Interrupted time series analysis after controlling for another variable (rate)

[sms1891]

Hi SAS experts,

I have a dataset with monthly hospital onset infection rates for 38 months with intervention variable and also community onset infection rates. How do I perform an interrupted time series for hospital-onset infection rates over time adjusting for intervention and also community onset infection rates? Any help with this would be greatly appreciated. I have attached my excel data sheet and the SAS code. I came up with a couple of SAS codes for this, but not sure if I am correct. The first one is for simple model without adjusting for Community_rate. But the main goal is to evaluate the trend in the HOV_RAte over time after adjusting/ controlling for Community_Rate.

proc autoreg data=TimeSeries;

model HOV_Rate = Baseline_Trend Intervention Trend_Change /

method=ml nlag=5 backstep dwprob;

run;


* Need to adjust for community_rate;
proc autoreg data=TimeSeries;

model HOV_Rate = Community_Rate Baseline_Trend Intervention Trend_Change /

method=ml nlag=5 backstep dwprob;

run;

Thank you,

SM

[sbxkoenk]

I have moved your post to the "SAS Forecasting and Econometrics"-board !

You should start doing this :

• Search the Communities (https://communities.sas.com/) for
  "interrupted time series" analysis
• Search https://www.lexjansen.com/
  https://www.lexjansen.com/search/searchresults.php?q=%22interrupted%20time%20series%22%20analysis
• Search blogs.sas.com (https://blogs.sas.com/)
  and bump into this (for example):
       Segmented regression models in SAS
       By Rick Wicklin on The DO Loop December 14, 2020
  https://blogs.sas.com/content/iml/2020/12/14/segmented-regression-sas.html
  Segmented regression models: This post shows how to create a segmented regression model in SAS. The breakpoints in the model are estimated from the data.

Koen