Spravato Medicare Outliers / Risk flags explained
What each Spravato risk flag means
Every provider in the Spravato Medicare outlier report carries a compliance risk score built from the flags below. Each flag is computed from the public CMS Part B file — nothing is inferred from a second source — and each section shows a real provider from the 2024 data as an example. A flag is a prompt to look closer, not an accusation.
Sell into this market? Create a free account to turn these providers into a territory call plan — 3 free runs, no card.
How the score works
The compliance risk score runs from 0 to 100. Up to about 45 points come from where a provider's sessions per patient sit against the 56-session label ceiling. Each hard flag — twice-weekly maintenance, above the weekly ceiling, and same-day duplicates — adds 20 points. Each soft flag adds 8. The bands are Low (under 25), Moderate (25–49), Elevated (50–69) and High (70 and above).
The thresholds come from the FDA esketamine label: twice weekly for the first four weeks, weekly through week eight, then every one to two weeks. A label-adherent year lands near 34 sessions per patient; never tapering anyone caps out around 56.
2x/wk maintenance
What it checks: A provider averages more than 70 Spravato sessions per Medicare patient in a calendar year.
The FDA label never calls for twice-weekly dosing beyond the first month. Sustaining more than 70 sessions per patient means most of the panel stayed on induction dosing all year — the exact profile federal auditors look for when they review esketamine claims.
Legitimate explanations: A very small panel of patients who each failed to taper could push the ratio up, but the label expects tapering to be attempted after response.
No provider in the 2024 data carries this flag.
Above weekly ceiling
What it checks: Sessions per patient exceed 56 — the ceiling for a provider who never tapers a single patient off weekly dosing.
Getting past 56 sessions per patient is not possible under the label without keeping patients on twice-weekly dosing past the induction window. It is the clearest single number that a program's dosing does not follow the label.
Legitimate explanations: Patient-count suppression by CMS can make the divisor imprecise, which is why the report also shows a low–high range rather than one number.
No provider in the 2024 data carries this flag.
Same-day duplicates
What it checks: The provider billed more Spravato service lines than the number of distinct patient-days — meaning at least one patient was billed twice for the same day.
One session per patient per day is the only pattern the label supports. Duplicate same-day billing can signal split claims, incorrect coding, or two sessions billed for one visit — all of which draw audit attention.
Legitimate explanations: Honest coding corrections and resubmissions can create duplicates in the raw file, so this flag is about patterns worth reviewing, not proof of double-billing.
No provider in the 2024 data carries this flag.
Not tapered
What it checks: Sessions per patient sit above the label-adherent ~34 but below the 56 weekly ceiling.
This is the mildest flag. It usually means patients stay on weekly maintenance instead of stepping down to every-other-week dosing, which the label permits only when clinical benefit is weighed against less frequent dosing.
Legitimate explanations: Many clinicians keep responders on weekly dosing for good reasons; the flag marks a billing pattern above the taper-adherent baseline, nothing more.
No provider in the 2024 data carries this flag.
Possibly above ceiling
What it checks: Depending on how patients receiving both dose strengths (56mg and 84mg) are counted, sessions per patient may exceed 56.
CMS reports patient counts per code, not per provider, so a patient on both doses is counted twice in the raw file. When the true divisor is ambiguous we show the range and flag the possibility instead of asserting it.
Legitimate explanations: This flag often disappears entirely once the true patient count is known — treat it as 'worth a second look', not a finding.
No provider in the 2024 data carries this flag.
Non-psych type
What it checks: The billing provider's specialty in the CMS file is not psychiatry, a psychiatric nurse practitioner, physician assistant or a similar mental-health credential.
Spravato is indicated for treatment-resistant depression, so most prescribing comes from psychiatric clinicians. Billing under an unexpected specialty can mean the rendering NPI belongs to someone other than the prescriber directing care.
Legitimate explanations: REMS-certified treatment centers and multi-specialty groups legitimately bill under internists, anesthesiologists or facility staff who supervise sessions. Common and usually fine — but worth understanding before you read the ratio.
Real example · 2024
David Boguslavsky · Bridgewater, NJ
1,214 sessions across 39 patients — 20.6 sessions per patient — with $1,187,074 in Medicare payment. Risk score 23.
Org NPI
What it checks: The billing NPI is an organization (Type 2) NPI rather than an individual clinician's NPI.
A group NPI blends every prescriber at the organization into one ratio. The sessions-per-patient number may reflect five clinicians following the label perfectly, or one outlier hidden inside the group — the file cannot tell them apart.
Legitimate explanations: Multi-prescriber clinics are supposed to bill this way. The flag is a reminder that the numbers describe an organization, not a person.
No provider in the 2024 data carries this flag.
Frequently asked questions
Does a flag mean a provider did something wrong?
No. Flags are computed from public Medicare billing patterns and are a prompt to look closer, not an accusation. Several flags have legitimate explanations — a group NPI is normal for a multi-prescriber clinic, and REMS-certified treatment centers can bill under credentials that are not a psychiatric specialty.
Where do the thresholds come from?
The FDA esketamine label. Induction is twice a week for four weeks, weeks five to eight are weekly, and week nine onward is every one to two weeks. A label-adherent year lands near 34 sessions per patient; never tapering anyone caps out around 56. Anything above 56 implies twice-weekly maintenance dosing all year.
How is the compliance risk score calculated?
The score runs from 0 to 100. Up to about 45 points come from where sessions per patient sit against the 56-session label ceiling; each hard flag (twice-weekly maintenance, above the weekly ceiling, same-day duplicates) adds 20 points, and each soft flag adds 8. Bands are Low under 25, Moderate 25–49, Elevated 50–69 and High 70 and above.
What data are the flags computed from?
Only the CMS Medicare Physician & Other Practitioners by Provider and Service file — the rows for Spravato codes G2082 and G2083, aggregated by rendering NPI. Nothing is inferred from a second source, and CMS suppresses any provider-code combination with fewer than 11 patients.
See the flags applied to every provider in the Spravato Medicare outlier report, or explore spend by state on the Spravato market dashboard. To see how scores moved between data years, open the risk score trends.
Build the list for one county and see whether the names are right.
Ask in plain language. RepVector finds the universe, enriches each target from its own website, scores it 0–100, and hands you a ranked call list with phones, addresses and named contacts.
3 free runs — no card required.