Alfred Steinschneider, MD, PhD, Aaron Curns, MPH, Cheryl Richmond, MD, MPH
American SIDS Institute, 2480 Windy Hill Road, Suite 380, Marietta, GA 30067 USA
Phone:770-612-1030. Fax: 770-612-8277. Email: prevent@SIDS.org

A retrospective study was conducted to evaluate the combination of epidemiologic and physiologic variables in defining an infant's risk for pathologic central apnea. Consistent with the NIH Consensus Development Committee's recommendation, pathologic apnea was defined objectively as apnea ³ 20-s or ³ 18-s if associated with bradycardia. Data were obtained from two groups of infants (N=828) followed on documented apnea/bradycardia monitors for two months. The ALTE/Asymptomatic group included ALTEs, those with a family history of SIDS, and infants with anxious parents. The Premie Apnea group consisted of premature infants with a history of apnea or bradycardia episodes. Physiologic data were collected during a feeding and nap in an environmentally controlled (90^o F) laboratory. The physiologic measures included a composite measure reflecting the frequency and duration of apnea pauses ³ 2-s (PSA₄), the relative percentage of apnea pauses ³ 6-s (A₆/D%), and the relative amount of respiratory difficulty during feeding (AO/D%). Using the laboratory data as well as epidemiologic data (concerning the mother, her pregnancy, and the infant) a separate logistic model was created for each group containing variables significant at p£ 0.10. Scores were calculated from the logistic regressions and cut-points determined to minimize the false negative rate; resulting in a statistically significant association (p<0.001) between the categorized logistic scores and the incidence of pathologic apnea for both groups. The risk of pathologic apnea for infants in the ALTE/Asymptomatic group with scores below the cut-point was 0.4% compared to 6.4% for those with higher scores (sensitivity = 91.7%, specificity = 58.2%, test(+) = 43.2%). No Premie Apnea infant with a score below the cut-point had an episode of pathologic apnea, whereas 23.0% of infants with higher scores did (sensitivity=100%, specificity = 47.8%, test + = 58.6%). The potential use of this objective approach in clinically managing high risk infants will be discussed.