Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Testing for fetal exposure to illicit drugs using umbilical cord tissue vs meconium



We assessed the agreement of testing for fetal exposure to illicit drugs contrasting paired specimens of meconium vs umbilical cord tissue.


We obtained paired samples of meconium and umbilical cord tissue from 118 pregnancies with high suspicion of illicit drug use by the mothers. Each specimen was tested for amphetamines, opiates, cocaine, cannabinoids, and phencyclidine using drug class-specific immunoassays.


The agreement of drug screening results between cord and meconium was above 90% for all drugs tested. Meconium identified 21 cases as positive for amphetamines. The paired cord identified 20 of these, and in addition identified three other positives that the meconium labeled as negative. Gas chromatography–mass spectrometry confirmed these three cord samples as methamphetamine positive. Meconium identified 97 samples that were negative for amphetamines, while the cord identified 94 of these as negative but three as positive. Agreement of cord with meconium for amphetamines was 96.6%. The concordance for opiates was 94.9%, for cocaine was 99.2%, and for cannabinoids was 90.7%.


Umbilical cord tissue performs as well as meconium in assessing fetal drug exposure to amphetamines, opiates, cocaine, and cannabinoids. Results of studies using the cord may have a more rapid return to the clinician, because waiting for meconium to be passed sometimes requires several days. Moreover, in some cases the meconium is passed in utero making collection impossible, whereas cord should always be available for drug testing.


Healthy People 2010 is a framework of health goals for the nation.1 Objective 16–17 of this framework is, ‘to increase abstinence from alcohol, cigarettes, and illicit drugs among pregnant women.’2 As a step toward meeting this goal, accurate data are needed on the incidence of illicit drug use in pregnancy. Urine testing and maternal reporting underestimate the incidence. Ostrea et al.3 reported that meconium testing could detect drug intake during pregnancy. They subsequently showed that meconium testing has advantages over urine testing, particularly when maternal drug administration occurred several days prior to delivery.4, 5, 6 However, sometimes meconium testing is not possible, or is not timely. For instance, sometimes the meconium is passed in utero making it unavailable for testing, and in other cases meconium is not passed for several days, particularly among preterm infants. Kintz and Mangin7 reported neonatal hair testing as an alternative to meconium. However, hair testing may not be as acceptable, because sufficient hair must be cut such that it is noticeable to parents, and some neonates have too little hair for accurate testing.

We hypothesized that cutting a 10 cm segment of umbilical cord at delivery, and homogenizing this tissue for drug testing, would give results comparable to that obtained with meconium. To test this, we obtained paired samples of meconium and umbilical cord from pregnancies with a high suspicion of illicit drug use by the mothers. Each of the pairs was tested for amphetamines, opiates, cocaine, cannabinoids, and phencyclidine (PCP), and the results obtained with cord tissue vs meconium were compared.


During an 18-month period (August 2003–February 2005) a segment of approximately 10 cm was cut from each umbilical cord of a delivery at McKay-Dee Hospital. The segment was lightly rinsed with sterile saline and placed in a sterile container in a laboratory refrigerator. If the neonatologist or pediatrician ordered MecStat™ testing on meconium of that neonate, the meconium and the umbilical cord segment were both submitted to the laboratory for testing. If no meconium testing was ordered after 1 week, the umbilical cord segment was discarded in the usual manner. MecStat™ testing was carried out if any of the following were present; history of a previous pregnancy where drug abuse was proven, maternal report of drug abuse during this pregnancy, no prenatal care, no permanent address, sexually transmitted diseases, mother or father appearing intoxicated, ‘high,’ or abusive or inappropriate.

The meconium was tested and the results reported to the physician as with any clinically ordered specimen. However, the umbilical cord specimens were batched for periodic research analysis, and the results of these were not reported to the attending physician, but rather were treated as a deidentified, anonymous dataset. The umbilical cord segments were stored at −20°C and were batched for analysis. No specimens were stored for more than 12 months. The aim of the study was to obtain 100–120 paired samples for testing from pregnancies at high risk for maternal drug abuse. The McKay-Dee Hospital Institutional Review Board approved the study protocol.

The umbilical cord specimens were extracted by the procedures reported by Le et al.8 for meconium, and the extracts were analyzed for amphetamines, opiates, cocaine, cannabinoids, and PCP by ELISA assays (Immunalysis Corp, Pomona, CA). Gas chromatography–mass spectrometry (GC–MS) confirmations were made following procedures modified from Le et al.8

To arrive at the agreement values between meconium and cord screening data it was necessary to determine cutoff values for the umbilical cord screening assays. This was carried out by using receiver-operating characteristic (ROC) plots to assign a screening cutoff value for amphetamines, opiates, cocaine, and cannabinoids.9


A total of 118 paired samples of umbilical cord and meconium were obtained. For the umbilical cord specimens, ROC plots were generated for amphetamines, opiates, cocaine, and cannabinoids, but could not be constructed for PCP because no PCP-positive samples were detected during this study. From the ROC plots, cutoffs were chosen that yielded the best sensitivity and specificity for each drug screening assay and these cutoffs then yielded the cord data for the agreement tables with meconium.

