First, it is helpful to stick with standard terminology. I think you refer to OTP and SC ("stabilisation centre"). CRENI is (to me) a confusing term as it can mean an old-fashioned TFC style intervention. OTP delivers the RUTF-based protocol. SC delivers (primarily) the F75-based protocol. Coverage is determined by recruitment and retention. Early treatment seeking (recruitment) in a beneficiary cohort can be investigated by plotting MUAC at admission. It can also be investigated by looking at the proportion of beneficiaries requiring inpatient stabilisation at an SC. This should be less then 5% after the initial phase of the program. You should also explore pathways to care in in-depth interviews with carers of beneficiaries, semi-structured interviews with program staff and other medical staff, interview with CBVs, community leaders, and informal group discussions with community members. You may also use data from a referral monitoring system if there is one in place. This is an example of triangulation by source and method that is used throughout SQUEAC. Triangulation by source and method means that the investigation is resistant to errors and biases in any single method. Now to your specific question ... I would include all children admitted (not transferred) into the program in the MUAC plot. Very extreme values are then identified as critical incidents. Investigation of critical incidents can tell you a lot about a program. There is an example of this in the technical reference. The principal analysis of MUAC at admission is based on the shape of the distribution. A handful of extreme values will not greatly effect this analysis If you have SC as a separate program to OTP rather than as just a separate program activity (this can happen) then you may want exclude direct admissions into the SC. When these children come to OTP they can be included as we do not expect much growth to occur in SC. Is this any help?