Thursday, March 10, 2011

CDC Study Ties Opioid Analgesics to Birth Defects

CDC LogoAccording to a newly published study from the U.S. Centers for Disease Control and Prevention (CDC), women who take opioid analgesics such as codeine or hydrocodone shortly before or during early pregnancy may have an increased risk of delivering babies with certain birth defects. While news media reports made this seem like a dire situation, careful examination of the data suggests that the problems are small and not yet clearly defined.

The CDC study, published online ahead of print in the American Journal of Obstetrics and Gynecology (AJOG), examined whether mothers using opioid analgesics between 1 month before pregnancy and during the first trimester was associated with certain birth defects in their newborns [Broussard et al. 2011]. Data were taken from the National Birth Defects Prevention Study (1997 through 2005), which is an ongoing population-based case-control investigation in the United States.

Women with diabetes and/or illicit-drug use were excluded, and data were adjusted to take into account maternal age, race/ethnicity, education, obesity, smoking status, and study center (data were collected at sites in 10 different states). “Cases” included women who gave birth to babies with birth defects of concern during the study period; whereas, “controls” consisted of a comparable group of women giving birth to babies without such defects during the same timeframe.

Therapeutic opioid use was reported by 2.6% of 17,449 case mothers and 2.0% of 6,701 control mothers, most commonly codeine (34.5%) and hydrocodone (34.5%) in combination with other ingredients, and to a much lesser extent meperidine (12.9%) and oxycodone (14.4%). Of the many birth defects examined, opioid therapy was significantly associated with greater odds of infants being born with the following: conoventricular septal defects (OR [Odds Ratio] 2.7; 95% CI [Confidence Interval] 1.1–6.3), atrioventricular septal defects (OR 2.0; 95% CI 1.2–3.6), hypoplastic left heart syndrome (OR 2.4; 95% CI 1.4–4.1), spina bifida (OR 2.0; 95% CI 1.3–3.2), or gastroschisis (OR 1.8; 95% CI 1.1–2.9).

The authors conclude that, consistent with previous investigations, their study shows an association between maternal use of opioid-analgesic therapy in early pregnancy and certain uncommon birth defects, including some types of congenital heart malformations that are important contributors to infant morbidity and mortality. They suggest that the opioid-related risks indicated by this study should be considered by women and their healthcare providers when making treatment decisions during pregnancy; however, they concede that the absolute risks for any individual woman and her baby are relatively small.

COMMENTARY: As might be expected, the news media created an alarming impression that all opioid analgesics cause birth defects of many types. However, there are a number of possible sources of confusion, as well as limitations, in this research study that are worth considering…
  • Of great importance, there is a strong possibility of recall bias, since opioid analgesic use data were based on retrospective self-reports by the mothers. The average time lag between delivery and the data-gathering interview was 9 to 11 months, with some interviews occurring 2 years postpartum. Thus, the women were expected to recall their use of specific medication during a particular timeframe long after the fact — up to 3 years in some cases.

  • The researchers did not attempt to collect specific information on opioid dosing, so possible dose-response relationships affecting malformations could not be assessed. Furthermore, many of the opioids were in multicomponent products, and teratogenic effects of the added ingredients may have played an undetermined role. Also, no distinctions were made between short- or long-acting formulations of the opioids in question.

  • Among the two-thirds of women in whom opioid use was linked to a specific reason, the most commonly reported were surgical procedures (41%), infections (34%), chronic diseases (20%), and injuries (18%). Therefore, morbidity and possible comedications associated with treating those conditions were potentially confounding factors that might have influenced fetal-development malformations leading to birth defects — the contribution of opioids in these cases is unknown.

  • It could be important that the lower limits of the confidence intervals for the Odds Ratios deemed as significant in this study were close to 1.0, which denotes even odds or no difference in effect. So, the statistical likelihood of opioids having an unequivocally strong influence on many of the noted birth defects might be questioned.

  • Toward the end of their article the authors stress, “an increased relative risk for any rare birth defect with an exposure [ie, opioids in this study] usually translates into only a modest absolute increase in risk above the baseline birth defects risk.”

    • Indeed, as an example, they provide data relating to hypoplastic left heart syndrome. The base-line, or expected, prevalence of this defect in the U.S. is 2.4 per 10,000 (0.024%) live births. In the CDC study there was a potentially 2.4-fold increased risk of this birth defect in the infants of opioid-treated women, and this would then represent a 0.058% relative risk or a modest absolute risk increase associated with opioid use of 3.4 per 10,000 births (0.034%). Another way of stating this is that, for every 2,900 women receiving opioid therapy just before or during early pregnancy, 1 additional case of this malformation might occur in their newborns beyond the frequency of this birth defect normally expected to occur in the population. [These are our calculations based on the authors’ data.]

    • Technical Note: Astute readers may notice that the CDC-report authors switch from discussing Odd and Odds Ratios, as they do throughout most of their article, and then refer to the same data as denoting Risks and Relative Risks. Normally, odds and risks are statistically quite different in magnitude and in some cases the given Odds Ratios overstate the comparable Risk Ratios. While the authors’ commingling of terms in their report is technically acceptable in most cases (since most event probabilities are extremely low in this study), this adds considerable confusion to the understanding and interpretation of results.
The study report, and subsequent news stories, give the overall impression that taking opioids produces multifold increases in birth defects, when, in fact, the relative and absolute increases in risk are small and could be strongly affected by sampling or methodological biases. Overall, the evidence in this study might be rated as being of rather low quality, and further investigations with more accurate, prospective data-gathering seem warranted.

Still, regardless of how modest or questionable the data are in this study, the numbers do represent human newborns possibly at some added risk; so, the CDC apparently felt the data were important enough to merit alerting the public. However, it is curious that, while the CDC first reported these data at an epidemiology conference in December 2008, they did not even submit the study for publication until July 2010, and it was not published online until about 8 months later. Ostensibly, the findings were viewed by the government and the journal as worthwhile but not of any great urgency.

Unfortunately, it was not the purpose of this study to provide advice on how healthcare providers, and their female patients, should respond to the concerns and what alternatives for pain management might be more appropriate. While nonsteroidal anti-inflammatory drugs (NSAIDs) — both prescription and over-the-counter — are frequently used for pain relief by women of child-bearing age or during pregnancy, there is some evidence that these drugs may negatively affect fertility or incur functional toxicities and structural anomalies in the developing fetus and newborn [eg, see Internal Medicine News]. Therefore, having raised a “red flag” of sorts about opioids and birth defects, it is hoped that the government will take the next steps by putting their findings into better perspective and providing relevant clinical guidance for practitioners.

REFERENCE: Broussard CS, Rasmussen SA, Reefhuis J, et al. Maternal treatment with opioid analgesics and risk for birth defects. AJOG. 2011(Feb 23); online ahead of print [abstract here].