Narcotic-exposed neonates in a First Nations population in northwestern Ontario

METHODS

This is a retrospective descriptive study of neonatal exposure to oxycodone. Maternal and neonatal chart reviews on all 482 deliveries at the SLMHC from January 2009 to June 2010 were undertaken by physicians and researchers involved in the maternity program at the SLMHC. We examined the most recent 18 months divided into 6-month periods, as we had noted a clinical increase in maternity patients who admitted to narcotic abuse. Ethics approval was obtained from the Research Review Committee of the SLMHC. Our First Nations Special Advisor (H.C.) was involved from planning to synthesis and approval of the final paper. Our findings were shared with our regional First Nations Health Authority before submission for publication.

Data were collected in Excel and imported into PASW Statistics, version 18.0, for statistical analysis. Descriptive analysis of the overall obstetric program was followed by bivariate analysis according to oxycodone exposure using χ² tests for categorical data (Pearson or Fisher exact tests, as appropriate) and independent-samples t tests for continuous data. Our chart review included summaries of charts from patients’ home communities, twice weekly prenatal clinics held in Sioux Lookout from 38 weeks’ gestation onward, and patient hospital records. We asked about daily use and occasional use (defined as from several times per week to monthly, which we described as episodic or “binge” use) of opioids. Infant Finnegan scores were recorded and categorized by whether or not they were greater than 7, as that is the protocol value above which more serious withdrawal is noted and pharmacologic treatment might be required. This scoring system is widely used but not yet validated.

A priori sample size calculations were not completed, as the parameters to estimate power and sample size did not yet exist. However, post hoc power calculations on the primary outcomes revealed that the power ranged from 64% (prematurity) to 84% (length of stay).

RESULTS

During our study period we had 482 live births; 61 of the neonates were exposed to oxycodone in utero. The incidence of oxycodone exposure during pregnancy increased significantly in the study period (P = .050) from an initial 8.4% in the first 6 months of 2009 to a high of 17.2% in the first 6 months of 2010 (Figure 1), with most of those who used oxycodone being occasional users (Figure 2). Narcotic-using mothers were typically smokers and used more alcohol than nonusing mothers did, and there was a trend toward increased parity. The 2 groups were otherwise similar with regard to comorbidities (Table 1).

• Download figure
• Open in new tab

Figure 1

Percentage of exposed neonates (n=61) over consecutive 6-month periods (January 2009 to June 2010): Pearson χ² for differences in proportions P=.05.

• Download figure
• Open in new tab

Figure 2

Patterns of narcotic use (n=61)

There were 61 narcotic-exposed infants in this cohort. Exposed and nonexposed groups of newborns showed equivalent birth weight and Apgar scores. Narcotic-exposed pregnancies had significantly more premature births (P = .001), had longer lengths of stay in hospital (P = .004), and were more likely to need transfer out (P = .005) to tertiary care facilities (Table 2). Among these 482 deliveries there were no neonatal deaths and no HIV or hepatitis C infections. Length of stay was longer among neonates with higher Finnegan scores. Infants with scores greater than 7 had a mean length of stay of 4.5 days versus 1.5 days for the total cohort (P = .004).

Twenty-one infants exhibited symptoms of NAS (4.3%) and had Finnegan scoring done; 18 of these neonates had been exposed to narcotics and 3 had been exposed to alcohol. Three required pharmacologic treatment of opioid withdrawal, 1 of whom was transferred to a tertiary care centre. Among narcotic-exposed infants the rate of NAS was 29.5%, and 11.5% were strongly affected, with at least 1 Finnegan score greater than 7. Daily maternal use of narcotics was associated with a higher rate (66.0%) of NAS-affected infants when compared with episodic users, even though occasional users were far more common (Figure 2). The exposed infants were more likely to be from remote First Nations communities and were also more likely to have been exposed to alcohol.

All our infants roomed in with their mothers, and all mothers were encouraged to breastfeed.

DISCUSSION

We have documented for the first time a very high rate of narcotic exposure during pregnancy in northwestern Ontario. Currently we are seeing a binge pattern of use, dictated perhaps by the geographic remoteness of many of our patients, where access to illicit drugs is episodic. The increasing incidence of oxycodone abuse during pregnancy is consistent with our observations and with media reports of a pervasive abuse of this specific drug in our region.

Methadone, often referenced as optimal treatment for opioid-addicted mothers, is not readily available in remote communities and might not even be the best maternal strategy for episodic users like those most commonly found in our area. The literature also supports long-acting morphine as equivalent maternal prenatal maintenance therapy if such is required.⁸ We generally use MS Contin, a long-acting opioid that can only be ingested orally and cannot be snorted or used intravenously. Only a few of our patients received it antenatally, usually self-identified daily or intravenous opioid abusers. Breastfeeding and rooming-in are standard care at our facility for these patients.^9–11

Oxycodone, the common drug of abuse in our region, is generally taken as a tablet or crushed and snorted, but it can also be injected intravenously. Our area generally has low rates of HIV and hepatitis C infection, but community physicians are seeing increasing intravenous injection use of this drug of abuse. In such closed communities, this raises concerns about a changing pattern of infectious disease, including risk of HIV, hepatitis C, and methicillin-resistant Staphylococcus aureus infections and Gram-positive sepsis, which will further challenge maternal and child health in this region. An urban-based Australian study noted rates of hepatitis C in intravenous narcotic–abusing mothers of NAS infants to be greater than 80%.¹²

Our low rate of NAS requiring pharmacologic treatment is likely a function of the pattern and dosage of oxycodone use. Half of our patients were self-described occasional users. Infants of methadone-treated mothers described in the literature traditionally have much higher rates of NAS. The infants of our daily users did have a higher rate of NAS (66%), which is similar to these studies of methadone-treated mothers.

Our typical narcotic-exposed neonate appears well at birth. The neonates in our study had normal Apgar scores and did not have significantly lower birth weight. These findings are common to most other recent studies. The normal Apgar score is expected, as withdrawal generally occurs some time after delivery, and the change in the infant’s clinical presentation is one of the hallmarks of NAS. Birth weight was low in studies of heroin users in the 1970s and 1980s, but such low birth weight has not been seen in more recent studies, likely owing to improved provision of obstetric care to addicted patients.

We did find an increased risk of prematurity in the narcotic-exposed group (8.2%, P = .001). Other NAS studies have showed similarly higher rates of prematurity.¹³

Follow-up care for mother and child at our centre is arranged through community-based medical, nursing, and counseling services and referral to regional programs, including the Children’s Aid Society when appropriate.

Physicians in our practices have ceased prescribing oxycodone-containing medications (ie, Percocet and OxyContin), and these medications have also been removed from the hospital formulary.