Short Blessed Test Calculator
Short Blessed Test Calculator
Six items, weighted error scores, 0 to 28, from Katzman’s 1983 short Orientation-Memory-Concentration Test. Higher is worse. The same six items and weights are what the separately copyrighted 6-CIT scores, which is why this page cites the 1983 original.
Short Blessed Test (short OMC)
6 weighted items, 0 to 28Year correct (0), month correct (0), time wrong (3), one error counting backwards (2), one error on months in reverse (2), two of the five address elements wrong (4)
Scoring
Maximum 4 + 3 + 3 + 4 + 4 + 10 = 28; higher is worse
- the weights and the error caps
- year ×4 (max 1 error), month ×3 (max 1), approximate time ×3 (max 1), counting 20 to 1 ×2 (max 2), months in reverse ×2 (max 2), recall of the five-element phrase ×2 (max 5). That last item alone carries 10 of the 28 points
- higher is worse
- this is an ERROR score, unlike the Abbreviated Mental Test Score, where higher is better. Both instruments are used in the same clinics and the direction is inverted between them. An alternative positive-scoring form exists in which 28 minus the error score is recorded and above 20 is normal; if your notes carry a bare number, establish which form produced it
- published bands
- 0 to 8 normal to minimal impairment, 9 to 19 minimal to moderate, 20 to 28 severe. A separate published cut-off is above 10 for cognitive impairment in Alzheimer disease, with sensitivity 88 per cent and specificity 94 per cent in a cohort of 133 people of mean age 64.5 years (range 37 to 91) with mixed diagnoses
- what a positive result is worth
- take those published figures and apply them. At a prevalence of 30 per cent — roughly what a geriatric inpatient population carries, and close to the 34 per cent abnormal cognition in the Abbreviated Mental Test’s own validity cohort — the positive predictive value is 86 per cent. At a prevalence of 5 per cent it is 44 per cent: more than half of positive results would be wrong, from exactly the same sensitivity and specificity. Negative predictive value goes the other way, 99 per cent at the lower prevalence. Sensitivity and specificity belong to the test; predictive value belongs to the population, and the 2 × 2 arithmetic is worth doing before acting on a screen
- test-retest, not inter-rater
- the only reliability figure found is Fuld’s 1978 test-retest in 18 patients with Alzheimer disease, retested at three weeks and scoring within 4 points of the original. No inter-rater reliability is recorded for this instrument in either registry read, and a test whose items are factual and whose scoring is arithmetic has less rater variance to measure than one requiring judgement — but “not measured” is not “good”
- unreachable totals
- 1 and 27 cannot occur. The smallest non-zero increment is 2, so 1 is impossible; and 27 would require dropping exactly one point from the maximum, which these weights cannot do. Every other integer from 0 to 28 is reachable
- licensing, and the 6-CIT
- Katzman and colleagues published this in the American Journal of Psychiatry in 1983. Two independent registries record it as free with no publisher, and the full instrument is reproduced without a permission notice. The Six-item Cognitive Impairment Test scores the SAME six items with the SAME weights to the same maximum of 28, but the Kingshill Research Centre owns the copyright to its Kingshill Version 2000 and allows free use to healthcare professionals — a grant limited to a class of user, which a public advertising-supported site cannot claim. This page therefore reproduces and cites Katzman’s 1983 original and not that version. The 6-CIT’s own published threshold is 8, with 0 to 7 normal, and overall sensitivity 90 per cent and specificity 100 per cent reported at the 7/8 cut-off, falling to 78 per cent sensitivity in mild dementia
Worked example
Year correct (0), month correct (0), time wrong (3), one error counting backwards (2), one error on months in reverse (2), two of the five address elements wrong (4)
0 + 0 + 3 + 2 + 2 + 4 = 11 points
11 sits in the published 9-to-19 band, minimal to moderate impairment, and above the separate Davous cut-off of more than 10
Watch what the weighting does. Getting the YEAR wrong and nothing else scores 4 — worse than making one error on both attention items put together, which scores 4 as well, and worse than a single address element missed, which scores 2. An unweighted count of six items would call all three of those patients identical
