Proficiency Testing Score Calculator
Enter the graded challenges for each analyte in an event from your CMS-approved PT program and get the analyte and event scores, whether each is satisfactory (80%, or 100% for ABO and D typing, antibody detection and compatibility testing), and whether the last three events add up to unsuccessful performance under 42 CFR 493. A second panel checks one result against its acceptance limit and works out its SDI.
PT score calculator — assaydeck.com
Score a PT event
Pre-filled with a made-up chemistry event. Enter each analyte's graded challenges from your evaluation report.
Pass mark 80% per analyte and for the event §§ 493.837, 493.841, 493.843, 493.845, 493.851
Testing event score
88%
| Analyte | Score | This event | History |
|---|---|---|---|
| Glucose5 of 5 | 100% | PassSatisfactory | PassClear |
| Potassium4 of 5 | 80% | PassSatisfactory | Watchthe next event |
| Sodium3 of 5 | 60% | FailUnsatisfactory | FailUnsuccessful |
| Creatinine5 of 5 | 100% | PassSatisfactory | PassClear |
| Cholesterol, total5 of 5 | 100% | PassSatisfactory | PassClear |
| Whole event22 of 25, pass mark 80% | 88% | PassSatisfactory | PassClear |
Check one result against its limit
Optional. Take the target value and peer SD from the program's evaluation report.
Limit: “Target value ±8% or ±6 mg/dL (greater)” § 493.931
- Glucose (not home-use meters): PassAcceptable
- Acceptable range 56.0 to 68.0 mg/dL around a target of 62, so ±6.0 (the fixed ±6 mg/dL is the greater part here).
- Your result is +5.0 from target, using 83% of the allowed error.
- SDI +2.08, in Bio-Rad's unacceptable band (2.0 or greater). CLIA grades PT on the limit above, not on the SDI.
Bench sheet
- Pass mark80% analyte and event
- Event score88%
- Acceptable responses22 of 25
- Unsatisfactory analytes1
- Unsuccessful analytes1
- Watch at the next event1
- Overall event historyClear
- Acceptance limits in force sinceJuly 11, 2024
How to use it
- Pick the specialty or subspecialty the event covered. That sets the pass mark: 80% for most, 100% for three blood bank tests.
- For each analyte, enter the challenges and the number marked acceptable exactly as your evaluation report scores them. Most events send five samples per analyte. Programs mark ungraded challenges in their own way (CMS notes that a non-consensus sample can carry an “artificial score of 100%”), and every ungraded result still needs your own accuracy check (see the guide to CLIA proficiency testing).
- Set the result of the last two events for each analyte. That is all the history the rule needs.
- If the results never reached the program, or reached it late, choose “Not sent or late”: CLIA scores that event 0.
- Optional: check one result against its acceptance limit, with an SDI if your report gives the peer group SD.
- Print the sheet for the PT binder, or copy the table into your corrective-action log.
Everything runs in this browser tab. Nothing you enter is sent to AssayDeck or stored.
What it calculates
Two scores, both written into Subpart I as formulas (§ 493.931(c)(4) and (5) for routine chemistry, the same in § 493.941 for hematology):
- Analyte score = acceptable responses for the analyte ÷ total challenges for the analyte × 100.
- Testing event score = acceptable responses for all challenges ÷ total number of all challenges × 100.
Each specialty's standard then sets the pass mark. A score equal to the mark passes: the rule asks for “a score of at least 80 percent”. So four of five is 80% and satisfactory; three of five is 60% and unsatisfactory.
| Specialty or test | Each analyte | Whole event | Rule |
|---|---|---|---|
| Chemistry, hematology, endocrinology, toxicology or general immunology | 80% | 80% | §§ 493.837, 493.841, 493.843, 493.845, 493.851 |
| ABO group and D (Rho) typing | 100% | 100% | § 493.859 |
| Unexpected antibody detection | event only | 100% | § 493.861 |
| Compatibility testing | event only | 100% | § 493.863 |
| Antibody identification | event only | 80% | § 493.865 |
Then the history. One failed event is unsatisfactory. The word that brings sanctions is unsuccessful, and § 493.2 defines it by a pattern across events: “Unsuccessful proficiency testing performance means a failure to attain the minimum satisfactory score for an analyte, test, subspecialty, or specialty for two consecutive or two of three consecutive testing events.” (42 CFR § 493.2)
So the tool looks at this event and the two before it. Unsatisfactory now and in either of the last two events means unsuccessful. Unsatisfactory now, or at the last event, means the next event decides: one more miss completes the pattern. That is the “Watch the next event” line. The same test runs on the overall event score, because the standards count both (§ 493.841(f), (g)).
