Measuring Famine – Scales, Indicators and Thresholds

The following two tabs change content below.
Jason Nickerson
Jason Nickerson is an operational researcher who focuses on health systems in humanitarian emergencies and fragile states. He trained as a respiratory therapist and maintains a clinical practice in anesthesia in Ottawa, and completed a PhD in population health. Beyond health systems issues, Jason focuses on access to essential medicines, surgical care, and the treatment of respiratory diseases. Jason is a Clinical Scientist with the Bruyère Research Institute in Ottawa, Canada.

The current food crisis in the Horn of Africa (consisting of the countries of Somalia, Eritrea, Djibouti and Ethiopia) is an evolving humanitarian emergency that is bordering on disaster. Millions of people face serious health risks related to malnutrition, fuelled by a drought in the region but largely compounded by a worsening security situation, conflict, population growth, over-utilization of land and poverty.

Today, the media has been abuzz with reports that the United Nations is going to officially declare the area to be a famine, a label that has been tossed around for weeks to describe the current situation, but that actually has a fairly specific meaning.

So, what does all of this mean? We often hear of “disasters” and “humanitarian crises” and so on, but are these simply adjectives or titles that describe the same phenomena? Can they be used interchangeably? How can you measure a disaster versus a crisis and at what point do those lines blur? These are questions that many of us in the world of public health in crises struggle with, though we do have some basis from which we draw our language and our conclusions. And, despite the chaos of it all, we do have some fairly standard practices that help us make sense of constantly evolving emergencies.

To put it bluntly – a famine is more than just a title or a label, it has a particular set of criteria and thresholds that are met before affixing this label. Same goes for humanitarian emergencies, and we are struggling with scales for severity assessments with disasters and crises.

Famine exists at the extreme end of a continuum of the assessment of food security – of which malnutrition, mortality (deaths),disease epidemics and other indicators are used to gauge a population’s ability to access and acquire food.The United Nations scale being used is the Integrated Food Security Phase Classification (IPC) scale that ranges from Level 1A (Generally food secure) to Level 5 (famine/humanitarian catastrophe).

Source: FAO

Moving from one level to another in absence of extremely obvious criteria is controversial and complex; each level corresponds to a number of indicators,the presence or absence of which could influence a decision to raise the alarm.

Consider this: Two key measurements on the IPC are mortality and malnutrition. How we evaluate both of these is methodologically complex, time consuming and expensive (I have touched on these issues in previous posts on health statistics in Libya and Japan). But what if the full impact of food insecurity hasn’t been reached and mortality has yet to climb? Rigidly sticking to categories of indicators doesn’t always yield the desired or necessary result, and we have to think about the contextual analysis of indicators, as well as their absolute values.

Malnutrition is assessed in a few ways – using something called Z-scores, mid-upper arm circumference (MUAC) measurements, and/or weight-for-height/weight-for-length measurements.  These are called anthropometric measures and are measured in children through a complex sampling scheme to assess the prevalence of malnutrition in a population. When analyzed, we get a measure called Global Acute Malnutrition (GAM) and the threshold at which an emergency is declared is when the prevalence of GAM in the population is >15%.

Similarly, we measure mortality to get an estimate of how many people have died in a given time period. Again, households are randomly sampled and a history is taken of how many people have died in a given time period. The output of this is known as the Crude Mortality Rate (CMR) and is presented as the number of deaths per 10,000 people per day (deaths/10,000/day). The accepted emergency threshold is reached when the baseline CMR is doubled, or when it is in excess of 1 death/10,000 people/day. CMR is measured in adults, whereas the under-5 mortality rate (U5MR) is measured in children under 5 years of age, and has a slightly higher threshold of 2 deaths/10,000 people/day.

So what does this all mean? Well, it means that when we talk about humanitarian emergencies and famines, we have a very particular language and set of indicators that tell us when something reaches a tipping point. These are not perfect systems – as I mentioned earlier, mortality is clearly a late stage indicator of the severity of a crisis (and so is malnutrition, for that matter). While the IPC famine scale has room for a number of indicators, if we adhere to this strictly we are waiting for the sky to fall before we intervene. We need to consider these indicators both individually and in context to understand how crises evolve and where the health priorities are. While labels have some value in our understanding of the severity of a crisis, they don’t always tell us the full picture.

For more information about standard indicators, check out the Sphere Project Humanitarian Charter and Minimum Standards in Humanitarian Response – this is the internationally accepted standard for many of these measures.

One Trackback

  • By This is all true. – Summits Guide on Dec. 12, 2016 at 07:29

    […] you know how many bodies are – according to the very real “Famine Scale” – a “cause for […]