non-type b. In a pneumonia surveillance study conducted in Pakistan by the CDC, over 90 percent of children bacteremic with H. influenzae were infected with nontypeable organisms.4 In addition, there is some evidence suggesting that nonencapsulated Hib strains can be responsible for serious disease in older children. 5

These and other similar data suggest that in certain settings, a conjugate Hib vaccine is likely to prevent only a moderate proportion of Haemophilus disease. Given the relative high cost of such vaccines, these data argue strongly for further epidemiologic testing to determine the prevalence of Haemophilus types in settings in which immunization programs are being contemplated.

It is worth noting that H. influenzae serotype data in the United States come primarily from isolates of cerebrospinal fluid from children with meningitis and from blood samples from children with high fevers, not pneumonia. In contrast, serotype data from the developing world derive mainly from children with cough who meet World Health Organization diagnostic criteria for pneumonia. These differences have important implications for vaccine testing and reinforce the need for rigorous vaccine trials in the developing world.6

In the case of S. pneumoniae, considerable data have accumulated in the United States over the past decade regarding infection with this pathogen. Studies in various areas of the country reveal that somewhere between 90 and 2,000 cases of pneumococcal disease occur each year per 100,000 persons, predominantly among the very young. In general, much less is known about pneumococcal incidence in the developing world. For example, in Nigeria, Africa ’s most populous nation, pneumonia accounts for some 30 to 40 percent of all visits to the doctor and roughly 12 percent of pediatric hospital admissions. About 30 percent of all childhood mortality can be attributed to pneumonia. Likewise, in the United Arab Emirates, a small developed nation at the eastern tip of the Arabian peninsula, ARI is a major cause of morbidity and mortality. In neither country, however, are there many hard data on the etiology of these infections.

In addition to variations in disease incidence between and even within certain regions of the world, there are differences in the pneumococcal serotypes present. A review of the published literature found that in the United States and a number of other developed nations, the seven most prevalent serotypes were (in rank order) 14, 6, 19, 18, 9, 23, and 7.7 In many developing nations, the top


Benjamin Schwartz.


Dan Granoff.


Joel Ward.


Sniadac D, Bogaerts J, Butler J, et al. Geographic variation on pneumococcal serotypes in children with invasive disease: Implications for conjugate vaccine use [Abstract 1651]. 33rd Interscience Conference on Antimicrobial Agents and Chemotherapy, New Orleans, Louisiana, 1993.

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement