PEDIATRICS Vol. 107 No. 2 February 2001, pp. 309-317

The Maternal Lifestyle Study: Drug Use by Meconium Toxicology and Maternal Self-Report

Barry M. Lester, PhD^*, Mahmoud ElSohly, Linda L. Wright, MD^§, Vincent L. Smeriglio, PhD, Joel Verter, PhD^¶, Charles R. Bauer, MD^#, Seetha Shankaran, MD^**, Henrietta S. Bada, MD, H. Chip Walls, BA^§§, Marilyn A. Huestis, PhD^||, Loretta P. Finnegan, MD^¶¶, and Penelope L. Maza, PhD^##

From ^* Brown Medical School, Women and Infants' Hospital and Bradley Hospital, Providence, Rhode Island;  ElSohly Laboratories, Inc, Oxford, Mississippi; ^§ National Institute of Child Health and Human Development, Bethesda, Maryland;  National Institute on Drug Abuse, Bethesda, Maryland; ^¶ George Washington University, Biostatistics Center, Rockville, Maryland; ^# University of Miami School of Medicine, Miami, Florida; ^** Wayne State University School of Medicine, Detroit, Michigan;  University of Tennessee Memphis, College of Medicine, Memphis, Tennessee; ^§§ Forensic Toxicology Laboratory, University of Miami, Miami, Florida; ^|| National Institute on Drug Abuse, Baltimore, Maryland; ^¶¶ Center for Substance Abuse Treatment, Rockville, Maryland; and ^## Administration on Children, Youth and Families, Washington, DC.

	ABSTRACT

Top Abstract Methods Results Discussion Conclusion References

Objective.  The objective of this study was to describe drug use by pregnant women participating in the 4-site Maternal Lifestyle Study of in utero cocaine and/or opiate exposure.

Methods.  Meconium specimens of 8527 newborns were analyzed by immunoassay with GC/MS confirmation for metabolites of cocaine, opiates, cannabinoids, amphetamines, and phencyclidine. Maternal self-report of drug use was determined by hospital interview.

Results.  The prevalence of cocaine/opiate exposure in the 4 sites was 10.7% with the majority (9.5%) exposed to cocaine based on the combination of meconium analysis and maternal self-report. However, exposure status varied by site and was higher in low birth weight infants (18.6% for very low birth weight and 21.1% for low birth weight). Gas chromatography/mass spectrometry (GC/MS) confirmation of presumptive positive cocaine screens was 75.5%. In the cocaine/opiate-exposed group, 38% were cases in which the mother denied use but the meconium was positive. There was 66% agreement between positive meconium results and positive maternal report. Only 2% of mothers reported that they used only cocaine during pregnancy and mothers were 49 times more likely to use another drug if they used cocaine.

Conclusion.  Accurate identification of prenatal drug exposure is improved with GC/MS confirmation and when the meconium assay is coupled with a maternal hospital interview. However, the use of GC/MS may have different implications for research than for public policy. We caution against the use of quantitative analysis of drugs in meconium to estimate the degree of exposure. Our study also highlights the polydrug nature of what used to be thought of as a cocaine problem.  Key words:  cocaine, opiates, polydrug use, pregnancy substance abuse, meconium, self-report, multisite, gas chromatography/mass spectrometry, low birth weight, prenatal drug exposure.

The accurate identification of prenatal cocaine exposure is critical for 2 reasons: 1) to fully understand the nature and magnitude of the problem, and 2) to determine appropriate medical and psychosocial intervention. Methods used to identify prenatal cocaine exposure vary and include interview, self-administered questionnaire, intake history, urine testing of mother and infant, and testing of infant hair and meconium, resulting in widely disparate estimates of rates of cocaine use by pregnant women. Maternal self-report of drug use is problematic because of the fear of the consequences of admitting to the use of an illegal substance, and the inaccuracy of recall, especially when details such as when, how often, and how much are asked. Underreporting of drug use by pregnant women has been documented in several studies.^1-3 In a sample of over 3000, in which 43% were positive for illegal substances, Ostrea et al¹ found that only 11% of the mothers admitted to illicit drug use. Frank et al³ found that self-report misclassified 24% of cocaine users identified by urine toxicology.

Biomarkers of in utero cocaine exposure in the neonate include measurement of the drug in blood, urine, hair, and meconium and, recently, in gastric aspirate and amniotic fluid.¹³ Urine has been the most widely used specimen; however, Ostrea et al^4,5 reported meconium to be more suitable than urine for detecting fetal exposure to cocaine. Urine detection underestimates exposure because cocaine metabolites are measurable in urine for no more than 96 to 120 hours after the last cocaine use in contrast to meconium, which can detect cocaine use throughout the second half of pregnancy. In addition, the concentration of drugs may be higher in meconium than in urine.⁵ A number of studies suggest that meconium is the most useful specimen for the detection of prenatal cocaine exposure.^1,6,7-11 Lewis et al¹² found that meconium detected 2.7 times more instances of cocaine use than urine. However, these studies have been criticized by Casanova et al¹³ who argue that the methods used for the meconium measurement differed in analytic sensitivity and specificity from the urine screening methods, precluding a direct comparison of the performance of urine versus meconium as matrices for the detection of cocaine. In addition to the choice of specimen, the detection of cocaine exposure is influenced by the choice of initial screening test and confirmation procedure. In several studies, urine toxicology was used to corroborate self-report.^14-16 The importance of using both self-report and a biomarker was illustrated by Frank et al.³ Although self-report misclassified 24% of cocaine users identified by urine toxicology, 51% of those who admitted use were negative for cocaine in the urine.