The agreement between results of cord and meconium testing ranged from 90.7 to 100% concordance, with sensitivity ranging from 75 to 95.24% where calculations were possible, and specificity from 91.18 to 100% where meconium samples were considered the Gold Standard. For instance, as shown in Table 1, 21 meconium samples screened positive for amphetamines while 20 of the matched umbilical cords also screened positive. Of the 97 meconium samples that screened negative for amphetamines, 94 of the matched cords also screened negative, but three screened positive. The three samples screening as positive for amphetamines in cord, but negative in meconium, were confirmed as methamphetamine-positive specimens using GC–MS. We observed 94.9% concordance for opiates, 99.2% concordance for cocaine, 90.7% for cannabinoids.

Table 1 Agreement between paired meconium (MEC) and umbilical cord (UC) specimens in drug screening assays


In 1980, Ostrea et al.3 reported tissue distribution of morphine at autopsy of six addicted monkey fetuses and two neonates of drug-dependent mothers. They observed significant drug concentrations in meconium and suggested that meconium screening might be a useful way to detect fetal morphine exposure. Ostrea later reported meconium specimens from 20 neonates of drug-dependent mothers and five control neonates, as analyzed by radio immunoassay for metabolites of heroin, cocaine, and cannabinoids.4 They reported that meconium of the controls had no drugs detected while that from neonates of drug-dependent mothers invariably showed the presence of at least one drug metabolite. Larger studies followed, showing that the meconium was a useful way to detect maternal drug use, and that meconium could be positive even if the urine tested negative.10, 11, 12, 13, 14

Neonatal hair has also been used as a tissue source for detecting fetal drug exposure.7, 14, 15, 16 Bar-Oz et al. reported a correlation between neonatal hair and meconium concentrations of cocaine, cannabis, and opiates. They found meconium to be slightly more sensitive than hair, but pointed out that collecting meconium may be more acceptable to some parents than cutting hair for testing.15

We hypothesized that drug testing could be carried out on umbilical cord tissue. If correct, this might have certain advantages over meconium or hair testing. For instance, cord tissue could be sent for testing immediately after delivery, while for preterm infants sometimes meconium is not passed for several days. Also, fetuses that are stressed often pass meconium in utero, making that source unavailable for testing. Testing hair may not be as acceptable to some as is testing meconium or umbilical cord tissue, which are otherwise discarded. Also, for widespread epidemiological testing, obtaining a piece of umbilical cord at delivery might be simpler than collecting meconium or cutting hair, and may be a method better suited to anonymous testing. However, meconium testing might have an advantage over cord testing, if the maternal drug abuse was not suspected at delivery and the cord discarded, yet meconium could still be collected.

Our level of agreement in 118 paired meconium vs umbilical cord samples is similar to the reports comparing meconium vs hair.14, 15, 17 Most clinical testing of meconium is carried out with an immunological screening method,4, 5, 6 and false positives can be a problem.6, 17 In the present studies, immunological screening resulted in an area under the ROC plot for amphetamine, opiates, cocaine, cannabinoids greater than 90%, indicating that the umbilical cord screening tests certainly have the ability to distinguish between the presence and absence of drug. GC–MS would be more specific and sensitive than our immunological methods, but they would be much more costly and time consuming.6, 17 When we found samples that were positive for amphetamines in cord tissues, which were negative in the paired meconium, we used GC–MS to resolve the issue of whether the cords gave a false-positive test, or rather, whether the cord was more sensitive than meconium. We observed that in all three cases, the methamphetamine-positive cord samples were confirmed as positive by GC–MS. Therefore, it is possible that umbilical cord tissue is more sensitive than meconium for detecting fetal methamphetamine exposure. Thus, this first study established the feasibility of using umbilical cord specimens for testing of maternal illicit drugs, but future studies will be needed where GC–MS is used for all of the umbilical cord specimens, not just those where a discrepancy between cord and meconium is found. This will be needed to firmly establish cutoff values for the cord specimens.

The question of how recent a maternal drug ingestion must be in order to be detected in meconium, hair, or umbilical cord, has not been settled. Studies of Silvestre et al.,18 using a pregnant rat model, illustrated the complexity of this issue. They observed that the size of the unit dose, the length of drug exposure, and the drug used, were among the variables influencing the drug concentration in meconium. Indeed, whether any given drug of abuse is detectable in meconium longer than it is in umbilical cord or hair is not completely clear. Drugs that are concentrated in bile might likely have higher levels in meconium than in nongastrointestinal tissues. Perhaps these drugs would be detectable in meconium longer than in umbilical cord or hair. For drugs that are not concentrated in bile, perhaps meconium, hair, and umbilical cord would have roughly equivalent concentrations, and would be similarly sensitive to distant exposure.

