I have database of 78706 resident incidents in aged care facilities (5 years of data). I want to to learn and implement a tool allowing analyzing these data using following attributes:

  1. Resident
  2. Date/Time
  3. Location
  4. Result
  5. Injury

I want to be able to get following assumptions from my system which will be passed to specialists for further research, decision making and action:

Examples of outputs:

  1. Most of incidents in facility A with residents X, Y and Z
  2. Falls occur in North Wing between 2am and 5am
  3. Skin tears happen during showering in facility B
  4. Most incidents in a facility C related to repositioning

My question is not what software package can help me but what type of statistical analysis solves this problem - regression, cluster etc.

Can you recommend some practical books for a starter too?

  • 2
    $\begingroup$ I may be completely wrong here, but all the example outputs seems like simple counting operations (additionally the word "assumptions" confuses me). Do you want to extract statistical measures or do you want to e.g. extract the combination of variable values which "cause" certain injuries and hence allow prediction? $\endgroup$ – steffen Dec 7 '10 at 7:55
  • $\begingroup$ I want to find where most incidents are in a particular facility if they can be grouped by attributes I have and I am asking about statistical apparatus which can help me to do it. I want to 'see' with statistics that people 'Fall' MOST OFTEN than they 'Burn' or have 'Wrong Medication' or I want to 'see' that there is nothing suspicious. Sorry for using non statistical language. It is not simple counting for sure. $\endgroup$ – Igor Dec 7 '10 at 23:41

You could consider association analysis. If your time is discretized appropriately and the data support your 2nd example (Falls occur in North Wing between 2am and 5am), one possible learned rule that comes out of the analysis might be {North Wing, 2AM-5AM} => {Fall}.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy