There is a workload that sits in a spreadsheet on most practice managers’ computers, gets discussed in the partners’ meeting, and rarely gets enough hours to be done properly. It is the outbound list: QOF recalls, flu campaign eligibles, chronic disease reviews, cervical screening invitations, post-discharge follow-up. The reason it gets squeezed is not a prioritisation failure. The reason is structural, and it has a specific operational shape that is worth naming.
This post is about the outbound workload, why it competes directly with inbound call handling for the same fixed reception capacity, and what AI outbound calling actually changes about that competition. It is written for practice managers running QOF, flu, and chronic disease workloads on a fixed reception team, and for PCN managers thinking about how to lift completion rates without adding headcount.
The Outbound Obligation Is Not Optional
QOF recall, flu uptake, and chronic disease review completion are contractual, clinical, and in QOF’s case directly tied to practice income. Across a 10,000-patient list, these obligations represent thousands of individual contacts per year, with sharp campaign windows in autumn for flu and rolling demand for QOF and screening.
The numbers worth pausing on are the time numbers. An outbound phone call from a receptionist takes three to eight minutes on average, depending on whether the patient picks up, whether they have questions, and whether the appointment is booked end of call. A flu campaign reaching 2,000 eligible patients with two contact attempts each is 100 to 270 hours of reception team time. That is three to seven weeks of one full-time receptionist’s working time, concentrated into a campaign window of six to eight weeks.
That time has to come from somewhere. In most practices it comes from the inbound queue, which is where the operational problem begins.
The Competition Between Inbound and Outbound
Inbound and outbound call work draws from the same fixed reception capacity. A receptionist spending the morning on flu campaign calls is not on the 8am inbound queue. A receptionist on the inbound queue is not making QOF recall calls. The two workloads are direct substitutes for one another in the practice’s daily resource allocation, and inbound is almost always allowed to win, because the queue is visible and the patient is on the line.
This is the structural reason QOF completion rates and flu uptake numbers vary so widely between practices with similar lists. Practices that hit higher completion rates are usually not better managed than practices that miss. They are usually either operating with more reception headcount than their list size requires, or running concentrated outbound sprints that displace other work for weeks at a time.
Before getting to what AI outbound calling changes, it is worth being precise about what the existing channels actually do, because the comparison only matters when the current state is named clearly.
What Letters, SMS, and Reception Calls Actually Achieve
Letters are low cost and require no live staff time at point of send, but they do not close the loop. A letter starts a process the patient has to complete, and for the patient cohorts most likely to miss recalls, complex needs, lower digital engagement, anxiety about attending, the letter does not produce the booked appointment without follow-up.
SMS reminders perform better than letters for the digitally confident segment of the list, but they share the same structural limitation. They require the patient to initiate the next step. For the same cohorts that letters underperform with, SMS underperforms with them too.
Reception team phone calls are the only existing channel that reliably closes the loop. The conversation can address hesitation, answer questions, and book the appointment in the call. The completion rate is high. The cost is also high, and the cost is paid in inbound queue time the practice does not have.
The current state in most practices is a layered approach: letters at scale, SMS to the digitally confident, and reception calls to the residual list when there is time. The completion gap that this approach leaves is the cohort of patients who do not respond to letters or SMS, and do not get reached by reception calls before the campaign window closes.
That cohort is where the QOF and flu completion numbers are actually decided.
What AI Outbound Calling Changes
AI outbound calling works on the same channel that closes the loop reliably. It is a voice conversation, not a one-way message. The system calls the patient on the list, verifies identity, explains the reason for the call, and books or records the response in the call itself.
The operational difference is not about better technology. It is about removing the substitution between inbound and outbound capacity. An AI running a flu campaign in the background is not on the inbound queue. A reception team running the inbound queue is not displaced by the flu campaign. The two workloads run in parallel rather than competing for the same hours.
For a 2,000-patient flu campaign that previously displaced 100 to 270 hours of reception team time, the time displaced becomes zero. The campaign runs. The inbound queue is unaffected.
Before looking at how this works in practice on a live deployment, it is worth understanding the second channel that the same system enables, because the dual-channel pattern is what actually moves completion rates.
The Dual Channel That Lifts Completion
The most effective deployment pattern is two streams running together.
The first stream is dedicated outbound. The patient list is uploaded, the AI calls the list, the conversations happen, and outcomes are logged. This is the primary mechanism for reaching patients who have not engaged with letters or SMS.
