Limitations of Estimating the Burden of Disease Associated with
Antibiotic- Resistant Bacteria
This report uses several methods, described in the technical appendix, to estimate the
number of cases of disease caused by antibiotic-resistant bacteria and fungi and the
number of deaths resulting from those cases of disease. The data presented in this report
are approximations, and totals, as provided in the national summary tables, can provide
only a rough estimate of the true burden of illness. Greater precision is not possible at
this time for a number of reasons:
- Precise criteria exist for determining the resistance of a particular species of
bacteria to a specific antibiotic. However, for many species of bacteria, there are
no standard definitions that allow for neatly dividing most species into only two
categories—resistant vs. susceptible without regard to a specific antibiotic. This
report specifies how resistance is defined for each microorganism. - There are very specific criteria and algorithms for the attribution of deaths to
specific causes that are used for reporting vital statistics data. In general, there
are no similar criteria for making clinical determinations of when someone’s
death is primarily attributable to infection with antibiotic-resistant bacteria, as
opposed to other co-existing illnesses that may have contributed to or caused
death. Many studies attempting to determine attributable mortality rely on the
judgment of chart reviewers, as is the case for many surveillance systems. Thus,
the distinction between an antibiotic-resistant infection leading directly to death,
an antibiotic-resistant infection contributing to a death, and an antibiotic-resistant
infection related to, but not directly contributing to a death are usually
determined subjectively, especially in the preponderance of cases where patients
are hospitalized and have complicated clinical presentations.
In addition, the estimates provided in this report represent an underestimate of the total
burden of bacterial resistant disease.
- The methodology employed in this report likely underestimates, at least for some
pathogens, the impact of antibiotic resistance on mortality. As described in the
technical appendix, the percentage of resistant isolates for some bacteria was
multiplied by the total number of cases or the number of deaths ascribed to that
bacterium. A number of studies have shown that the risk of death following
infection with a strain of resistant bacteria is greater than that following infection