Today, soft errors are one of the major design technology challenges in implantable medical devices (IMDs). IMDs such as pacemakers and implantable cardioverter defibrillators, relying on integrated-circuit technology, are susceptible to soft errors from energetic particles (neutron and alpha particles). Device failures by soft errors can have deadly consequences for patients. Moreover, as the impact of technology scaling on soft errors is increasing, the efficiency of error correction codes (ECCs) is decreasing. Accordingly, there is an urgent demand for new solutions that take into account that ensuring data reliability in IMDs is of the utmost priority. This paper presents a technique that provides high error-correction performance, high speed, and low complexity. The proposed technique uses the replication method and the extended Hamming code and guarantees data reliability in spite of the presence of up to three-bit errors. The extended Hamming code is modified in order to provide lower decoding complexity. The new technique and the associated analysis scheme for its implementation has been evaluated and is described. The technique is compared to ECCs with similar decoding complexity to better understand the overheads required and the new capabilities to protect data against three-bit errors. It is shown that this technique reduces the miss-detection probability of four-bit errors. This proposed technique will hopefully contribute to the persistent efforts to implement reliable IMDs.