Apply Davous's published 88 per cent sensitivity and 94 per cent specificity at a prevalence of 30 per cent: of 1,000 people, 264 true positives and 42 false positives, so a positive predictive value of 86 per cent
The same two figures at a prevalence of 5 per cent: 44 true positives and 57 false positives, a positive predictive value of 44 per cent. The test has not changed. More than half the positives are now wrong, and that is the single most useful thing to know before screening a low-prevalence population
Change the address item to all five elements wrong and the total becomes 0 + 0 + 3 + 2 + 2 + 10 = 17, still inside the same band. One item carries 10 of the 28 points
The six items, their weights and their error caps
| Item | Maximum errors | Weight | Maximum points |
|---|---|---|---|
| What year is it? | 1 | ×4 | 4 |
| What month is it? | 1 | ×3 | 3 |
| About what time is it, within an hour? | 1 | ×3 | 3 |
| Count backwards from 20 to 1 | 2 | ×2 | 4 |
| Name the months in reverse order | 2 | ×2 | 4 |
| Repeat the five-element name and address | 5 | ×2 | 10 |
Published figures, and what they are worth at two prevalences
| Figure | Value | Source and population |
|---|---|---|
| Score bands | 0–8 · 9–19 · 20–28 | Normal to minimal, minimal to moderate, severe impairment |
| Cut-off for cognitive impairment | above 10 | Davous et al. 1987, Alzheimer disease |
| Sensitivity | 88% | 133 people, mixed diagnoses, mean age 64.5 (range 37–91) |
| Specificity | 94% | the same cohort |
| Positive predictive value at 30% prevalence | 86% | this page’s arithmetic from the two figures above |
| Positive predictive value at 5% prevalence | 44% | the same two figures, a different population |
| Negative predictive value at 5% prevalence | 99% | the same two figures |
| Test-retest | within 4 points at 3 weeks | Fuld 1978, 18 patients with Alzheimer disease |
| Inter-rater reliability | not reported | neither registry read records one |
A weighted error score, and the test the 6-CIT is
Katzman and colleagues published this in 1983 as a short form of the Blessed Orientation-Memory-Concentration test: six questions, each error multiplied by a weight, totalling 0 to 28 with higher worse. The weights are the point. Getting the year wrong costs four points and missing one of five address elements costs two, so a patient disoriented in time and a patient with patchy recall are not interchangeable the way they would be on a count of right answers. The delayed-recall item carries ten of the twenty-eight available points, which is where most of the test’s discriminating power sits, and it only works if the five-element phrase is given at the beginning.
The published banding is 0 to 8 normal to minimal impairment, 9 to 19 minimal to moderate and 20 to 28 severe, with a separate cut-off from Davous and colleagues in 1987 of above ten for cognitive impairment in Alzheimer disease, at eighty-eight per cent sensitivity and ninety-four per cent specificity in a hundred and thirty-three people of mean age sixty-four. Those figures are the starting point for the only calculation that makes a screening score interpretable. At a prevalence of three in ten the positive predictive value is eighty-six per cent. At one in twenty it is forty-four per cent — more than half of positive results wrong, from exactly the same sensitivity and specificity, because predictive value is a property of the population and not of the test. Screening a low-prevalence group with any instrument produces mostly false positives, and that is arithmetic rather than a criticism of this test.
This page cites Katzman’s 1983 original rather than the Six-item Cognitive Impairment Test, and the reason is worth stating plainly because the two are the same six items. The 6-CIT scores the year at four, the month at three, the time at three, counting backwards at two to a maximum of two errors, months reversed at two to a maximum of two errors and address recall at two per element to a maximum of five — the same twenty-eight. But the Kingshill Research Centre owns the copyright to its Kingshill Version 2000 of the 6-CIT and allows free use to healthcare professionals, a grant limited to a class of user that a public advertising-supported site cannot claim. Katzman’s original is an earlier journal publication that two independent registries record as free with no publisher, and that is what is reproduced here.
