Two years ago, I asked, “Doctors diagnosing addiction: Are the blind leading the blind?”

The estimated rates of substance use disorder among clinical patients treated for pain vary greatly, depending on whom one believes. Studies also suffer from a widespread failure to address genetically mediated variability in minimum effective opioid dose between individuals. Thus, I have many times respectfully suggested that the entire body of literature on opioid safety and effectiveness needs to be thrown out and done over.

So far, the response of U.S. public health agencies has been a deafening silence.

I work consistently to find definitive data on this subject. Among other studies, one widely cited cohort study has long stood out to me. The study examined two years of clinical records for the entire residential population of North Carolina (n = 9,560,234). Results are summarized in the study abstract as follows:

“Opioid analgesics were dispensed to 22.8 percent of residents. Among licensed clinicians, 89.6 percent prescribed opioid analgesics, and 40.0 percent prescribed ER formulations. There were 629 overdose deaths, half of which had an opioid analgesic prescription active on the day of death. Of 2,182,374 patients prescribed opioids, 478 overdose deaths were reported (0.022 percent per year). Mortality rates increased gradually across the range of average daily milligrams of morphine equivalents.”

Mortality is, of course, not the only issue of concern in prescribing opioid analgesics to chronic pain patients. Estimates of the risk of opioid abuse in pain patients are also widely variable, depending on source. Fortunately, federal regulators seem to be waking up to the many nuances in this issue that were previously submerged in rhetoric.

An important distinction now recognized by the FDA but largely missed by anti-opioid zealots is that opioid use disorder cannot be defined in the same way for patients with chronic severe pain as for those who are not currently treated for pain. Some patient symptoms (opioid tolerance, physical dependence, withdrawal symptoms) are caused by treatment, rather than by addiction. When treatment-related symptoms are disregarded, the estimated cumulative risk of moderate to severe OUD in pain patients drops by about 80 percent, to 1.4 percent to 1.6 percent per year (see Page 76 in a recent FDA report).

(Note: The entire study document is worth reading for anyone who wants to understand details of diagnostic standards and risk factors associated with prescription opioid analgesics.)

Unfortunately for both doctors and their patients, the distinction between “substance use disorders” versus “substance use disorder in chronic pain patients” is not yet widely understood among clinicians or policy makers, and is opposed by some advocates. The result is a widespread clinician predisposition to undertreat pain or even to desert patients who need higher doses of prescription opioids.

That predisposition can kill patients if you indulge it.

Prevalence of this problem among clinicians can now be explored by artificial intelligence (AI)-based search agents, in ways that were impossible even a year ago. Here is what one such agent had to say on the subject in November 2025:

My question: “How do the standard DSM-5 criteria for substance use disorder differ from the modified DSM-5 criteria adjusted for pain patients?”

The answer: “The standard DSM-5 criteria for substance use disorder (SUD) include 11 criteria related to impaired control, social impairment, risky use, and pharmacological factors such as tolerance and withdrawal, with diagnosis severity determined by the number of criteria met. For pain patients who are taking prescription opioids under medical supervision, the criteria are adjusted: Tolerance and withdrawal symptoms are not counted toward a diagnosis of opioid use disorder, because these are expected outcomes of long-term, medically supervised opioid therapy, not signs of addiction or misuse.”

DSM-5 criteria for substance use disorder

• Diagnosis is based on a dimensional scale (two or more of the 11 criteria result in a diagnosis, with higher numbers indicating more severe disorder).
• Legal problems (a former criterion) were removed, and craving was added.
• Severity is graded as mild (2-3 criteria), moderate (4-5), or severe (6+).

Modifications for pain patients

• Tolerance and withdrawal are specifically excluded from the diagnostic count if the patient is taking opioids under appropriate medical supervision for pain, because these effects are physiological and not indicative of a substance use disorder in this context.
• The rationale is that these symptoms are iatrogenic (caused by medical treatment), not psychopathological (caused by misuse or addiction).
• Other criteria (such as loss of control, unsuccessful attempts to cut down, craving, and social or functional impairment) still apply for diagnosis in pain patients.

As a health care educator, I must wonder how many clinicians are aware of these distinctions. With that in mind, I asked a follow-up question: “What proportion of addiction specialists use or are aware of the modified DSM-criteria for substance use disorder in pain patients?”

The response noted in part:

“There do not appear to be solid data quantifying what proportion of addiction specialists are aware of or explicitly use the modified DSM-5 rule that excludes tolerance and withdrawal for patients taking opioids appropriately for pain. No survey was found that reports a specific percentage of specialists who know or apply this modification in clinical practice.

“DSM-5 and major clinical resources clearly state that tolerance and withdrawal should not be counted toward opioid use disorder (OUD) when opioids are taken solely under appropriate medical supervision. This wording is present in checklists and guidance distributed by professional organizations and public health agencies, so the modification is formally part of standard criteria rather than an optional add-on.

“Studies assessing clinicians’ knowledge of opioid prescribing and chronic pain management demonstrate substantial knowledge gaps about opioid risk assessment and diagnostic concepts in general, suggesting that correct application of nuanced DSM-5 rules (like excluding tolerance/withdrawal in appropriately treated pain patients) is likely imperfect and variable across clinicians and settings.” (Source: PMC article)

A parting note

Paraphrasing Lord Kelvin in an 1883 lecture (and others in more modern contexts), “We do not value what we do not measure.” Likewise, as humorists over the years have informed us, “It ain’t the things that a man don’t know that gets him into trouble. It’s the things he knows for sure that ain’t so.”

Perhaps nowhere in daily life are these statements truer, or more dangerous to human welfare, than in health care. If you are a clinician and you thought you were justified in withholding opioid treatment from people in agony because of “risks,” then it is time for you to think again.

Richard A. Lawhern is a nationally recognized health care educator and patient advocate who has spent nearly three decades researching pain management and addiction policy. His extensive body of work, including over 300 published papers and interviews, reflects a deep critique of U.S. health care agencies and their approaches to chronic pain treatment. Now retired from formal academic and hospital affiliations, Richard continues to engage with professional and public audiences through platforms such as LinkedIn, Facebook, and his contributions to KevinMD. His advocacy extends to online communities like Protect People in Pain, where he works to elevate the voices of patients navigating restrictive opioid policies. Among his many publications is a guideline on opioid use for chronic non-cancer pain, reflecting his commitment to evidence-based reform in pain medicine.