In the United States, health care costs approximately $3.2 trillion annually. Of that, 75% is attributed to diseases related to metabolic dysfunction, for example type 2 diabetes, hypertension, lipid problems, heart disease, non-alcoholic fatty liver disease, polycystic ovarian syndrome, cancer, and dementia. In the United States alone, metabolic diseases affect more than 100 million people, resulting in significant increases in medical costs.
In particular, globally, nearly 425 million adults live with diabetes with close to 325 million at risk of Type 2 diabetes, in particular. In 2017, diabetes alone was the cause of $727 billion dollars in health expenditure and that number has continued growing every year. The medical community has so far seen type 2 diabetes as a chronic and progressive disease. Once diagnosed, it is a life sentence. Medications improve the blood sugar levels (the symptom), but do not address the diabetes (the actual disease). Moreover, the management of diabetes treatments involve costly medications, painful insulin injections, constant finger pricks, dietary restrictions, and other factors that result in a general loss in quality of life. Worse, diabetic patients suffer from weakened immune systems, potential tissue death and amputation, numbness, increased risk of heart disease, and a multitude of other medical ailments.
Conventional disease management platforms or techniques either ignore or fail to fully understand important markers, such as blood sugar dysregulation, and root causes for these diseases, such as processed foods and lack of exercise. Traditionally, these platforms are designed to treat symptoms of the diseases as they arise rather than treating the root cause of the disease — the deterioration of a patient’s metabolic health. Platforms that have attempted to treat metabolic diseases focused on an “average” patient rather than tailoring their treatment and management regimens to the specific metabolic health of each patient. Accordingly, such platforms prescribe suboptimal treatments that have diminished efficacy or unwanted side effects (e.g., prescribing excessive medication) for individual patients.
Additionally, conventional disease management platforms struggle to address two challenges. First, they are unable to acquire relevant biosignal data in a timely manner, validate the accuracy of any acquired biosignal data, and ensure the completeness of ongoing data collection so that a resulting treatment recommendation may be trusted. Second, these platforms are unable to achieve high patient adherence to their prescribed treatment recommendations. Traditional approaches attempt to manually acquire such data and monitor patient adherence, which results in delayed, inaccurate and inconsistent results.
That is why, in Twins Digital, we introduce TAC (Timeliness, Accuracy, and Completeness) into our metabolic health management.
A patient health management platform for managing a patient’s metabolic diseases applies machine learning for precision treatment and analyzes a unique combination of continuous biosignals from one or more of the following sources: near real-time biological data recorded by wearable sensors, biological data recorded by lab tests, nutrition data, medicine data, and patient symptoms. The platform performs various analyses to establish a personalized metabolic profile for each patient by gaining a deep understanding of the combination of continuous biosignals of the patient. The platform generates a time series of metabolic states based on the biosignals continuously/regularly recorded for a period of time, which allows the platform insight into not only the patient’s current metabolic state at particular time points within a day/time period, but also a complete history of metabolic states that led to that current metabolic state (e.g., a collection of metabolic states at multiple time points across preceding days/time periods). These biosignals can be input into machine-learned model(s) that recommend personalized treatment based on a unique metabolic profile of the patient.
Based on the output of the machine-learned models, the patient health management platform generates personalized recommendations for a patient outlining a treatment plan for improving their metabolic health, for example a personalized nutrition plan, a medication plan, and an exercise and sleep regimen. The recommendations can be similarly detailed and time-specific, including recommending specific actions at particular time points (e.g., eating a specific amount of a specific food at 3pm). These recommendations may be reviewed by doctors and coaches to improve their accuracy and usefulness, but once approved, the recommendations are delivered to a patient via an application interface on a mobile device. Over time, as a patient follows these recommendations, the platform captures the resulting response in their biosignals to dynamically quantify the impact of each recommendation and serve as a feedback loop to refine and optimize the recommended treatments.
To confirm that a patient-based recommendation effectively addresses a patient’s metabolic health, the patient health management must evaluate the patient data recorded by a patient to confirm the timeliness, accuracy, and completeness of the recorded data. A timelines measurement evaluates the time delay between when event data, for example nutrition or symptom data, occurred and when it was recorded by the patient. An accuracy measurement evaluates whether data was misreported or is otherwise inaccurate, for example whether a food item was omitted from a record of entered nutrition data. A completeness measurement evaluates whether any critical data is missing from a record of patient data. To measure these aspects of a patient’s record of patient data, the patient health management platform compares a predicted metabolic state of a patient (predicted by the platform based on what information was recorded for the patient) with the patient’s true metabolic state and flags any inconsistencies. The flagged inconsistencies may, in some instances, be attributed to errors in a patient’s recordation of data. To encourage improvements in the above metrics, the patient health management platform may evaluate a patient’s overall record entries by assigning the patient a timeliness/accuracy/completeness (TAC) score and dynamically update the TAC score based on the patient’s subsequent patient data entries.
This is in a nutshell of Twins’ sophisticated TAC system. More and more features will be built based on this concept, and we are seeing great results from patients’ progress.