The Minimal-Eating Observation Form—Version II (MEOF-II) is an outcome measure developed to provide the clinician with a description of subject mealtime problems and to assess the need for possible interventions (Westergren, Norberg, Vallen, & Hagell, 2011). The assessment is an interactive observation tool composed of three domains: ingestion, deglutition, and energy and appetite, and has six subscales relating to (1) the presence of unintentional weight loss; (2) body mass index (BMI) or calf circumference (CC); (3) eating problems; (4) swallowing and mouth problems; (5) energy and appetite; and (6) the indication of clinical signs. Designed as a modification to the original MEOF by Axelsson (1996), changes included the replacement of opening/closing the mouth and aberrant eating time with chewing ability and appetite (Westergren, Lindholm, Mattsson, & Ulander, 2009). During assessment the clinician notes the presence or absence of particular adverse mealtime behaviors as well as the measurement of a person’s BMI or CC if BMI is unattainable. Items on the MEOF-II may include several observations that are scored as either 0 or 2 or 0 or 1. The measure can be completed in less than 10 minutes where a higher score is suggestive of more risk.
A study of 24 inpatients, median age 69 years, found MEOF-II total score inter-rater and intra-rater reliabilities to be r = 0.92 and r = 0.84, respectively (Westergren, Torfadóttir, & Hagell. 2014). Another study found inter-rater agreement between a trained observer and less trained observers (n = 20) to be r = 0.89 when testing 50 subjects post-cerebrovascular accident where agreement ranged from 0.82 for manipulation of food on the plate and mouth to 0.94 for ability to chew and swallow (Westergren et al., 2009). The same study determined that internal consistency for the trained observer was α = 0.76 and for the less-trained observers it was 0.71 (Westergren et al., 2009). Another study found that total values were above r = 0.80 for both inter- and intra-rater reliabilities using 8 raters with no prior experience among 24 subjects (Westergren et al., 2014). A study by Vallen, Hagell, & Westergren (2011) of various patient populations (n = 33 each) found that the number of those found to be at risk were 38.2% for orthopedic, 28.1% for cardiology, and 45.4% for stroke populations. When using the BMI version its results had exact agreement of 82% with the full Mini Nutritional Assessment—Long Form (MNA), which is much longer, and when CC was used agreement was 81%. A similar examination involving 87 persons receiving inpatient care found that out of 18 undernourished patients, according to gold standard (MNA), only 13 were considered to be at high risk according to MEOF-II resulting in a sensitivity (i.e., the proportion of people correctly identified according to the 18-item MNA) of 0.61, whereas the specificity and accuracy were 79% and 68%, (Westergren, Norberg, & Hagell, 2011). Furthermore, a positive MEOF-II result for undernutrition, was associated with an 82% probability that the individual really was undernourished according to the MNA and a negative result was associated with a 57% probability that the individual really was not undernourished (Westergren, Norberg, & Hagell, 2011).
The MEOF-II is an easy to use, relatively quick assessment that has shown both acceptable sensitivity and specificity in detecting undernutrition if used in the same manner as the MNA-SF (short form). For example, Vallen et al. (2011) found that it took, on average, only 8.8 minutes to complete the assessment. Administration also allows for substituting BMI with CC in situations where measures of patient height and weight cannot be easily obtained. Another advantage of using the MEOF-II compared to other quick screening tools is that it has the ability to identify actual problems for immediate intervention (Westergren, Norberg, Vallen, et al., 2011). Finally, the MEOF-II is not limited to only patient populations ≥ 65 years of age like the MNA-SF.
Differing inter- and intra-rater agreement between trained and untrained observers as well as the presence of outlier scores have been observed suggesting that instruction in performing eating observations may be required; however, the authors argue that only minimal training is needed (Westergren et al., 2009). Also, the amount of independent research (i.e., work that does not indirectly involve one of the original authors) is limited.
The MEOF-II is an interactive observational tool in which the practitioner notes the presence or absence of aberrant eating, feeding, or swallowing behaviors as well as defining a person’s BMI or CC. Each item is scored along a variable scale of 0 or 1 and yields a total score ranging from 0 to 8. A final score of 0 to 2 is considered low risk for under-nutrition, a score of 3 or 4 is a moderate risk, and a score ≥ 5 is considered a high risk for undernutrition (Valle et al., 2011).
The MEOF-II can be downloaded and used free of charge from the website listed in Contact however, the author asks that a simple form be filled out when downloading the material. Use of the MEOF-II for research or other purposes can be obtained by contacting its creators. The assessment can also found in its entirety in the following journal article:
Vallen, C., Hagell, P., & Westergren, A. (2011). Validity and user-friendliness of the Minimal Eating Observation and Nutrition Form—Version II (MEONF-II) for under-nutrition risk screening. Food and Nutrition Research, 55, 5801. doi: 10.3402/fnr.v55i0.5801
|POPULATION||Suspected malnutrition; general|
|TYPE OF MEASURE||Questionnaire/rating scale|
|WHAT IT ASSESSES||Malnutrition risk|
|TIME||< 10 minutes|
Albert Westergren, PhD
Kristianstad University School of Health and Society