The hard questions, answered straight.

We don't guess, and we don't second-guess your clinical judgement. The audit looks for internal contradictions in your own data — patients whose records point one way but whose coding points another. The classic example: someone on a long-term medication only ever prescribed for a specific condition, but with no diagnostic code for it on the register. Or repeat results, referrals or letters that imply a diagnosis the register never captured. These are patients who genuinely have the condition and are already being treated for it — the code simply never made it onto the register. We surface that list; we never decide it. A clinician reviews and confirms every single case before anything changes. The software finds the candidates; the doctor makes the call.

No. The audit is designed to run inside your own clinical system — we work on-site or in your screens, so the searches execute against your record and nothing patient-identifiable is exported. This is the lowest-risk delivery pattern and it's the default. If any step ever needed data handled off-site, it would be pseudonymised, minimised to only what the task requires, and governed by a written Data Processing Agreement — but for the core discovery work, the honest answer is simply: nothing leaves.

It's the opposite of padding, and we'd refuse to do padding. The whole method is register accuracy, not inflation: we only ever surface patients who genuinely have a condition but were never coded for it. That improves their care — they get on the right recall and review pathways — and it corrects income that was already earned. We never add a code to a patient who doesn't meet the criteria; that would be a clinical, audit and fraud risk we won't go near. A clinician validates every code, we work to genuinely-affected patients only, and we stay inside NHS England's guardrails. Implausible prevalence jumps are what invite scrutiny — coding real, evidenced patients does the reverse: it makes your register defensible.

Not when it's done as accuracy, which is the only way we do it. Clawbacks and reviews are triggered by figures that don't add up — sudden, unexplained prevalence spikes. Because every patient we surface is genuinely affected, clinically confirmed and backed by evidence already in their record, your numbers become more defensible, not less. If a commissioner ever asked, the audit trail of who-found-what and who-confirmed-it is exactly the evidence you'd want to produce.

Very little, and it's front-loaded. Realistically it's a kickoff call and a one-page data pull by the practice manager to get us started — we carry the analysis, the setup and the compliance paperwork. The partners' main commitment is reviewing and signing off the coding candidates we surface, which is clinical work only a clinician can do, presented in plain English so it's quick to work through. Nobody has to learn new software or operate any tools; we run those and hand you the reviews.

The initial MOT — the audit that tells you what's recoverable — is free, and commits you to nothing. No catch: we're confident enough in finding value that we'll show you the numbers before you pay anything. If the audit shows little to recover, you've spent an afternoon of the manager's time and you owe us nothing. If you proceed, you choose between a flat monthly fee (predictable, all-in, scaled to list size) or a share of what we actually recover in Year 1 against an agreed, jointly-signed baseline, after which you keep 100%. Consultancy is standard-rated for VAT, so we quote VAT-inclusive to avoid surprises.

Then we tell you that, plainly, and you pay nothing for the MOT. We'd genuinely rather be the people who were straight with you than the people who oversold. In practice it's rare to find a practice with nothing slipped through — but if that's you, that's a clean bill of health, not a sales problem, and we'll say so.

The lines are drawn deliberately. Your practice is, and remains, the data controller for all patient and staff data. DoctoriumGP acts only as a data processor on your documented instructions, under a written DPA — we never become the controller of patient data. Every tool we bring in is a named sub-processor with UK data residency. On the paperwork most practices never have time to complete: we author the DPIA, we produce the DCB0160 clinical-safety case and support the Clinical Safety Officer role, and we update your privacy notices. That's part of the service — and it's the single most common thing practices are missing when an inspector asks.

We name it openly rather than hope you don't notice. Dr Gemma Lewis is our Clinical Director and is married to the person who runs our commercial side, so where that relationship touches a decision in your practice it gets declared and recused in the normal way — recorded on your conflicts register, with clinical sign-off resting with your own partners, not with us. The service is contracted and invoiced at arm's length through PHW Ltd, keeping the commercial and clinical roles cleanly separated. And if it's ever cleaner for you, we'll structure the formal case study around a neighbouring practice instead. The principle: any conflict is surfaced and handled transparently, never glossed over.

A current, live gap for most practices — and the one we'll often check first, because it's free. Practices are already using AI — ambient scribes, triage, coding and document AI — usually bought tool-by-tool with no governance wrapper. CQC has published guidance on AI in general practice and now considers AI governance at inspection, so being able to show it's governed is no longer optional. Our half-day AI Readiness sweep produces five things you can hand an inspector: a register of every AI tool, a governance check per tool (DTAC, DPIA, clinical-safety, lawful basis, human-oversight), a one-page staff acceptable-use policy, a named owner with a review date, and a CQC-evidence pack. To be clear on scope: we organise and evidence — we don't give legal advice or certify medical-device compliance. This area moves fast, so we re-verify the position before each engagement.

Keep both — we work alongside them, not instead of them. Ardens optimises coding inside your clinical system; a strong manager runs the operation; your accountant does the books. None of them join your compliance to your income, carry an independent outside audit across the whole picture, or recover what has already slipped through. We're the accountable layer that connects those tools and finds what falls between them. And rather than making the manager's job redundant or heavier, the aim is to take the repetitive chasing, coding and spreadsheet-juggling off their plate and give them a live picture to manage from — which makes their work easier and their value to the partners more visible.

A medical accountant works on the books after the income has (or hasn't) come in — they don't go into the clinical record to find the patients who were never coded, and they can't clinically validate a diagnosis. A pure recovery firm chases a number, often on hard contingency, with no clinical oversight and no interest in whether your registers are actually right. We sit upstream of the accountant and we're clinically fronted: a serving GP partner validates the clinical findings, we author the data-protection and clinical-safety paperwork, and we treat register accuracy as a care-quality and inspection issue, not just a revenue line.

There are NHS support programmes — Time for Care / GPIP and similar — and we work alongside them, not against them. They improve your flow (triage, demand and capacity) for free, then leave. They don't recover lost income, build or integrate your automation, author your DPIA and DCB0160, or measure partner workload. If you've already been through one, you're often exactly the practice that benefits most from the layer it was never designed to deliver.

You keep clinical control throughout: a clinician signs off every code, so nothing changes in a patient record without a doctor deciding it should. We advise and enable; you decide. On our side, we carry professional indemnity for the consultancy, the controller/processor split keeps the data liabilities where they belong, and the contract sets all of this out plainly so there's no ambiguity if a question is ever raised.

To the MOT itself, nothing — it's free and you can walk away. If you proceed, it's a rolling arrangement (monthly or annual) with a sensible notice period, not a lock-in. On the share-of-results model you keep 100% of the recovered income after Year 1; most practices then move to a light monitoring retainer so the gains don't drift back over time, but that's optional and entirely your call.