Between 2005 and 2014, an annual mean of 35 (range 31–40) armed conflicts occurred, of which six (four to 11) major conflicts (>1000 people killed annually) were active globally,1 directly affecting about 172 million people in 2012, the sole year for which global estimates including displaced and affected non-displaced people were available.2 Between 2006 and 2015, an annual mean of 393 (range 341–462) natural disasters (geological and hydrometeorological hazards only) occurred, affecting a mean of 169 (97–260) million people annually.3
In these crises, robust and timely public health information is crucial to rapidly establish public health needs and priorities, and thus an appropriate package of public health services; quantify and mobilise funds and resources required to deliver this package, given the population in need and the required intensity of support to the local disrupted health system; and monitor the performance of the humanitarian response, by identifying and reacting in real time to substandard quality of health services, new threats (eg, an epidemic), gaps in service availability, and other changes (eg, improvements in food security or reduced use of health services). A secondary function of public health information is to enable advocacy and documentation of the crisis' impact (including for legal purposes).4
War and disasters, however, disproportionately occur in countries where public health information systems are already weak. Crises compound these weaknesses by further disrupting government services. Other challenges specific to crises include the short timeframe and high frequency (days or weeks) within which data should be generated so as to monitor fast-developing health events such as epidemics or detect deteriorations in malnutrition and mortality, practical or statistical constraints of data collection methods in displacement situations, paucity of readily available expertise and resources for robust data collection, and security constraints, particularly where data collection or publication are perceived as threatening by combatants.
Epidemiologists and demographers have partially developed adapted methods5 to respond to these challenges, but it is unclear to what extent these methods are used consistently, and major methodological evidence gaps remain. In this paper, we distinguish the different domains of public health information in crisis settings, map existing methods against each of these domains and review the evidence supporting their use, suggest a minimum set of public health information products during the acute phase of the crisis and investigate their actual availability in recent crises, and, lastly, we propose an agenda for methods development and systemic measures to make accurate, timely public health information more consistently available in future crises.
Key messages
- •
Timely robust public health information is essential to guide an effective response to crises (armed conflicts and natural disasters). This response encompasses establishing needs and priorities, strategic planning and deciding on appropriate service packages, and reacting in real time to insufficient health service performance or new public health threats. Public health information is also important for advocacy and documentation purposes.
- •
Various methods exist to measure relevant aspects of demographics, public health risks, and status and services in crisis settings; however, many of these methods are not strongly evidence based.
- •
Actual, timely application of existing methods has been inadequate in recent large crises, even when considering a minimal set of essential public health information services.
- •
Far greater investment and institutional buy-in are required to make an advancement from the present unsatisfactory baseline. We propose that a dedicated body might need to be established to do core functions of public health information generation and analysis in crisis settings.
- •
The research agenda for development of more robust methods needs to be consolidated, and priorities tackled collaboratively across academic and operational agencies.
For the purpose of this paper, our definition of crisis encompasses sudden unplanned displacement, direct exposure to armed conflict resulting in heightened public health risks or reduced public health services (or both), sudden deterioration in nutritional status (as opposed to trends over multiple years), natural or industrial disaster, or a sudden breakdown of key administrative and management functions in a country (appendix p 13). We consider both the acute and protracted phases of a crisis in this review. We omit large epidemics or pandemics, because they have very specific public health information requirements that, moreover, largely depend on the pathogen and context.