What happens to an inquiry after hours
A new patient's first contact can come by phone, web form, fax, portal message or, more and more often, through an AI assistant acting for them. This article covers what agents do at that first contact, what stays with your staff, which rules apply to automated calls and texts, and what we measure. The best-known evidence on response speed is old and comes from outside healthcare, and we say so where we use it.
Why the first hour matters
The most cited study of response speed looked at sales leads across US companies. Firms that tried to reach a potential customer within an hour of an inquiry were nearly seven times as likely to qualify the lead as firms that tried even an hour later (Harvard Business Review, 2011). The data is cross-industry and more than a decade old, so we use it to show direction only.
Healthcare call centers have their own standards. The Veterans Health Administration's goal is an average speed of answer of 30 seconds or less and an abandonment rate of 5% or less. A study of VHA call centers found that longer waits on the phone were linked to patients reporting worse access to urgent appointments, while abandonment rates showed no clear link (American Journal of Managed Care, 2019).
Every way an inquiry arrives
Patients, families and referrers reach you through:
- your main phone line and the numbers on your ads and map listings
- web forms, including insurance-check forms
- faxed and emailed referral packets, ID cards and insurance-card images
- portal and chat messages
- AI assistants that call or fill in forms for a patient
Fax is still common. In a 2019 MGMA Stat poll with 1,581 responses from medical practice leaders, 89% said their organization used fax (MGMA, 2019). So your fax line stays, agents read what arrives on it, and referral partners keep sending the way they do now.
What agents do at first contact
An agent answers a call, or replies to a form, within seconds of it arriving, at any hour. It says it is an automated assistant and that the call is recorded, then follows the script your intake lead wrote. It records the details in your CRM or EHR, reads faxed packets and card photos into the right fields, checks coverage and offers a time for the next step. A field the agent reads with low confidence goes to a person to confirm before it is used.
What stays with your people
Any clinical question goes to your on-call clinician under your own escalation protocol. Agents are configured not to give clinical advice, and we test that limit before go-live. A caller can ask for a person at any point. Your staff confirm anything quoted to a patient as their cost before it is sent. How that review works before and after go-live is set out on the human review page.
Checking coverage at first contact
Eligibility checks are the most frequent administrative task that CAQH tracks. Its 2024 Index counted 31.5 billion medical eligibility and benefit verifications in 2023, with $44 billion spent on them. It put the savings from moving the remaining manual and portal checks to electronic ones at $11.7 billion, the largest opportunity among the medical transactions it measures (CAQH Index 2024, published 2025).
An agent runs the electronic check when the inquiry arrives and again before the visit. It reads the benefits in the response, flags a missing secondary plan, and queues a call to the payer when the electronic response leaves a gap. What happens to that coverage record in billing is covered on our revenue cycle page.
Consent and disclosure for automated calls and texts
In February 2024 the FCC ruled that AI-generated voices count as "artificial" voices under the Telephone Consumer Protection Act (FCC, 2024). As a result, AI-voice calls generally need the called party's prior express consent (Hunton Andrews Kurth, 2024), and telemarketing calls need prior express written consent (FCC, 2024). A person can revoke consent by any reasonable means, and the request must be honored within 10 business days (47 CFR 64.1200(a)(10), via Cornell LII).
Agents say they are automated, contact people only where consent covers it, and are designed to act on an opt-out as soon as it is given. Your compliance team approves the scripts and consent language, and your counsel decides which state rules apply. The compliance page covers these rules in more detail.
Being reachable by AI assistants
Some first contacts now come from software. In July 2025 Google began rolling out AI calling that phones local businesses for a user to ask about pricing and availability (TechCrunch, 2025). OpenAI's ChatGPT agent, launched the same month, was built to complete online tasks for a user, including filling in forms (OpenAI Help Center, 2025). People also ask chatbots about health: in a KFF poll fielded from 24 February to 2 March 2026, 32% of US adults said they had used AI chatbots for health information in the past year (KFF, 2026).
An intake line built for this answers structured questions, such as hours, accepted insurance and the next open slot, and asks for the patient's own confirmation before any health information is shared or a visit is booked. Forms need clear labels that software can fill in. Whether assistants name you in the first place is a marketing question, covered on the marketing page.
What we measure
Before anything goes live, we agree a dated baseline with you. Then we track:
- median minutes from form to first contact attempt, after hours and daytime separately
- call abandonment rate
- the share of inquiries with a booked next step
- how many calls go to a person, and why
Every call and message is logged with its transcript, so each figure traces back to the records behind it.