Apologies for overlooking this question. From our experience in non emergency CTC/CMAM programmes Sphere standards are normally being reached for cure and death rates but there are places where defaulter rates are higher. But Sphere standard coverage rates are generally the hardest to achieve in non emergency contexts. Outcomes for integrated non emergency programmes vary depending on resources and local priorities, however we still find Sphere standards a useful guide for monitoring and evaluating a programme.

In the CTC programs that I have seen in both emergency and "integrated" setting (many in many settings), SPHERE minimum standards for for cure-rates are usually met and exceeded. A big problem is defaulter rates. There are many reasons for this. The most common reasons that I have observed are periodic stock-outs of RUTF (which seems only to occur in "integrated" programs and reflects the poor supply chains common in many settings), excessive waiting times, and highly judgemental attitudes of staff towards carers. Acceptable levels of coverage appear the most difficult to achieve in all settings. The most common reason for coverage failure are usually lack of attention to the community-based aspects of the program, poor siting with regard to proximity (i.e. too few sites that are too far apart) and the use of W/H admission criteria. In short, I have seen emergency and "integrated" programs meet and exceed SPHERE standards for cure, death, default, and recovery. I have also seen emergency and "integrated" programs fail on some (and in a couple of cases) all of these standards. This only partly answers the question. It is possible for "integrated" programs to meet SPHERE standards. This does not mean, however, that the SPHERE standards, developed for emergencies, are appropriate for "integrated" programs. I think that some work is required on this. I also think that "integrated" programs should adopt an audit approach to improving performance over time (as is done with the SQUEAC assessment method). With regard to NGO roles in "integrated" programming. This seems to be either absent or, in post-emergency, a quick hand-over to the local MoH. I would like to see NGOs exploring support roles in which the MoH is responsible for services but the NGO provides expertise in audit of programs and coverage evaluations, training, back-up RUTF stocks, &c. Just my tuppence.

Hi, With reference to what have Ann mentioned, again i wonder if the adapted SPHERE standards could be applied at national scale rather than emergency scale , now Sudan will have it's CMAM implemented soon nationally so can we use SPHERE indicators for emergency,post emergency and non emergency setting? THANK U

With regards to using Sphere Minimum Standards nationally, for all contexts: I would say yes, the standards could be used in all contexts as a tool to monitor CMAM. However, normally in an emergency there are far more resources and attention given to treating acute malnutrition, therefore meeting and exceeding the standards fairly rapidly is achieveable. The reality of meeting and exceeding the standards in a non emergency context will certainly be more challenging and may well to take longer, but they are still acheiveable and relevant.

Anne makes an important point. I think it is assumed in SPHERE documentation that standards for (e.g.) coverage can be achieved very rapidly in emergency contexts. This has been shown to be, on occasion, true (many CTC programs fail to achieve satisfactory coverage due to inattention to the community-based aspects of programs). With emergency CTC / CMAM we can open many centres very quickly and support them with emergency context resources (e.g. vehicles, staff, training, per-diems, RUTF supplies, essential drug supplies, paper forms, program analyses, and intensive expatriate consultancies) often outside of statutory systems. In non-emergency contexts the whole process is slower and less well resourced. There are also the problems of poor supply lines, unpaid salaries, under-staffing, weak infrastructure, &c. In such settings I think it might be useful to nuance the SPHERE standards with the addition of realistic timescales in which standards are to be achieved. I think that many iNGOs are good at the sort of vertical programming that characterises emergency response but very few are good at the sort of programming that will be required to make CMAM universally successful. This is, I think, a major challenge. It seems to me that we are coming to the close of the period of intensive technical work on CTC / CMAM (e.g. case-defintions, discharge criteria, protocols, and even standards and methods for evaluation / audit) and that there is now a need to work out how we (i.e. iNGOs, UNOs, governments, academics, and the private sector) can roll out these programs. I believe that iNGOs should play a major role in this but that many of them will need to learn new skills and change the way they work. Just my tuppence.

Hello All, Just some thoughts I'd like to share Given the fact that the two approaches (CMAM or CTC and the traditional TFC) are programs for management of severe acute malnutrition, these standards were set for TFCs which has the two phases (phase one, transition and phase two in one compound) and the management is taking place as one unit and reporting as well in one monthly reporting format (the reports here are the tool which is capturing the monthly trends of the performance indicators), in the CMAM/CTC approach the two phases are separated and sometimes being run by different partners which in turn makes the reports comes for the same child at two treatment stages from two centers. What I'm trying to say is that applying the same standards to be met by the SC and the OTP does not look making sense due to the changes in the conditions for the same child, the issue need to be revisited from that angle instead of looking at it from emergency and non emergency situations, as the severely malnourished with complication child remain severely malnourished with complications in both situations and the difference will be in the numbers of admissions which does not require to amend the standers accordingly , so am suggesting to amend the standards of the OTP indicators to set somewhere in the middle between the SFC and the SC/TFC for the cure , defaulters and death rates. Thanks

interesting discussion. If then, the SPHERE targets are easy (and i use the term easy loosely) to hit in an emergency, shouldn't one use 100% coverage when calculating the numbers to be treated based on the case load (rather than use the 70%) in this case. If one bases their budgets and supplies on the 70%, isn't there a risk of running into a shortfall? does anyone have any practical experience of this especially where there is heavy reliance on donor rather than organisational funding?

100% coverage has proven elusive. In the CTC research program we manages to get 70% to 80%. A CHW-delivered CMAM program in Bangladesh go just under 90%. This is the best coverage recorded to date. Most CMAM programs do not achieve coverage anything like these levels. A lot (NoH and INGO run) of CMAM programs never exceed 30%. In my experience the SPEHER standards (although minimum standards) remain challenging for MoHs and INGOs to achieve). BTW ... the coverages give above are for direct period coverage estimates by CSAS, SLEAC, or SQUEAC.

that's very surprising. is the same level of coverage (~30%) also being seen in emergencies? i was looking at data from an INGO working in the neighboring district to where we are setting up (in the Sahel) and the numbers they treated for one month was about a third of our expected caseload for the year. and they have a program covering half a district while we are aiming to set up CMAM for the whole district. hence the debate between budgeting for 70 vs 100%.

The < 30% figure is what we have seen in many programs over the past decade. The range is from just over 8% to just under 90%. As a general rule we see coverage at about 20% - 30% when a program does nothing very wrong but also does nothing very right (i.e. they get he clinical side OK but the case-finding, sensitisation, mobilisation, messages, spatial reach &c. components of the program are weak). The term "emergencies" covers a multitude of circumstances. The figures in my original reply (above) are mostly from what might be termed "complex humanitarian emergencies" and are direct coverage estimates. If you treat 33% of your annual expected caseload in one month then you should definitely suspect your expected caseload estimate. It is likely that your early caseload is made up of a lot of prevalent cases who have been SAM for some weeks or months. A three to four month expected caseload is not, therefore, impossible. It is, however, extremely unlikely since it would mean something like 100% coverage in the first month. Achieving this level of coverage so quickly is outside of all experience with the CTC / CMAM model. The expected caseload approach is an INDIRECT approach usually based on a historic point estimate from a survey with an urban bias (PPS), a sample size too small to estimate SAM with useful precision (e.g. n = 2850 is required for a relative precision of 30% on a 2% estimate using a design effect of 1.5), and using a population estimate that is likely to be inaccurate in emergency settings. This is a very poor way to estimate coverage. The figures in my reply (above) are for DIRECT estimates of coverage using the CSAS, SLEAC, SQUEAC, and (for one) the S3M method.