Portal de Periódicos da CAPES

Need for comprehensive and timely data to address the opioid overdose epidemic without a blindfold

2022; Wiley; Volume: 117; Issue: 8 Linguagem: Inglês

10.1111/add.15957

1360-0443

Nora D. Volkow, Redonna K. Chandler, Jennifer Villani,

Poisoning and overdose treatments

Communities need real-time, disaggregated data to identify, which groups of individuals are most at-risk so that prevention and treatment services can be targeted efficiently to address the opioid crisis at the local level. We are amid a public health crisis. Over 88 000 people died worldwide because of opioid use disorder (OUD) in 2019 [1]. That year, the rate of deaths attributed to OUD per 100 000 individuals in the United States (US) was 14.43, which far surpassed countries with the next highest rates of 8.0 in Estonia, 5.11 in Canada, 5.01 in Lithuania, 4.93 in United Arab Emirates and 3.96 in Finland [1]. In the United States, the increased presence of fentanyl in the illicit drug markets account for the continued rise of deaths from opioid overdoses estimated at nearly 80 000 in 2021 [2]. These preventable deaths demand a coordinated response to deliver evidence-based prevention and treatment strategies to address substance use disorders and overdoses. However, like any public health crisis, real-time data are needed to understand who is at risk for overdosing, what services are available and who has access and how to provide broader availability of evidence-based prevention and treatment interventions to all in need. Although abundant administrative data are collected at national and local levels, severe constraints in availability, timeliness, harmonization, sharing and linkage limit its usefulness to develop and guide effective response strategies. We are mostly fighting the drug overdose crisis blindfolded. The value of real-time data was made abundantly clear during the coronavirus disease (COVID)-19 pandemic. Data reporting lags prevented us from understanding the impact of mitigation strategies on drug use, overdose deaths and access to lifesaving interventions like naloxone, medications for opioid use disorder (MOUD), or recovery support services. It was not until 2021 when we fully understood the unprecedented increase in drug-related overdose deaths that occurred early in the pandemic. Access to real-time data on drug overdose deaths are particularly problematic for they rely on medicolegal investigations, toxicology testing and death certifications that take months to complete. Reporting is so lagged that public health officials must rely on prior year data on overdose deaths to inform current response strategies. This approach is insufficient in an ever-evolving drug market with more potent substances driving increases in overdose deaths and the urgent need to mount interventions to prevent them. The situation is further compounded at a global scale by differences between countries in post-mortem toxicology testing and the lack of harmonized coding practices [3]. Although vast amounts of administrative data exist at local and national levels regarding substance use disorders and care, these reside in unlinked siloed systems; lack demographic information (i.e., age, sex and race/ethnicity) needed to effectively monitor trends; and are either not available to health officials or access is delayed for months to years. Incomplete information interferes with the optimal deployment of resources to provide needed interventions to those at highest overdose risk. For example, if 18- to 24-year-olds are experiencing more opioid overdoses, then an effective strategy should provide naloxone and other harm reduction strategies in venues these individuals frequent like college campuses, vocational training programs, concerts, cafes, sports events, etc. If data show a higher incidence of neonatal opioid withdrawal syndrome, then efforts should target linking pregnant and post-partum women to MOUD through their obstetricians and gynaecologists. Additionally, the absence of demographic information is not only a major blind spot for public health surveillance, but one that perpetuates structural racism and constrains efforts to achieve health equity [4]. Specifically, in the United States, some racial/ethnic groups are disproportionately dying of drug-related overdoses with the highest rates observed among American Indian and Alaska Native individuals and the fastest rate of rise among Black/African American individuals [5, 6]. The lack of data consigns us to limited explanations and leaves us wondering if these differences are because of exposure of these groups to a more potent drug supply or greater barriers to naloxone access or MOUD treatment. We do not have specific answers regarding why these groups do not seek medical care (e.g. structural racism, economic inequalities, or fear of stigma). Innovation in data collection is needed, which should include timely information on emerging drugs of abuse and data from non-traditional settings with high numbers of people who use drugs like the justice system and homeless shelters. This could be achieved via rapid drug testing by first responders, in emergency departments and in jails. Some jurisdictions around the world have deployed new technology to surveil community wastewater for drugs of abuse with high sensitivity [7-9]. Fatal overdoses are mapped in Italy