Substance use during pregnancy is of particular concern because of the increased risk of adverse maternal and infant health outcomes.1 Cocaine abuse among pregnant women increased rapidly during the early 1990s, fueled by the widespread availability of an inexpensive and highly addictive form of cocaine called “crack.”2 Initially concentrated in the Western and rural United States,3 the morbidity associated with the use of amphetamines, particularly methamphetamine, has recently begun to grow rapidly,4,5 causing concern about potential abuse during pregnancy.6
Although the medical literature on the adverse maternal and infant outcomes of cocaine abuse during pregnancy is substantial, less research has been done on amphetamine abuse during pregnancy; however, the health effects reported are similar.7 Both drugs have been associated with little or no prenatal care, fetal death (spontaneous abortion; stillbirth), intrauterine growth restriction, fetal distress, placenta previa, abruptio placentae, premature rupture of membranes, premature labor and birth, precipitate labor, pregnancy-induced hypertension, and complications of anesthesia.7–10
In addition to adverse obstetric outcomes, amphetamine or cocaine use is associated with medical disorders that may complicate the treatment of pregnant women, such as cardiovascular disorders, stroke, seizures, psychiatric disorders, nutritional deficiencies, transmission of infectious and sexually transmitted diseases, and trauma-related injuries.11–15 However, differences in patterns of neural degeneration,16 self-reported adverse effects,17,18 and use of psychiatric inpatient services19 suggest that the severity of the health effects may differ by amphetamine or cocaine use.
Administrative data are frequently used to examine maternal morbidity. Useful information about amphetamine abuse among pregnant women can be produced by examining a nationally representative sample of hospital discharges. Using the Healthcare Cost and Utilization Project Nationwide Inpatient Sample (NIS), we examined trends in pregnancy hospitalizations with a diagnosis of amphetamine or cocaine abuse and compared the prevalence of medical conditions associated with such abuse among pregnancy hospitalizations.
MATERIALS AND METHODS
Hospital discharge data were obtained from the Healthcare Cost and Utilization Project NIS. The NIS is a research database produced annually through a partnership between the Agency for Healthcare Research and Quality and public and private state-level data-collection organizations to provide national estimates of inpatient care in the United States.20 It is the largest collection of all-payer data on inpatient care in the United States and provides demographic data, diagnostic/procedural data, and information on facilities.
Using a stratified probability design, the NIS is constructed to approximate a 20% sample of all U.S. community hospitals as defined by the American Hospital Association. The American Hospital Association defines community hospitals as all nonfederal short-term (average length of stay less then 30 days) general and specialty hospitals whose facilities are open to the public. The sampling frame consists of state-specific hospital discharge data provided to the Healthcare Cost and Utilization Project. The number of participating states has varied from year to year; the latest release of NIS in 2004 contained data from 37 states. The sampling frame for the NIS 2004 included approximately 90% of all hospital discharges in the United States. Subject to sampling restrictions specified by the states, hospitals are selected on the basis of five characteristics for stratification: rural or urban location, bed size, geographic region, teaching status, and ownership. HCUP classifies hospitals in metropolitan statistical areas as urban and hospitals in a non-metropolitan statistical area as rural. The NIS includes all inpatient data from sampled institutions. When weighted, the NIS produces national estimates of hospital inpatient stays.
The Centers for Disease Control and Prevention determined that this project did not require human-subject research review because of the use of de-identified information from a publicly available administrative data set. The unit of analysis was the hospitalization, not the individual person. Analysis was conducted on hospitalizations of pregnant women aged 15–44 years using NIS data from 1998 to 2004. Data analysis was performed using SUDAAN 9 (Research Triangle Institute, Research Triangle Park, NC) to account for the complex survey design and to produce accurate standard errors. Programming and data results were confirmed by two independent researchers.
