In Puerto Rico, excess deaths following Hurricane Georges persisted for three months

The interruption in basic services such as electricity, drinkable water, and exposure to atypical circumstances following climate disasters increases mortality risk within the settings affected by these events. Recently, some members of academia have argued that no methodology exists to study excess deaths attributable to climate disasters. This study uses death records for Puerto Rico between 1990 and 1998 to assess excess deaths following Hurricane Georges by comparing death counts for 1998 with patterns of variation from the previous eight years. Because no population shift occurred in that decade, other than expected ones based on historical information, the average number of deaths is indicative of expected deaths and the confidence intervals are the ranges of accepted variation. If a count following a climate disaster exceeds the upper limit of the confidence interval these deaths could be considered above the historical ranges of variation and this excess could be associated with the climate disaster of interest. Death counts for September-November 1998 indicate that 819 deaths were in excess of historical ranges of variation. When the year in which Hurricane Hortense is excluded from the construction of the ranges of variation, the excess is 945 deaths. A total of 811 or 937 are missing in comparison to the official death count for this Hurricane. Considering that death counts data structures are comparable across the countries of the world, this method can be used to analyze the effect of other climate disasters.