In this study, we demonstrated the feasibility of using umbilical cord tissue as a means of assessing fetal drug exposure. We recognize that additional confirmatory testing is needed, particularly for opiates, cocaine, and PCP, since so few of the 118 cases were positive for these. We predict that the cost of performing such studies using umbilical cord segments will be similar to meconium testing, and that laboratories now set up to analyze meconium could do these determinations on umbilical cord tissue. We speculate that umbilical cord testing would sometimes have an advantage over meconium testing because it could generally be sent for analysis sooner than could meconium. Moreover, umbilical cord tissue might be a more suitable screening method of anonymous epidemiologic testing, with fewer inherent problems than with hair or meconium collection, in order to acquire the data needed for the Healthy People 2010 goals of reducing illicit drug use during pregnancy.


  1. 1

  2. 2

  3. 3

    Ostrea Jr EM, Lynn SM, Wayne RN, Stryker JC . Tissue distribution of morphine in the newborns of addicted monkeys and humans. Clinical implications. Dev Pharmacol Ther 1980; 1: 163–170.

    CAS  PubMed  Google Scholar 

  4. 4

    Ostrea Jr EM, Brady MJ, Parks PM, Asensio DC, Naluz A . Drug screening of meconium in infants of drug-dependent mothers: an alternative to urine testing. J Pediatr 1989; 115: 474–477.

    Article  Google Scholar 

  5. 5

    Ostrea Jr EM, Romero A, Yee H . Adaptation of the meconium drug test for mass screening. J Pediatr 1993; 122: 152–154.

    Article  Google Scholar 

  6. 6

    Moore C, Negrusz A, Lewis D . Determination of drugs of abuse in meconium. J Chromatogr B Biomed Sci Appl 1998; 713: 137–146.

    CAS  Article  Google Scholar 

  7. 7

    Kintz P, Mangin P . Evidence of gestational heroin or nicotine exposure by analysis of fetal hair. Forensic Sci Int 1993; 63: 99–104.

    CAS  Article  Google Scholar 

  8. 8

    Le NL, Reiter A, Tomlinson K, Jones J, Moore C . The detection of oxycodone in meconium specimens. J Anal Toxicol 2005; 29: 54–57.

    CAS  Article  Google Scholar 

  9. 9

    Zweig MH, Campbell G . Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin Chem 1993; 39: 561–577.

    CAS  PubMed  PubMed Central  Google Scholar 

  10. 10

    Ostrea Jr EM, Brady M, Gause S, Raymundo AL, Stevens M . Drug screening of newborns by meconium analysis: a large-scale, prospective, epidemiologic study. Pediatrics 1992; 89: 107–113.

    PubMed  PubMed Central  Google Scholar 

  11. 11

    Casanova OQ, Lombardero N, Behnke M, Eyler FD, Conlon M, Bertholf RL . 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.

    CAS  PubMed  Google Scholar 

  12. 12

    Moriya F, Chan KM, Noguchi TT, Wu PY . Testing for drugs of abuse in meconium of newborn infants. J Anal Toxicol 1994; 18: 41–45.

    CAS  Article  Google Scholar 

  13. 13

    Buchi KF, Zone S, Langheinrich K, Varner MW . Changing prevalence of prenatal substance abuse in Utah. Obstet Gynecol 2003; 102: 27–30.

    PubMed  Google Scholar 

  14. 14

    Ostrea Jr EM, Knapp DK, Tannenbaum L, Ostrea AR, Romero A, Salari V et al. Estimates of illicit drug use during pregnancy by maternal interview, hair analysis, and meconium analysis. J Pediatr 2001; 138: 344–348.

    CAS  Article  Google Scholar 

  15. 15

    Bar-Oz B, Klein J, Karaskov T, Koren G . Comparison of meconium and neonatal hair analysis for detection of gestational exposure to drugs of abuse. Arch Dis Child Fetal Neonatal Ed 2003; 88: F98–F100.

    CAS  Article  Google Scholar 

  16. 16

    Vinner E, Vignau J, Thibault D, Codaccioni X, Brassart C, Humbert L et al. Neonatal hair analysis contribution to establishing a gestational drug exposure profile and predicting a withdrawal syndrome. Ther Drug Monit 2003; 25: 421–432.

    CAS  Article  Google Scholar 

  17. 17

    Moore C, Lewis D, Leikin J . False-positive and false-negative rates in meconium drug testing. Clin Chem 1995; 41: 1614–1616.

    CAS  PubMed  Google Scholar 

  18. 18

    Silvestre MA, Lucena JE, Roxas Jr R, Evangelista ES, Ostrea Jr EM . Effects of timing, dosage, and duration of morphine intake during pregnancy on the amount of morphine in meconium in a rat model. Biol Neonate 1997; 72: 112–117.

    CAS  Article  Google Scholar 

Download references


The authors thank the nurses of the Labor and Delivery Unit and the Neonatal Intensive Care unit at McKay-Dee Hospital for their valuable assistance in collecting the umbilical cord and meconium specimens. This work was supported by grant R43 DAD017412.

Author information



Corresponding author

Correspondence to R D Christensen.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Montgomery, D., Plate, C., Alder, S. et al. Testing for fetal exposure to illicit drugs using umbilical cord tissue vs meconium. J Perinatol 26, 11–14 (2006).

Download citation


  • drug abuse
  • amphetamines
  • cannabinoids
  • meconium
  • umbilical cord

Further reading


Quick links