How Auxilium scores claims integrity
Auxilium gives a ministry an integrity score out of 100. This is how it is calculated, in full — the method is published because a score nobody can check is a score nobody should act on.
Start at 100 and subtract named findings
Every ministry starts at 100. Each rule that matches subtracts a stated number of points and records why, in a sentence with the actual figures in it. The deductions sum to the score — you can add them up by hand and get the same answer.
Rules fall into four groups: where the money went, where it went instead, whether denials follow your published guidelines, and whether claims actually move. Each rule also stores the documented failure it was written from, so a ministry that disagrees with a finding can argue with a specific case rather than an abstraction.
What moves the score most
The heaviest deductions, and why they are weighted that way.
| Capability | Auxilium | Alternative | Notes |
|---|---|---|---|
| Months with contributions received and nothing shared | ✓ | ✗ | Heaviest single finding. A run of these is the pattern behind federal action in this category. |
| Share ratio far below the ACA floor | ✓ | ✗ | Weighted by severity — below 25% is treated far more seriously than 45%. |
| Denials applying guidelines published after the member joined | ✓ | ✗ | Holding someone to rules they never agreed to. |
| Related-party payments | ✓ | ✗ | Scaled by share of the pool. Small and disclosed is a note; large is a finding. |
| Overhead exceeding sharing | ✓ | ✗ | |
| Claims past their turnaround commitment | ✓ | ✗ |
One calibration rule worth knowing
Rate-based findings require a minimum sample before they are treated as rates. "One of one denials had no guideline citation" is a 100% rate carrying almost no information, and without a floor a small ministry with a single lapse would score identically to an organization diverting most of its pool.
Below that floor, Auxilium scores the count instead of the rate. A score that cannot distinguish a bad week from systemic diversion is one nobody will trust twice, and a compliance number people stop believing is worse than no number at all.
Bands
Healthy is 85 and above. Watch is 70 to 84. Concern is 50 to 69. Critical is below 50.
The bands matter more than the exact number. A ministry at 72 and one at 78 are in the same situation and should do the same things about it; a ministry at 45 is in a different situation entirely. Treating the score as a precise ranking rather than a band invites exactly the kind of number-chasing that makes a metric useless.
What the score deliberately does not do
It does not accuse anyone of anything. Every finding describes something observable in the ledger or the claims record — a ratio, a run of months, a citation that does not hold up. Intent is not something software can assess, and a system that implied otherwise would be both wrong and dangerous.
It does not compare you to other ministries. There is no percentile, no league table, and no benchmark against peers. Partly that is because reliable cross-ministry data does not exist in this category, and partly because a ratio that is merely better than average is not actually a defence if the average is poor.
And it does not weight recency the way an operational dashboard would. A single bad month barely moves the score, because the trailing window exists precisely so that ordinary variation does not read as a crisis. The findings that move it most are patterns sustained across months.
How to use it
Look at the findings, not the number. The score is a summary; the findings are the work. Each one names the amounts and the members involved, and most can be resolved directly — assign the unowned claims, re-review the denials that cite nothing, decide the overdue appeals.
Recompute after you act, and watch the specific finding disappear. A score that moves because you fixed something is the only kind worth tracking.