Considerations for assessment of measurement quality of mid-upper arm circumference data in anthropometric surveys and mass nutritional screenings conducted in humanitarian and refugee settings.

ABSTRACT

Despite frequent use of mid-upper arm circumference (MUAC) to assess populations in humanitarian settings, no guidance exists about the ranges for excluding implausible extreme outliers (flags) from MUAC data and about the quality assessment of collected MUAC data. We analysed 701 population-representative anthropometric surveys in children aged 6-59 months from 40 countries conducted between 2011 and 2019. We explored characteristics of flags as well as changes in survey-level MUAC-for-age z-score (MUACZ) and MUAC means, SD and percentage of flags based on three flagging approaches ±3 and ±4 MUACZ z-scores from observed MUACZ survey mean and a fixed interval 100-200 mm of MUAC. Both ±4 and 100-200 flagging approaches identified as flags approximately 0.15% of records; about 60% of all surveys had no flags and less than 1% of surveys had >2% of flags. The ±3 approach flagged 0.6% records in the data set and 3% of surveys had >2% of flags. Plausible ranges (defined as 2.5th and 97.5th percentiles) for SD of MUACZ and MUAC were 0.8-1.2 and 10.5-14.4 mm, respectively. Survey-level SDs of MUAC and MUACZ were highly correlated (r = 0.68). The average SD of MUACZ was 0.96 using the ±4 flagging approach and 0.94 with ±3 approach. Defining outliers in MUAC data based on the MUACZ approach is feasible and adjusts for different probability of extreme values based on age and nutrition status of surveyed population. In assessments where age is not recorded and therefore MUACZ cannot be generated, using 100-200 mm range for flag exclusion could be a feasible solution.