Procurement conversations about outsourced call handling tend to flatten into two numbers: cost per call and capacity. Both are easier to compare than the thing that actually determines whether the spend pays back, which is what happens between the call being answered and the patient record being updated.
The cost-per-call comparison hides a structural difference between two delivery models that are often presented as interchangeable. A generic outsourced contact centre answers the phone. Dedicated reception staff trained on a practice’s own protocols, working inside the practice’s own clinical system, do the work a directly employed receptionist would do. The output of one is a message in a queue. The output of the other is a completed action in EMIS or SystmOne.
That distinction sounds operational. It is. It is also the difference between a service that reduces practice workload and a service that displaces it.
What the Generic Call Centre Model Actually Delivers
Generic outsourced call handling is built for any sector that has inbound voice traffic. The training framework is portable. Tone of voice, message capture, basic escalation if a caller becomes distressed. The staffing model is shared pools, often with handlers working across multiple clients in a single shift. Performance is measured on calls per hour and average handling time.
For a utility company or a retail returns line, that model is appropriate. For a GP practice, it has three structural problems that show up in the work the practice has to do *after* the call.
The handler does not have access to your clinical system. Without it, they cannot verify the patient against the record, cannot see whether there is an existing appointment, cannot book directly, and cannot add a note where it needs to live. Whatever happens on the call has to be transferred manually by your team. The call has been answered. The work has not been done.
The handler does not know your practice. They do not know which clinicians take which appointment types. They do not know which conditions you route to your first-contact physiotherapist or social prescriber. They do not know your local pharmacy’s Pharmacy First arrangements. They are running a generic script and producing generic output, which your reception team has to translate into specific actions.
The incentive structure pushes against thoroughness. A handler measured on calls per hour and average handling time has a reason to wrap calls quickly. In a retail context, that is fine. In a primary care context, where the information captured at first contact shapes what the clinician sees, brevity becomes a downstream cost. Your team makes the callbacks the handler did not make. Your clinicians chase the details that did not get logged.
The cost per call looks low because it is measured on the wrong unit. The unit that matters to a practice is cost per resolved patient contact, and that calculation includes the work your staff do to clean up after a generic call has been handled.
What Dedicated Reception Staff Working Inside Your Systems Looks Like
The alternative model is structurally different. It is not a contact centre with a healthcare overlay. It is reception capacity, hired and managed by an external provider, trained on your protocols, and given secure access to your clinical system so the work happens where it needs to happen.
The mechanics matter, so they are worth being specific about.
System access via GPITC. The team accesses EMIS or SystmOne through GP in the Cloud, with the practice retaining ownership of credentials, access levels, and audit logs. Two-factor authentication, twelve-character passwords, no copy-paste out of the clinical system, auto-logout after five minutes, accounts deleted within twenty-four hours of offboarding. NHS DSP Toolkit compliant, GDPR ready, ISO 27001 aligned. The handler is working in your system the way a directly employed receptionist would, with the same restrictions and the same audit trail.
Trained on your in-house protocols. The practice owns the protocols. The clinical lead defines what reception is permitted to do without clinical input, which appointment types route where, how to handle Pharmacy First referrals, what triggers an escalation. The external team learns those protocols and operates within them. The protocols do not get rewritten by the provider, and clinical decisions are not transferred to the provider. The reception layer is the layer that gets staffed externally. The clinical and protocol layers stay where they are.
Dedicated to your practice, not a shared pool. The same team handles your calls every shift. They learn the names of your regular complex patients. They know which of your clinicians prefers which kind of summary. That institutional knowledge is what turns a competent handler into someone who reads as part of the practice on the call. Patients do not know they are talking to an outsourced team, because operationally they are not. They are talking to your reception, just with the staffing managed externally.
Outputs land where they need to land. Appointments are booked in your diary. Notes are added to the patient record. Tasks are assigned to the right clinician. Outbound calls go out in your name. The handler completes the action. Your team does not pick up the trail.
The Numbers That Actually Tell You What You Are Buying
The reporting question is where the difference between the two models becomes most visible, and where practices most often get sold something that is not what it sounds like.
A generic contact centre will report on call volume, answer rates, and abandonment. So will dedicated reception. The difference is whether those numbers reflect the work that needs doing in your practice.
The metrics that do reflect that work, and that any reception staffing model worth the spend should be able to give you, are these.
Call volume by day, hour, and site. This is the baseline. If you cannot see when your peaks fall and where they fall hardest, you cannot staff against them.
Answer rates and speed to answer against your agreed targets. The targets are yours. The reporting is against your targets, not the provider’s averages.
Abandonment rates and patterns. Not just the headline number but when in the day and at which queue points patients drop off.
Handling time trends. Useful as a stability indicator. A handling time that is creeping up can indicate a shift in case mix, a protocol gap, or a training issue. It can also indicate, depending on the period, that more elderly patients or patients in distress are calling, which legitimately changes call patterns and is not in itself a problem to fix. The trend tells you to look. It does not tell you what you are looking at.
Performance against your own service KPIs. Whatever your practice has agreed as the standard, that is what the reporting tracks.
