This paper from 2012 uses data from 2008 to study the Geographic Variation in Opioid Prescribing in the U.S.
Some excerpts (refer to the original paper for full details):
Some excerpts (refer to the original paper for full details):
.....Geographic variation in prevalence of prescribed opioids is large, greater than variation observed for other healthcare services. Counties having the highest prescribing rates for opioids were disproportionately located in Appalachia and in Southern and Western states. The number of available physicians was by far the strongest predictor of amounts prescribed, but only one-third of county variation is explained by the combination of all measured factors......
Wide geographic variation that does not reflect differences in the prevalence of injuries, surgeries, or conditions requiring analgesics raises questions about opioid prescribing practices. Low prescription rates may indicate under-treatment, while high rates may indicate overprescribing and insufficient attention to risks of misuse.....
...Regression analysis was conducted to identify the correlates of prescribing prevalence at the county level....
Table 4
Estimated Association of County-Level Attributes and Amounts of Prescription Opioids (Morphine Equivalents) Prescribed During 2008
| Coefficient | SD | |
|---|---|---|
| Resident Population Characteristics | ||
| County population size | 0.162 *** | 0.026 |
| Percent of county classified as urban | 0.095 * | 0.043 |
| County is zero percent urban | 0.530 | 0.277 |
| Percent college graduates | −0.248 | 0.216 |
| Percent completed HS, but not college | −0.192 * | 0.096 |
| Education variable not reported | −0.023 | 0.058 |
| Percent white non-Hispanic | 0.844 *** | 0.072 |
| Percent African American | 0.046 *** | 0.010 |
| Percent Hispanic | −0.015 | 0.020 |
| Persons in poverty per 1,000 residents | 0.336 *** | 0.083 |
| Household income inequality (Gini coefficient) | 0.100 | 0.689 |
| Percent <65 insurance="" old="" td="" without="" yrs=""> | 0.220 ** | 0.079 |
| Percent Medicaid eligible <65 old="" td="" yrs=""> | 0.102 | 0.077 |
| Health Care Utilization & Rx Monitoring | ||
| Surgeries per 1,000 residents | 0.023 | 0.014 |
| In-patient days in short-term hospitals per 1,000 residents | 0.020 | 0.022 |
| Emergency department visits per 1,000, short-term hospitals | 0.038 | 0.029 |
| Prescription drug monitoring program in state, 2008 | 0.067 | 0.039 |
| Availability of Prescribers | ||
| Active MDs per 1,000 residents | 0.347 *** | 0.035 |
| Surgeons as percent of all MDs | 0.083 ** | 0.027 |
| Psychiatrists as percent of all MDs | −0.060 * | 0.024 |
| Pediatricians as percent of all MDs | 0.054 * | 0.027 |
| Emergency MDs as percent of all MDs | −0.001 | 0.020 |
| Constant | −0.191 | 1.165 |
Notes.
*p < .05,
**p < .01,
***p < .001.
Readers may remember this story about increased mortality rates among non-Hispanic whites, with these graphs:
Mortality Rate for 45-to-54-Year-Olds, By Country
Quote: "Deaths from drug overdoses among people aged 45 through 64 increased 11-fold between 1990 and 2010, and nearly 90 percent of people who try heroin for the first time these days are white. (Most found heroin through prescription painkillers, which treat the chronic pain this age group struggles with, but can also make it worse.)"
IMO, these tragedies are connected.
Readers may remember this story about increased mortality rates among non-Hispanic whites, with these graphs:
Mortality Rate for 45-to-54-Year-Olds, By Country
 | |||
| Key: U.S. White non-Hispanics (USW), US Hispanics (USH), and six comparison countries: France (FRA), Germany (GER), the United Kingdom (UK), Canada (CAN), Australia (AUS), and Sweden (SWE). (PNAS) Causes of Increased Mortality |
 |
| Among white non-Hispanics ages 45–54 (PNAS) |
IMO, these tragedies are connected.
