Substance use monitoring in supportive oncology: surveillance is not a clinical outcome, and policing is not a therapeutic framework.

Sousa and colleagues recently published a review of substance use in supportive oncology ostensibly addressing its impact on clinical outcomes. The paper is substantively a monitoring review; no clinical outcomes data are presented or analyzed. I identify 11 domains of concern: (1) the title's unmet promise of outcomes evidence; (2) the conflation of complex opioid dependence with addiction; (3) the absence of demonstrated benefit from opioid monitoring in cancer populations; (4) the poor diagnostic accuracy of risk screening instruments as quantified by likelihood ratio analysis; (5) regulatory and guideline prohibitions against opioid discontinuation or practice dismissal on the basis of aberrant urine drug screens; (6) the systematic inadequacy of clinician competence in urine drug screen interpretation; (7) an unfavorable cost-benefit ratio for universal surveillance in supportive oncology; (8) the stigmatizing effect of universal screening and its contribution to disparities in cancer pain management; (9) the failure to distinguish patients with advanced cancer from cancer survivors, populations with fundamentally different risk horizons and therapeutic priorities; (10) the risk that evidence-free monitoring recommendations will propagate into institutional policy and payer requirements, generating measurable financial toxicity without demonstrated clinical return; and (11) the absence of any management pathway for positive screening results - a logical incompleteness that renders the entire surveillance framework clinically inoperable, since neither dismissal from practice nor opioid discontinuation is appropriate in most circumstances, and an absent opioid on UDS does not constitute evidence of diversion in isolation. I argue that the dominant failure mode in oncology opioid management is undertreatment, not diversion, and that any framework proposing expanded surveillance must first demonstrate that it does not worsen this problem.