|
*Please be aware that SOME RECENT forum threads may not yet be indexed by Google. |
|
View Full Version : Number of Hospital Admissions - Monitoring with an Attribute Chart? D.Salman 11th October 2006, 12:52 AM Dear Experts, Could I employ the attribute charts to monitor the number of admissions (monthly) in our hospital? If yes, which chart p or np and what are the advantages from this job? Many thanks. Omar. Jim Wynne 11th October 2006, 10:50 AM Dear Experts, Could I employ the attribute charts to monitor the number of admissions (monthly) in our hospital? If yes, which chart p or np and what are the advantages from this job? Many thanks. Omar. The method of graphing used is dependent on what you want to know. Neither of the charts you mention would be appropriate, because (A) there are no "defects" or "defectives" involved, and (B) there is (apparently) no binary (good/bad, yes/no) choice involved. A simple line graph might suffice, with the number of admissions on the y-axis and chronology on the x-axis, and that data can be used for further statistical analysis. Tim Folkerts 11th October 2006, 11:20 PM How about just a Xbar (or Xbar, R) chart? Strictly speaking, the number of admissions is not a continuous variable, but if the number large enough (say 100+), then you could for all intents and purposes treat it as continuous. One caution is that most source suggest at least 30 points before really assessing control charts. If you are plotting one point per month, that's 2.5 years! Tim F D.Salman 12th October 2006, 01:42 AM How about just a Xbar (or Xbar, R) chart? Strictly speaking, the number of admissions is not a continuous variable, but if the number large enough (say 100+), then you could for all intents and purposes treat it as continuous. One caution is that most source suggest at least 30 points before really assessing control charts. If you are plotting one point per month, that's 2.5 years! Tim F Dear Mr. Tim, Thanks for sending help. May I know the advantages from using the Xbar (or Xbar, R) chart? Many thanks in advance.:thanx: D.salman. Atul Khandekar 12th October 2006, 05:21 AM D.salman, I agree with Jim and would go with a simple run chart as suggested. Unless there is some more info you can give us, I do not see where control charts will fit in. What is the purpose of charting, do you have any spec limits (upper or lower or both) for admissions to hospital, if you construct control limits how are you going to react to out-of-control situations, do you need to predict something based on charts, are you going to categorize the admissions (say by treatment / disease type...), will it be better to plot weekly numbers instead of monthly.... |
|