Before I start – I just want to emphasize that everyone is entitled to at least one boring Blog post. Well, this is mine. Apologies up front. I did have good intentions, however. I was even going to start it off with some experiences I had in the jungles of South America. But I changed my mind. This topic just had no way to interrelate - It is that boring. Then I thought I would share some useless knowledge on termite mounds and anteaters. But again, there was just no sufficient segue. Finally – I made a heroic attempt to tie some amateur knowledge I had with drug metabolism into the topic – but – failed again. So – I guess I should just start. But please be warned up front – while this knowledge is useful, I never thought about it in the past until I was asked to design a process around it. So be careful. Now then…lets begin…

I am sure most of the companies you work for have a rather defined process for the TMF. Wait – what the heck is a TMF? Ok – if you do not know, the TMF is the “Trial Master File”. Oh, ok. Wait…what the heck is the trial master file?

Great question, let me expand on that.

The Trial Master File – as I put it – is a one stop shopping experience for an agency reviewer. It is a document that spans all departments – from clinical, lab, site, monitor, stats, data management, etc. It contains a list of the minimum set of required sensitive documentation that reviewers love to look at.   These documents serve to demonstrate the compliance of the investigator, sponsor and monitor with the standards of Good Clinical Practice and with all applicable regulatory requirements.

All those SOPs and WIs that we train on, and read (??), and rewrite, and retrain. All those referenced documents that you signed-off on – you know those?  Right…well, those certain milestones that we completed – are recorded and eventually added to the trial master file. Typically – the TMF is started at study start-up. For ease of reading – I am going to stick with the required TMF artifacts needed for Biostats and Programming. We have it easy. Data Management – not so. Actually – if you are a data manager or if you know one – ask about the TMF artifacts. I am sure this will lead to some evil laughter.

Like I mentioned – the TMF is typically set up at study startup and usually managed by a separate group or your Clinical Project Manager – or even sometimes by the data manager. First things to go in are usually the sign off forms post protocol and database creation, etc. Hopefully your SOPs and WIs mention how and where these are stored.

The actual TMF will contain links or a description of the archive locations. For example – your SAP is a required artifact and will typically live in your electronic project notebook or Documentum or Livelink – what ever you use. Also – as this document is created at study startup, it is amended through out the study with the required artifacts. You will not have final TMFs and ADaMs until much later in the study. Note that a study cannot close until the TMF is final. This is a very important point – so know where those documents are – especially if you are outsourcing work to a vendor – you, as the sponsor are still required to finalize the TMF.

The TMF itself – again, we are using Biostats and Programming artifacts as an example - will contain information on the Lead Clinical Programmer and Lead Biostatistician. Next – artifacts related to programming – including – Data Definitions for ADaM (DDTs), Analysis QC Documentation , Raw Data Downloads – for interim and Final Submissions, SAS Programs, SDTM and ADaM datasets, TLFs, and the SAP. You may be realizing why we will have one document instead of files with all artifacts present. Well – there is just too much stuff. So a document with links is fine. Again – think one stop shopping.

Next we have the randomization artifacts, which includes the Randomization Plan, Randomization programming and validation. In addition to the links for these items – Randomization Parameters are usually required in the TMF as well.

Finally – we have artifacts related to the decoding Procedures for Blinded Trials. These include Study Unblinding requests, emergency unblinding requests and documentation, end of study unblinding requests and results and the subject evaluability documentation.

All these artifacts are important and it includes the required minimum set. At all times – we need to know or have an idea on what and where these items are. Again – think of the auditor, or rather, think like an auditor.

There is an added complexity that must be addressed as soon as possible – especially when dealing with older studies and outsourced studies. I will not go into detail on these, but with outsourced work – ensure that TMF artifacts are worked into the arrangement. Sooner or later the sponsor will be hunting for these.

I will leave you with this. Recently – I received a frantic call from a colleague of mine. He was concerned about the TMF – as our compliance group was prepping for an internal audit in about a week. The concern had more to do with the unknown of the TMF. What is it, what do I need, etc. Having a clear and defined TMF process is crucial. I was able to alleviate my colleagues fear by presenting him with a simple template. All he needed to do was fill in the blanks. It took him half an hour for each protocol.  So keep it simple. Have a process. Work with your compliance group on this and communicate.

