The Sleep Clinic's Referral for a Patient Who Won't Use CPAP

The referral framing ('address nonadherence') risks colluding with a motivation-deficit hypothesis that the data do not support. Mr. Delgado's description of feeling suffocated plus claustrophobia points to a desensitization/behavioral problem (conditioned anxiety and escape behavior), not low motivation - he is already attempting use and expresses embarrassment, which suggests he holds himself responsible. Dr. Okafor should (1) conduct a functional assessment of the antecedents and consequences of mask removal (time of night, sensation, panic, escape relief), screen for panic and for insomnia (the 1.5-hour pattern may reflect sleep-onset use followed by removal), and (2) set a collaborative target of gradual exposure with shaping (e.g., graded mask-on time while awake, then titration), not a blunt 'hours per night' goal imposed by the clinic. Because the clinic has attached a 4-session expectation, she should clarify that adherence is a behavioral outcome influenced by device fit, pressure settings, and anxiety, and that a 4-session contract may be insufficient if comorbid insomnia emerges. She should communicate to the pulmonologist that the working formulation is claustrophobic avoidance, request a mask-fit review and consideration of a different interface, and note that a motivational framing would likely increase shame and disengagement. Documenting the reformulation protects the patient from being labeled noncompliant.

The clinical issue is a predictable relapse within an exposure paradigm - an early success followed by an aversive experience that reinforces avoidance. Dr. Okafor should normalize this as a shaping setback rather than treatment failure, review the panic episode's context, and adjust the hierarchy (smaller steps, perhaps practicing with the device off the face first, addressing catastrophic cognitions about suffocation). The ethical issue is the request for confidentiality within an integrated care setting. Dr. Okafor should explain at the outset (and again here) the limits of confidentiality in team-based care: clinically relevant information needed for safe medical decision-making is shared, while details not relevant to the medical plan can remain in the psychotherapy note. She should negotiate what will be shared - that panic occurred and the behavioral plan is being adjusted - without disclosing content the patient considers sensitive, unless safety requires it. Withholding a panic episode from a physician managing severe OSA and cardiovascular risk could compromise care, so the patient's preference should be honored where possible but not at the cost of the referring physician's ability to make safe decisions. She should document the discussion and the patient's preferences.

Dr. Okafor should push back on a single dichotomous outcome applied at a fixed session count. That criterion is a reimbursement/payer threshold, not a clinically meaningful index of treatment response, and it ignores baseline severity, device factors, and the staged nature of behavioral change. A defensible evaluation plan would include: (a) a baseline and repeated measure of mean nightly use and percentage of nights used, (b) process measures (sessions attended, exposure steps completed, panic frequency), (c) patient-reported outcomes (Epworth Sleepiness Scale, insomnia severity, claustrophobia ratings), and (d) a defined responder criterion stated in advance but interpreted with the caveat that 4 sessions may be a dose insufficient for some patients. She should also note that reporting only successes creates a selection/survivorship bias and that patients who do not meet the threshold are informative for program improvement. She can offer to report both the payer-aligned metric and the broader outcomes, with a note on the limits of attributing change to her intervention without a comparison condition. This is program evaluation, not research, but the same measurement logic applies; if the administrator wants causal claims, an IRB-reviewed design would be needed.