Estimates of cocaine exposure also vary depending on the population studied. The 1988 National Maternal and Infant Health Survey,¹⁷ a mail survey, reported cocaine use during pregnancy by .4% of women. The more recent National Pregnancy and Health Survey¹⁸ based on questionnaire responses estimated cocaine use by pregnant women at 1.1%. However, the National Pregnancy and Health Survey also included urine test results from a subsample and found significant underreporting of cocaine use. Higher prevalence rates of cocaine use during pregnancy were reported in 2 high-risk population hospital surveys. In a 36-hospital review of discharge records, 10% to 17% of the women used cocaine during pregnancy.¹⁹ The US Government Accounting Office²⁰ reviewed medical records in 10 hospitals and reported prenatal cocaine exposure ranging from .3% to 11.6%. Prevalence rates of cocaine use during pregnancy in 5 state studies^21-25 ranged from .8%²² to 2.6%.²³ Studies conducted at the local level reveal a broader range: .4% in 26 hospitals in 1 county in California,²⁶ 3.4% in 1 county in Florida,²⁷ and 18.4% to 31% in single hospitals.^1,28 Other factors that may account for the variability in estimates of cocaine use by pregnant women are whether the sample is from a clinic,²⁷ inner-city,³ or private hospital,²³ or whether samples were highly selected to include women in drug treatment,²⁹ or whether only women enrolled in prenatal care³ were included.

We report the results of drug use by pregnant women from a study conducted in 4 sites that varied in racial and ethnic characteristics, social class, geographic region, and that included public and private hospitals. We used self-report and a meconium assay with confirmation to determine exposure status to cocaine and opiates, and we included the use of other licit and illicit drugs. We also studied low birth weight (LBW) infants and report our results by birth weight and by study site. Our intent is not to provide prevalence estimates for public health or public policy purposes but to provide a description of prenatal drug exposure in the sites and to discuss some of the issues in the detection of prenatal cocaine exposure.

	METHODS

Top Abstract Methods Results Discussion Conclusion References

The Maternal Lifestyle Study is a prospective multisite longitudinal study on the association of drug use during pregnancy with acute neonatal events and long-term neurodevelopmental child outcome conducted under the auspices of the National Institute of Child Health and Human Development (NICHD) Neonatal Research Network. The NICHD developed a fiscal and scientific collaboration with the National Institute on Drug Abuse, the Administration on Children, Youth, and Families, and the Center for Substance Abuse Treatment to design and fund the study. Four clinical centers were selected: Brown University, the University of Miami, the University of Tennessee at Memphis, and Wayne State University. George Washington University served as the Biostatistical Coordinating Center. The study was approved by the institutional review board at each center and recruitment occurred between May 1993 and May 1995. In addition, a National Institute on Drug Abuse Certificate of Confidentiality was obtained by each center that ensured confidentiality of information regarding the participant's drug use. The certificate superseded the mandatory reporting of illegal substance use that was in effect in the Florida and Rhode Island centers. The certificate was explained to the mother during the recruitment and informed consent procedure including the fact that the certificate did not exclude reporting of evidence of child abuse or neglect. Testing of the infant's meconium for drugs was included as part of the informed consent.

Mothers were recruited shortly after delivery and recruitment strategies varied among the 4 centers to maximize enrollment. The recruitment strategy was designed to identify and recruit all women delivering very low birth weight (VLBW) infants of 501 to 1500 g. For mothers delivering LBW (1501-2500 g) and normal birth weight (NBW; >2500 g) infants, recruitment varied by center. Maternal exclusion criteria included age <18 years, identified psychosis, history of institutionalization for retardation or emotional problems, or language barriers that prevented her from understanding the study. Infant exclusion criteria included outborn birth, multiple gestation, birth weight <501 g, gestational age >42 weeks based on obstetrical estimate, or if in the judgment of the attending physician, the infant was unlikely to survive.