Ok – so that wasn’t too bad, right? If you have any further questions about the TMF, I am quite the expert now, so feel free to contact me.

With the new PhUSE/FDA collaboration another project starts which will be open source (Working Group “Development of Standard Scripts for Analysis and Programming”). The group currently prepares the basics required like technical platform, guidance, ideas and processes how the open source project will be started, ongoing and much more. Never mind, but will this project have a chance?

One main problem is the motivation. Yes, people like to promote themselves, being seen and be able to make the life of others easier, but who will benefit from scripts for TLFs? In the end it is not just the companies and authorities? And do these companies even have a motivation when they should share their knowledge?

In my point of view this project has a great chance. And the reason behind is not an organized startup and maintenance, but the FDA being involved. We would not have achieved the current CDISC standardization without FDA pressure. The clinical standardization started with STDM. Then ADAM came. So the next step is about tables, listings and figures. An FDA vision presented on the PhUSE/FDA meeting is that there should be a repository of programs to be used by academia, industry and authorities. FDA will work out a motivation paper or something similar to explain the mission which should motivate the companies and hopefully works.

If something similar like SDTM and ADAM is now build up for TLFs, companies do have the motivation to be involved. Then they know this project and can influence its direction. The industry problems can be addressed and solved as well. Many people would like to participate in such a project and hopefully they will get agreement and some resources from their companies so they do not have to ask themselves whether it is ok to participate with their technical knowledge received by their company.

Another question which could arise with this open source initiative for TLFs is where the future goes with all the standardization: Do we lose our jobs? Just look back at SDTM. There is a standard available since years including many new processes around in the companies. Is there less work to do? It is not by now and will not be within the near future. It is likely that the job role will change as it did in the past, but due to always growing requirements I cannot see a danger there. And like mentioned in the last blog … there are always “I’m special” studies around.

To over come “Frühjahrsmüdigkeit” (01MAR2012 PhUSE Blog - Shafi), I tend to pretend I am something else – a surgeon, operating on a patient, a flight controller – safely bringing those planes to land; an intelligence analyst – making sense out of confidential data; an architect – building a skyscraper; a doctor, learning about and treating a patient; the list goes on and on – and it is mainly driven by my desire to “make my day positive and rewarding.” From a paycheck perspective, it is. I am not complaining, but lately, I have been feeling more like a painter: strip the old coat, prime the walls, one coat, two coats, and there you go – next. As I write this, I am actually waiting for my house painters to arrive. They are a happy bunch of guys – ex Navy, relaxed – this is what they like to do. No complaints – they make their own hours, take on only the number of jobs they need to in order to live, etc. Perhaps it is just mid life creeping up, but – it doesn’t sound that bad – with the exception of the pay. Actually – it was this that really got me to thinking about happiness and work life balance. With gas prices soaring, long commutes, horrible traffic – is it necessary to come into the office? Most will say yes – especially when we talk about team dynamics, etc. But, I thought we were progressing into a new world – a new way of life that is “virtual”. Don’t get me wrong, the office is important – it builds team morale, especially when one is a manager – you need to see your people. You need to interact with other groups, other teams. But – my guess is that, as companies continue to redefine their business models – becoming more operationally efficient, trimming layers of fat and what they call “redundant” positions, etc. – you will see a more distant, less communicative and more authoritarian – structure. Let me be clear – my use of authoritarian – is not in a negative tone – I am using it from a valid leadership style connotation (see Kurt Lewin).

Most of our teams are global. I have personally worked for many years with programmers and statisticians that I have never meet face to face – some just being a few floors up! I am not sure if the average person thinks about this, but our way of work is virtual – we spend our day coding, emailing, interpreting, logging on and logging off - sitting. Most of our IT budgets include operation and management of citrix farms and VPN/SSL connectivity. All this work, all this money, all this emphasis on technology and the virtualization of operations, on the “cloud”, on sharepoint, on communications,  all this emphasis on building the 21^st century platform – and what has it changed? I do understand that the correct answer to this is in reduced drug development cycle time and time is money, yes, I agree! But, there are residual benefits to these platforms as well– and that is changing the way we work.

