Relationships of Prescription Opioid Sales to Opiate Treatment Admissions and Overdose Deaths: A Reanalysis of Relationships -- United States 2010-2019

ABSTRACT Background: In 2016, the United States Centers for Disease Control and Prevention (CDC) issued CDC Guideline for Prescribing Opioids for Chronic Pain. The guideline used data from 1999 to 2010 that related Opioid Sales to Opiate Treatment Admissions (a measure of drug addiction rates) Opioid Deaths. The CDCs analysis indicated that opioid prescriptions are directly correlated1-8 with drug addiction and overdose mortality outcomes. As a result, prescription opioid sales have been curtailed, making it difficult for sufferers of severe chronic pain to access an effective treatment option. In this paper, the relationships between opioid sales, drug addiction, and opioid deaths are reexamined using data from 2010 to 2019. Methods: Linear regression models were fit to each response separately. Opioid sales (measured as MME per capita) was the independent variable. Total overdose deaths (TOD), any opioid overdose deaths (AOD), prescription opioid overdose deaths (POD) and opiate treatment admissions (OTA) were the dependent, response variables. The models were assessed using three criteria: the statistical significance of the model (Overall P-Value), the quality of the fit (R2), and the sign of the slope coefficient. Results: The analyses revealed that the direct correlations (significant, positive slopes) reported by the CDC based on data from 1999 to 2010 no longer exist based on data from 2010 to 2019. The relationships either have reversed (i.e., significant negative slopes) or are non-existent (no significant model). Conclusions: The basis for the CDC guideline no longer holds. The guideline should be updated or annulled based on the current relationships that have existed for a decade. This is especially relevant for people suffering from severe, chronic pain, whose access to prescription opioids has been severely curtailed under the current guideline and to ensure that government resources are directed to an effective approach to reducing overdose deaths.


44
In 2015, the U.S. Department of Health and Human Services (HHS) declared "There is a 45 clear correlation between opioid prescribing rates and overdose death rates in the United 46 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 1, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022 Deaths (AOD), Prescription Opioid Deaths (POD), Opiate Treatment Admissions (OTA) and 48 prescription opioid sales (POS) from 1999-2010 (see Figure 1) led the CDC to conclude that 49 prescribed opioids are the determinant for overdose deaths and opiate/opioid treatment 50 admissions. 1-8 Since the 2016 "CDC Guideline for Prescribing Opioids for Chronic Pain" 51 (guideline), cutting prescription opioid sales has been CDC's, DEA's, legislative policy makers ', 52 and practitioners' solution to cut overdose deaths and opiate treatment admissions. 10 reports "the CDC Guideline has harmed patients" 14 , "72% of pain medicine specialists said that 61 they-or their patients-have been required to reduce the quantity or dose of medication they 62 have prescribed" 14 as a result of the guideline. 63 "Almost 18 million Americans are currently taking long-term prescription opioids" 15 to 64 treat chronic pain. A "Health and Human Services Pain Management Best Practices Inter-65 Agency Task Force has determined that tens of millions of Americans rely on legal prescription 66 opioids to treat acute or severe chronic pain, including pain arising from cancer as well as 67 terminal or degenerative illnesses. The Food and Drug Administration (FDA) long ago approved 68 opioid medications for these purposes, and doctors throughout the country lawfully prescribe 69 them." 16 92% of physicians and patients believe that opioids reduce pain and 57% report better 70 quality of life. 17 71 72 With drug overdose deaths "growing, inexorably and exponentially, for four decades" 18 to 73 a record 93,000 in 2020; FY2019 federal funding to combat the opioid crisis increasing to a 74 record $7.6 billion 19 and "drug overdoses are now costing the United States approximately $1 75 trillion annually", 20 it is critical that public health policy guidance not be based on out-of-date 76 and/or misleading information. 77

78
The present paper updates correlations between POS to OTA, POD, AOD and TOD 79 from 2010 to 2019 and shows that the direct relationships that existed from 1999 to 2010 are no 80 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 1, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022 longer present --in fact, in some instances the relationships have reversed. Consequently, we 81 find that continuing the policies of cutting prescription opioid sales to reduce TOD, AOD, POD, 82 and OTA is unfounded. sources necessarily contains inaccuracies. It is widely understood that it is impossible to 90 ascertain the specific cause of an overdose death in all cases. This may be due, for example, to 91 poly drug use with alcohol, and an inability to reliably determine the source of opioids detected 92 in postmortem blood toxicity screens (e.g., prescription vs. illicit), among other confounding 93 issues. ,21,22 However, the authors have employed the same data sources that CDC guideline 94 appears to be based upon. As such, the results of analyses presented here are at least as 95 reliable as what the CDC would be able to obtain from their own analyses, if they chose to 96 undertake them. Thus, the following sources have been applied. Consistent with the methods used by the CDC, simple linear regression models were fit to the 168 data. Separate models were fit to each of the four dependent variables (TOD, AOD, POD, and 169 OTA) using Annual Opioid Sales (i.e., MME per Capita) as the independent variable. Two 170 models were fit to each dependent variable. One model covered the years presented in the 171 original CDC chart (for which MME per Capita data were available) (2006)(2007)(2008)(2009)(2010)  Annual Prescription Opioid Sales (i.e., MME per Capita) are either non-existent or significantly 201 negative/inverse (Figures 2 and 3). Results for all regression models are presented in Table 1. 215 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 1, 2022. Starting in 2010 opioid MME per Capita (POS) does not have a "clear correlation" 9 or 221 move "in parallel" 3 or "in lockstep" 2 with OTA, POD, AOD or TOD. The relationship changed 222 from direct to inverse. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 1, 2022. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 1, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 1, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 1, 2022.  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 1, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022