Measles, a highly infectious vaccine-preventable viral disease, is characterized by clustering of cases that occur during cyclical epidemics [1
]. In many parts of the world, measles is also a seasonal disease with fewer cases found during the non-measles season [2
]. Clinically, the infection is expressed as a maculopapular rash accompanied by fever and at least one of the three "c's": cough, coryza and conjunctivitis; virtually all cases of measles are clinically expressed [3
]. Measles is a potentially fatal disease [1
]. The World Health Organization (WHO) defines a measles-associated death as one occurring within 30 days of rash onset, not obviously due to another cause such as trauma [5
Historically, measles case fatality ratios (CFRs) have been reported to vary from 0.1% [1
] in the developed world to as high as 30% among refugee populations [6
]. Current estimates of CFRs used by WHO in endemic countries range between 0.05% - 6% [8
]. Factors thought to affect CFR include age [11
], intensity of exposure to measles virus (for which household crowding may be seen as a surrogate) [12
], measles vaccination status [13
], nutritional status [14
], immunodeficiency [15
] and access to appropriate case management [16
]. Studies conducted in the late 1980 s demonstrated that supplementation of measles case-patients with vitamin A could decrease measles mortality by as much as 64% [17
] leading to recommendations by WHO and United Nations Children's Fund (UNICEF) in 1987 to treat all measles case-patients with vitamin A in areas where measles CFRs were greater than 1% [19
]. These recommendations, in conjunction with the rollout of Integrated Management of Childhood Illness (IMCI) guidelines in the mid 1990 s [20
] which target pneumonia and diarrhea, might be anticipated to have decreased measles CFRs since the 1990 s. However, few data exist on the extent to which these interventions are used in health facilities, particularly in countries that are highly endemic for measles. Anecdotal evidence from outbreak investigations in Niger, Sudan and South Africa indicate that these interventions are underused [21
Further decrease in CFRs in the past decade may have occurred due to the renewed political will and creation of the Measles Initiative in 2001. WHO and UNICEF developed a comprehensive strategy for sustainable measles mortality reduction with the goal of a 90% reduction in global measles deaths (compared with 2000 levels) by 2010.
The four-pronged strategy focuses on improved routine immunization, providing all children with a second dose of measles vaccine delivered either through periodic SIAs or routine services, improved measles case management and careful measles surveillance. This strategy has contributed to reducing the overall burden of measles and has potentially led to decreased CFR linked to earlier detection and improved case management.
Recently the global burden of mortality from measles, critical in prioritizing measles vaccination relative to other health interventions, has been an area of much discussion. A published point estimate from a 2003 analysis of childhood mortality using a proportional mortality model [25
] differed by hundreds of thousands of deaths from the WHO point estimate for the same period using a static natural history model [27
]. A major factor contributing to this discrepancy has been disagreement over which CFRs are appropriate to use in models seeking to estimate global burden of measles deaths, in particular what CFRs are appropriate to use in countries that are highly endemic for measles. In these settings, disease reporting, death surveillance and vital registration tend to be incomplete or non-existent, requiring the collection of primary data to determine disease-specific CFRs.
The recent controversy surrounding measles CFRs in the developing world and the paucity of up-to-date data on this topic have led us to conduct six published studies to determine measles CFR in five countries (Niger [21
], Sudan [22
], Nepal [28
], Chad [24
] and Nigeria [24
], Table ). Although a recent publication reviewed existing literature to better estimate probable CFRs by geographic region [10
], there is little published guidance on the conduct of field studies to determine measles CFR. This paper aims to fill this gap by summarizing both the challenges that are inherent in such studies, and the approach that we advocate given these challenges. Improving the rigor of future studies in this topic area will contribute to more accurate burden of mortality estimates.
Summary of published studies retrospectively estimating measles case fatality ratios and conducted by the authors