with details on the decedent's age, sex, nationality and type of drug based on daily scans of online national and local newspapers [10]. In the United States, the Overdose Detection Mapping Application Program (ODMAP) links first responder data to a mapping tool to provide near real-time location data of suspected overdoses to public safety and public health agencies [11]. In addition, machine learning methods could be used to estimate fatal drug overdose surveillance data more rapidly and with high positive predictive value and sensitivity [12]. Efforts are also underway to successfully link health data collected by the justice system to community organizations providing substance abuse treatment in the Justice Community Opioid Innovation Network in the United States [13]. These advancements in data collection can provide communities with more rapid information on where to reach at-risk individuals. Some states in the United States have successfully enhanced their drug overdose surveillance by enacting laws requiring faster and more complete reporting. For example, Arizona requires first responders and health care facilities to report all suspected opioid overdose events and deaths to the state health department within 5 days. The state uses this near real-time reporting to update their online data dashboard on a weekly basis showing counts of fatal and nonfatal overdoses stratified by individuals' age, sex, race/ethnicity and geographic location [14]. Kentucky improved the quality of its fatal opioid overdose data through a state law mandating all death investigations must use toxicology testing for controlled substances [15]. As a result, the frequency of which specific drugs are mentioned on the death certificates for drug overdose deaths increased and improved Kentucky's surveillance. Linking individual-level records across public health data systems has also been undertaken to gain a more complete understanding of the opioid crisis in some jurisdictions. An overdose cohort study in Canada linked data from hospital visits, physician visits, ambulance calls, prescriptions dispensed and deaths to understand risk factors for fatal and nonfatal overdose [16]. Another study in Australia linked data from hospital visits, mental health services, criminal justice and deaths to measure the frequency of adverse outcomes among individuals receiving MOUD [17]. These large-scale data linkage projects are very informative, but they are rife with ethical concerns about confidentiality and require complex data use agreements among data stewards. This approach may not be feasible in countries lacking universal health care access or where there is insufficient trust from the community to link health records. As with public health, community-based research studies also endure major challenges when population-level data are lagged, incomplete, or not interoperable. The HEALing Communities Study (HCS) is a cluster randomized controlled trial launched in 2020 to test whether community-tailored action plans and the provision of local data on OUD helps communities reduce opioid overdose deaths and increase individuals with OUD in treatment [18]. Sixty-seven communities highly affected by the opioid crisis in four US states are participating in the study. The study requires research sites to have data partnerships to collect and monitor a vast array of administrative and de novo data standardized across the four states. Even with a high level of resources, there are still major gaps and delays in administrative data that impact study implementation (Table 1). For example, we have very little information on who is receiving naloxone kits from pharmacies or state health departments. We also cannot describe the group of individuals for whom ambulances and emergency medical services (EMS) personnel respond and administer naloxone and depending on the state, we have found that race and ethnicity information from government-funded insurance claims data can be 30% incomplete, data on overdose events are lagged up to 18 months and despite herculean efforts toward improvement, data on drug-related overdose deaths remain lagged up to 11 months. Resolving this unprecedented drug overdose public health crisis will require a data-driven approach like the world has built for COVID-19. In the United States, states and municipalities quickly developed public health surveillance systems and online data dashboards to track new COVID-19 cases, hospitalizations, vaccinations and deaths daily. These data included demographic information to help monitor trends by age, sex and race/ethnicity thereby enhancing the ability to deploy prevention strategies for highly impacted subpopulations. It is time to realize the power of a parallel data system to address drug overdose deaths. Localities could track real-time data on drug overdose events and deaths, including type of drug involved, use of naloxone, MOUD and recovery support services with necessary demographic information about those impacted. Giving communities the ability to understand who needs prevention and treatment services will remove the blindfold and ensure efforts and resources are finally targeting those who are most at-risk. None. Nora Volkow: Conceptualization, writing, editing. Redonna K. Chandler: Conceptualization, writing, editing. Jennifer Villani: Conceptualization, writing, editing.