https://data.pa.gov/Opioid-Related/Overdose-Information-Network-Data-CY-January-2018-/hbkk-dwy3
The opioid epidemic that has proliferated across the United States has hit Philadelphia especially hard, with the Kensington and Old Kensington neighborhoods being the city’s epicenters. To combat the increase in overdose deaths from opioids in September 2014 Pennsylvania made naloxone, a prescription medicine that reverses opioid overdoses and sold under the brand names of Narcan and Evzio, available under a standing order. However, it was not until 2018 with the creation of the Overdose Information Network Data dataset by the Pennsylvania State Police that policy makers and the public could see how often naloxone was being administered to save lives. This dataset is available for free to the public via download, in various file formats (CSV, JSON, XML), or through Open Data PA and Socrata API. Developers and researchers are highly recommended to create a Socrata account and register an API token, however, it is not required but requests are severely limited and throttled for unregistered users.
The data itself is voluntarily submitted to the PA State Police by state, county and local first responders (police, fire, ems, etc.). The State Police update the dataset monthly. As of the last update, which occurred April 16th, 2021, there are just under 26,000 records and 34 features in the dataset. The features capture information on the individual who overdosed, the suspected overdose drug, location of the overdose, who responded and naloxone administration (OpenDataPa has a full and comprehensive data dictionary). There is no personable identifiable information (PII) in the data set, however, information about an individual’s gender, age range, race, ethnicity, and county of residence are included. Privacy concerns and data security are always present surrounding data, especially healthcare related data, but since there is no PII in this dataset it is a moot issue.
The opioid epidemic has touched every community in the United States. Recent lawsuits and settlements brought by local, state, and federal governments against opioid manufacturers and drug distributors are beginning to fund the work needed to stem this epidemic. This dataset can be used by law enforcement, healthcare professionals, social workers, policy makers and the public to make sure these funds are being allocated in areas with the most need as well as to determine the efficacy of intervention and treatment programs. Further, this data set demonstrates the effectiveness of naloxone in saving lives as individual survival is over twice as likely when naloxone was administered than when it was not. While there is not historical data available prior to 2018 to objectively quantify how big of an impact the standing order for naloxone in PA was on overdose survival, should the order ever be rescinded or expire this data can serve as a baseline. Law enforcement and public health officials could also use this data to identify spiked batches of drugs. A monthly update cadence probably makes this hard to pinpoint in real time and a more frequent update cadence would be needed to make the data actionable for this use case. One of the issues with this data is that it is self-reported by first responders. As the workload on first responders has increased in the past year plus due to COVID19, racial unrest and increases in crime across the state the bandwidth for other tasks, especially data collection may decrease. Secondly, since these stats are self-reported they are underrepresenting the issue, by how much may be difficult to measure if reporting is not mandated by local or state governments or agencies. Lastly, the lack of historical data prior to 2018, as previously mentioned, is a downside of this data. The opioid epidemic existed in Pennsylvania prior to 2018 and having the data is an asset for those involved in public health and law enforcement, having more historical data just enrichens the view and gives more context for current trends.