After informed consent, a maternal interview was conducted in hospital to determine past and current drug use and sociodemographic information. The interview was structured, based on a written script, and conducted by social workers or nurses who were trained and certified to administer the interview. The interview was conducted in privacy after the mother had delivered and was comfortable. A physical examination of the infant was conducted and the infant's meconium was collected and refrigerated. Specimens were collected from more than 1 diaper and pooled if the quantity of meconium was small, as often occurred in the preterm infants. Before discharge, maternal and infant charts were abstracted to collect selected medical data. Meconium samples were processed, refrigerated, batched, and express shipped to a central laboratory (ElSohly Laboratories, Oxford, MS), where they were frozen at 20°C for analysis of metabolites of illicit drugs. The assay consisted of an initial enzyme-multiplied immunoassay technique (EMIT) screen for cocaine, opiates, cannabinoids (THC), amphetamines, and phencyclidine (PCP) with cutoff concentrations of 20 ng/g for THC and PCP, and 200 ng/g for the cocaine, opiate, and amphetamine metabolites. Presumptive positive screens were confirmed with gas chromatography/mass spectrometry (GC/MS) with cutoffs at the limit of detection of the assays. These were 50 ng/g for the cocaine, opiate, and amphetamine metabolites, 5 ng/g for THC-COOH, and 3 ng/g for PCP. Details of the assay method are reported by ElSohly et al.³⁰ The cocaine metabolites included in the GC/MS confirmation were cocaine, benzoylecgonine (BE), cocaethylene (CE), and m-hydroxybenzoylecgonine (HBE). HBE was added midway through the screening phase. The opiate metabolites included morphine, codeine, 6-monoacetylmorphine (6MAM), hydromorphone (HYM), and hydrocodone (HYC). THC, amphetamine, and methamphetamine and PCP were also confirmed by GC/MS, although amphetamines and PCP analysis were discontinued because of low prevalence.

A mother-infant dyad was defined as exposed if either the mother admitted using cocaine or opiates during the pregnancy hospital interview or if the infant's meconium was positive for one of the indicated metabolites of cocaine or opiates in the EMIT screen and by GC/MS confirmation. A dyad was categorized as nonexposed if the mother denied using both cocaine and opiates during the pregnancy on the hospital interview and the infant's EMIT meconium screen was negative for both cocaine and opiates. Maternal use of alcohol, marijuana, and nicotine during the pregnancy were treated as background variables in both the exposed and unexposed groups.

	RESULTS

Top Abstract Methods Results Discussion Conclusion References

During the 2-year recruitment period, 19 079 mothers were screened in the 4 centers. Of these, 16 988 met the eligibility criteria, and 11 811 (70%) consented to participate in the study. The 5177 eligible mothers who were not included are those who refused consent, those not available, and those with language barriers. Meconium was collected and centrally assayed on 8805 of the infants. Of the 8805 meconium samples collected 277 (3.1%) had quantity not sufficient (QNS) to perform the EMIT screen. QNS varied for the 3 birth weight groups from 7.1% for the VLBW group to 2.7% for the NBW group. Birth weight was missing for 1 infant, resulting in a final sample of 8527 specimens on which all analyses are reported. The demographic characteristics of the 8527 mothers in the final sample and the 3284 mothers who had consented but meconium was not collected or QNS are shown in Table 1. The demographic characteristics were similar between these 2 groups, suggesting that the sample in which the meconium was collected was representative of the sample of mothers who consented to participate in the study.

Screening Results

The results of the EMIT meconium screen by clinical center and birth weight group are summarized in Table 2. The overall percentage of presumptive positive cocaine screens by EMIT was 9.5%. However, this varied considerably between centers and birth weight groups. For VLBW infants, the EMIT screen was positive for cocaine in 18.2% of infants at Wayne State, but only 4.8% at Miami. A similar variation was seen for each of the 2 other birth weight groups: 26.6% to 12.0% in the LBW group and 10.0% to 4.3% for the NBW group.

Although the prevalence of opiate use was less (2.3%), we still observed center variation (7.2% at Brown to 1.6% at Tennessee for LBW infants). The results for the cannabinoid screen also showed both center (Wayne State was the highest and Miami was the lowest) and birth weight variation (VLBW infants showed the lowest positive rate except at Brown). The overall EMIT estimated prevalence of THC use was 7.2%; however, only 36.5% of these were confirmed positive by GC/MS. This low confirmation rate resulted in too small a sample for proper evaluation and the THC screen was dropped midway through the recruitment period. The EMIT-estimated prevalence of amphetamine was .8% and no positive screens were confirmed. Three percent of the meconium specimens screened positive for PCP but only 4% of these were confirmed by GC/MS.

GC/MS Results

The meconium samples for 809 infants screened positive for cocaine use. Of these, 22 (2.8%) were QNS for GC/MS confirmation. The data in Table 3 summarize the results of the GC/MS testing on the remaining 787 samples. Of these, 556 (70.6%) were positive for at least 1 of the 3 metabolites: BE, cocaine, and CE. In addition, after ~50% of the screens had been completed we added a fourth metabolite (HBE) to the confirmation process. For the 382 infants where confirmation was possible for all 4 metabolites, 287 (75.1%) were confirmed for BE, cocaine, or CE. An additional 38 (10%) were confirmed only for HBE. Thus, in the subset in which all 4 metabolites were assessed, 325 of 382 or 85.1% were confirmed by GC/MS.

As in the screening results, we report both a center and birth weight variation in the type(s) of metabolites that were confirmed. For all birth weight groups, BE and later, HBE was more likely to be the confirming metabolite. CE was least likely (<10%) to be identified in all birth weight groups. In each of the birth weight groups, the use of BE alone would have identified 86.5% (VLBW), 98.7% (LBW), and 98.6% (NBW) of all samples confirmed positive for cocaine exposure.

