Evaluation of death count estimates for Puerto Rico following Hurricane Maria
This descriptive finding examines estimates of death counts following Hurricane Maria in Puerto Rico for September and October 2017. We evaluate the monthly death count estimates and estimates of excess deaths in Puerto Rico based on historical patterns of variability by month for the 2010-2016 and published official death counts for 2017. Official death records from the Puerto Rico Vital Statistics Systems by month and year (2010-2016) were used to produce means and 95% confidence intervals (95% C.I.) for each month. Death count and excess death estimates for September and November 2017 are employed to: (1) illustrate the estimation process and (2) assess the accuracy of these estimates when compared to official death counts for the same period. Estimates produced with incomplete information were 2,987 (95% C.I. 2,900-3,074) and 3,043 (95% C.I. 2,995-3,091) for September and October 2017, respectively. Corresponding official death counts for the same months for 2017 were 2,928 and 3,040. Using estimated death counts, 1,085 excess deaths (95% C.I. 950-1,220) were estimated in November 2017. Using official counts yielded 1,023 excess deaths (95% C.I. 956-1,090). Despite initially overestimating the number of deaths in September and October by 1.04%, subsequent estimate of excess deaths using official death counts was within the 95% C.I. of the initial estimate. Our findings demonstrate the timely production of death count estimates following climate disasters using historical death records and a thorough study of previous experiences.