Bioequivalence and Bioavailability Forum • SAS Programming

venu
☆

Mumbai, India,
2007-02-23 20:42
(6649 d 03:46 ago)

Posting: # 536
Views: 11,539

SAS Programming [Software]

Dear HS,

Being a biostatistician in this field from past 3 years, I have seen recently an ample of growing opportunities in the SAS programming in Clinical Trials and BA/BE studies.

I like doing sas programming and learning new techniques, procedures and updating the skills every day. :-)

Many companies have biostatistician but they are not skilled in the field of sas programming.

They do the analysis from SAS, WinNonlin and other statistical softwares and generate tables, figures and listings manually i.e. in MS Excel and MS Word typing and entering the results and making reports.

But I always used to generate the tables, figures and listings required in the Clinical Trials and BA/BE studies by sas programming and preparing the templates which can be used in future studies. :-)

My 1st question is:- Is it mandatory to do SAS programming in Clinical trials and BA/BE studies?

My 2nd question is:- Does any guidelines give some relevant information on sas programming?

My 3rd question is:- Can the FDA make some principles on sas programming in Clinical Trials and BA/BE studies?

I have read in the many guidelines that for the role and responsibilities of a biostatistician an equally qualified/trained/experienced are required in the Clinical Trials and BA/BE studies.

My 4nd question is:- Can a non-statistician (for eg:- B.Com., B.A., M.Com, MBA) do the statistical analysis, generate tables, figures and listings, SAS programming and preparation of statistical report?

I know the answer to this 4 question and the answer is NO.

But I still want to reconfirm, since I have seen these situations in some CROs.

Best Regards,
Venu

—
Best Regards,
Ms. Vandana Panchal
Biostatistician
Mumbai
India

drshiv
★

India,
2007-02-25 17:07
(6647 d 07:21 ago)

@ venu
Posting: # 540
Views: 9,672

SAS Programming

Post reply

Dear Venu,

I am tempted to answer your queries out of curiosity!!

❝ Being a biostatistician in this field from past 3 years, I have seen recently an ample of growing opportunities in the sas programming in Clinical Trials and BA/BE studies.

Correct assessment.

❝ I like doing SAS programming and learning new techniques, procedures and updating the skills every day. :-)

You are a good learner and will be bright star one day in this field if you continue to do what you are doing.

❝ Many companies have biostatistician but they are not skilled in the field of sas programming.

Again correct assessment of the ground realities.

❝ They do the analysis from SAS, WinNonlin and other statistical softwares and generate tables, figures and listings manually i.e. in MS Excel and MS Word typing and entering the results and making reports.

This is the standard practice in most of the CROs and pharmaceutical companies.

❝ But I always used to generate the tables, figures and listings required in the Clinical Trials and BA/BE studies by SAS programming and preparing the templates which can be used in future studies. :-)

You are really smart!!

❝ My 1st question is:- Is it mandatory to do sas programming in Clinical trials and BA/BE studies?

No its not. You can do all the calculations in a piece of paper using calculator if you are good biostatistician. But, SAS is hard to buy and easy to use. The choice is yours.

❝ My 2nd question is:- Does any guidelines give some relevant information on sas programming?

You mean BABE studies?. There is a SAS program in the appendix of FDA guideline on Statistical issues in BABE studies. The book supplied by SAS to users called Statistical analysis of clinical trial data is good. Don't look for guidelines for everything. Look for solutions, use them in your milieu and justify its use.

❝ My 3rd question is:- Can the FDA make some principles on SAS programming in Clinical Trials and BA/BE studies?

Oops its hot!!!

❝ I have read in the many guidelines that for the role and responsibilities of a biostatistician an equally qualified/trained/experienced are required in the Clinical Trials and BA/BE studies.

It's safe to have sombody like that. If you dont.. you know what to do...

❝ My 4nd question is:- Can a non-statistician (for eg:- B.Com., B.A., M.Com, MBA) do the statistical analysis, generate tables, figures and listings, SAS programming and preparation of statistical report?

