IPC Famine Classification - Special Additional Protocols

IPC promotes accountable famine classification, and thus specific protocols have been adopted to ensure the technical rigour, neutrality and quality of analysis. While IPC Famine classifications follow all regular IPC protocols, special protocols also need to be observed in all four Functions, as detailed below.

As a best practice, a national IPC Technical Working Group that foresees the possibility that its upcoming or ongoing IPC analysis might result in a classification of Famine or Famine Likely is strongly encouraged to consult the IPC Global Support Unit to clarify the way forward in terms of support and the review process. 

While this section provides an overview of the special protocols for Famine and Famine Likely classifications, more detailed guidance is included in IPC Resources.

Function 1: Build Technical Consensus

  • When a Famine classification is being considered, it is imperative that the analysis team include food security experts, nutritionists experts with advanced knowledge in analysis of mortality data and, optimally, communication experts. Additionally, given the high profile of the classification, it is strongly advised that global and regional experts be invited to support the analysis.

Function 2: Classify Severity and Identify Key Drivers

  • Evidence requirements for Famine classification are different from those of other phase classifications. The amount and reliability of evidence will determine if a Famine or Famine Likely classification is allowed, with less strict requirements for areas with limited or no humanitarian access. The criteria are described in Figure 61 and detailed below.
    • Famine classification requires R2 direct evidence on all three outcomes (food consumption and livelihood change, nutritional status and mortality), with the following notes and exceptions:
      • Evidence for Food Consumption and Livelihood Change should optimally include the Household Hunger Scale (HHS), since this is typically the only collected indicator with a cut-off for Phase 5. However, other pieces of evidence on the other indicators included in the IPC Acute Food Insecurity Reference Table can be counted towards meeting the minimum evidence requirements for Famine classification. In cases where direct reliable evidence is available for mortality or acute malnutrition, a classification can still be performed without relying on direct evidence on food consumption and livelihood change, provided that analysts document the analytical process of inference for food consumption or livelihood change, which needs to be based on    at least four pieces of evidence on outcomes and/or contributing factors and rely on at least two of the three recognized inference approaches, i.e. calibration, extrapolation or causal pathways. The inference should indicate the proportion of households expected to be in Phase 5 Catastrophe, and in order to support Famine classification, at least 20 percent of households should be in IPC Phase 5 Catastrophe.
      • Evidence for Nutritional Status only includes reliable data on GAM based on WHZ or oedema.
      • Evidence for Mortality includes the CDR and the U5DR from representative surveys of good method. If the CDR is below the Famine threshold but the U5DR is higher, the latter can be used to classify the Famine if the 95 percent confidence interval of CDR includes the Famine threshold (i.e. 2/10,000/day). The recall period for the CDR should optimally be around 90 days during the recent past; however, in the event that recall periods are longer, evidence can be still used but analysts should assess trends in deaths and provide an explanation on how death rates reflect recent conditions. Death rates should reflect deaths in the areas being classified. Death rates need to be directly attributable to outright starvation or to the interaction of food consumption deficits and disease; all deaths due to trauma should therefore be discounted from death rates.
    •  Famine Likely classifications can be performed when evidence requirements for a Famine classification are not met but there is at least R1 (+ or -) direct evidence on outcomes, or other evidence as described in Figure 62 and detailed below:
      • Evidence for Food Consumption and Livelihood Change  optimally includes direct evidence, but in the absence of direct evidence, indirect evidence including inference of outcomes can be used. For inference to meet the requirements, it must be based on at least four pieces of evidence on outcomes and/or contributing factors and should indicate the proportion of households expected to be in Phase 5 Catastrophe. At least two of the three recognized inference approaches, i.e. calibration, extrapolation or causal pathways need to be used to conduct the analysis. In order to support Famine Likely classification, at least 20 percent of households should be in IPC Phase 5 Catastrophe.
      • Evidence for Nutritional Status includes GAM based on WHZ or MUAC, including oedema. The cut-off for GAM based on WHZ for Famine Likely classification is 30 percent, whereas for GAM based on MUAC the cut-off is 15 percent as per the IPC Acute Food Insecurity and Acute Malnutrition Reference Tables.
      • Evidence for Mortality includes the CDR and the U5DR, following the same cut-offs and guidance as for Famine classification. In addition to accepting mortality data from reliable household surveys, Famine Likely classifications can also use R1 mortality data collected through monitoring systems such as hospital records, community-based surveillance systems and vital registration records.
    • Classifications of areas with limited or no humanitarian access can rely on evidence with a reliability score of R0 even for Famine classification, provided that the data adhere to general IPC guidance for collecting evidence on these areas as per special protocols for areas with limited or no humanitarian access. 
    • For projections, in addition to the requirements specified above, evidence should not be older than 12 months at the end of the projection period, and at least four supporting pieces of evidence should be inferred for the projection period. Historical evidence used to classify Famine Likely conditions is exempted from the 12-month rule.
  • All current, projected or inferred evidence needs to be at or above Famine thresholds for Famine or Famine Likely classifications (i.e. at least 20 percent of households with extreme food gaps, at least 30 percent of children acutely malnourished identified through GAM based on WHZ, and in the absence of GAM based on WHZ,  at least 15 percent of children acutely malnourished identified through GAM based on MUAC for Famine Likely classifications; and a CDR of at least two deaths per 10,000 per day, or a U5DR of at least four deaths per 10,000 per day). For a projection of famine, the current situation can still be below famine thresholds, but through a critical analysis, it is concluded that the condition will pass the famine cut-offs in the projection period. 
  • There may be situations when circumstances suggest that a Famine may be occurring or will occur but the evidence available is insufficient to meet the minimum requirements described in Figure 62. In such cases, in the presence of substantial evidence, which can include both indirect and lower reliability evidence, an in-depth consultation process between the Famine Review Committee and the Technical Working Group can be put in place. This process can result in a Famine Likely classification based on the whole body of evidence and on expert judgment.
  • Any unit of analysis can be classified as Famine, including household groups or any geographical area, provided that they add up to at least 10,000 people.

