Scoring the CORE – Outcome Measure
The CORE-OM measurement tool is a broad measure of psychological distress. It is applicable across a wide range of situations. A higher score indicates greater psychological distress.
a) Understanding the domains
CORE-OM has 34 questions across four different aspects of psychological wellbeing:
- Wellbeing (four items)
- Problems and Symptoms (12 items)
- Functioning (12 items)
- Risk (six items)
These different aspects are referred to as domains.
Each question in the CORE outcome measure refers to one of these core domains. The domain is indicated by the letter to the right of the furthest box (W=Wellbeing; P=Problems and symptoms; F=Functioning; R=Risk)
b) Scoring Range
The scoring range of the CORE-OM is between 0 and 4, with 4 being the highest level of severity. As can be seen from the image above, some items are positively framed (items 3 and 4). For these items the scoring range is reversed.
The survivor considers each statement and how often they have felt like this over the past week. They are asked to then tick the box which corresponds most closely to their experience, using the categories at the top of each column from Not at all to Most or all the time.
c) How to calculate the mean total score
- To obtain the mean total score, add the scores for all questions to derive a total score between 0 and 136.
- Divide the total score by the number of completed items (normally 34) to obtain a mean score between 0 and 4. (E.g., a total score of 56 / 34 will yield a mean score of 1.64).
- Multiply this mean by 10 which gives you the total clinical score between 0 and 40. Round up/down to a whole number. This would give 16 in the example above.
d) How to calculate the mean score for each domain
In the same way it is possible to use the steps above to obtain total and mean scores for each of the four domains. For a mean score for the wellbeing domain, for example, simply add the four wellbeing items and divide by four.
What do the CORE-OM scores mean?
A higher score on the CORE measures, domains or individual items indicates a higher level of distress or symptom severity.
Bandings are routinely used to help assess severity of the clinical score in the CORE-OM:
9 or less– Healthy/ low level problems
10 to 14 – Mild psychological distress
15 to 19 – Moderate psychological distress
20 to 24 – Moderate-to-severe psychological distress
25 or above – Severe psychological distress
Clinical cut-off for CORE-OM is 10. Therefore, a score below 10 is technically sub-clinical.
How do you know if the change is due to trauma treatment or chance?
The CORE-OM is a validated tool to measure psychological distress. It has been thoroughly tested and checked. The reliable change index (or change required to be reliable and not just chance) for this measurement tool is 5.
For example, in the chart below, a survivor began treatment with a score of 21. By session six they have improved and are now scoring 6 (within healthy / low level problems band). This is an improvement of 15 points. This is a change greater than could have been achieved by chance. The survivor has achieved reliable and clinically significant improvement.
Adapted from here.
How to deal with missing data
Sometimes people decide they don’t want to complete a question, or simply miss one out. It is helpful to scan the form as soon as it is completed for any missing data and to encourage full completion.
CORE-OM is a comprehensive tool, and some adaptations can be made if there are a small number of gaps.
Total Score adaptations for missing data
- If data is missing for less than 10% of questions (10% = 3 missing questions) the results can be pro-rated i.e., a mean calculated across a minimum of 31 questions.
- Alternatively, a non-risk score can be calculated. This calculates the score on all questions except those in the risk domain. The non-risk score (“all minus R” on the bottom of the form) can be reported as the means across up to 26 items
Domain score adaptations for missing data
- Pro-rating is possible in the Problems and/or Functioning domains if data from only one question is missing. In this case the mean would be calculated across 11 instead of 12 questions.
- Pro-rating is NOT recommended if there is data missing in the well-being and risk
Illustrative example where no data is missing:
Illustrative example where 2 data items are missing
In this example one data point is missing from both the Functioning and Problems domains and the wellbeing and risk domains are complete. The data can be pro-rated.
Illustrative example where some key data is missing.
In this example one data point is missing from both the Function and Problems domains. A domain score can be pro-rated by dividing by 11 instead of the usual 12.
Two data points are missing from the Wellbeing domain. This means that a score cannot be calculated for this domain as there is too much data missing.
Similarly, it is not possible to calculate an overall score or all item score minus R as more than 10% of the data is missing.
Importance of clinical judgement
Clinical scores should always be interpreted alongside clinical judgement.
For example, if a survivor has a ‘healthy’ score (i.e., CORE-OM mean score of less than 6) the chance of further improvement is negligible (they cannot recover as they are already “healthy”)
Similarly, a survivor with a ‘low level’ score (e.g., CORE-OM mean score between 6 and 10) there is less scope to demonstrate improvement.
Consideration could be given to not accepting survivors with below cut-off scores for treatment. It is important to note however, that some survivors who suffer quite significant (and chronic) mental health difficulties develop coping strategies which result in them presenting with a low CORE-OM score. Some people may also exaggerate their difficulties (Barker at al, 2002)
Recommendation:
“The clinician should be free to make their own judgement regarding the patient’s condition in determining whether to accept them for treatment and not rely entirely on the CORE score” (CORE Partnership, 2007).
Previous
Section 5B Introduction
Next
Scoring the CORE-10 measure