THE IMPACT OF DIFFERENT CLASSIFICATION OF PNEUMONIA ON THE COST-EFFECTIVENESS OF A PNEUMOCOCCAL CONJUGATE VACCINE IN THE GAMBIA

Purpose: Despite recent consensus by WHO on the criteria for radiographic evidence of primary endpoint pneumonia, investigators have cited the need to define criteria that includes radiographic evidence that does not meet the standard for primary endpoint pneumonia.  While previous studies of pneumonia have distinguished between diagnoses based on clinical criteria only and diagnoses confirmed by radiograph, recent studies have distinguished between evidence for primary endpoint pneumonia and other radiographic abnormalities.  Our objective was to examine the impact of different definitions of clinical versus radiological pneumonia on the cost-effectiveness estimates of a 9-valent pneumococcal vaccine (PCV9) in the Gambia.

Method: Our differing case definitions for “radiological” versus “clinical” pneumonia were based on the inclusion (an approach used by previous investigators) or non-inclusion (WHO approach) of radiographic evidence not within WHO standards in defining radiologic pneumonia.  We developed a comprehensive Markov model that simulates natural histories of various pneumococcal diseases (clinical and radiological pneumonia, meningitis, and sepsis), reflecting the most recent epidemiological data. We applied the model to the 2006 Gambian birth cohort, and for each of the two classification systems, estimated the numbers of epidemiological outcomes and incremental cost-effectiveness ratios (expressed in 2005 international dollars per disability-adjusted life years [DALY] averted) under vaccination using PCV9 compared to no vaccination, over the first 5 years of life. We conducted deterministic and probabilistic sensitivity analyses over plausible ranges and distributions assigned to key uncertain parameters.

Result: The estimated numbers of radiological versus clinical pneumonias were substantially different, while the epidemiological outcomes of other pneumococcal diseases were very similar, across the two different classification systems. For the classification system based on the WHO standards, under base-case assumptions ($5 per vaccine dose and 91% coverage), the incremental cost-effectiveness ratio was Int$ 1,830. Based on the conventional approach, the corresponding value was Int$ 3,340 per DALY averted. Under both classification systems, model outcomes were most sensitive to vaccine price per dose, discount rate, and diagnostic costs.

Conclusion: Our findings show that the cost-effectiveness estimates of PCV9 can be influenced by different criteria for designation of cases as either radiological or clinical pneumonia. Our results highlight the importance of using a standardized classification system across different studies that address the same decision problem on whether to introduce pneumococcal vaccines into a resource-poor setting.