Of the 200 positive screens for opiates, 18 (9.0%) were QNS for GC/MS confirmation. Overall, 70% of the 182 presumptive positive samples were confirmed positive by GC/MS. Morphine accounted for 46% to 62% of the birth weight-specific confirmations, while codeine was confirmed in 43% to 47% of the cases. The 6MAM, HYM, and HYC metabolites were identified in only 5 of the cases (2.7%). The numbers of cases at the specific sites and birth weight groups were too small to draw inferences about variability in drug exposure.

The variation in cocaine exposure by birth weight based on the meconium results is shown in Table 4. Dividing the infants into 500-g birth weight groups shows that the rate of cocaine exposure increases through 2500 g both for the positive screens and for those with positive GC/MS confirmation.

Maternal Report

Table 5 summarizes the results of the in-hospital maternal interview reported use of cocaine and/or opiates during the pregnancy. Of the 8805 mothers for whom there was an infant's meconium available, 619 (7.3%) reported using cocaine and 100 (1.2%) opiates. As noted above for the meconium results, the mothers' report of use also varied by center and birth weight. Mothers at Wayne State reported the highest use of cocaine for VLBW and LBW infants. Mothers at Wayne State reported the highest use of opiates for all VLBW and NBW infants, although the prevalence of opiate use was considerably less than that for cocaine. Opiate use was highest at Brown for LBW infants and was virtually absent at Miami and Tennessee.

Toxicology and Maternal Report

For this study, an algorithm that used the results of both the mothers in-hospital interview and the meconium assays defined exposure status. Table 6 shows the prevalence of exposure status using both toxicology and maternal report. Overall, 915 of the dyads (10.7%) were classified as exposed to cocaine and/or opiates. The majority (809) were exposed to cocaine (9.5%), with or without exposure to opiates. Again, note that the dyads at Wayne State showed the highest exposure prevalence for all 3 birth weight groups. By comparing Tables 5 and 6, we can determine, for each site and birth weight group, the number of additional cocaine or opiate dyads identified by the addition of the meconium assay to the screening process. For example, for VLBW infants, 8 additional infants were identified at Wayne State, 6 at Tennessee, 3 at Miami, and 6 at Brown. In total, the meconium assay resulted in a 34.3% increase in the number of VLBW exposed dyads for the study.

In summarizing Tables 3, 5, and 6, we find that for the 3 birth weight groups, maternal self-report provided an estimated prevalence of exposure of 13.9% (VLBW), 16.4% (LBW), and 5.3% (NBW). Results from the meconium analysis yielded prevalence rates of 9.2% (VLBW), 16.7% (LBW), and 5.6% (NBW). Using both the self-report data and the meconium results, we find that the prevalence rates are 18.6% for VLBW infants, 21.1% for LBW infants, and 7.8% for NBW infants. The overall prevalence rate of exposure in this cohort is 10.7%.

Conditional Probabilities

To better understand the utility of GC/MS confirmation for cocaine in this study, conditional probabilities were computed comparing the GC/MS results with the EMIT screen and with maternal report from the hospital interview. The probability of GC/MS confirmation for cocaine given a positive EMIT screen was .755 (n = 787). The probability of GC/MS confirmation for cocaine given that the mother reported that she used cocaine and meconium was available was .662 (n = 458). The probability of GC/MS confirmation for cocaine given that the mother denied use of cocaine on the interview was .024 (n = 7893).

Polydrug Use

Polydrug use findings are shown in Table 7. The table shows the rates of marijuana, alcohol, and tobacco use from the maternal hospital interview for the 4 centers by birth weight group. Marijuana use was reported by 8% of the mothers, alcohol use by 35%, and tobacco use by 26%. Of the 619 mothers who reported using cocaine, only 16 (2%) reported use of no other drugs (opiates, marijuana, alcohol, or tobacco). One hundred eighteen mothers (19%) reported the use of cocaine and one other drug. Two hundred sixty-two mothers (42%) reported using cocaine and 2 other drugs, 72% of which were alcohol and tobacco. Two hundred eleven mothers (35%) reported using cocaine and 3 other drugs, 86% of which were marijuana, alcohol and tobacco. Only 3 of 100 opiate users (3%) did not use at least 1 of the other 4 drugs; 12 used 4 other drugs.

Odds ratios (ORs) and 95% confidence intervals (CIs) were computed comparing cocaine use with other drug use during the pregnancy. Mothers who used cocaine were more likely to use cigarettes (OR: 23.1; CI: 19.5-27.4), alcohol (OR: 6.2; CI: 5.2-7.3) and marijuana (OR: 14.2; CI: 12.2-16.5), than mothers who did not use cocaine. In addition, mothers who used cigarettes were more likely to use alcohol (OR: 3.3; CI: 3.0-3.6) and marijuana (OR: 10.6; CI: 9.1-12.3) than mothers who did not use cigarettes, and mothers who used alcohol were more likely to use marijuana (OR: 4.0; CI: 3.5-4.7) than mothers who did not use alcohol. The odds of using at least 1 of the other 4 drugs if the mother admitted to using cocaine was 37.7:1, while the odds for a mother who did not admit to using cocaine was .77:1. This results in an OR of 49, indicating that a mother who admits to using cocaine is 49 times more likely to also use opiates, alcohol, marijuana, or tobacco than a mother who did not admit to using cocaine.

	DISCUSSION

Top Abstract Methods Results Discussion Conclusion References

This is the first large-scale report of cocaine and other drug use during pregnancy using a meconium assay with confirmation and maternal history. In this sample of over 8500 dyads from 4 different sites, we found an overall prevalence of cocaine/opiate exposure of 10.7% with the majority (9.5%) exposed to cocaine, based on the combination of maternal self-report and meconium analysis. However, exposure status varied by site and birth weight; higher at Wayne State and Tennessee than at Miami or Brown and higher in the 2 lower birth weight groups at each site. The higher prevalence of cocaine or opiate use among women delivering prematurely has been previously reported.^29,31,32

The meconium assay that we used included GC/MS confirmation. Historically, immunochemical or thin-layer chromatographic screening methods were thought to be sufficient for clinical and emergency toxicology use, whereas forensic applications require confirmation of positive screening results by GC/MS.³³ However, in clinical settings, the accurate detection of prenatal drug exposure may be compromised when presumptive positive screens are not confirmed. Moore et al³⁴ found up to a 43% false-positive rate for cocaine when drug screens were used without confirmation. In our study, 75.5% of the positive screens were confirmed. Therefore, had we relied on screening alone, we could have incorrectly included an additional 25% of mothers in the cocaine group. In our study, mothers who screened positive but were not confirmed with GC/MS were not included in either group. Therefore, we probably have a more conservative estimate of cocaine exposure than some other studies. Other studies may have included positive screens that are not confirmed in the control group. Lack of confirmation could also help explain some of the variability across studies in prevalence rates. For example, Ostrea et al¹ reported 30.7% of infants positive for cocaine in a large-scale epidemiologic study based on a meconium screen without confirmation and a fourfold increase when the meconium screen was compared with self-report. It is difficult to evaluate these findings without knowing the false-positive rate of the meconium screen.

The addition of the HBE metabolite of cocaine, first reported by Steele et al⁸ was important for our study. In one study,¹² HBE was the only metabolite found in 23% of the meconium specimens. In our study, the addition of HBE resulted in a 10% increase in GC/MS confirmation to 85% when all 4 cocaine metabolites were used. Additional metabolites of cocaine may help reduce false-negative results and perhaps explain some of the variability in cocaine exposure by birth weight. For example, urine assays predominantly test for BE, the major urinary metabolite of cocaine. This method can only detect cocaine if it is present in high concentration. In a VLBW sample Browne et al¹⁰ found 7.5% positive for cocaine metabolites in the urine, compared with 19.8% in the meconium. However, BE was not detected in any of the meconium samples. We did find BE in our VLBW infants and it is interesting that BE identification was lowest for VLBW at Wayne State (Table 3), although BE alone would have identified the majority of exposed infants in all 3 birth weight groups. It is also interesting that the rate of cocaine exposure increased through 2500 g (Table 4). It is possible, as suggested by Browne et al,¹⁰ that the chemical breakdown and fetal metabolism of cocaine are a function of gestational age and birth weight.

Some investigators have advocated the use of quantitative analysis of drugs in meconium as a means of estimating the degree of drug exposure during gestation. Ostrea et al³⁵ reported that the concentrations of nicotine metabolites in meconium were directly related to the degree of active smoking by the mother. In a similar manner, Stolk et al³⁶ observed a correlation between the maternal methadone dose and the concentrations of methadone and EDDP, its primary metabolite, detected in meconium. It has even been suggested that serial analysis of meconium specimens may reflect both the chronology and degree of drug exposure. Ostrea et al³⁷ proposed that meconium collected 0 to 10 hours after birth reflects gestational exposure up to 20 weeks, 11 to 20 hours 21 to 30 weeks of gestation, and 21 to 36 hours >30 weeks of gestation. The analysis of drugs in meconium and interpretation of test results are complex processes. Overinterpretation of meconium data are a dangerous practice, especially at the current stage of development of this new and useful technology. It is clear that there are matrix effects associated with the analysis of meconium, as there are with each biological fluid or tissue. It is essential that standards and controls be prepared in meconium to prevent false-positive and false-negative test results especially when using a urine- or different matrix-based immunoassay. The importance of a second confirmation test that is based on a different analytical method to improve the accuracy of test results has been stressed, as well as the importance of the analytical cutoff or sensitivity of the method. However, there are other factors that contribute to the difficulty of interpreting quantitative meconium results.

