The main issues here are simplicity, rapidity, low cost, and standardisation. Simplicity, rapidity, and low cost are essential if the system is to run for a reasonable length of time. Standardisation must be maintained over time because you want to be sure that you can make comparisons between survey rounds. If you change methods or case-definitions then you can never be sure whether what you see is real change or due to changes in method. Typical survey requirements such as bias are of secondary importance if your main interest is trend (this is because the bias remains constant between survey rounds). Rapidity refers to the period of fieldwork required and to the time between fieldwork and reporting. This should be within a few days of data collection. Most surveillance systems fail because they are expensive and overcomplicated (so keep it cheap and simple), because of a lack of timely reporting (so make it rapid), and because of a poor dissemination of results (so don't keep results to yourself). I was involved with the design of an "open cohort" surveillance system for SC-UK / ACF known as "Listen Posts" (originally "HUMS"). I have wrapped up some of the design documents and put them [url=http://www.brixtonhealth.com/HUMS.LP.ENN.zip]here[/url]. I was also involved in the design of a repeated survey surveillance system (the "Community Nutrition Surveillance" (CNS) system) in Sudan. This used RAM type surveys with PROBIT estimators (typically 50% the cost of SMART with similar precision). Previous discussion of nutrition surveillance systems of these forums are (in no useful order) ... [url=http://www.en-net.org/question/61.aspx] Digit preference[/url] [url=http://www.en-net.org/question/72.aspx]Which nutrition surveillance system ...[/url] [url=http://www.en-net.org/question/245.aspx]How often to measure MUAC for ...[/url] [url=http://www.en-net.org/question/1260.aspx]Establishment of Nutrition Surveillance System[/url] [url=http://www.en-net.org/question/1201.aspx] Setting up of Nutrition serveillance [/url] I hope this is of some use.