The QA Compliance Report

Compliance Overview is a bar per criterion, sorted by rate with the lowest at the top and the exempt ones at the bottom. You are meant to read the first three rows and stop. Nothing below them is going to be the conversation.

What the figure on the right means

Each bar carries a count that says how the rate was arrived at. The shape tells you which kind of criterion it is:

The overall figure

Overall Caseload Compliance is the average of the criterion percentages — here is how that is worked out. The subtitle says how many went into it and names anything exempt: Average across 16 compliance criteria · 2 exempt, not scored. When nothing could be scored at all you get a dash and No compliance criteria applied to this caseload, which is not a zero.

Below the chart is the evidence: every active case, with its case type, client name, referral date, service-confirmation date and initial-session date, then a column per criterion. The client column stays put while you scroll sideways.

The five cells

Two columns that are not criteria

• Assessment names which instrument answered for that case — CANS or FAST. It is information, not a score.
• The dates at the left are what every deadline is counted from. If one of them is blank, the criteria that depend on it read N/A rather than No — there is nothing to count days from. That is worth checking first when a case looks oddly empty.

Column headings are shortened to fit — Scheduling Attempts (3 Wit… is the full criterion with its middle cut out. Hover the heading to see all of it.

Four things that are true of every criterion

1. Only active cases are scored. Exited and archived cases are out of the report, with one exception in the closure rules below.
2. Each one is measured from a date on the case — the referral (RRA), the service confirmation (SC), or the first in-person session (IFF/IFFT). If that date is missing the criterion reads N/A, never No. There is nothing to count days from.
3. Nothing is marked red before it is due. While a case is still inside the window it reads N/A. It turns to No on the day the window closes with nothing recorded. Initial Intake Session is the one exception, flagged below.
4. This week and this month are never scored. They show on the calendar so you can still act on them, and they stay out of the percentage — nobody is marked down for a week they have not finished.

How the percentage is worked out

Two kinds of criterion, counted two different ways — and that difference is the commonest reason a number looks wrong.

Overall Caseload Compliance is then the plain average of those criterion percentages. If Weekly Update to SW came out at 84% and Intake Documents at 88%, each contributes one share of the average — it is not everything that passed divided by everything that was checked.

Intake — the first days of a case

The first row is measured from the referral; everything else here from the service confirmation.

Counting the four business days

The confirmation day itself does not count, and weekends are skipped. A service confirmation on a Monday is due that Friday; on a Friday, the following Thursday; on a Wednesday, the next Tuesday. The seven days in Scheduling Attempts are plain calendar days, weekend included — so a Thursday confirmation gives you until Wednesday for the intake and Thursday for the attempts.

The missed-visit allowance does not extend the four days; it lets a later session still pass. A missed visit in the fortnight after the confirmation, followed by a session at any point, scores Yes. A missed visit with no session eventually is still a No.

Assessment — either CANS or FAST

These five still say CANS F in the data, but the headings read CANS F / FAST and either instrument passes them. A case is assessed with one or the other, and the Assessment column tells you which one answered for that family. Where a case has both, FAST wins — an assessment that was carried out beats a face sheet that may have been opened and never filled in.

Three other things also count as the assessment being done: an activity logged as Sent FAST … to SW or Sent CANS … to SW; a session under the FAST or CANS code; or, on a CANS case, the date on the CANS face sheet. For the transition it is a FAST-R or CANS-R session, or an activity naming a transition FAST or CANS.

The eighteen days are calendar days from the first in-person session, not from the referral or the confirmation. A first session on 1 March means the assessment is due by 19 March — the 19th passes, the 20th does not, and before then the row reads N/A.

Note what Submission Email does, because it looks inconsistent until you see the rule: the waiting period protects work that is not due yet, never work that is already finished. Mark the assessment done and the email is due with it, so that row goes red immediately rather than waiting out the eighteen days.

When the transition is due

Three different deadlines, depending on the instrument and the case type. The row tells you which one it used.

Public holidays are not in the data, so a deadline counted in business days can land a day or two early. That can flag a case that was actually on time — which you can see and correct — rather than letting a real miss through.

A case referred by a county that does not use CANS reads N/A on all five and says so above its calendar — Snohomish — CANS exempt. That is decided per case from the referring email on the face sheet, not by a setting, which is why one family can be exempt while the rest are not.

Every week the case is open

Three of these four share one clock: Sunday to Saturday weeks, starting one week after the service confirmation, with this week left out of the score. That one-week runway is deliberate — nobody is measured on correspondence for the week the case opened. Safety plan review uses its own weeks; see the note under the table.

On a single case each of these reads Yes when every week passed, No when none did, and Inconsistent when some did — which is what the grid cell shows. The bar in the chart is different: it adds up the weeks across the whole caseload, so 6/7 wks is everybody’s weeks, not one family’s.