❝ I know the answer to this 4 question and the answer is NO.

Bravo!!! correct answer. You get marks 100 out of 100 for this sixer.

❝ But I still want to reconfirm, since I have seen these situations in some CROs.

There are some funny CROs still do that. One CRO gave the offer of Head-Biostatistics job to my secretary who is a B.com. funny ah!!

venu ☆ Mumbai, India, 2007-02-26 06:02 (6646 d 18:27 ago) @ drshiv Posting: # 541 Views: 9,586	SAS Programming Post reply
	Dear Sir, It seems you have answered all my questions and have given me an inspiration by replying to all my doubts. Thank you. Best Regards, Venu — Best Regards, Ms. Vandana Panchal Biostatistician Mumbai India

vernon gray ● 2007-03-09 05:57 (6635 d 18:31 ago) @ venu Posting: # 570 Views: 9,548	SAS Programming Post reply
	I am not sure about this. The Statistician must have adequate training for the job - he/she could take the requisite courses after completing a BA, MBA, etc. or even through online courses and producing a certificate that qualifies him/her for the job. I doubt that any regulatory agency will say you need a Master's degree in Stats. Vernon

venu ☆ Mumbai, India, 2007-03-09 07:03 (6635 d 17:26 ago) @ vernon gray Posting: # 571 Views: 9,550	SAS Programming Post reply
	Dear Vernon Gray, Thanks for the input. Regarding the sentence "I doubt that any regulatory agency will say you need a Master's degree in Stats." I will suggest you to read all the guidelines thoroughly. You can also search in the Google on this topic and you will get many articles and information on this. Best Regards, Venu. — Best Regards, Ms. Vandana Panchal Biostatistician Mumbai India

Helmut
★★★

Vienna, Austria,
2007-03-09 13:33
(6635 d 10:56 ago)

@ venu
Posting: # 572
Views: 9,746

Qualification for Biostatistics

Post reply

Dear Venu,

excuse me for cutting in.

❝ Regarding the sentence "I doubt that any regulatory agency will say you need a Master's degree in Stats." I will suggest you to read all the guidelines thoroughly. :-)

❝ You can also search in the Google on this topic and you will get many articles and information on this. :-)

OK, let’s go to the bottom line. Which guidelines are you referring to? Which keywords in Google did give you ‘many articles and information on this’?

I’m not qualified in your sense either, because I’ve got no master’s degree in statistics myself (though I’m a member of the International Biometric Society, the International Society for Clinical Biostatistics, and the European Network for Business and Industrial Statistics).

Just to give you some examples:

First programming experience in 1976, coding a software package for PK/NCA/BE (60,000+ lines of source code in 1981 – 1983), evaluation of 500+ PK/BA/BE-studies, international lecturer on BE, co-author of a textbook on BE, never getting a deficiency letter questioning my qualifications,…

not qualified (that’s me)
Unplanned interim analyses (if planned, without α]-adjustment), improper handling of missing values, violating statistical assumptions (ANOVA of values from a discrete distribution like t_max), violation of statistical analysis plans (if existing at all), 1000 pages of SAS-output in a BE-study including an ANOVA-breakdown for eye-color (OK, the drug was atropine ), ignoring multiplicity issues,…

qualified (PhD in mathematical statistics)

Unfortunately I’ve seen far too many ‘qualified’ Push-the-Button-Statisticians in the last years, and I’m always happy if someone goes beyond (you would probably say below) SAS, like in this post.

Since the European Concept Paper on BA/BE for HVDs/HVDPs is open for comments from ‘International scientific societies in statistics and in pharmacokinetics’ only, I tried to interest my colleagues of the IBS to no avail. They considered the topic ‘trivial’; their lowest entry point most likely is the Kullback-Leibler divergence…

So please reconsider your prejudices! Three years of SAS programming experience do not impress me at all.
:lamer:

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes