Association ≠ Causation
Myron Yaster MD, Elliot Krane MD, and Steven J. Weisman MD
Today’s PAAD reviews an article published by Dr. Tony Anderson and his colleagues at Stanford University looking at adolescents with Prolonged Opioid Use after Surgery (POUS).1 Dr. Elliot Krane and I previously reviewed an article by the same group that looked at machine learning as a predictor (PAAD September 1, 2021
) of POUS.2 The basic premise is that POUS, defined as ≥1 opioid prescription in the 90–180 days after surgery is a marker of opioid abuse.3 Many of my concerns with that article remain the same in today’s article. I’ve asked Dr. Steve Weisman, a good friend and Medical director, Pain Management, Children's Wisconsin, as well as Dr. Krane to assist. Myron Yaster MD
Original article
Alice Kate Cummings Joyner, Michael R King, Conrad Safranek, Gomathy Parvathinathan, Elizabeth De Souza, T Anthony Anderson. Health Care Burden Associated With Adolescent Prolonged Opioid Use After Surgery. Anesth Analg. 2023 Feb 1;136(2):317-326. PMID: 35726884
“The long-term use of opioids is associated with poor health outcomes, including reduced quality of life, cognitive dysfunction, overdose, and death. Opioid-naive patients who are introduced to opioids in the period before, during, or after surgery are at risk of developing prolonged opioid use after surgery (POUS, defined as ≥1 opioid prescription filled between 90 and 180 days after surgery). The development of POUS has been associated with a significant increase in health care costs in the 3 to 6 months after surgery.”1 The authors “designed a retrospective analysis of a large, national insurance claim, de-identified database (Optum Clinformatics Data Mart Database, Optum Insight, Eden Prairie, MN) to determine the association between POUS and health care-related costs and utilization in adolescents. They hypothesized that patients with POUS would have increased total and individual health care costs and increased health care visits compared to patients without POUS”.1 Not surprisingly, they found that in their adolescent patients, POUS was associated with increased total health care costs and utilization in the 730 days after their surgical encounter“.1
We believe that the very premise of this article is invalid, namely, the presumption that the need and prolonged use of opioids following surgery is unjustified and leads to abuse and negative health care consequences. In fact, the very definition of “prolonged opioid use” as the filling of as few as one prescription in 6 months seems biased to the point of absurdity.
The analysis of healthcare visits as well as health care costs is also a tautology. In our medical economic system one cannot exist without the other, therefore emphasizing both merely amplifies one finding into two, lending the finding a false sense of truth by confirmation.
There are many reasons other than opioid abuse to explain why patients may need and therefore obtain one or more opioid refills, and why they may independently of that have increased visits to an ER or require inpatient hospitalization. Attributing this to abuse is, in our opinion, reductive and an example of confirmation bias, the interpretation of information by looking for, or interpreting, information that is consistent with an author’s preexisting beliefs. Furthermore and more importantly, we remind our readers that statistical association is not causation, and this paper offers a perfect example of this.
There are more probable explanations for why patients who receive one or more opioid prescriptions in a 6 month period following surgery have Increased health care utilization and costs. Here is a partial list:
· There persists moderate to severe pre-surgical pain that is not resolved or is increased or is compounded by the surgery (e.g. severe trauma, deep infection, malignancy, burns, spinal cord injury, amputation of a limb malignancy)
· There are painful surgical complications that are slow to resolve (e.g. MRSA wound infection, hematoma)
· There are painful medical complications of surgical care (e.g. C. difficile hemorrhagic colitis secondary to antibiotic use)
· The surgical procedure is one known to be associated with slow recovery and lengthy postoperative pain (e.g. multiple level spine fusions, the Nuss procedure, small bowel transplantation)
· The patient had preoperative opioid dependency, either therapeutic or abuse related (e.g. cancer pain, the pain related to a terminal condition, or preoperative OUD controlled with methadone or buprenorphine)
Indeed, we don’t know if the patients in this study had ongoing opioid use at all, because all the data on opioid prescription refills were based solely on insurance databases and not on medical record reviews or patient interviews. The assumption that the need for a refill is a sign of abuse simply may simply be a reflection of confirmation bias.
While the opioid “crisis” in this country remains a concern amongst the public and our profession as well, it should be remembered that now, as in the past, there is an unproven and very hypothetical link between the prescription of postoperative opioid analgesics to children and adolescents and the later development of opioid abuse. This paper, in our view, does not add any clarity to this discussion because Association ≠ Causation.
References
1. Cummings Joyner AK, King MR, Safranek C, Parvathinathan G, De Souza E, Anderson TA. Health Care Burden Associated With Adolescent Prolonged Opioid Use After Surgery. Anesthesia and analgesia. Feb 1 2023;136(2):317-326. doi:10.1213/ane.0000000000006111
2. Ward A, Jani T, De Souza E, Scheinker D, Bambos N, Anderson TA. Prediction of Prolonged Opioid Use After Surgery in Adolescents: Insights From Machine Learning. Anesthesia and analgesia. Aug 1 2021;133(2):304-313. doi:10.1213/ane.0000000000005527
3. Harbaugh CM, Lee JS, Hu HM, et al. Persistent Opioid Use Among Pediatric Patients After Surgery. Pediatrics. Jan 2018;141(1)doi:10.1542/peds.2017-2439