Develop Case Definitions
Last Updated: March 2022
Cases of overdose that will be part of an investigation should ideally include everyone involved in the event, whether they survived the overdose or not. Morbidity data will naturally only include those who survived the overdose.
Event Case Definition
This section is distinctive from determining if a true outbreak has occurred. If your state has a system that generates alerts automatically, each should be assessed for data quality and other potential errors that may lead to the false generation of an alert (ex: a usual or expected number of overdoses followed by a period of no overdoses).
In an overdose outbreak cause investigation, any available specifics such as substance ingested or event circumstances should be considered along with the date and time interval spanned by the cases and distribution of patient ages, genders, and locations. Poison center call records often have these details, and sometimes free-text fields in EMS and ED records include them. The following steps can be useful in doing so:
- Assess similarities in demographics and characteristics (age, sex, race/ethnicity, area of residence, prescription history, prescribers, where overdosed)
- Obtain records of interest: EMS, ED and hospital records, law enforcement reports, toxicology testing results, and death certificates, if applicable
- Patient disposition (i.e., mortality and morbidity), if available, can be leveraged to assess lethality
- If narratives are available, look for indications that the same substance was used
- Check retrospective data to see if there have been previous instances of the same event (same county, same level, etc.)
- Check neighboring areas, such as counties
- If available, look at granular geographic data for potential clustering and proximity to a major highway
- Use mapping to identify potential clusters of cases in the jurisdiction
Establish mechanisms such as windows showing line listings of related cases, so that linkages among these cases can be investigated. This can be used to identify who overdosed, when they overdosed, and where they overdosed to identify hotspots. Consider leveraging data submitted to the state’s SUDORS program to track data on those who overdosed. This can also be useful to identify communities at risk for overdose outbreaks as well.
Engage stakeholders at this point to make them aware of the situation so they can identify other cases and begin to take public health actions.
Conduct enhanced surveillance using data available from PDMPs, if appropriate. Public health response team should work with the requesting jurisdiction to understand what data are available. If data are available, the jurisdiction should identify high-risk individuals for follow-up.
Emergency Department Case Definition
Applying definitions to ED data will include grouping chief complaints and/or diagnosis codes into queries to classify visits as overdoses. For example, in the NSSP, a chief complaint of “overdose” or “heroin overdose” or a T40 drug poisoning ICD-10-CM code may be sufficient, but many overdoses may be detected from a combined symptom profile, such as ‘loss of consciousness’ AND opioid abuse ICD-10-CM codes in the F11 category together classify a visit as an overdose.
Emergency department data coding should be interpreted cautiously, and standard case definitions should be considered. Most ED visits classified as overdose are not lab-confirmed by toxicology. Also in most systems, “cases” of overdose are not lab-confirmed, fully vetted individually investigated cases. Overdose case detection is not an exact count of all those patients.
ED ICD-10-CM Injury Codes for Drug-related Poisonings
The related ICD-10-CM codes are available in CDC’s 2018 Annual Surveillance Report of Drug-Related Risks and Outcomes. Hospital discharge ICD-10-CM definitions used will differ from syndromic definitions, which include additional unstandardized text to assist in identification of an overdose.
EMS Nonfatal Opioid Overdose Case Definition
Jurisdictions may already have overdose case definitions in place for identifying probable opioid overdoses. Most of these definitions utilize key EMS data elements listed in the table below:
| Data Source | NEMSIS v3 Variable | NEMSIS v3 Variable Definition | NEMSIS v3 Variable Response Options |
|---|---|---|---|
| Patient Care Report Narrative | eNarrative.01 – Patient Care Report Narrative | The narrative of the patient care report (PCR). | State, E, 0:1 String values 1-10000 characters |
| Medication administered by EMS | eMedications.03 – Medication Given | The medication given to the patient |
National, State, R, 1:1
|
| Response to medication | eMedications.07 – Medication Given | The patient’s response to the medication |
National, State, R, 1:1
|
| Primary impression (including ICD-10-CM codes) | eSituation.11 – Provider’s Primary Impression | The EMS personnel’s impression of the patient’s primary problem or most significant condition which led to the management given to the patient (treatments, medications, or procedures). | National, State, R, 1:1 ICD-10-CM codes: (R[0-6][0-9](\.[0-9]{1,4})?|(R73\.9)|(R99))|([A-QSTZ][0-9][0-9A-Z])((\.[0-9A-Z]{1,4})?) |
| Secondary impression | eSituation.12 – Provider’s Secondary Impressions | The EMS personnel’s impression of the patient’s secondary problem or most significant condition which led to the management given to the patient (treatments, medications, or procedures). | National, State, R, 1:M String values based on ICD-10-CM codes, 1-255 characters |