These are call handling metrics. They are not clinical workflow metrics. The reporting tells you how the staffing layer is performing against the targets you set. It does not tell you how to redesign your triage, what your escalation rate should be, or what your data means about patient demand. Those are practice-side questions, and they are answered by your clinical and operational leadership, who have the context the call data alone does not contain.
A reception staffing provider that promises analyst-level interpretation of your access patterns is selling you something that sits well outside the staffing layer. That is not the spend that pays back. The spend that pays back is reception capacity that performs against your targets, in your system, on your protocols, with reporting clean enough that you can see what is happening.
Why the Human Layer Still Matters
The other distinction worth being explicit about is the case for human handlers in primary care reception, and where AI fits.
AI-led call handling has a use case in primary care. Routine, structured tasks where the patient knows what they are calling for and the system can confirm against the record without needing context can be handled efficiently by AI. That has real value at peak volume.
The case for human handlers is the long tail. Elderly patients who need time to explain what is going on. Patients with mental health presentations where the call is doing more than booking an appointment. Patients in emotional distress where what they are asking for is not what they need. Patients whose first language is not English. Patients with cognitive difficulties. The calls where the handler has to listen for what is not being said, and where misreading the call has consequences.
These calls are not edge cases. They are a meaningful share of any general practice’s inbound volume, and they are the calls where a generic, throughput-optimised model produces the worst output. They are also the calls where a dedicated team that knows your practice, operating within protocols agreed with your clinicians, produces the best.
The position is not anti-AI. It is that AI deployed in primary care reception works best when it sits alongside human handlers who know the practice, with clear rules about which calls go where. Human-led, with AI in the right places, rather than AI-led with humans on exception.
What to Ask Before You Sign Anything
For practice managers, PCN leads, and federation operations leads evaluating outsourced reception or call handling, the questions that separate one model from the other are these.
Will the team have secure access to our clinical system, or are they passing messages back to us? If the answer is the second, you are buying message-taking, not reception.
Are they trained on our protocols, or on a generic healthcare framework? If the framework is generic, the practice-side adaptation cost lands on you.
Is the team dedicated to our practice or shared across clients? Shared pools cannot accumulate practice knowledge. Dedicated teams compound it over time.
What does the reporting actually cover? If the answer is volume metrics only, you have no operational visibility. If the answer is “we will analyse your clinical workflows and tell you what your data means,” you are being sold consultancy you did not ask for and probably do not need.
How long does implementation take? A realistic onboarding timeline for protocols, training, and system access is in the region of seven weeks. Significantly shorter usually means corners cut on training. Significantly longer usually means the provider does not have a settled process.
These are the questions that surface the structural differences. The cost-per-call comparison does not.
Frequently Asked Questions
What is the difference between an outsourced call centre and dedicated reception staff for a GP practice?
A generic outsourced call centre answers calls and takes messages, with handlers typically working across multiple clients on a shared pool basis. They do not have access to the practice’s clinical system, so any action arising from the call has to be completed manually by practice staff afterwards. Dedicated reception staff, by contrast, are trained specifically on the practice’s protocols, work inside the practice’s clinical system through secure remote access, and complete the action on the call. The output of the first is a message in a queue. The output of the second is a booked appointment, an updated record, or a task assigned to the right person.
How does external reception staff get secure access to EMIS or SystmOne?
Through GP in the Cloud (GPITC), a secure remote desktop gateway. The practice retains full control of credentials and access levels. Standard safeguards include two-factor authentication, twelve-character passwords, no copy-paste out of the clinical system, automatic logout after five minutes of inactivity, full activity logging, and account deletion within twenty-four hours of offboarding. The setup is NHS DSP Toolkit compliant, GDPR ready, and ISO 27001 aligned. Operationally, the external handler works inside the practice’s system the same way a directly employed receptionist would.
How long does it take to onboard an external reception team into a GP practice?
A realistic timeline from agreement to go-live is around seven weeks. The phases are typically a discovery meeting to confirm hours, start date, and current call volumes, followed by protocol and workbook development with clinical sign-off on red flags and escalation. From there, team assembly and training on the practice’s specific workbook, system access setup through GPITC, and completion of NHS mandatory learning and DBS checks. Then a gradual go-live with close monitoring and quality audits. Significantly shorter timelines usually indicate corners cut on training. Significantly longer usually indicates the provider has no settled process.
Can dedicated reception staff handle calls across multiple sites in a PCN or federation?
Yes. A dedicated team can be configured to handle calls across multiple practices in a PCN or federation, with practice-specific protocols and system access for each site. The operational advantage is consistency in how calls are handled across the network, which makes performance data comparable site to site and reduces the variation in patient experience between practices. The team still operates within each practice’s own protocols, signed off by each practice’s clinical lead, rather than applying a network-wide generic standard.
Talk to LineIn About a More Stable Model
LineIn provides dedicated reception staff for UK GP practices, PCNs, and federations, trained on your in-house protocols and working inside your clinical system through GPITC at £19.45 per hour. The reporting covers call volume, answer rates, abandonment, handling time, and performance against your service KPIs.
If you are weighing up whether the call handling capacity you have today matches the access standard you want to hit, and you would like to compare how a dedicated team at that price point would sit against what you are doing now, get in touch. Happy to walk through the model, the reporting, and the implementation in detail at your pace.