Preferred route of administration varied by substance and by rural/urban status. Among urban participants, oral use (swallowing whole or chewing and swallowing) was the most common route of administration. This contrasted sharply with substance-specific variation in routes of administration among rural participants. For example, snorting was the most frequent route of administration for hydrocodone, methadone, OxyContin®, and oxycodone, while injecting was most commonly used for hydromorphone and morphine administration. After adjustment for age, race, and gender, rural users had significantly higher odds of snorting hydrocodone, OxyContin®, and oxycodone compared to urban participants.
The increased odds of rural participants to use alternative routes of administration warrant consideration. Previous research has demonstrated that multiple routes of administration are involved in nonmedical prescription opioid use [40, 41, 48]. In fact, our finding on the frequency of snorting OxyContin® compared to swallowing and injecting is consistent with the findings of another Kentucky study . That study, conducted in a clinic-based sample from central Kentucky, found that methadone, morphine, and hydromorphone were being administered through various alternative routes, including snorting, chewing, and injecting .
Previous literature has posited that the decreased availability of heroin in rural areas may contribute to rural-urban differences in prescription opioid use [11–13]; however, this trend is not apparent in this sample, as nearly twice as many rural participants reported lifetime use of heroin than did urban participants (data not shown). Rather, differences in the prevalence of alternative routes of administration is likely to be more intimately linked to differences in drug problem severity. Previous substance use [23, 25] and frequency of current substance use [26, 27] are known risk factors for transitioning to injection from other routes of administration. Scores from the Addiction Severity Index  indicate that rural participants had much higher drug problem severity than did urban participants, which may have contributed to the rural/urban differences in route of administration evident in this study.
The routes of administration for buprenorphine use among rural participants in this study are consistent with other studies [37, 49–52]. For example, the relative frequency of buprenorphine snorting compared to injecting in this study is interesting with implications for preventing diversion. Strategies intended to prevent buprenorphine intravenous misuse, like Suboxone®, may not prevent misuse by alternative routes of administration. The opiate antagonist naloxone contained within Suboxone® "guards" against misuse by causing withdrawal symptoms in those who inject or snort it; however, the data are conflicting .
The routes of fentanyl administration by rural study participants are also noteworthy. Over 70% of rural fentanyl users administered the drug orally. Oral administration of fentanyl has been identified within other populations [38, 54–56]; however, these studies have generally found oral administration to be rare in comparison with other routes of administration. Oral fentanyl administration can result in a wide range of concentrations in the blood, depending on whether the substance is retained in the oral cavity or swallowed [56, 57]. Nevertheless, oral fentanyl administration can have fatal consequences, as demonstrated by findings from post-mortem studies of fentanyl-related deaths [55, 56]. Injecting fentanyl, found among 42% of the fentanyl users in this study, has also been reported in other populations [55, 58, 59]. The frequency of fentanyl injection in this study is concerning given its implications for toxicity and overdose. A fentanyl dose that is survivable following transdermal administration may result in death if administered intravenously . Deaths due to fentanyl overdose following injection can occur at low blood concentrations (2.0 μg/L - 3.0 μg/L) [55, 59–61]. These results are especially disconcerting given that ambulance response times are significantly slower in rural areas , which may increase the likelihood of fatal overdose.
Perhaps most concerning about the high prevalence of alternate routes of administration is the potential for transmission of blood-borne infections such as HIV and hepatitis B and C. While HIV and hepatitis C (HCV) in particular are transmissible by injecting [63–65], it has also been demonstrated that HCV can be transmitted by sharing equipment used to snort drugs, such as straws [65–67]. A seminal review by Strang and colleagues (1998) discusses various health implications for route of drug use, including nasal ulceration from snorting and respiratory and thrombotic complications, abscesses, and endocarditis from injecting . The health consequences of nonmedical prescription opioid use, as delivered by any route of administration can be severe, entailing potential for physical dependence and addiction, severe respiratory distress, and fatal overdose . Overdose risk, in particular, is compounded by the route of administration . Reports have noted that this is especially problematic in OxyContin® use, which was designed to be a slow-release formulation .
While this study broadens understanding of rural substance abuse and alternate routes of administration for prescription opioids, it is not without limitations. The data in this study are self-reported and are subject to response bias. This study is also limited by sample size, which prohibited making statistically meaningful rural-urban comparisons for buprenorphine and fentanyl, as well as statistically precise point estimates for certain routes of administration of other substances. The rural-urban comparisons were also complicated by the baseline demographic differences between the two groups. Race-, gender-, and age-adjusted analyses were used in an attempt to isolate the influence of rurality on the outcome of interest; however, a number of unmeasured social, economic, and structural factors may have also influenced the comparison. Also, given the influence of ecological factors such as drug availability and drug price on determining routes of administration , the study would have been strengthened by an examination of these characteristics in the rural and urban settings involved.