Hospitalizations were classified using International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnosis codes.21 Discharge records with any primary or secondary ICD-9-CM diagnosis code of 640–677, V22–V24, V27, V28, or 792.3 were defined as pregnancy hospitalizations. Pregnancy hospitalizations with a diagnosis of pregnancy loss were excluded (ectopic and molar pregnancy [ICD-9-CM codes 630–633] and other pregnancy with abortive outcome [ICD-9-CM codes 634–639]). Delivery hospitalizations were identified by ICD-9-CM code V27. Hospitalizations with a diagnosis of amphetamine abuse were identified by ICD-9-CM codes 304.4 and 305.7, and cocaine abuse was identified by codes 304.2 and 305.6. Hospitalizations with a diagnosis of both amphetamine abuse and cocaine abuse were excluded because of small numbers (unweighted sample size 194 hospitalizations).
Trends in pregnancy hospitalizations with a diagnosis of amphetamine or cocaine abuse were investigated using hospitalization ratios. Hospitalization ratios were defined as the number of hospitalizations per 100 deliveries. The SUDAAN analytical procedure PROC Ratio was used to calculate hospitalization ratios for amphetamine and cocaine abuse and to test for linear trends over time.
Subsequent to the trend analysis, we created a comparison group of pregnancy hospitalizations without a diagnosis of substance abuse. Pregnancy hospitalizations with ICD-9-CM diagnosis codes for substance abuse (291, 292, 303–305, 648.3, or 648.4) were excluded. Amphetamine-abuse hospitalizations were compared with cocaine-abuse hospitalizations and the non–substance-abuse hospitalization comparison group. A χ2 analysis was used to compare hospitalization characteristics. Characteristics of hospitalizations examined included age group (15–24, 25–34, and 35–44 years), primary expected payer (publicly funded insurance [Medicaid or Medicare], private insurance [eg, health maintenance organization], or other [including self-pay and no charge]), hospital location (urban or rural), geographic region (Northeast, Midwest, South, or West), and delivery status (yes or no).
We examined medical conditions commonly associated with amphetamine or cocaine abuse. Medical conditions were classified using Clinical Classifications Software codes and ICD-9-CM codes. Clinical Classifications Software coding is a diagnosis categorization scheme developed at the Agency for Healthcare Research and Quality that collapses multiple ICD-9-CM codes into a smaller number of clinically meaningful categories.22 Clinical Classifications Software codes were used to define nonobstetric complications, whereas ICD-9-CM codes were used to supplement these groupings with pregnancy-related diagnoses and define obstetric complications. Select obstetric and nonobstetric complications were evaluated and presented among nondelivery hospitalizations, whereas obstetric complications were accessed among delivery hospitalizations.
Logistic regression was used to build multivariable models with each medical complication of interest as the dependent variable and with an indicator variable for the comparison groups (amphetamine, cocaine, and non–substance abuse) as an independent variable. Estimates were adjusted for age group, primary expected payer, hospital location, geographic region, year, and diagnosis of alcohol, opioid, cannabis, or tobacco abuse. Conditional probabilities estimated by logistic regression were used to calculate adjusted prevalence ratios comparing amphetamine-abuse hospitalizations with cocaine-abuse hospitalizations and non–substance-abuse hospitalizations.23
From 1998 to 2004, an estimated 130,389 pregnancy hospitalizations included a diagnosis of cocaine abuse (unweighted sample size 26,375 hospitalizations), and an estimated 36,062 pregnancy hospitalizations included a diagnosis of amphetamine abuse (unweighted sample size 7,478 hospitalizations). During this interval, the hospitalization ratio for cocaine abuse decreased 44%, from 0.74 per 100 deliveries in 1998 to 0.41 per 100 deliveries in 2004 (linear trend, P<.001) (Fig. 1). The hospitalization ratio for amphetamine abuse doubled, from 0.11 per 100 deliveries in 1998 to 0.22 per 100 deliveries in 2004 (linear trend, P<.001) (Fig. 1).
