Epidemiology: Looking At Hysterectomy Rates Over 11 Years - 1 Need Help!

[jeremy2010]

Hi,

I'm new here. Anyway, I'm looking at hysterectomy rates in a state in north eastern USA and my aims are:

1. To look at overall rates over 10 years and determine whether they increased, decreased or remained stable.

2. To determine if a new type of hysterectomy procedure has increased overall rates in the 11 years period.

3. To determine if morbidity among recipients of the new type of procedure is higher that among recipients of the older procedure.

My data source is the local State hospital discharge data for non federal hospitals.

I'll be using data from the Census Bureau for the 2000 Census, specifically for the State involved to provide the denominator for calculating the hysterectomy rates.

I'd appreciate advise on which SAS procedures are best suited for the above 3. Suggestions on which SAS codes to use are most welcome.

Thanks for your help. Message was edited by: jeremy2010

[deleted_user]

To better help answer this question more details are needed.
For example what database are you using?

For example if you are using the National Inpatient Sample database from HCUP you will be need to be using PROC SURVEY with weight cluster and strata.

Trend data is always fun considering coding and sampling may have changed over the 11 year period. Also, why not round that off to 10 years, Titles with past decade sound better than past 11 years.

~Josh

[jeremy2010]

Thanks for your input Josh. I'll make the clarifications (revisions) accordingly!

[Doc_Duke]

Taken statewide, the census data is just a constant and 1) can be done with FREQ. You can test the proportions against a hypothetical distribution (like uniform). If you have populaiton by year, you will get a more precise estimate.

Ditto 2). You will need to know when the new procedure started.

To look at 3) , I would use LOGISTIC regression; it does not require the Census data. You could use year as a continuous predictor. Then the type of procedure could be another binary predictor.

You may not have enough sample size to estimate hospital effect, let alone provider effect, in 3.

Doc Muhlbaier
Duke