Two practical points. The direction is inverted relative to the Abbreviated Mental Test Score, where higher is better, and both are used in the same clinics; an alternative positive-scoring form of this test also exists, in which twenty-eight minus the errors is recorded and above twenty is normal. And a cognitive screen of any kind answers the wrong question in an acutely unwell patient: it cannot distinguish dementia from delirium, and an instrument with an acute-change item can.
A screening score is not a diagnosis: a published sensitivity is a property of the instrument in the population it was validated in, not a statement about this patient. This page reports published figures and recommends no action. Every weight, cut-off and outcome figure here comes from a named derivation cohort, and cohorts differ in case mix, era, coding and outcome definition; where your own institution’s protocol or analysis plan differs, it takes precedence.
Frequently asked questions
How is the Short Blessed Test scored?
As a weighted error count. Each error on the year scores 4 (maximum one error), the month 3 (one), the approximate time 3 (one), counting backwards from 20 scores 2 per error to a maximum of two, months in reverse the same, and the five-element name and address 2 per element to a maximum of five. Total 0 to 28, and higher is worse.
What score indicates impairment?
The published bands are 0 to 8 normal to minimal, 9 to 19 minimal to moderate, 20 to 28 severe. A separate published cut-off is above 10 for cognitive impairment in Alzheimer disease, at 88 per cent sensitivity and 94 per cent specificity in 133 people of mean age 64.5 years. A screening score is not a diagnosis: a published sensitivity is a property of the instrument in the population it was validated in, not a statement about this patient.
If the test is 88 per cent sensitive, is a positive result usually right?
Only where the condition is common. With 88 per cent sensitivity and 94 per cent specificity, the positive predictive value is 86 per cent at a prevalence of 30 per cent and 44 per cent at a prevalence of 5 per cent. The instrument is identical in both cases. Predictive value belongs to the population being screened, which is why screening a low-prevalence group produces mostly false positives.
Is the Short Blessed Test the same as the 6-CIT?
The same six items, in the same order, with the same weights and the same maximum of 28. They differ in their thresholds and in their licensing. The 6-CIT’s own published cut-off is 8, with 0 to 7 normal, and 90 per cent sensitivity with 100 per cent specificity reported at that cut-off. Its Kingshill Version 2000 is copyright of the Kingshill Research Centre, free to healthcare professionals; this page reproduces and cites Katzman’s 1983 original instead.
Which totals are impossible?
1 and 27. The smallest non-zero increment is 2, so a total of 1 cannot occur; and 27 would mean dropping exactly one point from the maximum, which these weights cannot do. Every other integer from 0 to 28 is reachable.
Related calculators
References
- Katzman R, Brown T, Fuld P, Peck A, Schechter R, Schimmel H. Validation of a short Orientation-Memory-Concentration Test of cognitive impairment. Am J Psychiatry. 1983;140(6):734–9.
- Internet Stroke Center. Short Orientation-Memory-Concentration Test of cognitive impairment, reproduced by the Shirley Ryan AbilityLab Rehabilitation Measures Database (sralab.org), citing Katzman et al. 1983 (accessed 9 October 2026).
- Shirley Ryan AbilityLab. Short Orientation-Memory-Concentration Test of Cognitive Impairment, Rehabilitation Measures Database. sralab.org (accessed 9 October 2026).
- University of Utah College of Health. Short Blessed Test assessment rating form, citing Katzman et al. 1983. health.utah.edu (accessed 9 October 2026).
- NHS Scotland Right Decision Service. 6-CIT (Six-item Cognitive Impairment Test). rightdecisions.scot.nhs.uk (accessed 9 October 2026).
Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/