A method of macro-auditing and assessing the preventability of infant mortality using large volume computerized files.

We present a method for auditing and evaluating infant mortality with the aid of a preventability grading system, based on national computerized files of livebirths and infant deaths. Diagnostic categories and specific causes of deaths were classified into one of the following three preventability grades: Preventable (P), Possibly Preventable (PP) and Non Preventable (NP). This classification was then applied to two different scales: Preventability of Condition (PC) and Preventability of Death (PD) from which a third scale--Preventability of Mortality (PM)--was derived. The method was then applied to matched records of 39,786 livebirth and 452 infant death certificates between 1977 and 1984, in a semi-urban region in Israel encompassing 220,000 inhabitants. Comparison of mortality rates, according to the proposed preventability scores, demonstrated that higher infant mortality rate in non-Jewish population, or in Jewish mothers with a lower educational, was present only in the preventable categories (P or PP), while death rates due to non preventable causes were identical for all groups. The suggested macro-auditing method facilitates the assessment of large scale infant mortality rates in terms of preventability.