CMS's PT and PT referral brochure puts it the same way: unsatisfactory performance “for the same analyte in either 2 consecutive testing events, or 2 out of 3 testing events”.
Takeaway
Worked example
The tool opens on this made-up routine chemistry event: five analytes, five challenges each. The history columns read oldest first (U = unsatisfactory, S = satisfactory).
| Analyte | Acceptable | Score | Two events ago, last, now | Result |
|---|---|---|---|---|
| Glucose | 5 of 5 | 100% | S, S, S | Clear |
| Potassium | 4 of 5 | 80% | S, U, S | Watch the next event |
| Sodium | 3 of 5 | 60% | U, S, U | Unsuccessful |
| Creatinine | 5 of 5 | 100% | S, S, S | Clear |
| Cholesterol, total | 5 of 5 | 100% | S, S, S | Clear |
The event score is 22 ÷ 25 × 100 = 88%, satisfactory. Sodium at 60% is unsatisfactory, and with a miss two events back it is unsuccessful: two of three consecutive events. Potassium passed this time at exactly 80%, but it missed last time, so one more miss at the next event makes it unsuccessful too. CMS's PT brochure asks for exactly that after an 80%: find out why the one sample fell outside the range, even though the score passed, and document it.
The result check opens on glucose, whose limit is “Target value ±8% or ±6 mg/dL (greater)”. At a target of 62 mg/dL, 8% is 4.96 mg/dL, so the fixed 6 mg/dL is the greater part and the range is 56 to 68 mg/dL. A result of 67 is acceptable. With a peer SD of 2.4 its SDI is +2.08, which Bio-Rad's QC statistics workbook would call unacceptable for peer data. The grade is a pass; the bias is still worth a look at your QC: plot the level on a Levey-Jennings chart (see also how often to run QC).
The acceptance limits the check uses
These are the limits from the July 2022 final rule (87 FR 41194), in force since July 11, 2024, quoted as the regulation words them. Where two parts are given, the greater one applies at that target. The same limits serve as total allowable error for a sigma metric.
| Analyte | Criteria for acceptable performance | Rule |
|---|---|---|
| Glucose (not home-use meters) | Target value ±8% or ±6 mg/dL (greater) | § 493.931 |
| Potassium | Target value ±0.3 mmol/L | § 493.931 |
| Sodium | Target value ±4 mmol/L | § 493.931 |
| Cholesterol, total | Target value ±10% | § 493.931 |
| Cholesterol, HDL | Target value ±20% or ±6 mg/dL (greater) | § 493.931 |
| Creatinine | Target value ±10% or ±0.2 mg/dL (greater) | § 493.931 |
| Leukocyte count | Target ±10% | § 493.941 |
| Hemoglobin | Target ±4% | § 493.941 |
| Platelet count | Target ±25% | § 493.941 |
| Prothrombin time (seconds or INR) | Target ±15% | § 493.941 |
| Partial thromboplastin time | Target ±15% | § 493.941 |
The target is the program's, not yours. For a peer group of 10 or more, § 493.2 makes it the mean of all participant responses after removing outliers beyond three SD, or a reference-method mean, or the peer group's mean. The tool counts a result exactly on the limit as acceptable; for a result that close to the edge, the program's evaluation report is the one that counts. The full list of limits is in § 493.931 and its sister sections. To set two of the programs side by side, see CAP vs API proficiency testing.
What 42 CFR 493 says
A missed or late event scores 0 (§ 493.841(c), (d)). CMS may give consideration only when patient testing was suspended for the whole PT window, the lab told its inspecting agency and the program within that window, and it took part in the two events before.
Every unsatisfactory score needs documented remedial action, kept “for two years from the date of participation”; if the cause was not a missed event, the lab must also “undertake appropriate training and employ the technical assistance necessary” (§ 493.841(e)). Your own QC for the analyte, summarised in the QC statistics calculator, is a natural first place to look.
Unsuccessful performance brings sanctions under § 493.803. For a first unsuccessful performance CMS may direct training or technical help instead, unless there is immediate jeopardy, no evidence of correction, or a poor compliance history. A lab suspended for the analyte needs “sustained satisfactory performance on two consecutive proficiency testing events” to come back, and the suspension runs “not less than six months” (§ 493.807).
An ungraded challenge is not a pass. When a score “does not reflect laboratory test performance”, the lab “must verify the accuracy” itself and document it (§ 493.1236(b), (d)).
Never send a PT sample out
A lab “must not send proficiency testing samples or portions of proficiency testing samples to another laboratory” for testing it is certified to do itself; intentional referral can cost the certificate for at least a year (§ 493.801(b)(5)). The full rules, from enrollment on, are in our guide to CLIA proficiency testing.