Information about the CORE-OM

Page created for 2015 site, last updated 11/8/19, author CE, licence: Attribution 4.0 International (CC BY 4.0) .

“CORE-OM” please, not “CORE-34”!

Keeping to the original naming, which has the CORE-OM as the primary measure, and the CORE-10 showing that the latter is a shortend form with 10 items, keeps things clear on the web and in the literature.

The CORE-OM (Clinical Outcomes in Routine Evaluation – Outcome Measure) was the central measure in the CORE system, linking with the CORE-A, the complementary practitioner completed context setting tools to provide what we hoped would be a sound core for evaluation of change in psychological therapies.

The CORE-OM has 34 items, all with the same five level response choice, and covers the last seven days.  It was designed, and this was led partly by the commissioning specification, to cover four main domains: wellbeing, problems, functioning and risk and has 4, 12, 12 and 6 items focused on those domains respectively.  Of the risk items, four address risk to self and two address risk to others.  Overall, 18 items were “tuned high” and 16 tuned to a lower range of disturbance/distress.  Eight items are positively cued and reverse scored.  The scores on the domains can be reported separately but were never expected to form neatly separate factors and psychometrically have never fitted such models in any populations.  In every sample explored so far, the risk items are, predictably, relatively distinct from the other items forming a very clearly separable response space from that covered by the other items.  The “non-risk” score of the 28 items other than the risk items forms a good aggregate score if the risk issues are to be separated but most studies report the total score.

Scoring

The original scoring was the mean across the items, i.e. between 0 and 4 as the scoring of the item responses is from 0 to 4. After that, as some users didn’t like using two decimal places, scores have often been reported as a “clinical score” by multiplying that by ten, giving scores between 0 and 40.  Caution should be used to make sure whether results are reported with that x10 multiplier; however, the difference is so large that it is rarely likely that it could be unclear which scoring has been used. Occasionally, and not in line with any recommendations from the CORE teams over the years, people report total scores with a theoretical range, for all 34 items, from zero to 136, again, there should be no difficulty identifying when this scoring has been used.

Missing items and pro-rating

Slightly harder to identify can be what the practitioners or researchers have done about missing items.  We recommend prorating up to 10% missing items, so a total score can be reported as the mean across up to 31 items, the non-risk score on up to 26 and the problems and functioning scores as the means across 11 of the items on those domains.  Prorating is not recommended for shorter scores (i.e. for the wellbeing and risk scores).  It is very rare that so many items are omitted that even a prorated total score cannot be reported.

References

The main references to the CORE-OM are as follows.

Download

CORE-OM English language, PDF (22 downloads)
Share this: