Hi there i think the article attached with my comment may help you which is written by Bruce Hollingsworth

Article Non parametric efficiency measurement. Econ J 114: F307-F311

I found this article on the topic, but I am not necessarily a believer in this approach. I need to read through the paper more thoroughly.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3562895/

Sultan Khan: the problem is that I haven't found any review of the efficiency measures -including the one you listed- that classify Pabon Lasso model as one of them, and yet, it has been used a lot to measure efficiency.

Ariel Linden: I'm even confused because sometimes it's used as a performance measure and sometimes as an efficiency measure!

like this report of WHO:

https://www.aho.afro.who.int/en/ahm/issue/14/reports/assessing-efficiency-hospitals-malawi-application-pab%C3%B3n-lasso-technique

where it's measuring efficiency, and the one you have listed that measures performance.

And can you explain more why you don't believe in this appreoach please?

Hi Manar,

For me the issue is that I don't believe that these components are measures of efficiency. There are many reasons why average length of stay will vary, and most of them have nothing to do with efficiency (in other words, they cannot be modified in a manner that will make a hospital appear "efficient"). These models do not account for patient severity or case-mix! Moreover, these measures are dependent on each other. Obviously a longer length of stay will result in less bed turnover. Why would someone measure them both?

Finally, when compared at the aggregate level (ie., between hospitals in a region or country), I would expect that those with initial high levels of "efficicency" will drop in the next measurement, while those in the lowest levels of "efficiency" will increase in the next measurement. This is called regression to the mean, and it has been demonstrated to be the cause of perceived improvements in readmission rates (another measure typically used to compare hospital performance).

I hope this helps...

Ariel

As a follow up, I just spoke with a colleague who specializes in hospital quality, and he has never heard of this metric.

Hi Ariel,

In my studied sample of hospitals, sometimes the extreme values of the indicators ALOS, BTO and BOR may be caused by reasons that do relate to efficiency, like unnecessary long stays for example, or unnecessary admissions. and such problems, when diagnosed, can be modified to improve efficiency, this is what I am proposing in my research.

Do you have any recommendations for an efficiency measure that is as simple as Pabon Lasso and doesn't need too much data? because I'm a little bit restricted to the limitation of data availability in the hospitals of my sample.

See the attached paper. This is an excellent source that discusses efficiency vs quality, and provides a comprehensive assessment of all the existing measures

Thanks a lot Ariel

I'll check it