The move from GPMP/TCA to the GPCCMP items changed the claiming, but not the real problem: a good chronic condition management plan takes twenty minutes of thinking and two hours of typing if the workflow is wrong. The fix isn't typing faster — it's a workflow where the writing happens once, in the right stage, with the clinician as reviewer rather than transcriber.
Key takeaways
- The GPMP (721) and TCA (723) items were consolidated into the GPCCMP items — 965 (standard) and 967 (Aboriginal and Torres Strait Islander patients) — with nurse/AHP support under 10997. Confirm current descriptors on MBS Online.
- Run plans as a four-stage workflow: pre-consult data pull, in-consult capture, structured drafting, clinician review and sign-off.
- The nurse does the heavy lifting on history, measures and goal prompts; the GP confirms, decides and approves.
- Document structuring tools can draft — they cannot decide. Docsert AI works on de-identified clinical content only, and the clinician reviews and approves every document.
- File it the same day. An unfiled plan is a plan that will be repeated.
What changed with the GPCCMP
Under the MBS changes, the old GPMP (item 721) and Team Care Arrangement (item 723) were consolidated into a single General Practice Chronic Condition Management Plan, claimable as item 965, with item 967 for Aboriginal and Torres Strait Islander patients. Practice nurses and Aboriginal health practitioners continue to support preparation and review under item 10997. The paperwork still has to cover what the old plans did — patient goals, problems and needs, actions the practice will take, and referrals to allied health where relevant — so the workflow below works regardless of which descriptor era your templates came from. Item numbers and rules change; check the current descriptors on MBS Online before claiming.
Stage 1 — Pre-consult: the nurse assembles the raw material
Most of a plan is history and measurements that don't need the GP in the room. Before the plan consult, the practice nurse pulls together: recent path results, current medications and allergies, last readings (BP, weight, HbA1c where relevant), existing specialist and allied health involvement, and the patient's own goals in their words. Ten minutes of nursing time here saves double that in GP time — and it's exactly the support item 10997 was designed for.
Stage 2 — In-consult: capture once, in the room
The plan consult itself is a conversation. Capture it as structured notes — problems, goals (patient-first: "walk the dog without stopping" beats "improve exercise tolerance"), agreed actions, referral intentions — rather than prose. Whether the capture happens in a template, a dictation app or typed notes matters less than one rule: what the patient said gets written while they're still saying it. Recreating the consult from memory that night is where the errors and the 9pm paperwork both come from.
Stage 3 — Drafting: structure the notes, don't author them
Turning raw notes into the practice's plan template is mechanical work — headings, sections, formatting — and it's the stage most suited to help. This is where Docsert AI fits for Australian practices: the clinician supplies the notes and Docsert structures them into a GP management plan or care plan draft that works with Best Practice, ready for review. Two boundaries are non-negotiable in our view. First, Docsert works with de-identified clinical content only — entering patient-identifiable information (PHI) is strictly prohibited, by design. Second, it structures documents; it doesn't make clinical decisions. The clinician reads, edits and approves every document before anything is filed or acted on. A draft is a suggestion with good formatting.
Stage 4 — Review, sign, file — same day
The GP reviews the draft against the consult notes, corrects anything that isn't theirs, and signs. Then it's filed in the record and (where applicable) the referral or shared-care copy goes out — the same day, while the consult is still fresh. The nightly catch-up isn't a workload problem, it's a queueing problem: plans that queue overnight get re-read, second-guessed and duplicated. A practice-wide habit of "drafted, approved, filed before close" is worth more than any template.
The nurse-GP split, in one line each
- Nurse: assembles data, captures history and measures, prompts goals, prepares the draft material.
- GP: confirms the content is right, makes the clinical decisions, approves and signs.
Practices that blur this in either direction get into trouble — plans the GP never really endorsed, or nurses doing unpaid transcription at 8pm. Named stages keep the roles clean, and they map neatly onto the run-sheet habits from our task run-sheet article.
Audit-readiness is a by-product
A plan done this way already contains what a documentation audit looks for: patient-voiced goals, measured baselines, actions with owners, dated review intervals, and evidence the clinician approved the final document. No extra compliance work — the workflow produces the evidence as it goes. For a fuller picture of who can claim what, our MBS items guide for practice nurses walks through the GPCCMP, health assessment and 10997 items in detail.
This article is general information for Australian general practice teams, not medical or billing advice. MBS item descriptors, numbers and claim rules change — always confirm against the current MBS Online item descriptors and your practice software's billing guidance before claiming. Patient care decisions remain the responsibility of the treating practitioner.
Frequently asked questions
What replaced GPMP items 721 and 723?
The GPMP (721) and TCA (723) items were consolidated into the General Practice Chronic Condition Management Plan (GPCCMP) items — 965 for standard plans and 967 for Aboriginal and Torres Strait Islander patients — with practice nurse and Aboriginal health practitioner support under item 10997. Always confirm current descriptors and rules on MBS Online before claiming.
Does Docsert AI make clinical decisions?
No. Docsert AI's intended use is document structuring only: it turns notes the clinician has supplied into a structured draft in the practice's template. It does not diagnose, decide or recommend — the clinician reviews and approves every document before anything is filed. Clinical decisions remain the treating practitioner's.
Is it safe to put patient information into Docsert AI?
Docsert AI is designed to work with de-identified clinical content — entering patient-identifiable information (PHI) is strictly prohibited, and the tool is non-PHI by design. You supply the clinical content; the plan is drafted, reviewed and approved by the clinician before it goes anywhere near a patient record.
Which document formats does Docsert AI produce?
Docsert AI produces structured GP management plan and care plan drafts that work with Best Practice, ready for the GP to review, edit and approve. The output follows the practice's own template structure rather than a locked proprietary format.
Turn consult notes into plan drafts, not homework. Docsert AI structures clinician-supplied notes into GP management plan and care plan drafts for your review — de-identified content only, non-PHI by design.