Many questions remain to be answered about the disposition of drugs in meconium. It is not yet known which are the most appropriate drug analytes to measure in meconium. It is clear that measurement of HBE is important for cocaine analysis and ElSohly et al³⁸ have recently described the importance of 11-hydroxy-⁹-tetrahydrocannabinol and 8,11,dihydroxy-⁹-tetrahydrocannabinol for detection of marijuana exposure. Meconium from 11 neonates born to cocaine-abusing mothers were found to be positive for cocaine, benzoylecgonine, norcocaine, and/or benzoylnorecgonine.³⁹ Qualitative and quantitative differences in the compounds detected were observed in the different infants. If a correlation between the concentration of drug in meconium and maternal drug exposure is attempted, should it be based on the total concentration of all drug analytes, the parent drug, or a specific metabolite? The fetus may metabolize drugs differently at specific gestational ages. Could the presence of a particular analyte be a marker for drug exposure at a certain gestational age? How important is enzymatic or acid/base hydrolysis of the meconium sample to accurate and sensitive drug measurement? Fetal maturity also affects the efficiency of phase II metabolism and the production of glucuronide and other metabolites. False-negative results may occur if the immunoassay does not cross-react with the appropriate analytes or if the analyte is not free to react with the antibody, ie, if it is the glucuronidated metabolite. Moore et al⁴⁰ found that the concentration of hydrocodone, hydromorphone, and codeine increased markedly after acid hydrolysis with a more moderate increase noted in the morphine concentration. This is in contrast to a report by Becker et al⁴¹ in which hydrolysis of meconium was not found to significantly increase the morphine concentration. Also, fetal maturity was found to be a factor in the detection of methamphetamine in meconium from guinea pigs after intraperitoneal injection of the drug at different gestational ages.⁴² Furthermore, gestational age was determined to be an important consideration in interpreting quantitative methamphetamine results.

There also is uncertainty about possible reabsorption of drug from the fetal gastrointestinal track during gestation. Meconium is not a homogeneous specimen but a series of layers formed in the intestine during gestation. It is possible that drugs may diffuse through the meconium over time. Most analytical methods advise a thorough mixing of the specimen before analysis because the distribution of drugs throughout a sample may not be uniform. Another important confounding factor is the possible contamination of the meconium specimen by urine. Different analytes may be present in urine at different concentrations. Contamination of meconium by urine could complicate the interpretation of quantitative results. It is clear that additional research is necessary to address these important issues and to improve our understanding of the disposition of drugs in meconium.

Exposure status in this study was based on an algorithm that used the results of meconium assays and maternal self-report from an in-hospital interview. The advantage of using both drug toxicology and maternal self-report has been shown in several studies.^1,3,27,28 It is also important to distinguish between maternal report based on a structured questionnaire as we used and information collected about the mother from medical record review. The latter is less reliable¹⁹ and may not be appropriate to compare with toxicology results.¹ The importance of using both meconium and maternal self-report is to identify mothers who deny use but who did use as evidenced by positive meconium confirmation. It is generally assumed that mothers are less likely to say they used illegal drugs when they did not. In our study, 661 mothers admitted to cocaine/opiate use (619 to cocaine). An additional 254 mothers denied use but their infant had a positive meconium confirmation for cocaine/opiates (190 for cocaine). Therefore, we identified an additional 38% of the cocaine/opiate exposed infants (31% of the cocaine-exposed) based on the meconium assay alone. More importantly, had we relied only on maternal self-report, these 254 infants would have been eligible for inclusion in the unexposed group.

From a public health perspective one could argue that these 254 infants only represent 3.2% of the 7866 of the total number of mothers in our sample who denied cocaine/opiate use (190 of 7908 or 2.4% of mothers who denied cocaine use) and that the effort and expense involved in collecting and conducting the meconium assay (including GC/MS confirmation) may not be effective. In contrast, these 254 infants represent a substantial (28%) proportion of the sample of the 915 cocaine/opiate users (or 190 of 809), or 23% of the sample of cocaine users. Therefore, the use of the meconium assay in the forensic gold standard way in which it was used in this study may have different implications for research than for public policy. Confirmation of presumptive positives is essential for the accurate identification of women whose infants may be removed by State authorities. Also, in the research environment our use of the Certificate of Confidentiality could have influenced informed consent and maternal report of drug use, an option not available in the public health arena. Therefore, under public health guidelines the underreporting of maternal drug use could be substantially higher than that found in this study.

The probability that the mother reported that she used cocaine given a positive GC/MS confirmation for cocaine in this study was .66. Casanova et al¹³ also found a 66% agreement between positive toxicology results and positive maternal report. This is not surprising because the meconium assay, at best, can only provide a record of drug use during the second half of pregnancy. Infants of mothers who report that they used cocaine but not in the second half of pregnancy will have a negative meconium for appropriate reasons. Therefore, it would not be advisable to rely on the meconium assay exclusively and not include a structured interview.

We also addressed the issue of polydrug use in this study. It is becoming increasingly clear that women who use cocaine during pregnancy are also likely to use other licit and illicit drugs. In our database of published studies on cocaine and child outcome, only 13% of the studies reported that cocaine was the only drug used.⁴³ Chasnoff¹⁹ found that 32% of the positive meconium specimens were positive for at least 2 drugs. In our study, only 2% of the mothers reported that they used cocaine and no other drugs, and mothers were found to be 49 times more likely to use another drug if they used cocaine. The cocaine problem has truly become a problem of polydrug use. This has implications for drug toxicology to develop improved methods to identify drug-exposed infants, especially those exposed to multiple drugs. There are also implications for developmental follow-up studies because little is known about the effects of drug interactions on child behavior and development.

