collect Digital Biomarkers in Assisted Living Settings

PROJECT SUMMARY—Toi Labs is developing TrueLoo®, an Internet-connected replacement toilet seat that continually captures digital biomarkers from urine and stool. Older adults are at increased risk for urinary tract infections (UTI), clostridium difficile infections, gastrointestinal bleeding, dehydration, constipation, colorectal cancer, and other conditions that cause suffering and excess morbidity, mortality, and healthcare costs. Often, the first signs of these conditions are changes in urine or stool. When changes go undetected or detection is delayed, the underlying health conditions may worsen, leading to higher costs and worse outcomes. The current standard of care in skilled nursing facilities is excreta logging to document the frequency and appearance of residents’ urine and stool. This approach misses up to 76% of toileting events, and most changes are not identified. For example, in a 24-hour study of 54 residents in a skilled nursing facility, staff failed to document 93% of urinations and 60.5% of formed stools, and they failed to identify any of the multiple instances of red color or cloudiness in urine. To address this challenge, Toi Labs developed TrueLoo, an Internet-connected replacement toilet seat that images toilet contents, analyzes images to detect urine and stool changes, and flags concerning sessions for expert review. In a Phase I SBIR, Toi Labs a) demonstrated the feasibility and acceptability of deploying TrueLoo in care facilities and b) characterized TrueLoo’s performance for detecting nine digital biomarkers: red color with urine, black stool, red color with stool, red color on toilet paper, red color in isolation, loose stool, cloudy urine, frequency of urination or bowel movement, and oil-like droplets in stool. In pilot evaluations, TrueLoo increased documentation and timely reporting of clinically-relevant changes in urine and stool, decreased staff excreta logging time (average 10 to 30 minutes per staff member per day), and improved multiple patient outcomes (e.g., 32%-50% reduction in falls and an 11% reduction in hospitalizations due to UTIs). In this Phase II SBIR, Toi Labs proposes a two-arm randomized controlled trial to (Aim) Evaluate the effectiveness of TrueLoo for identifying clinically-relevant digital biomarkers from urine and stool and prompting clinically-appropriate follow-up compared to the current standard of care (i.e., manual excreta logging) in skilled nursing facilities. Primary Endpoint: Rate of identifying clinically-relevant digital biomarkers with TrueLoo is equal to or greater than the rate with manual logging. Secondary Endpoints: 1) Performance of a machine learning algorithm for identifying each digital biomarker; 2) initiation rates for follow-up; 3) time to initiation of follow-up; 4) staff time devoted to logging; 5) fall rates; and 6) hospitalization rates for UTI. Exploratory Endpoints: 1) catheterization rates, 2) enema rates, and hospitalization rates for 3) GI bleed and 4) constipation / fecal impaction. Impact—Demonstrating TrueLoo is at least as effective as skilled staff would support its deployment across many contexts as a low-cost, scalable approach for identifying and prompting intervention for changes associated with increased risk of falls, hospitalizations, morbidity, and mortality in older adults.