One note on Safety/Crisis Plan Reviewed Weekly: its weeks are simply the weeks that contain an IFF, FFC or FFCT. There is no runway and no empty weeks — ten sessions spread over six weeks is a score out of six, and a case with none of those sessions reads N/A. Read it as weeks with a session in them, never as weeks open.

Switching something off part-way through the year removes only the weeks from that date forward. The earlier weeks keep their result, so you still see a real percentage over the time it applied.

Every month

Which month a monthly report counts for

It depends on the day you date the MS, against the due day — the 3rd unless your agency has changed it:

Move the due day to the 5th and that last one counts for March instead. A change applies from the day you make it forward; it cannot re-score a month that has already closed, and it never changes which months are counted — that comes from the BILL activities alone. A case with no billing is scored on no months at all and reads N/A, however long it has been open.

Monthly Practice Supervision is counted per clinician, not per case: every month from the case’s confirmation month to the end of last month, and a month passes if you logged a PSUP anywhere that month. Two supervision sessions in one month still pass one month, and every case on your list shows the same months passing or failing.

When the case ends

Case Closure asks two separate questions about the most recent IFFA, and a miss on either is a No: did contact resume within fourteen days, and is there an ES_NC on the case at all. The second is not time-boxed — it just has to exist. A case with no IFFA ever passes outright, because there is no no-contact clock running. So the two usual failures are a case that went quiet and was never formally closed, and one that was closed properly but has no ES_NC logged against it.

What counts as the end of service

Any activity whose code starts with ES_ — ES_NC for no contact with the client, ES_SG for completed service goals, ES_OTH for anything else. The code is what ends the case here, not the case status. A case usually stays open in your practice system while it winds down, so the status would not tell this criterion anything useful, and the ES_ row is written the day you actually finish.

The bill is the BILL activity — the one typically named like 5.2024 Billing and Report Sent. The criterion simply asks whether those two land within ten days of each other.

Where the report looks

Four ways something leaves the score

• N/A — not due yet, or no date to measure from. That one family drops out of that one criterion. The criterion still scores across everyone else.
• Exempt — switched off in Preferences. Work your agency does not do. For the weekly criteria only the weeks from that date forward come out; everything else greys out completely.
• Exempt — nobody uses PSUP. Monthly Practice Supervision disappears from the score for the whole practice.
• N/A — the referring county does not use CANS. All five assessment rows drop out for that family only.

So the number of criteria behind a score is not fixed. Two clinicians at the same agency can be scored on different numbers of criteria in the same week. The line under the overall figure always says how many it used, and the exported spreadsheet prints the exemptions in its header — which is what lets you reproduce a figure weeks later.

Everybody at once

The tab opens on All clients: one column per week across the months, and three rows per family.

Families are ordered worst first on the weekly-update rate, so the top of the list is where to look. Hovering any cell says what it found and what it wanted — including No SWE/WU sessions logged this week, which is the honest version of a blank square.

One family, month by month

Pick a client and it becomes a month calendar with a week column down the left: a tick for a passed week, a cross for a missed one, a tilde where one of the two passed and the other did not, and a dash for a week outside the scored span — before the service confirmation, or after discharge.

The tinted bands behind the days are the windows the criteria are measured against:

If it is empty

• No active cases to display. — nothing open for this person.
• No eligible weeks across the active caseload yet. — the cases are open but none has passed its service confirmation, so there is no week to score yet.

What this agency does carries a switch per capability, at two scopes — Whole practice, and this one therapist. A therapist setting overrides the practice for that person only.

CANS is not a switch here. It is decided per case from the referring county on the face sheet, which is why one family in a caseload can be CANS-exempt while the rest are not.

Monthly report deadline

Monthly report deadline sets the day of the month after which a monthly report counts as late — a whole number from 1 to 28, settable for the practice and overridable per therapist. The criterion renames itself to match, which is why it reads Monthly Reports to Admin by 3rd by default.

If the tab says the table has not arrived

This install has not received the QA preferences table yet. means exactly that — the next definitions update or app restart brings it. Scoring carries on unaffected in the meantime; every metric is being measured the way it always has been.

Download is on the Score tab, and only when there is at least one case. It writes a spreadsheet named for the therapist and the date, built on this machine — nothing is uploaded and no server is asked.

It carries the chart, the full grid with the colours intact, and a header block most exports would leave out: who it is for, how many active cases, when it was generated, the exemptions in effect, and the monthly report due day. The exemptions line prints even when there are none.

• It does not compare therapists. One caseload at a time, by design. A column failing across several people is still visible — you just have to open them.
• It is not a second source of truth. Everything here is read from your practice system. A session logged against the wrong week is wrong here because it is wrong there, and the repair is there.
• It measures documentation, not care. Every criterion is about a record existing in a window. That is worth saying out loud before the number is used in a performance conversation.
• A long-running trend can cross a rule change. The scoring spec is versioned. If you are comparing this quarter to last year, ask us whether the rules moved inside it.