The categories "discharged as not-cured" and "discharged as non-reponder" (the same thing described slightly differently) are standard categories for routine monitoring statistics in CMAM programs. They apply to cases that have been in the program for a long time without being cured and are discharged using a maximum length of stay rule (usually 16 weeks). You have something a bit different. I think it is a compliance issue, The category "default" is, I think, the nearest client side compliance indicator to what you describe. If the program starts and stops with the consequence of discharging non-cured cases then I think we can treat these are "program defaults" (i.e. the program fails completely to comply with the CMAM protocol). You could then have: Client default - The standard definition Program default - What we have above All defaults - Both client and program defaults You need to include these exist or you will show you program as being a lot better than it really is. Remember to report these as a proportion of all exits. Is this any help?

I think that treating them as "transfers" could be closer than "defaulter". Did you reffer any of these children for continued rehabilitation elsewhere or you just left them hanging? These information may guide on deciding the disharge outcome

I agree. If you transferred them to another program of appropriate intensity then they are transfers. If you "just left them hanging" then they are program failures and should be recorded as such. If they were transferred to (e.g.) SFP whilst still SAM cases then they are also program defaulters.

Thanks Mark and Florence for feedback. These cases were actually left hanging (not refered anywhere). I reckon we have a consenses here that it should be reported as proportion of total discharges. Thanks

Dear Anonymous 1063, I was wondering if you clarify your question so that other readers can have better understanding. What do you mean when you say “each time the program halted”? In each episode, How was your admission procedure? When you reopened, did you start from the scratch just new registration whether new or old cases? Why I am asking is because it could have implications on other indicators as well. The other point I want comment is on Mark Myatt’s suggestion on using the term Program default. What is the problem if we use the terminology for those left hanging children just as "childrenremained in the program"? As far as I know, through out the program implementation period, up to the 11th hours, usually we do receive new admission as far as it is targeted programme. Even if the program is not halted, at the end of the program official period, it is possible to get left hanging children. These are children who are admitted at the last period of the program implementation (around last 3.9 months?). Given this fact, I was wondering if Mark explains the rational why theses cases are to be categorized as defaulters, program defaulters. If these groups are included, I think the whole globally agreed performance indicators will be artificially distorted and could give wrong impression. Programs can be halted for various reasons. As to me, during reporting, I would keep them separate.

I don't know about "children remained in the program" as an agreed performance indicator. The term is a deceit. I like to ask "What if this were my child?". Let's play along with this ... My child is very sick (SAM mortality is about 200 / 1000 / year - that's about six times baseline risk) and was enrolled in a treatment program but was discharged before she was cured. Am I supposed to be happy about that? I'm not! This is a clear program failure and a clear breach of medical ethics (we are obliged to treat the case in front of us and especially if we say we will). The program is failing these children whether this happens because the program has dreadful logistics and RUTF keeps running out (or whatever is behind the stop-go sequence) or because an emergency has passed and it is time to pack our bags. These children are not "children remained in the program" (in what sense do they "remain" in the program?). They are prematurely discharged current and recovering SAM cases. A "patient default" occurs when the patient terminates treatment prematurely. The situation here is that the program terminates treatment prematurely. Logic demands that we classify this as a "program default". Treating these cases as anything other program failures will distort performance indicators. It may be that we will inevitably fail these children but I think we should be honest about it and count them as failures when we judge how well we've done. Anything else is, in my opinion, deceitful. Just my tuppence.

Dear Mark, This is just to clarify. I didn’t say "children remained in the program" as an agreed performance indicator. What I said is this “If these groups are included, I think the whole globally agreed performance indicators will be artificially distorted”. When I said “these groups” I am referring those who are left hanging children. So, when those left hanging children are categorized as defaulters, then the conventional defaulter rate indicator will be distorted. The other comment is regarding the way how you defined “program failure”. Before you conclude as “program failure”, I think you need to clarify the nature of the program; emergency where the management is beyond the capacity of the local government, non emergency where the program is institutionalized within the existing government structure. In principle, I am in favour of your position; “we are obliged to treat the case in front of us and especially if we say we will” however, in actual world it is not easy.

The standard approach to calculating the routine monitoring statistics is to use the total number of program exits in the reporting period. Your approach treats these "dumped" children as if they were not program exits. For this to be true then they must still be in the program or have been transferred to a program of appropriate intensity. It seems to me that you want to exclude these children from being counted as special cases. I think that distorts performance indicators by hiding failure. My suggestion is that we should (1) included these children, and (2) be honest about their status as failures (i.e. us having failed them). It is a case of default but a program default rather than a patient default. Perhaps the use of the term "default" is confusing. In that case I think we need a term that carries an adequate sense of failure. I suggest "dumped" or "premature discharge". It does not matter if this is an MoH or an NGO or an emergency or a development program. These kids are prematurely discharged and we should be honest about that. As I wrote before "It may be that we will inevitably fail these children but I think we should be honest about it ...". Only by being honest about it will we see this as something wrong and, hopefully, make attempts to fix it. As for the "actual world" ... I recently became slightly involved with an emergency CMAM program run by MSF. This program lasted about 10 months and at closure they left behind a small core logistics and clinical team to augment the local MoH so that the "remainders" could be transferred and treated appropriately. That is an actual program running in the actual world. It seems that ethical programming is possible in the actual world.