Assembling the Predetermination Packet: Evidence, Codes, and One Submission a Dentist Actually Signs

A dental administrator assembling a predetermination for a crown opens four things: the practice management software for the treatment plan and codes, the radiography store for the images, the periodontal chart for supporting measurements, and a blank box where the clinical narrative goes. None of the four knows the others exist. The packet takes twenty minutes to assemble and thirty seconds to reject if one radiograph is the wrong date.

This is the transformation we would run for that group. It is presented as a concept, built on the same method we use in client engagements, so the before, the after, and the path between them are all visible.

What does assembling a predetermination actually cost?

The honest answer for Canadian dentistry is that nobody has published a defensible number, and this study will not invent one. The widely circulated figures on dental no-show rates, recall percentages and reactivation revenue all trace back to vendors selling software into the category, which makes them marketing rather than evidence. What can be said with a source is the size of the surrounding problem.

The CAQH Index 2025, published in February 2026 and built on data from more than 600 provider organizations and health plans representing 63% of insured lives, puts annual US spending on staff time for routine administrative transactions between providers and plans at $83 billion, with a further $21 billion available from automating transactions that are still manual or partly manual. That index covers medical rather than dental transactions and is US data, so it sizes the category rather than pricing an Ontario practice. It is quoted here for exactly that, and no further.

The Canadian Dental Association has been more direct about the direction of travel, describing frequent claim denials as creating an administrative burden for dentists, and the predetermination EOB transaction only became available for certain carrier IDs in March 2026. The evidence assembly behind the transaction stayed manual throughout.

Before and after: the packet

Drag the handle. The before is the four-window assembly an administrator does by hand; the after is the drafted packet designed in its place.

The insurer’s requirements do not change and neither does the dentist’s judgment. What changes is that the evidence arrives already gathered, matched to the coded procedures, and shown in one place where a clinician can disagree with it in one click.

What we built

The concept above is the output of the same engagement steps we run in client work, and two of them are decisions a competent peer could argue with.

• We index the image store against the chart first, because that is where the real cost hides. Practices assume the bottleneck is typing the narrative; it is usually finding which of eleven radiographs from three visits actually shows the fracture. Indexing by tooth number and capture date is unglamorous and it is where most of the twenty minutes goes.
• We draft the narrative from the chart note and leave it explicitly unfinished. The system produces the factual scaffolding, dates, findings, prior treatment and the coded procedures, and stops before the clinical justification. This is the decision most likely to be challenged, because a fluent complete narrative demos better. It is also the one that keeps a clinical assertion attributable to a clinician.
• We refuse to auto-submit, even where the transaction would allow it. The packet is assembled, reviewed and submitted by a person.
• We build a hard gate on attachment matching: the packet cannot be marked ready while any attachment is unmatched to a coded procedure, and anything below the confidence floor is flagged rather than guessed.
• We wire evals from day one against the practice’s own historical predeterminations, scored on whether the assembled evidence matched what was actually submitted, and separately on whether the packets that got denied were denied for evidence reasons the system could have caught.

How it works

The practice management software stays the source of truth. When a treatment plan reaches a procedure class that requires predetermination, the system pulls the coded procedures, matches each one to the radiographs and periodontal measurements that support it, and drafts the factual portion of the narrative from the chart note. The dentist adds the clinical justification, confirms or replaces the attachments, and submits. Nothing transmits without that confirmation.

Because charts, images and narratives stay in Canadian infrastructure and processing runs with zero data retention, the design lines up with what PHIPA expects: a completed privacy impact assessment before deployment, an agreement naming the processor as an agent under section 17, and an audit log on every access to personal health information.

Why this pays back

The return is administrative hours converted into either fewer administrative hours or more chair-side support, and the practice decides which. Break-even for the first slice is about twenty-five predeterminations a month at a typical loaded administrative hourly cost, and the practice should substitute its own cost rather than trust that figure.

The limit worth saying before anyone signs: this does not reduce denials that are clinical disagreements. If an insurer does not accept that a crown is warranted, a better-assembled packet argues the same case more clearly and still loses. It reduces the denials caused by missing, mismatched or stale evidence, which in most practices is the larger share but is never the whole of it. A practice whose denials are mostly clinical should spend the money elsewhere.