Correspondence
Perinatal Substance Exposures in California
N Engl J Med 1994; 330:794March 17, 1994
- Article
To the Editor:
In the Special Article “Prevalence and Magnitude of Perinatal Substance Exposures in California” (Sept. 16 issue), Vega et al. report an unusually high prevalence of illicit drug and alcohol use among black women as compared with women of other racial groups1. This finding is surprising. Other large studies have not described such significant interracial differences. Anonymous urine toxicology studies by Chasnoff et al.2 and Vaughn et al.3 found positive results with equal frequency in black and in white women. A statewide study of illicit drug use among pregnant women in labor in Rhode Island4 reported a prevalence of 7 percent among whites and 10 percent in other groups. A recent review5 found that the prevalence of alcohol consumption in pregnant women was equal among blacks and whites.
The selection of participating hospitals may have influenced the outcomes reported. Without careful stratification, it is difficult to obtain a representative population in California, even if the sample is large. The study population may contain a considerable selection bias.
The prevalence of alcohol use was determined by urine testing. Urine is not the body fluid of choice for evaluating alcohol intake. The women were patients presenting to hospitals in labor. Given normal behavior, they probably prepared themselves at home before leaving for the hospital. In addition, they had to get to the hospital and complete registration procedures before submitting a urine sample. Considering the elapsed time, it is unlikely that substantial numbers of patients would have maintained detectable urine alcohol levels. They would have had either to drink alcohol for a number of hours before leaving for the hospital or to consume alcoholic beverages en route to the hospital or in the labor and delivery suite.
The prevalence of tobacco use was determined from the patients' own reports at the time of admission. Different groups of patients may have been questioned differently. Members of groups for which there are stereotyped notions of behavior are likely to have been questioned more rigorously than others.
Errors in ascertaining the prevalence of alcohol and tobacco use among black women may have inflated the prevalence of exposure to “any drug” in this group, which Vega et al. reported as more than twice as high as that of any other racial or ethnic group.
5 ReferencesEllen Mason, M.D.
Cook County Hospital, Chicago, IL 606121
Vega WA ,Kolody B ,Hwang J ,Noble A . Prevalence and magnitude of perinatal substance exposures in California. N Engl J Med 1993;329:850-854
Full Text | Web of Science | Medline2
Chasnoff IJ ,Landress HJ ,Barrett ME . The prevalence of illicit-drug or alcohol use during pregnancy and discrepancies in mandatory reporting in Pinellas County, Florida. N Engl J Med 1990;322:1202-1206
Full Text | Web of Science | Medline3
Vaughn AJ ,Carzoli RP ,Sanchez-Ramos L ,Murphy S ,Khan N ,Chiu T . Community-wide estimation of illicit drug use in delivering women: prevalence, demographics, and associated risk factors. Obstet Gynecol 1993;82:92-96
Web of Science | Medline4
Statewide prevalence of illicit drug use by pregnant women -- Rhode IslandMMWR Morb Mortal Wkly Rep 1990;39:225-227
Medline5
Serdula M ,Williamson DF ,Kendrick JS ,Anda RF ,Byers T . Trends in alcohol consumption by pregnant women: 1985 through 1988. JAMA 1991;265:876-879
CrossRef | Web of Science | Medline
Author/Editor Response
Dr. Vega replies: To the Editor: A cluster-stratified sampling procedure was used in our study, with the California counties divided into 20 strata. For each ethnic group, subsample sizes were estimated for a 1 percent error term, with a 12 percent prevalence of drug exposure assumed for each county stratum. Lower prevalence levels for exposure to individual drugs would have even more precise error terms. Twenty separate statewide samples were drawn, with separate weighted estimates available for each ethnic group according to drug type1. To further decrease the error term for statewide estimates, two thirds of all eligible hospitals were sampled. Comparisons showed no variance according to hospital type between hospitals that were sampled and those that were not1.
It is not surprising that our results differ from those reported in more limited studies conducted in other areas. Estimates of illicit drug exposure among black women in California range from 4 percent to 18 percent within the 20 county strata. The overall prevalence of such exposure was high because the rates were highest in large urban counties, where most black women reside. The 1991 Household Survey by the National Institute of Drug Abuse reported that among major U.S. cities, Los Angeles had the highest rates of illicit drug use and that those 18 to 34 years of age were responsible for 80 percent of all such use reported in the preceding month2. Our study should be more accurate than studies based on self-reports because its anonymous design eliminated refusals to participate, thereby reducing the selection bias, and because it did not rely on self-reported answers to ensure veracity.
Dr. Mason correctly notes that urine testing is a limited approach to establishing to prevalence of alcohol use, a fact that is emphasized in the article. However, this is not a source of systematic error that would inflate the rates for blacks. Dr. Mason also questions the self-reported nature of our information on tobacco use. As the article stated, this information is collected routinely in California hospitals and was not obtained specifically for our study.
2 ReferencesWilliam A. Vega, Ph.D.
University of California, Berkeley, Berkeley, CA 947201
Vega WA, Kolody B, Noble A, et al. Profile of alcohol and drug use during pregnancy in California, 1992: scientific report. Sacramento: California Department of Alcohol and Drug Programs, September 1993.
2
Substance Abuse and Mental Health Services Administration. National household survey on drug abuse. Washington, D.C.: Government Printing Office, 1993.