Notes: 

  1. The three outcomes refer to: (i) food consumption and livelihood change; (ii) acute malnutrition; and (iii) mortality. 
  2. Direct evidence for Food Consumption and Livelihood Change includes the Dietary Energy Intake, the Household Dietary Diversity Score (HDDS), the Food Consumption Score, the Household Hunger Scale (HHS), the reduced Coping Strategy Index (rCSI), the Household Economy Analysis (HEA), the Food Insecurity Experience Scale (FIES) and the Livelihood Coping Strategies -indicator. Direct evidence should ideally be available for indicators that have thresholds assigned for IPC Phase 5 in the IPC Acute Food Insecurity Reference Table, such as the HHS. If direct evidence is available for mortality and acute malnutrition, a Famine classification can still be performed without relying on direct evidence for food consumption and livelihood change, provided that analysts document the analytical process of inference for food consumption or livelihood change from at least four pieces of R1 direct or indirect evidence on outcomes and/or contributing factors. Analysis should indicate that at least 20 percent of households are/will/would be in IPC Household Phase 5 Catastrophe.
  3. Direct evidence for Acute Malnutrition includes GAM based on WHZ. The Famine threshold for GAM based on WHZ is 30 percent and above. In the absence of data on WHZ, for Famine Likely classifications, data on MUAC can be used together with an understanding of the relationship between WHZ and MUAC in the area of analysis supported by at least two indicators with R1 evidence on the immediate causes of malnutrition to confirm MUAC findings. Note that in both indicators the presence of oedema is included.
  4. Direct evidence for Mortality includes the CDR and the U5DR. The CDR should be calculated for non-trauma deaths only. The famine thresholds for the CDR are more than two deaths per 10,000 people per day. If the CDR is below 2 deaths per 10,000 people per day but the U5DR is greater than four deaths per 10,000 children-under-five per day, this evidence can be used to classify Famine if the 95 percent confidence interval of CDR includes 2.
  5. The Evidence Reliability Assessment should follow Protocol 2.4, with indicators on food consumption and livelihood change adhering to the protocol detailed under the IPC Acute Food Insecurity classification, and indicators on acute malnutrition adhering to the protocol for the IPC Acute Malnutrition classification. 
  6. Other evidence allowed for Famine Likely classifications includes inferred outcomes of food consumption and livelihood change, GAM based on MUAC from a method of limited soundness (M1) or limited time relevance (T1), and the CDR or U5DR from a method of limited soundness (M1) or limited time relevance (T1).
  7. If available evidence does not meet the minimum criteria for the evidence requirement and the Technical Working Group is concerned about the existence of Famine, the Technical Working Group should contact the Global Support Unit for further guidance on how to proceed.
  8. In extreme circumstances, a Famine Likely classification can be made using substantial indirect or lower reliability evidence in consultation with the Famine Review Committee (see page 85 for details). 
  9. Famine and Famine Likely can be projected even if no Phase 5 classification is reached for the current analysis period. In these cases, the amount and reliability of evidence used to classify the current situation need to correspond to the criteria included in Evidence Levels of Famine and Famine Likely.