| NOTE: This list is meant to be a tool to highlight NEMSIS data elements that may be useful for opioid overdose surveillance. It is not exhaustive in nature and should not replace any standing jurisdictional guidance. |
Data Element Key
| National and State Element Indicator | NEMSIS v3 Variable Response Options |
| National = Yes, is an indication that the data element is required to be collected at the local EMS agency level and submitted to the state State = Yes, is an indication that the data element is recommended to be collected at the “State” level. |
M = mandatory data field, does not allow NOT values R = must be completed and allows NOT values [7701001 – Not Applicable; 7701003 – Not Recorded] E = recommended field 0:1 = element is not required and can occur only once 1:1 = element is required and can occur only once 1:M = element is required and can repeat multiple times |
[Text Wrapping Break] CSTE, in conjunction with national EMS partners conducted a series of field tests and validations to determine standard queries for determining nonfatal opioid-related overdoses. As noted in the table above, some data elements are required for reporting to the National EMS Information System (NEMSIS). Further, some elements are coded, or have a drop-down selection menu associated when EMS personnel select these options. The guidance recommends jurisdiction utilize a combination of six coded elements and selected free text elements. The coded data elements are listed below:
- Provider’s Primary Impression
- Provider’s Secondary Impression
- Primary Symptom
- Other Associated Symptoms
- Medication Administered
- Response to Medication Administered
For the free-text elements, CSTE recommends using selected terms from the narrative field, though this recommendation depends on each jurisdiction’s access to that field, available resources for any necessary analysis, and local restrictions regarding use of the narrative. In the CSTE field test project, specific terms were recommended and revised, and these terms yielded improvements without sophisticated natural language processing methods.
NOTE: CSTE Nonfatal Opioid Overdose Standardized Surveillance Case Definition will be updated with a standard guidance for determining nonfatal opioid-involved cases from EMS data.
NSSP Syndrome Definitions
NSSP has developed case definitions that can be used to assist state, local, tribal, territorial, and federal public health practitioners in monitoring emergency department (ED) visits for suspected overdoses. These queries include four that are drug-specific (stimulants, heroin, opioid, and benzodiazepines, and one that includes all drugs, illicit and prescription. The queries are updated regularly by the NSSP Community of Practice through a vetting process. Current versions of the queries are linked below.
Additional information about how to use and build ESSENCE queries is located on the CDC NSSP website and the CSTE Knowledge Repository.
Mortality Data
There are a number of data sources that are useful in examining fatal overdose cases including vital statistics data, SUDORS, and ME/C’s reports.
National Vital Statistics System Data
Drug overdose deaths can be identified in the National Vital Statistics System multiple cause-of-death mortality file or locally coded state vital statistics files using ICD-10-CM, based on ICD-10-CM underlying cause-of-death codes X40–44 (unintentional), X60–64 (suicide), X85 (homicide), or Y10–Y14 (undetermined intent).
Among deaths with drug overdose as the underlying cause, the type of drug or drug category is indicated by ICD-10-CM multiple cause-of-death codes. Commonly used codes include: opioids (T40.0, T40.1, T40.2, T40.3, T40.4, or T40.6); natural/semisynthetic opioids (T40.2); methadone (T40.3); heroin (T40.1); synthetic opioids other than methadone (T40.4); cocaine (T40.5); and, psychostimulants with abuse potential (T43.6). These ICD-10-CM definitions are taken from Overdose Deaths Involving Opioids, Cocaine, and Psychostimulants — United States, 2015–2016.
Some deaths involved more than one type of drug; these deaths are included in the rates for each drug category. Therefore, categories are not mutually exclusive. Also, small wording differences on death certificates can lead to misclassification of opioid overdose deaths as not an overdose death.
State Unintentional Drug Overdose Reporting System (SUDORS)
The SUDORS system captures full toxicology of opioid overdose deaths, including whether a drug was listed as contributing. Participating health departments identify cases with ICD-10-CM codes or through identification systems used by ME/Cs. For instance, some deaths may only list drug overdose as the underlying cause of death but review of the ME/C report indicates an opioid was the drug causing the death. See example: SUDORS analysis using full toxicology data.
Medical Examiner or Coroner (ME/C) Reports
In some states or cities, ME/Cs produce their own reports on drug overdose deaths based on their investigation findings or forensic toxicology findings. These reports often can provide much better detail than reports based on the death certificate. For instance, Florida and Virginia have these types of reports.
Toolkit Resources
Get more insights by using our toolkit resources.
Go to Resources
Glossary
Learn the definition of the key words being used.
Go to Glossary
Thanks
Thank you, to all of our contributors.
View our contributors
the public’s health.
Contact Us
Have any questions or recommendations, you can contact us at overdose@cste.org