Introduce confidence interval to the injury model

[usr110]

For the underlying poisson injury model, using Standard Errors (SE), include confidence interval in the model and follow it through to the final outputs.

For the 95 % interval, use the equations below:

> lower_bound = predicted_value - (1.96 * standard_error)
> upper_bound = predicted_value + (1.96 * standard_error)

Use type = 'link' for the predict function and linkinv function to predict fitted and lower and upper values

fit_resp = ilink(fit_link),
upper = ilink(fit_link + (2 * se_link)),
lower = ilink(fit_link - (2 * se_link)))

where ilink is

ilink <- family(model)$linkinv

Reference: https://fromthebottomoftheheap.net/2018/12/10/confidence-intervals-for-glms/