Pemodelan Angka Kematian Bayi di Provinsi Sumatera Utara Menggunakan Regresi Poisson Invers Gaussian

Abstract

Poisson inverse Gaussian regression modeling is one of the regressions that can handle data containing overdispersion. Overdispersion is a condition where the variance value in the data is smaller than the average. In this study, it was found that there was overdispersion in the infant mortality rate data in North Sumatra Province in 2022. Infant Mortality Rate (IMR) is a number that shows the number of infant deaths that occur at the age of 0-11 months in a given year, infant mortality rates are often used as a determinant of health status in a country. In this case, NorthSumatra Province is ranked 13th with an infant mortality rate of 18 per 1,000 live births out of 38 provinces in Indonesia. In this study, the number of infant mortality cases in North Sumatra Province in 2022 is used as the dependent variable (y) and the independent variables (x1) are the variable of low baby weigh, (x2) variable percentage of deliveries of pregnant women assisted by medical personnel, (x3) namely the percentage of the poor population, the percentage of pregnant women implementing the K6 Program as (x4) (x5) variable percentage of infant health services, (x6) variable number of medical personnel (x7) variable percentage of handling obstetric complications. From these variables, the best model is obtained, namely:μ= exp (2,969714 − 0,015514X2 + 0,041468X3 + 0,008090X6 +0,00367X7), By testing the parameters partially, it is found that several variables affect the Infant Mortality Rate (IMR) in 2022 in North Sumatra, namely variable (X2) which is the delivery of pregnant women assisted by medical personnel, (X3) which is the percentage of poor people, (X6) which is the number of medical personnel.personnel.