The second stream is the inbound campaign prompt. When a patient who is on an outbound list calls in for another reason, a repeat prescription query, an appointment cancellation, a sick note request, the AI completes the original request and then offers the campaign booking opportunistically in the same call. A patient calling about a prescription who is also on the flu list finishes the call with a flu jab booked.
The second channel matters more than the headline suggests. Inbound campaign prompts convert at a higher rate than standalone outbound, because the patient has already chosen to engage. The combined effect is that every inbound call from a patient on any campaign list becomes a campaign touchpoint, without any additional outbound effort.
The Park Street and Greensand Surgery deployments are both running the inbound channel of this pattern in their live operations. The dual-channel outbound model is the next stage of deployment for practices that have established the inbound base.
How the Pricing Model Aligns With Campaign Outcomes
AI outbound calling is typically priced per effective call, not per attempt. The operational point is that the cost only registers when the contact actually achieves something the practice needed.
If the AI calls a patient who does not answer, the practice pays nothing. If the patient answers but declines, the practice pays the per-call rate because the contact has been completed and the decline is logged against the patient record. If the patient answers and books, the practice pays and gains the appointment.
The reason this matters operationally is that flu and QOF work historically carries a cost regardless of outcome. Receptionist time spent on unanswered calls or unsuccessful contact attempts is paid for at the same rate as time spent on successful bookings. The performance-based pricing model removes the cost of failed attempts entirely, which changes the maths on running larger or more aggressive campaigns.
For a practice running a 2,000-patient flu campaign with historical reception team time of 100 to 270 hours, the equivalent AI outbound cost is tied to the number of contacts actually completed, not the number of attempts made to complete them.
What This Looks Like in Practice Operations
For the practice manager, the operational shift is specific. Campaign work no longer competes with inbound capacity. The flu sprint that previously displaced three to seven weeks of receptionist time no longer displaces anything. QOF recall completion can run on a continuous basis rather than in concentrated bursts that depend on quiet inbound weeks. The dual-channel inbound prompt means that every inbound call from a patient on a campaign list becomes an opportunity to close that patient’s outstanding contact in the same call.
The practical effect is that completion rates move because the workload is no longer rationed against inbound demand. Practices that have historically had the discipline to run strong campaigns at the cost of inbound capacity recover the inbound capacity. Practices that have historically had to choose between inbound queue management and outbound completion no longer have to choose.
The free four-week pilot Auxilis runs with practices typically establishes the inbound base first, then layers in the outbound channel once the inbound deployment is stable. The pilot covers existing telephony, no hardware, no setup fee, and produces live data from the practice’s own patient list.
Book a 20-minute demo at auxilis.ai
Frequently Asked Questions
What is QOF recall and why does it depend on outbound calling?
The Quality and Outcomes Framework rewards GP practices for achieving clinical care standards for patients with long-term conditions, and QOF completion directly affects practice income. Recall calls bring patients with conditions like diabetes, hypertension, or COPD in for periodic reviews. Letters and SMS reach a portion of the list. The residual cohort that does not respond to those channels is reached by phone or not reached at all, which is where most of the completion variation between practices is decided.
How does an AI handle patients who are hesitant or have questions on the call?
A well-built AI outbound system holds a real two-way conversation rather than reading a script. If the patient asks why they are being called, the AI explains the reason. If the patient is unsure, the AI can offer basic reassurance and a choice of times. If the patient declines, the response is logged and the patient is not called again for that campaign.
Does AI outbound calling meet NHS data protection and clinical safety standards?
GP-specific AI outbound systems are built to comply with GDPR, the NHS Data Security and Protection Toolkit, and DCB0129 clinical safety standards. Patient identity is verified at the start of each call using two-factor authentication against NHS PDS. Data collected during the call sits inside the practice’s existing data governance framework. The practice remains the data controller.
Does AI outbound calling integrate with EMIS and SystmOne?
Yes. Booked appointments from outbound campaigns are logged directly in the clinical system. Patient lists can be uploaded from existing recall management workflows. No separate data export or manual entry is required from the reception team.
What happens to patients who explicitly decline to be contacted?
A decline is recorded on the patient record against that campaign and the patient is not called again for that campaign cycle. The decline is treated as a completed contact for reporting purposes, so the campaign data reflects the actual state of the list rather than carrying the patient as outstanding.
Can outbound run at the same time as inbound on the same system?
Yes, and this is the deployment pattern that produces the strongest completion lift. Outbound campaigns run in the background while the inbound system continues to handle the 8am queue and routine inbound traffic. The dual-channel inbound campaign prompt then layers additional campaign touchpoints onto inbound calls from patients on the outbound list.