Appendix I The Number of Drug-Exposed Infants May Be Seriously Underestimated The Number of Drug- The identification of infants who have been prenatally exposed to drugs is key to understanding the magnitude of the problem and providing effective medical and social interventions for these infants. However, there is no consensus on the number of drug-exposed infants born in the United States each year. A comprehensive nationwide study to specifically determine the incidence of drug-exposed births has not been done. Additionally, hospitals' procedures allow many drug-exposed infants to go undetected. Based on data from the National Center for Health Statistics' National Hospital Discharge Survey, which includes a representative sample of all births, an estimated 9,202 drug-exposed infants were born in 1986 in the United States.' By 1988, the latest year that data were available, the number had grown to 13,765 infants.2 However, this is likely to be a substantial undercount of the problem. At present, physicians and hospitals do not routinely screen and test all women and their infants for drugs. Recent studies have found that when screening and testing are uniformly applied, a much higher number of drug-exposed infants is identified. One study found that hospitals that assess every pregnant woman or newborn infant through a medical history and urine toxicology had an incidence rate that was three to five times greater than hospitals that relied on less rigorous methods of detection.3 The average incidence of drug-exposed infants born at hospitals with rigorous detection procedures was close to 16 percent of all births as compared with 3 percent of births at hospitals with no substance-abuse assessment. Likewise, our work indicates that the National Hospital Discharge 'The estimate ranged from 7,178 to 11,226 at a 95-percent confidence interval. 2The estimate ranged from 8,259 to 19,271 at a 95-percent confidence interval. 3Ira J. Chasnoff, "Drug Use and Women: Establishing a Standard of Care," Prenatal Use of Licit and Illicit Drugs, ed. Donald E. Hutchings. New York: New York Academy of Sciences, 1989. 4Appendix V provides more detailed information on the degree of drug-exposed infants identified at the 10 hospitals. Hospitals Lack We also found that the wide range in the number of drug-exposed infants we identified at the different hospitals in our review may be associated with the effort taken by hospitals to identify drug-exposed infants. For example, one of the 10 hospitals did not have a protocol for assessing drug use during pregnancy. This hospital had the lowest incidence of drug-exposed infants. Protocols at the remaining 9 hospitals did not require systematic screening and testing of every mother and infant for potential substance use or exposure. Instead, the protocols primarily required testing if the mother reported her drug use or if drug withdrawal signs became manifest in the infant. Hospital officials acknowledge that these screening criteria allow many drug-exposed infants to remain unidentified in the hospital. For example, women often deny using drugs because they do not want to be The Number of Drug-Exposed Infants May Be reported to the authorities for fear of being incarcerated or having their children taken from them. In addition, many cocaine-exposed infants display few overt drug withdrawal signs. Some will show no signs of drug withdrawal, while for others withdrawal signs may be mild or will not appear until several days after hospital discharge. The visual signs of drug exposure vary from severe symptoms to milder symptoms of irritability and restlessness, poor feeding, and crying. Since these milder symptoms are nonspecific, maternal drug use may not be suspected unless urine testing is conducted. Even when hospitals do conduct urinalysis, drug use may go undetected if drug concentrations within the body are too low. Urinalysis can only detect drugs used within the past 24 to 72 hours. According to recent studies, hair analysis and meconium analysis, two testing methods for detecting drug use, have advantages over urinalysis because they are more accurate or can detect drug use over a longer period of time after drug use has occurred.5,6,7 One of the studies, conducted at a large urban hospital in Detroit accounting for over 7,000 births annually, used meconium analysis to detect drug use during pregnancy. Preliminary results revealed that 42 percent of infants were found to be drugexposed in 1989.9 However, the hospitals in our review that conducted testing for drug exposure relied exclusively on urinalysis. 8 When an infant does not show signs of drug withdrawal or the mother does not self-report drug use, a physician may consider other factors as presumptive of drug exposure during pregnancy and recommend that drug testing be conducted. Such factors or characteristics have been found to occur more often among drug-exposed infants than infants not exposed to drugs and include (1) inadequate prenatal care (defined as four or fewer prenatal care visits for a pregnancy of 34 or more 5 Meconium is the first 2- to 3-days' stool of a newborn infant. 6 Karen Graham and others, "Determination of Gestational Cocaine Exposure by Hair Analysis," Journal of the American Medical Association, Vol. 262 (Dec. 15, 1989), pp. 3328-30. 7Enrique M. Ostrea, Jr., A Prospective Study of the Prevalence of Drug Abuse Among Pregnant Women. Its Impact on Perinatal Morbidity and Mortality and on the Infant Mortality Rate in Detroit. (July 13, 1989, preliminary report.) Ostrea, A Prospective Study of the Prevalence of Drug Abuse Among Pregnant Women. "The 42 percent of births identified as drug exposed using meconium testing compares with 8 percent identified based on the mother's self-reporting drug use. The Number of Drug-Exposed Infants May Be weeks),10 (2) low birth weight (defined as less than 5.5 pounds), and (3) low gestational age or prematurity (defined as less than 38 weeks),11,12 (See table I.2.) We were able to obtain data from 9 of the 10 hospitals in our review on the degree to which infants had these characteristics. We identified an estimated 4,391 infants with two or more characteristics of possible drug exposure. The last column of table I.2 shows the number of infants with two or more drug-exposure indicators who were not tested for drug exposure at the 9 hospitals where we obtained data. We estimate that at these hospitals during 1989, there were 2,791 potentially drug-exposed infants who were not tested, based on our review of hospital medical records. 10Institute of Medicine, Infant Death: An Analysis by Maternal Risk and Health Care. Contrasts in Health Status, ed. D.M. Kessner, Vol. 1 (Washington, D.C.: National Academy of Sciences, 1973), pp. 58-59. 11Gestational age refers to the period of time, normally 40 weeks, from conception to an infant's birth. 12 Maternal demographic characteristics and socioeconomic status effect birth outcomes. Infant mortality and low birth weight rates are higher for young, uneducated, unmarried, non-white women with limited financial resources. We also found that some hospitals where we identified low percentages b |