The number of traced case children who had been immunized with DPT within 7 days prior to the acute illness was quite small (n = 18). Thus, the authors were not able to analyze whether long-term outcomes were different depending upon the initial diagnosis. Neither could they determine whether case children who had received DPT prior to the acute illness had experienced long-term outcomes of a different naturethan the case children who had notreceived DPT prior to the acute illness.
DISCUSSION
The NCES is the most ambitious study of chronic nervous system dysfunction in children to date. Detection of rare events poses particular methodologic problems. The case-control method used in the NCES is generally considered to be an appropriate design. However, the interpretation of studies of very rare events, even a study as large as the NCES, often rests on a small number of cases. Reclassification of even a few cases can sometimes change the relative risk estimate. This might encourage some critics to re-classify individual cases, which effectively turns a controlled epidemiologic study into a case series. This is not considered sound epidemiologic practice, and although the committee believes valuable clinical information can be determined from investigating in full all cases in such a study, the committee discourages such post-hoc changes in study protocol. The implications these changes would have for the NCES have been addressed by many, including the authors.
The new data from the NCES demonstrate that children who experience a serious acute neurologic illness (as defined—however problematically —by the NCES) are at risk for chronic nervous system dysfunction (in the broad areas of neurologic, motor, sensory, behavioral, educational, or self-care dysfunction) as defined in the 10-year follow-up report of the NCES (Madge et al., 1993; Miller et al., 1993) or death. Children who experience a serious acute neurologic illness within 7 days after receiving DPT might sustain these same outcomes and appear to be indistinguishable from children who had similar serious acute neurologic illness not in temporal relation to receiving DPT.
The nonspecific nature of the neurologic illnesses associated with DPT exposure concerned the committee. One would be more certain of a causal relation between DPT and neurologic illness if there were a well-identified “post-DPT” syndrome. Instead, this nonspecificity suggested to the committee that DPT might somehow “trigger” acute neurologic illness in children with underlying brain or metabolic abnormalities, that is, children who might have experienced subsequent chronic nervous system dysfunction regardless of vaccine administration. Concerns about the possible influence of underlying brain abnormalities on the occurrence of neurologic illness are not unfounded. For example, there are data suggesting that children who are neurologically