6When the Device Fails: Hard and Soft Failure
Sometimes the implant simply stops, and sometimes it works on the bench yet fails the patient. This module draws the line between hard and soft failure, explains the integrity test and why a normal result does not exonerate the device, and follows the cumulative-survival reporting and consensus statements that let centres and families weigh the decision to reimplant.
TTwo kinds of failure
Hard failure is an objectively confirmed loss of device function - no telemetry, a failed integrity test, or a device shown to be out of specification - and is unambiguous grounds for reimplantation. Soft failure is a working, communicating device with an unexplained decline in performance, aversive symptoms (pain, shocks, non-auditory stimulation) or intermittency, that cannot be proven faulty by current in-vivo tests - and that resolves after reimplantation. Soft failure is a diagnosis of exclusion: programming problems, electrode issues, medical causes and central factors must first be ruled out before the internal device is blamed. In revision series, soft failures tend to present earlier after implantation but take longer to reach revision, reflecting the difficulty of proving them.[2005][2013]
CThe integrity test and its limits
The integrity test (manufacturer-specific) interrogates the internal receiver-stimulator and electrodes - telemetry, impedances, and the device's response to a known input - to confirm whether the hardware meets specification. A failed integrity test confirms hard failure; the difficulty is the false-negative - a normal integrity test does not exclude a true device fault and is not reliable for predicting whether revision will help. Reported false-negative rates are substantial (around 60% in one soft-failure cohort), which is exactly why the soft-failure construct exists. Because in-vivo testing is imperfect, the explanted device should be returned to the manufacturer for failure analysis to close the loop and feed reliability statistics.[2005][2010]
CCumulative survival and standardised reporting
Manufacturers report device reliability as a cumulative survival percentage (CSP) - the proportion of a given device model still functioning at a stated time after implantation, plotted as a survival curve (method adapted from the cardiac-pacemaker ISO 5841-2 standard). An International Consensus Group standardised definitions of device failure, survival time, clinical benefit and the categorisation of explants so that CSP figures are comparable across manufacturers. Modern receiver-stimulators are highly reliable, with CSP for current models typically in the high-90s percent at several years - but historical recalls and lot-specific failures shaped today's reporting discipline. Standardised reporting matters clinically: it lets a centre quote honest, comparable failure rates to families and detect early signals of a problem device.[2010][2007]
CCounselling and the decision to reimplant
Hard failure -> reimplant: outcomes after reimplantation are generally good, often matching or exceeding the original device, because the cochlea and array tract are already established. Suspected soft failure is harder: reimplantation usually helps when the device truly was the cause, but immediate auditory recovery can be poorer and a small group does not improve - so realistic expectations are essential. The decision weighs symptom severity (aversive shocks and pain push strongly toward revision), the trajectory of decline, exclusion of non-device causes, and the patient's/family's tolerance of a second operation. Counselling should cover the surgical risk of revision (re-exposure, the same intraoperative complications), the chance of incomplete benefit in soft failure, and the value of returning the device for analysis.[2013][2005]
How is this best classified, and what is the implication?
What distinguishes a hard failure from a soft failure?
What does the cumulative survival percentage (CSP) describe?
Why was an international consensus group convened around cochlear implant reliability?