The amphetamine-abuse group had a larger proportion of women aged younger than 24 years than the cocaine-abuse and non–substance-abuse groups (P<.001) (Table 1). Pregnancy hospitalizations with a diagnosis of amphetamine abuse were more likely to be located in rural areas than were pregnancy hospitalizations with a diagnosis of cocaine abuse (P<.001). The amphetamine-abuse group was more likely than the non–substance-abuse group to list public insurance as the primary expected payer (P<.001); however, the proportion of hospitalizations with public insurance was not significantly different between the substance abuse groups. The vast majority of hospitalizations with a diagnosis of amphetamine abuse (82.2%) were concentrated in the Western region of the United States.
Nondelivery hospitalizations with a diagnosis of amphetamine abuse had a higher prevalence of several medical conditions (anemia, genitourinary tract infections, hepatitis, psychiatric disorders, epilepsy or convulsions, cardiovascular disorders, injury, hypertension complicating pregnancy, premature rupture of membranes, placenta previa, and placental abruption) as compared with nondelivery hospitalizations without a diagnosis of substance abuse (Table 2). In contrast, the amphetamine-abuse group had a lower prevalence of premature labor without delivery than the non–substance-abuse group. When the amphetamine-abuse and cocaine-abuse groups were compared with each other, psychiatric disorders were less prevalent in the amphetamine-abuse group, whereas cardiovascular disorders, hypertension complicating pregnancy, and placenta previa were more frequent in the amphetamine-abuse group.
For obstetric conditions among delivery hospitalizations, several diagnoses were more prevalent in the amphetamine-abuse group than in the non–substance-abuse group: hypertension complicating pregnancy, premature rupture of membranes, placenta previa, placental abruption, premature delivery, infection of the amniotic cavity, intrauterine death, poor fetal growth, and precipitate labor (Table 3). The amphetamine-abuse group had a lower prevalence than the cocaine-abuse group for premature delivery, infection of the amniotic cavity, and poor fetal growth. The prevalence of hypertension complicating pregnancy was higher in the amphetamine-abuse group than in the cocaine-abuse group.
Our results comparing hospitalization characteristics are similar to those of previous reports regarding amphetamine and cocaine abuse. Amphetamine users tend to be younger, located in rural areas, and living in the Western United States,24 whereas cocaine users are older and live in urban areas where the drug is readily available.25 The association we found between insurance status and hospitalization with a diagnosis of amphetamine or cocaine abuse may reflect socioeconomic status, because both drugs are associated with poor social conditions.6,25
As expected, we found that most medical conditions we evaluated were more common among hospitalizations with a diagnosis of amphetamine abuse as compared with non–substance-abuse hospitalizations; however, some results did not conform to the pattern we expected. Premature labor without delivery was more common in the non–substance-abuse group than the amphetamine-abuse group, whereas the reverse was true for premature delivery. Because amphetamine use is associated with less prenatal care,6 it is possible that cases of premature labor that may be managed during the prenatal period are not diagnosed until delivery is imminent among amphetamine users. Therefore, the prevalence of this condition is artificially low among nondelivery amphetamine-abuse hospitalizations.
Although the health effects of cocaine and amphetamines are often assumed to be similar, we found some significant differences in the prevalence of particular outcomes between these groups. Obstetric conditions that may influence infant morbidity, including premature delivery and poor fetal growth, were more common in the cocaine-abuse group. The prevalence of psychiatric disorders was higher in the cocaine-abuse group than the amphetamine-abuse group, contrary to some reports that suggesting that amphetamine abuse may have a greater and more long-lasting effect on mental health than cocaine.17,19 The prevalence of disorders due to the vasoconstrictive properties of stimulant abuse (cardiovascular complications and hypertension complicating pregnancy) was higher in the amphetamine-abuse group than the cocaine-abuse group. Cocaine abuse is often investigated as a causative factor for cardiovascular incidents in young adults,26 yet amphetamine abuse may also play a significant role in cardiovascular incidents among pregnant women.
