Alabama's Opioid Crisis, 2006–2012

Introduction

This site displays data on Alabama's opioid crisis during the period from 2006–2012, when the crisis was spinning out of control Our project is based on data from the US Drug Enforcement Agency (DEA) ARCOS database for this time period, which was obtained by the Washington Post after a year-long legal struggle, and published in July 2019. ARCCOS refers to the DEA's Automation of Reports and Consolidated Orders System.

The raw data that we obtained from the Washington Post for Alabama consists of transaction records. Each record gives data on a purchase of opioid pills by a pharmacy from a distributor, and includes the name and address of the pharmacy, the name and address of the distributor, the name and address of the manufacturer, the type of opioid purchased, the dosage unit (for example, 5 mg) and the total dosage (number of pills). The raw data for Alabama consists of nearly four million records, making up a file of nearly two gigabytes in size.

The primary metric used by the Washington Post in its analysis of the opioid crisis is the number of pills per person per year (abbreviated PPY in our project) for a given population—either a county or an entire state. We also use this metric, but in addition we compute the number of pills per household per year (PHY). Moreover, we aggregate the data not just for counties but also for several other geographic population groups:

• Communities. A community is classified by the US Census Bureau as either a city, a town, or a place. Communities are also classified as either rural or belonging to a Metroppolitan Statistical Area (MSA). A community belonging to an MSA is considered urban.
• Counties. Alabama has 67 counties.
• Core-Based Statistical Areas (CBSAs). A CBSA consists of a county or a group of adjacent counties with close economic and commuting ties to a large metropolitan area. Alabama has 13 CBSAs. For this project, the non-CBSA counties are grouped together into a rural area, so that the CBSAs together with the rural area partition the state.
• Primary Care Service Areas (PCSAs). The PCSAs are the same ones used in our study of primary care in Alabama. Each PCSA is based on a centroid: a single general admission hospital in most cases, or a cluster of such hospitals in a few large urban areas, or the center of a small town without a hospital in a few rural cases. The PCSA corresponding to the centroid consists of the region of the state that is closer to that centroid than any other, as measured by drive time through the road network. We have 79 centroids distributed throughout the state so that the PCSAs generally correspond to drive times of 30 minutes or less to the centroids. Like the counties, the PCSAs partition the state.

We also give ancillary data for our various geographic populatiion groups: rural/urban classification, total population, number of households, area, number of pharmacies, total dosage, and total records

User Interface

Our data are displayed in the form of interactive maps and interactive tables. The maps generally have point or area layers that can be added or removed with the layer control on the map. The user can zoom in and out of a map and move about in the usual way. Zooming in reveals additional features such as small towns and then streets and roads. A click on the home button returns the map to its origional location and scale. The zoom to area button allows the user to zoom to a selected rectangular area of the map. Clicking on an object in a layer shows summary data for that object.

An interactive table can be sorted according to a data field by clicking on the name of the field in the header. The table can also be searched for a specified text string, and the data can be copied to the clipboard in tab-separated text format or downloaded in various formats.

For area data (counties, PCSAs, and the US), table rows can be selected by clicking and then the corresponding areas are highlighted in the map.

Background

The following basic statistics are from the Washington Post:

• There have been over 200,000 deaths due to opioid overdose since 1996. Over 100,000 deaths occurred during the period covered by this study, 2006–2012.
• Over 76 billion opioid pills were distributed throughout the US from 2006–2012.
• The national death rate from opioids was 4.6 per 100,000 persons, but in counties with the highest distribution of opioids, the rate was two or three times higher.

Some general statistics for Alabama are hard to come by, but here are a few:

• In 2012, Alabama had the highest rate of opioid prescriptions in the country—1.438 prescriptions per person.
• From 2006 to 2014, there were 5128 deaths due to drug overdoses, a rate of about 11.87 per 100,000 persons per year. Most are believed to be due to opioids, but Alabama's reporting is insufficient.