	CONCLUSION

Top Abstract Methods Results Discussion Conclusion References

In this large, prospective multisite study, we found that the prevalence and observed metabolites of cocaine showed considerable variation among the 4 sites and according to birth weight. Accurate identification of prenatal drug exposure is likely to be improved with GC/MS confirmation and when the meconium assay is used in conjunction with a maternal hospital interview. However, the use of GC/MS may have different implications for research than for public policy. Finally, the problem of cocaine use has been redefined as one of polydrug use, which has implications for the fields of drug toxicology, pediatrics, and child development.

	ACKNOWLEDGMENTS

This study was supported by the National Institute on Child Health and Human Development through Cooperative Agreements U10 HD 27904, U10 HD 21397, U10 HD 21385, U10 HD 27856, U10 HD 19897, NICHD Contract 5-U10-HD27904-02, and intraagency agreements with the National Institute on Drug Abuse, Administration on Children, Youth and Families, and the Center for Substance Abuse Treatment.

	FOOTNOTES

Received for publication Dec 10, 1999; accepted Jun 1, 2000.

Reprint requests to (B.M.L.) Infant Development Center, Women and Infants' Hospital, 101 Dudley St, Providence, RI 02905-2499. E-mail: barry_lester{at}brown.edu

	ABBREVIATIONS

LBW, low birth weight; NICHD, National Institute of Child Health and Human Development; VLBW, very low birth weight; NBW, normal birth weight; EMIT, enzyme-multiplied immunoassay technique; THC, cannabinoids; PCP, phencyclidine; GC/MS, gas chromatography/mass spectrometry; BE, benzoylecgonine; CE, cocaethylene; HBE, m-hydroxybenzoylecgonine; 6MAM, 6-monoacetylmorphine; HYM, hydromorphone; HYC, hydrocodone; QNS, quantity not sufficient; OR, odds ratio; CI, confidence interval.