Some limitations arise when using administrative data to study substance-abuse morbidity among pregnant women. We were unable to evaluate additional outcomes of interest, such as human immunodeficiency virus (HIV) or sexually transmitted diseases, because of small sample sizes in the amphetamine-abuse group. Substance abuse is underreported in hospital discharge data27; therefore, this is a conservative measure of the prevalence and effect of amphetamine or cocaine abuse on pregnancy hospitalizations. Although psychiatric units within U.S. community hospitals are included in the NIS, the NIS does not include data from alcoholism or chemical-dependency treatment facilities, long-term institutions, or psychiatric hospitals; therefore, we are not able to estimate the burden of care for amphetamine or cocaine abuse among pregnant women who present at these facilities. Because the unit of analysis was individual discharge records and all personal identifiers were removed, women hospitalized repeatedly during the study period may have been counted more than once. Race or ethnicity data in the NIS are incomplete because of differences in state procedures for collecting and reporting this variable; therefore, we did not include it as a covariate. In addition, other variables that may be appropriate covariates for our analysis, such as history of prenatal care and gestational age at admission, were not available in this data set.
Nonbiased, long-term monitoring of the health effects on children born to women who abuse amphetamines is needed. Substantial evidence indicates that despite fears at the time, an excessive burden on society from an epidemic of “crack babies” did not develop when cocaine use among pregnant women was increasing rapidly in the early 1990s and that punitive measures are ineffective in reducing illegal drug use among pregnant women.28 Clinicians who serve reproductive-age women have an ethical obligation to provide complete and comprehensive medical care and should be aware of evidence-based guidelines for universal screening, brief interventions, and referral to treatment for substance abuse.29 Brief physician advice has been shown to be cost-effective and successful in reducing substance abuse and improving overall patient health status.29 Substance-abuse treatment during pregnancy is associated with increased fetal growth, which may decrease the risk of poor neonatal outcomes.30 In addition, programs are needed to address factors that contribute to substance abuse and adverse outcomes, such as poor mental health, poverty, and broken social environments.
In conclusion, more scientific data are needed to quantify the health effects of substance abuse during pregnancy; however, there are few nationwide studies on this topic. The use of the NIS allowed a national estimate of the effect of amphetamine abuse among pregnancy hospitalizations. Although the burden of cocaine abuse is higher than that of amphetamine abuse, the latter problem is rising and is more common among younger women, so clinicians must familiarize themselves with the adverse consequences of amphetamine abuse during pregnancy.
1. Greenfield SF, Manwani SG, Nargiso JE. Epidemiology of substance use disorders in women. Obstet Gynecol Clin North Am 2003;30:413–46.
2. Cocaine in pregnancy. ACOG Committee Opinion: Committee on Obstetrics: Maternal and Fetal Medicine Number 114—September 1992 (replaces no. 81, March 1980). Int J Gynaecol Obstet 1993;41:102–5.
3. Rawson RA, Anglin MD, Ling W. Will the methamphetamine problem go away? J Addict Dis 2002;21:5–19.
4. Substance Abuse and Mental Health Services Administration, Office of Applied Studies. Treatment episode data set (TEDS): 1992–2002. National admissions to substance abuse treatment services. DASIS Series: S-23, DHHS Publication No. (SMA) 04-3965. Rockville (MD): 2004.
5. Substance Abuse and Mental Health Services Administration, Office of Applied Studies. Emergency department trends from the Drug Abuse Warning Network, final estimates 1995–2002, DASIS Series: D-24, DHHS Publication No. (SMA) 03-3780. Rockville (MD): Substance Abuse and Mental Health Services Administration; 2003.
6. Arria AM, Derauf C, Lagasse LL, Grant P, Shah R, Smith L, et al. Methamphetamine and other substance use during pregnancy: preliminary estimates from the Infant Development, Environment, and Lifestyle (IDEAL) study. Matern Child Health J 2006;10:293–302.
7. Bell GL, Lau K. Perinatal and neonatal issues of substance abuse. Pediatr Clin North Am 1995;42:261–81.
8. Kuczkowski KM. Anesthetic implications of drug abuse in pregnancy. J Clin Anesth 2003;15:382–94.
9. Faiz AS, Ananth CV. Etiology and risk factors for placenta previa: an overview and meta-analysis of observational studies. J Matern Fetal Neonatal Med 2003;13:175–90.