Function 3: Communicate for Action

  • Develop the IPC Famine Alert as a summarized version of the IPC Standard Communication Brief to provide a clear and concise explanation of the situation.
  • Adhere to the following procedures:
    • Communicate the classification clearly, using the guidance in Figure 64, stating the name of the classification (Famine or Famine Likely), the key message (including severity, number of people and evidence level), and linking the classification to decision-making (calling for urgent action to prevent widespread deaths and reduce malnutrition and starvation   and to strengthen data collection as relevant).
    • Specify areas/groups classified in the Famine/Famine Likely and the time frame (Figure 63). 
    • Provide a critical reasoning for classification, including reference to actual supporting evidence and source(s). 
    • Include separate map(s)/zoom-in(s) for the areas classified in the Famine/Famine Likely as relevant. 
    • Include the definition of Famine: “According to the IPC, ‘Famine’ exists in areas where at least one in five households suffers from an extreme deprivation of food. Starvation, extreme critical levels of acute malnutrition (at least 30 percent of children malnourished) and significant mortality, directly attributable to outright starvation or to the interaction of malnutrition and disease (at least 1 person for every 5,000 dies each day) are occurring.” 
    • State the Famine review process followed.

Function 4: Quality Assurance

A Famine Review is mandatory for any Famine classification (Famine or Famine Likely). The Review focuses on assessing the plausibility of Famine classification so that the classification can be validated or disproved, and includes two main activities:

  1. The Famine Review preparation, which is led by the Global Support Unit with direct inputs from experts from IPC Global Partners who have not been involved in the IPC analysis. This review consists of a preliminary screening of the Famine classification in order to verify adherence to IPC protocols and provide the Famine Review Committee with recommendations.
  2. The Famine Review by the Famine Review Committee is led by a four- to six-member team of independent international food security and nutrition experts who are objective concerning the IPC outcome and who have the relevant technical knowledge and experience in the specific crisis context. The Famine Review by the Committee aims at validating (or disproving) the Famine classification, including when the body of evidence available does not meet the minimum requirements for Famine Likely classification.

Famine Reviews are mandatory for both IPC products and IPC compatible products and are to be conducted before the release of findings. These Reviews are activated by the Global Support Unit based on a request from the Technical Working Group or, in case of a breakdown in the technical consensus relating to a (potential) Famine classification, by IPC partner(s) or by the Global Support Unit, based on the evidence available. The Famine Review conclusions and recommendations are communicated by the Global Support Unit to the country Technical Working Group members and shared with the IPC Global Steering Committee. The Famine Review Committee report is made publicly available on the IPC website together with the IPC country report.