Table 1.
Hospital discharges in the HRAEB grouped by year.
Fig 1.
Time series of the monthly percentages of HRAEB hospital discharges due to death.
Fig 2.
Trends and structural change in the monthly percentage of HRAEB hospital discharges due to death.
(A) Trend modeling using a linear model. (B) F-statistics from the Chow test (alpha = 0.05; red horizontal line). (C) Segmented trend modeling using two linear models.
Fig 3.
Additive decomposition of the monthly percentage of HRAEB hospital discharges due to death.
Fig 4.
Monthly percentage of HRAEB hospital discharges due to death.
(A) Original values. (B) Autocorrelation function (alpha = 0.05; blue bands). (C) Partial autocorrelation function (alpha = 0.05; blue bands).
Fig 5.
First differences in the monthly percentage of HRAEB hospital discharges due to death.
(A) First differences. (B) Autocorrelation function (alpha = 0.05; blue bands). (C) Partial autocorrelation function (alpha = 0.05; blue bands).
Fig 6.
Forecasts from ARIMA(0,0,0)(1,1,2)12 models with 90% prediction intervals.
(A) Estimated predictions considering no structural change. (B) Estimated predictions considering the identified structural change at t = 24.
Table 2.
Predictions and observed values (test dataset) of the monthly percentage of HRAEB hospital discharges due to death.
Table 3.
Forecast-accuracy metrics for the monthly percentage of hospital discharges due to death at the HRAEB.