10. Kandall SR. Perinatal effects of cocaine and amphetamine use during pregnancy. Bull N Y Acad Med 1991;67:240–55.
11. National Institute on Drug Abuse. Research Report Series. Cocaine: abuse and addiction. NIH Publication No. 99-4342. Rockville (MD): National Institute on Drug Abuse; 2004.
12. National Institute on Drug Abuse. Research Report Series. Methamphetamine: abuse and addiction. NIH Publication No. 06-4210. Rockville (MD): National Institute on Drug Abuse; 2006.
13. Gowing LR, Henry-Edwards SM, Irvine RJ, Ali RL. The health effects of ecstasy: a literature review. Drug Alcohol Rev 2002;21:53–63.
14. Schermer CR, Wisner DH. Methamphetamine use in trauma patients: a population-based study. J Am Coll Surg 1999;189:442–9.
15. Macdonald S, Anglin-Bodrug K, Mann RE, Erickson P, Hathaway A, Chipman M, et al. Injury risk associated with cannabis and cocaine use. Drug Alcohol Depend 2003;72:99–115.
16. Ellison G, Switzer RC 3rd. Dissimilar patterns of degeneration in brain following four different addictive stimulants. Neuroreport 1993;5:17–20.
17. Rawson R, Huber A, Brethen P, Obert J, Gulati V, Shoptaw S, et al. Methamphetamine and cocaine users: differences in characteristics and treatment retention. J Psychoactive Drugs 2000;32:233–8.
18. Williamson S, Gossop M, Powis B, Griffiths P, Fountain J, Strang J. Adverse effects of stimulant drugs in a community sample of drug users. Drug Alcohol Depend 1997;44:87–94.
19. Leamon MH, Gibson DR, Canning RD, Benjamin L. Hospitalization of patients with cocaine and amphetamine use disorders from a psychiatric emergency service. Psychiatr Serv 2002;53:1461–6.
20. Steiner C, Elixhauser A, Schnaier J. The healthcare cost and utilization project: an overview. Eff Clin Pract 2002;5:143–51.
21. U.S. Public Health Service and Health Care Financing Administration. International Classification of Diseases. 9th Revision, Clinical Modification. DHHS Publication No. (PHS) 80-1260. Washington (DC): U.S. Department of Health and Human Services; 1980.
22. Healthcare Cost and Utilization Project (HCUP). Clinical Classifications Software (CCS) for ICD-9-CM. Rockville (MD): Agency for Healthcare Research and Quality; 2007. Available at: http://www.hcup-us.ahrq.gov/tools_software.jsp
. Retrieved November 20, 2007.
23. Greenland S. Model-based estimation of relative risks and other epidemiologic measures in studies of common outcomes and in case-control studies. Am J Epidemiol 2004;160:301–5.
24. Wu LT, Schlenger WE. Psychostimulant dependence in a community sample. Subst Use Misuse 2003;38:221–48.
25. Ogunyemi D, Hernandez-Loera GE. The impact of antenatal cocaine use on maternal characteristics and neonatal outcomes. J Matern Fetal Neonatal Med 2004;15:253–9.
26. Cregler LL. Cocaine: the newest risk factor for cardiovascular disease. Clin Cardiol 1991;14:449–56.
27. Smothers BA, Yahr HT. Alcohol use disorder and illicit drug use in admissions to general hospitals in the United States. Am J Addict 2005;14:256–67.
28. Armstrong EM. Drug and alcohol use during pregnancy: we need to protect, not punish, women. Womens Health Issues 2005;15:45–7.
29. ACOG Committee on Ethics. ACOG Committee Opinion. Number 294, May 2004. At-risk drinking and illicit drug use: ethical issues in obstetric and gynecologic practice. Obstet Gynecol 2004;103:1021–31.
30. Little BB, Snell LM, Van Beveren TT, Crowell RB, Trayler S, Johnston WL. Treatment of substance abuse during pregnancy and infant outcome. Am J Perinatol 2003;20:255–62.