National polydrug use patterns among people who misuse prescription opioids and people who use heroin. Results from the National Household Survey on Drug Use and Health

Abstract

Background: Polysubstance use among people who misuse opioids (PWMO) is highly prevalent, but understudied. We defined, estimated, and analyzed national polysubstance use patterns among PWMO using National Household Survey on Drug Use and Health data (2017-2019).

Methods: We obtained estimates of past-month patterns of polydrug use using cluster analysis and latent class/profile analysis. We considered misuse of prescription opioids and use of heroin, cocaine (including crack), marijuana, alcohol, and "other" substances.

Results: We identified a five-cluster solution for binary indicators of past-month use and a six-cluster solution for frequency of use. The largest binary cluster (37%) included misuse of prescription opioids and use of alcohol. The second-largest cluster (15%) included misuse of prescription opioids, alcohol, marijuana, and "other" substances. Among those who used heroin, 36% used methamphetamine. In terms of frequency of use, the largest cluster among people who misuse opioid who used multiple substances (almost 40%) misused prescription pain relievers, alcohol, and marijuana infrequently. The second-largest cluster (23%) used marijuana almost daily and misused prescription pain relievers an average of 6.6 days. PWMO in a cluster of almost daily heroin use indicated use of methamphetamine, marijuana, and prescription opioids. Those who used methamphetamine, were using it more than 15 days a month.

Conclusions: We have developed reference measures of polydrug patterns among US household population and estimated their demographic characteristics. We identified clusters of high-risk polydrug use. These findings have implications for the development of prevention and treatment solutions in the United States.

Keywords: Cluster analysis; Latent class analysis; National Survey on Drug Use and Health (NSDUH); Opioid misuse; Polydrug patterns; Polysubstance.

Fig. A.1.

Class-Conditional Response Probabilities Output From the Latent Class Analysis Model for the Seven Substance Categories, 2017–2019 NSDUH. Note: MARIJ = Marijuana; METH = Methamphetamine; RX PR = Prescription Pain Relievers.

Fig. A.2.

Patterns of Past-Month Substance Use Prevalence Among PWMO and Multiple Substances; Results of Cluster Analyses with Four-Cluster Solution for the Seven Substance Categories, 2017–2019 NSDUH.

Fig. B.1.

Model-Estimated Number of Days Used/Misused Substances in Past Month for Each Profile from the Latent Profile Analysis, 2017–2019 NSDUH. Note: MARIJ = Marijuana; METH = Methamphetamine; RX PR = Prescription Pain Relievers.

Fig. 1.

Past-Month Use of Alcohol, Marijuana, and Remaining Drugs Among All People Aged 12 or Older Who Used Illicit Drugs or Alcohol in the Past Month (N = 83,302), 2017–2019 NSDUH Note: The Ns in this figure are unweighted counts. The percentages in this figure are weighted to the national population.

Fig. 2.

Past-Month Use of Alcohol, Marijuana, and Remaining Drugs Among PWMO and Multiple Substances Aged 12 or Older, 2017–2019 NSDUH Note: The Ns in this figure are unweighted counts. The percentages in this figure are weighted to the national population.

Fig. 3.

Patterns of Past-Month Substance Use Prevalence Among PWMO and Multiple Substances, 2017–2019 NSDUH.

Fig. 4.

Patterns of Polydrug Use Among PWMO and Multiple Substances Based on the Number of Days Used Drugs in Past Month, 2017–2019 NSDUH.

Fig. 5.

Patterns of Substance Use Among PWMO and Multiple Substances Based on Number of Days Used Substance in Past Month Conditioned on People Who Used the Substance in the Past Month, 2017–2019 NSDUH.