In the United States, 20.2 million adults or 8% of the population is estimated to suffer from a substance use disorder (SUD). SUDs include a wide array of substances such as alcohol, opioids, methamphetamine, and other substances and are characterized by an inability to decrease use, despite severe social, economic, and health-related consequences to the individual. In 2017, the US Department of Health & Human Services declared a public health emergency to combat what has been termed as “the opioid epidemic” and highlighted five cricital strategies:

1. Improving access to treatment and recovery services;
2. Promoting use of overdose-reversing drugs;
3. Strengthening our understanding of the epidemic through better public health surveillance;
4. Providing support for cutting edge research on pain and addiction; and
5. Advancing better practices for pain management.

Computational support may contribute to each of these strategies by mobilizing a new set of systems, algorithms, and tools to understand and combat substance use disorders. These technologies may provide scalable and accessible complementary approaches to traditional methods and services.

The goals of this workshop are to:

• Create a cohort of junior investigators and established mentors who use relevant methods, leverage relevant technologies, or work with relevant communities to address SUD challenges
• Foster the creation of interdisciplinary research working groups, including investigators using computing, design, policy, and clinical research approaches
• Enable future partnerships and exchange of ideas by connecting academics, industry practitioners, government agencies, service providers, and amplifying the voices of individuals experiencing SUDs 
• Jointly initiate a white paper and a special issue of the Journal of the American Medical Informatics Association, articulating key research and technology opportunities and risks in the space of computational support for SUDs

Some areas of opportunities for computational support to address the substantial challenge of SUDs include:

Prevention: Prevention includes universal, selective, and targeted interventions which can include programs to reduce social and behavioral risk factors that contribute to the development of problem behaviors (universal), efforts to address individuals with significant risk factors for the development of SUDs (selective), and efforts to assist  those with high rates of drug-related harm prior to the occurence of a SUD (targeted). Computational support might contribute to identifying those at risk in order to better identify individuals for selective and targeted interventions, reduce costs to make preventive efforts, especially universal prevention, more scalable and cost-effective, or improve the long-term monitoring of individuals who undergo preventive interventions. 

Detection: Detection involves ways to identify individuals who may be at risk of experiencing SUDs, before individuals even acquire full-fledged symptoms of SUDs. In recent years, there has been considerable success in employing machine learning and other statistical predictive techniques to both explain and forecast imminent risk, leveraging a variety of naturally occurring digital and physical behavioral traces. Early detection may also be helpful in the course of treatment of SUDs, as it can reveal risk markers of adverse episodes and outcomes like relapse and vulnerability. Better detection techniques may enhance prevention and treatment of SUDs, improve individuals’ functional outcomes, and reduce the burden of these illnesses in the broader population. 

Treatment: Effective treatment approaches for SUDs typically address “the whole person,” including support during initial withdrawal, providing medical and/or psychotherapeutic treatments, and developing skills for preventing relapse. Some examples of computational opportunities for supporting treatment include designing systems to support the medical team’s decision-making, developing personalized treatment plans, and enhancing the conventional mental health interventions.

Recovery: Recovery refers to the long-term process by which an individual maintains abstinence and regains control over their life. The current prevalent model of SUDs are as a chronic condition with relapse rates of 40-60% (comparable to asthma, hypertension, etc.). Thus, long-term recovery requires similar considerations as other chronic conditions, including adjusting treatment plans as necessary, facilitating behavior change, and connecting with social support. Among other opportunities, computation provides a compelling way to scale long-term access to care, incentivize and track behavior change, and support the development and maintenance of social support networks.

These four areas each require collaboration across disciplines both within and outside of Computer Science to develop novel and effective solutions to address the challenges of SUDs.