Biological samples, such as blood, urine, saliva, etc., are the gold standard for monitoring chronic diseases in clinical trials and real world settings. When analyzed frequently, they provide valuable information about response to therapy and adverse events. However, frequent blood testing is impractical due to its high cost and the burden it imposes on patients, particularly in underserved populations. Our non-invasive, digital solution continuously measures clinically significant variations in biological-sample biomarkers using wearable data combined with our proprietary algorithms. Starting with blood NT-proBNP concentration (a leading indicator of cardiovascular risk), we have built a unique, proprietary dataset of longitudinally paired digital and blood biomarkers to develop a validated digital NT-proBNP prediction algorithm. Our algorithm predicts heart failure decompensation at a 75% sensitivity and a 99.5% specificity, matching the performance of expensive invasive, but highly regarded, predictions of implanted systems like CardioMEMS and HeartLogic. Our platform is biological sample-type agnostic and we are expanding to include digital biomarkers for side effects of oncology medications and a variety of cardiorenal and pulmonary conditions, including CKD, hyperkalemia, and PAH.