to facilitate growth in children with severe scoliosis and other conditions that limit chest development. In certain orthopedic repairs, surgeons may remove implanted fixation devices to accommodate children’s growth when they might leave such devices undisturbed in mature patients. For yet other implanted devices such as neurostimulators placed in the brain or bone cements used to repair bone defects, concerns about growth and development lead manufacturers and FDA to advise against use of the device with children who have not completed all or most of their growth in areas (e.g., brain, skeleton) where the device is placed.
In addition, pediatric issues or problems may be revealed during clinical testing of a device with children prior to marketing approval. For example, as described earlier in this chapter, studies of deep brain neurostimulators to treat dystonia revealed that use of the device in children, compared to adults, required adaptations in the placement of the device. If two neurostimulators are implanted, they must be implanted at least 8 inches apart to minimize interference.
Problems may be recognized as experience with a device accumulates following its entry into the market. In some cases, problems are identified or confirmed through systematic postmarket clinical or epidemiological studies. For example, in 2002, a manufacturer of cochlear implants reported to FDA 15 cases of meningitis in implanted patients. Subsequently, other manufacturers reported meningitis cases, mostly in young children. Some clinicians had already become concerned about the risk based on conversations at meetings about their experiences following their patients (Niparko, 2004). Based on a review of the adverse event reports, FDA worked with CDC and health departments in many states and three cities on an epidemiologic study that attempted to assess risk factors for meningitis among implant recipients compared to a control group (Reefhuis et al., 2003).
Some problems with devices may be identified soon after they begin to be used with children. To cite an example, the measurement of transcutaneous partial oxygen pressure in arterial blood (measured with an oxygen electrode) was thought to provide a reliable proxy for arterial blood oxygenation (Huch et al., 1977). Soon, however, clinicians recognized that there was a marked disparity between the two values when an infant’s perfusion (blood flow into tissues) is poor (Peabody et al., 1978a,b). As this problem was becoming more widely recognized, the pulse oximeter was developed to measure arterial blood oxygenation using a different and superior strategy (Jennis and Peabody, 1987).
Other device problems are uncovered only as clinicians follow patients for extended periods. One example involves problems with fracturing of the peritoneal catheters for cerebrospinal fluid shunts (Langmoen et al., 1992; Cuka and Hellbusch, 1995). The experiences cited above with the