No. SleekAI is configured with explicit refusal language for every diagnostic, treatment, dosage, and symptom-interpretation question. The instruction includes a hard refusal: never diagnose, never recommend specific treatments. Respiratory symptoms have a uniquely time-sensitive urgency profile, so the bot is configured with extra sensitivity on phrases like 'reliever inhaler isn't working', 'can't speak full sentences', or 'chest pain with shortness of breath', which all route immediately to call 911 or the local emergency number.

Configured with extra-sensitive defaults. Pulmonology has a higher cost of missed urgency than most outpatient specialties because severe asthma exacerbations, pulmonary embolism, and acute COPD exacerbations can deteriorate fast. The system prompt errs generously toward escalation: 'when in doubt, call now', with 911 routing for any combination of shortness of breath, chest pain, inability to speak sentences, peak flow well below personal best, or signs of severe respiratory distress. Audit-test quarterly to confirm the escalations hold.

Yes. Pulmonary function testing has a specific prep pattern (typically a short-acting bronchodilator hold for some hours and a long-acting bronchodilator hold per your protocol, no smoking for an hour beforehand, comfortable clothing). The bot books the PFT on the right calendar with prep instructions auto-included in the confirmation, sourced from your published prep page. The bot does not advise on whether to hold any specific medication outside your published protocol; that conversation stays with the prescribing physician.

Yes, end-to-end logistics. Sleep apnea evaluation typically runs as a consultation, then either a home sleep study or in-lab polysomnography depending on patient factors, then a follow-up to discuss results and CPAP titration if needed. The bot walks patients through your practice's specific pathway, confirms which sleep lab you partner with, and books each step. It does not interpret any sleep study data; that conversation belongs to the sleep medicine pulmonologist.

Yes. Pulmonology has meaningful subspecialty division (interstitial lung disease, pulmonary hypertension, sleep medicine, interventional pulmonology with bronchoscopy, sarcoidosis). The bot reads each pulmonologist's subspecialty profile and routes patients accordingly. 'I was referred for a pulmonary nodule' goes to general pulmonology or interventional depending on your workflow; 'I have suspected interstitial lung disease' goes to the ILD-focused pulmonologist; 'I need a sleep apnea workup' goes to sleep medicine.

HIPAA compliance is a stack question, not a single-plugin question. SleekAI runs on your WordPress install with an API key you provide. For US pulmonology practices handling PHI, you need a HIPAA-eligible API vendor configuration, a BAA in place with that vendor, encryption at rest on the WordPress database, role-based access on conversation logs, and the chat channel documented in your Notice of Privacy Practices.

Yes, in your WordPress database. Retention is configurable. For PHI we recommend short retention windows (days, not years), encryption at rest on the database, role-restricted access on log readers, and BAAs across the stack including your API vendor (OpenAI, Anthropic, Google, OpenRouter) and hosting provider. The system prompt is configured to avoid inviting sensitive clinical detail into chat, keeping logged content operational.

Yes. Patients writing in any common language get responses in their language, with operational details translated cleanly and the clinical-refusal language translated faithfully. The respiratory urgency escalations also translate, so the safety boundary holds for patients writing in Spanish, Vietnamese, Mandarin, or any other language you serve. This is especially important for pulmonology because COPD and asthma populations include many older patients with limited English.