The present invention provides diverse methods and apparatus for in vivo monitoring, detecting and/or predicting potential failure modes or deleterious trends of chronically implanted medical electrical leads prior to actual failure of said leads. Certain embodiments of the invention involve applying a relatively increased data sampling rate at various time intervals (e.g., periodically, randomly, and/or manually-triggered and the like) prior to actual detection of a deleterious trend, to thereby increase the probability of detecting one or more parameters indicative of a potential lead performance issue. At least some of the parameters are utilized because they are not typically reliably detected at relatively lower data sampling rates. In addition to an initial relatively increased data sampling rate, certain embodiments provide for adjusting the sampling rate and storing and/or adjusting data pattern template-based data to execute pattern related triggers so that additional information regarding a medical electrical lead can be obtained.
