How to plot bar chart with a layered background?

I have generated a bar graph consisting of 3 categorical variables, and I would like to "classify" the height of the bar graph by altering the background color (e.g. height less than 10 is green, and height greater than 20 is red). I would also like to set the xlim to be from [0 4] since this would make the graph look more asthetic, but this does not seem to work?

x_groups = categorical({'Jump Height', 'RSI','Stiffness'});
y_height = [30, 15, 26];

b1 = bar(x_groups, y_height);
b1.FaceColor = 'flat';

% Generate the relative patch 'boxes' for classification
box1 = [0 0 4 4]

box1 = 1×4

0 0 4 4

box_y1 = [0 10 10 0]

box_y1 = 1×4

0 10 10 0

box_y2 = [10 20 20 10]

box_y2 = 1×4

10 20 20 10

patch(box1, box_y1, [0 1 0], 'FaceAlpha', 0.2)
patch(box1, box_y2, [1 0 0], 'FaceAlpha', 0.2)

I would like to have the bars be in front of the patches. Please help.

[Geoff Hayes]

@Mark Kramer - you may want to look at the accepted answer at https://www.mathworks.com/matlabcentral/answers/8350-how-send-to-back-patch-objects-in-a-graph to get an idea on how to change the order of objects (patches, bars) so that the patch objects are moved to the back. As for changing the x-axes limits, I'm not sure that you can do that when the categories are used for the x-axis.

[Mark Kramer]

@Geoff Hayes thank you very much that helps a ton. I managed to find the solution for setting the x-axis limits when there is categorical data for anyone that might be interested:

hAx=gca;

hAx.XAxis.Limits = categorical({'Jump Height','Stiffness'});

Thank you again for the help!