Investigate Possible Explanations

Once you’ve gathered potential explanations for an overdose, the next step is to evaluate them systematically. Use data to test your hypotheses, uncover patterns, and determine the most likely drivers of the outbreak.
Analyze the Data

Start with descriptive statistics to summarize what you know. Break down the data by key variables such as:

  • Demographics (age, race/ethnicity, sex)
  • Geography (neighborhood, ZIP code, proximity to known hotspots)
  • Substance involved (e.g., fentanyl, xylazine, counterfeit pills)
  • Time (hour of day, day of week, seasonality)

Use visualizations (e.g., heat maps, line graphs) to identify clusters or alerts.

Test Risk Factors

Test Risk Factors

Use statistical methods to explore potential associations and risk factors:

  • Do certain age groups or neighborhoods show a disproportionate increase in cases?
  • Are overdoses linked to a specific drug formulation or combination?

Note: Lack of statistical significance doesn’t mean a factor isn’t important—use qualitative data and local knowledge to complement your findings.

If your data allows, consider formal study designs such as:

  • Case-control studies to identify exposures linked to overdose
  • Cohort studies to track individuals at high risk over time
Analyze Multiple Data Sources

To fully understand the anomaly, analyze and triangulate data from a variety of sources:

  • Interviews with people who use drugs
  • Emergency medical services (EMS) and ambulance run records
  • Emergency department (ED) data
  • Police reports and arrest data
  • Law enforcement drug testing results
  • Medical examiner or coroner reports
  • Medical record abstractions
  • Site visits and community observations

Cross-referencing data sources can help verify trends and provide context behind the numbers.

Real-World Examples: Outbreak Investigation in Action

Explore how other jurisdictions have evaluated and investigated overdose anomalies:

For more real-world examples, view Use Cases in the Resources page.

Mapping and Responding to Overdose in Rhode Island

Rhode Island is divided into 11 overdose regions, each with its own anomaly threshold based on the previous year of data. RIDOH tracks these regions daily and uses a Levels of Response plan to guide the resources deployed and the partners and outreach involved, ranging from alerts for sudden spikes to strategies for sustained high burden. When thresholds are exceeded, RIDOH activates its Overdose Spike Alert Notification System, which alerts local leaders, public safety, healthcare providers, and community partners via email list serv about a sudden increase in nonfatal overdoses. Responses may include expanded outreach, harm reduction efforts, and emergency community meetings. Higher rates of overdose burden in a region warrant a greater response level. For instance, a region with a ‘Sustained High Rate of Overdose Burden’ may involve community partners for increased outreach and convene an emergency community meeting. This response plan drives timely action and helps integrate overdose anomaly data into real-time public health action. 

RIDOH established a 7-day threshold system to send alerts when an overdose region’s nonfatal overdoses exceeded a level set at three standard deviations above the annual average count for that region. RIDOH developed a more prospective metric that examines overdose rates over the past six months. This approach compares each region’s overdose rate per 100,000 residents to the statewide rate, highlighting areas with sustained burden. By shifting focus from short-term spikes to longer-term trends, RIDOH can better target resources and prevention strategies where they’re most needed. 

For more information: https://ridoh-drug-overdose-surveillance-iss-rihealth.hub.arcgis.com/ 

 

Continue The Toolkit

< Previous: Generate Possible Explanations

Next: Mitigation >

Prepare

Respond

Evaluate

Resources