	REFERENCES

Top Abstract Methods Results Discussion Conclusion References

1. Ostrea EM, Brady M, Gause S, Drug screening of newborns by meconium analysis: a large scale, prospective, epidemiologic study. Pediatrics 1992; 89:107-113 [Abstract/Free Full Text]
2. Schutzman DL, Frankenfield-Chernicoff M, Clatterbaugh HE, Singer J Incidence of intrauterine cocaine exposure in a suburban setting. Pediatrics 1991; 88:825-827 [Abstract/Free Full Text]
3. Frank DA, Zuckerman BS, Amaro H, Cocaine use during pregnancy: prevalence and correlates. Pediatrics 1988; 82:888-895 [Abstract/Free Full Text]
4. Ostrea EM, Parks P, Brady M Rapid isolation and detection of drugs in meconium of infants of drug dependent mothers. Clin Chem 1988; 34:2372-2373 [Free Full Text]
5. Ostrea EM, Brady MJ, Parks PM, Asenio DC, Naluz A Drug screening of meconium in infants of drug dependent mothers: an alternative to urine screening. J Pediatr 1989; 115:474-477 [CrossRef][Medline]
6. Wingert WE, Feldman MS, Kim MH, A comparison of meconium, maternal urine and neonatal urine for detection of maternal drug use during pregnancy. J Forensic Sci 1994; 39:150-158 [Medline]
7. Clark GD, Rosenzweig IB, Raisys V, The analysis of cocaine and benzoylecgonine in meconium. J Anal Toxicol 1992; 16:261-263 [Medline]
8. Steele BW, Bandstra ES, Niou-Ching Wu, Hime GW, Hearne WL m-Hydroxybenzoylecgonine: an important contributor to the immunoreactivity in assays for benzoylecgonine in meconium. J Anal Toxicol 1993; 17:348-352 [Medline]
9. Angus J, Greenglass E, Bermes E, Kahn S Benzoylecgonine in the meconium of neonatal infants: an analysis by three methods. Clin Chem 1992; 38:1016
10. Browne S, Moore C, Negrusz A, Detection of cocaine, norcocaine, and cocaethylene in the meconium of premature neonates. J Forensic Sci 1994; 39:1515-1519 [Medline]
11. Ryan RM, Wagner CL, Schultz JM, Meconium analysis for improved identification of infants exposed to cocaine in utero. J Pediatr 1994; 125:435-440 [CrossRef][Medline]
12. Lewis DE, Moore CM, Leikin JB, Koller A Meconium analysis for cocaine: a validation study and comparison with paired urine analysis. J Anal Toxicol 1995; 19:148-150 [Medline]
13. Casanova OQ, Lombardero N, Behnke M, Detection of cocaine exposure in the neonate: analyses of urine, meconium, and amniotic fluid from mothers and infants exposed to cocaine. Arch Pathol Lab Med 1994; 118:988-993 [Medline]
14. Madden JD, Payne TF, Miller S Maternal cocaine abuse and effect on the newborn. Pediatrics 1986; 77:209-211 [Abstract/Free Full Text]
15. Bingol N, Fuchs M, Diaz V, Stone RK, Gromisch DS Teratogenicity of cocaine in humans. J Pediatr 1987; 110:93-96 [CrossRef][Medline]
16. Oro AS, Dixon SD Perinatal cocaine and methamphetamine exposure: maternal and neonatal correlates. J Pediatr 1987; 111:571-578 [Medline]
17. Davis M. NMIS Fact Sheet. Hyattsville, MD: National Center for Health Statistics; 1988
18. National Institute on Drug Abuse. National Pregnancy, Health Survey. Rockville, MD: National Institutes of Health; 1996
19. Chasnoff IJ Drug use and women: establishing a standard of care. Ann N Y Acad Sci 1989; 562:208-210 [Medline]
20. US General Accounting Office. Drug-Exposed Infants: A Generation at Risk. Washington, DC: US General Accounting Office; 1990
21. Vega WA, Kolody B, Hwang J, Noble A Prevalence and magnitude of prenatal substance exposures in California. N Engl J Med 1993; 29:850-854
22. Nalty D. 1991 South Carolina Prevalence Study of Drug Use Among Women Giving Birth. Columbia, SC: South Carolina Commission on Alcohol and Drug Abuse; 1991
23. Hollinshead WH, Griffin JS, Scott HD, Burke ME Statewide prevalence of illicit drug use by pregnant womenRhode Island. MMWR Morb Mortal Wkly Rep 1990; 39:225-227 [Medline]
24. George KD, Price J, Hauth JC, Barnette DM, Preston P Drug abuse screening of childbearing-age women in Alabama public health clinics. Am J Obstet Gynecol 1991; 165:924-927 [Medline]
25. Madry K, Fredlund E, Wallisch, Spence R. 1990 Texas Survey of Postpartum Women and Drug-Exposed Infants. Austin TX: Texas Commission on Alcohol and Drug Abuse; 1991
26. DeVane D, Garite T, Hunt L, et al. Prevalence and Outcomes of Perinatal Drug and Alcohol Use by Women in Orange County, California. Costa Mesa, CA: March of Dimes Birth Defects Foundation; 1991
27. Chasnoff IJ, Landress H, Barrett M The prevalence of illicit drug or alcohol use during pregnancy and discrepancies in Pinellas County, Florida. N Engl J Med 1990; 322:1202-1206 [Abstract]
28. Zuckerman B, Frank DA, Hingson R, Effects of maternal marijuana and cocaine use on fetal growth. N Engl J Med 1989; 320:762-768 [Abstract]
29. Chasnoff IJ, Burns WJ, Schnoll SH, Cocaine use in pregnancy. N Engl J Med 1985; 313:666-669 [Abstract]
30. ElSohly MA, Stanford DF, Murphy TP, Immunoassay and GC-MS procedures for the analysis of drugs of abuse in meconium. J Anal Toxicol 1999; 23:436-445 [Medline]
31. Kliegman RM, Madura D, Kiwi R, Relation of maternal cocaine use to the risks of prematurity and low birthweight. J Pediatr 1994; 124:751 [CrossRef][Medline]
32. Doberczak TT, Thornton JC, Bernstein J, Impact of maternal drug dependency on birthweight and head circumference of offspring. Am J Dis Child 1987; 141:1163 [Abstract/Free Full Text]
33. Lombardero N, Casanova O, Behnke M, Eyler FD, Bertholf RL Measurement of cocaine and metabolites in urine, meconium and diapers by gas chromatography/mass spectrometry. Ann Clin Lab Sci 1993; 23:385-394 [Abstract]
34. Moore C, Lewis D, Leikin J False-positive and false-negative rates in meconium drug testing. Clin Chem 1995; 41:1614-1616 [Abstract/Free Full Text]
35. Ostrea EM, Knapp D, Romero A, Montes M, Ostrea AR Meconium analysis to assess fetal exposure to active and passive maternal smoking. J Pediatr 1994; 124:471-476 [CrossRef][Medline]
36. Stolk LM, Coenradie SM, Smit BJ, van As HL Analysis of methadone and its primary metabolite in meconium. J Anal Toxicol 1997; 21:154-159 [Medline]
37. Ostrea EM, Knapp DK, Tannenbaum L, Serial meconium drug analysis can estimate the chronology and degree of the infant's in utero drug exposure. Pediatr Res 1993; 33:229A
38. ElSohly MA, Feng S ⁹-THC metabolites in meconium: identification of 11-OH-⁹-THC, 8, 11-diOH-⁹-THC, and 11-nor-⁹-THC-9-COOH as major metabolites of ⁹-THC. J Anal Toxicol 1998; 22:329-335 [Medline]
39. Murphey LJ, Olsen GD, Konkol RJ Quantitation of benzoylnorecgonine and other cocaine metabolites in meconium by high-performance liquid chromatography. J Chromatogr 1993; 613:330-335 [Medline]
40. Moore CM, Deitermann D, Lewis D, Leikin J The detection of hydrocodone in meconium: two case studies. J Anal Toxicol 1995; 6:514-518
41. Becker J, Moore C, Lewis D, Leikin J Morphine detection in meconium: Hydrolyzed v. nonhydrolyzed specimens. Clin Chem 1995; 41:S114
42. Nakamura KT, Ayau EL, Uyehara CJ, Methamphetamine detection from meconium and amniotic fluid in guinea pigs depends on gestational age and metabolism. Dev Pharmacol Ther 1992; 19:183-190 [Medline]
43. Lester BM, LaGasse LL, Brunner SM Data base of studies on prenatal cocaine exposure and child outcome. J Drug Issues 1997; 27:487