I am just going to say it – work life (in our industry/line of work) would be amazing if we were given the freedom to log on where ever. A number of years ago – when I reported to one of the best bosses I had ever worked for, I needed to spend some time in Florida as my mother-in-law was being treated for cancer. My father-in-law had just passed, and we needed to create a support structure for her. I had asked if I could work remotely for a month. When I arrived – I realized I needed to adapt to my new working situation. I felt it was important to mimic the office, so I rented an office in downtown old St. Augustine. It was nice – the office actually came with an administrative assistant! I actually was able to work more hours – as I felt in control, and more “responsible” – and this was during critical path activities for a regulatory filling. There are a few other factors that made this a success – 1. There was a clear and well documented process for our operations – macros, work flow, communications, etc.; 2. I needed to understand, and adapt to my work environment – which I did by renting office space; 3. I became more responsible when it came to managing myself – not that I wasn’t before – but I felt fluid, free, but connected. I call these factors – PAM, or Process, Adaptation, Manage. I think these steps are crucial when one is considering or has the flexibility to work remotely.

Anyway – my point is, as we work to feed our IT infrastructure, I think the resounding theme of some of these blog posts is to also work to feed the future of stat programmers, statisticians, DMs, etc. We live in an ever virtual life – and because of this, I think it is about time we share the fruits of this to the workers. Consider virtual work pilot programs if you are in a position to do so – which includes trainings in virtual teams. Obviously, it’s not for all. There is a level of responsibility needed, that frankly, some people do not have. I have seen it – and these are the ones that “ruin it for the rest of us.” There are a lot of issues/risks to consider as well, but I think the pay off would be huge. In this age where we have come so far, where our world is more complex with working families, children, gas prices, mortgages, less vacation time, more work, etc., let us truly understand what work/life balance is.

This all started about 8 years ago, when I was just getting used to being thirty something, in the prime of my health, a sharp and active mind, no data issue or programming errors were getting past me. Just what you would expect from an experienced programmer. Then suddenly out of nowhere at the start of March it hit me, like the November fog of South Germany, where visibility is barely more than a few metres, I was suddenly in the middle of it. Frühjahrsmüdigkeit!

Of course I had no idea what it was, all I knew was that I was suddenly feeling tired, very tired, very sleepy! I couldn't explain it.  I still had the programming and data challenges to face like before, no more long meetings than at any other time of the year, I was even doing exercise! but no luck, I was just so tired. After a few days of watching me trying to beat this by emptying the coke machine and finishing the day's coffee by mid morning, my wise colleague sharing the office with me finally explained it to me. Frühjahrsmüdigkeit she explained. That is what it was, and drinking all the coke in the world was not going to help, except perhaps to make me go to the Gents a few more times than on an average day.

Ever since, I have noticed during the early spring period, I am tired at work, very tired, not the fresh spring feeling you think I would have. It also means that my creativity is perhaps not at its sharpest during this time. So accepting projects where I have to be creative is perhaps not the best thing during this period. Of course our daily work is not very creative. In fact I am not sure how much things have changed over the last 15 years when it comes to programming analysis datasets, tables, listings and figures. I feel that very little has changed, and if I am honest, then I don't really know why that is.

A friend of mine mine from the IT side told me about 8 years ago that there would be no need for programmers like myself in 6 years time because they would have standardised everything, and the programs would be done centrally, with minimal effort required at the local level to run these programs and produced the desired output. I tried to assure him that things were standardised, at least within the company, but the routine programming still does not seem to go down. In fact, the more we move towards centralised programming, the more programmers we seem to need at a central location, without a reduction in the number of local programmers required to do the same task. In any case, it has been more than 6 years and I am still programming, so I guess there is still a bit more standardisation and centralisation to go.

CDISC is being hailed as the great saviour. It doesn't matter what the problem is, CDISC will make it easier. Cancer, diabetes, we will find a cure as soon as we all accept and adopt CDISC standards. My feeling is that it is just another standard, different from the company standards we now have, and may help to reduce some paperwork, but when it comes to managing project programming and trial programming, I am not seeing any reduction in the resource required. I think the dream is that over time everyone will be used to working with this new data and somehow new tools will be developed (by others of course as the internal programmers are too busy trying to meet the day to day timelines), which will make creating analysis datasets, tables, listings and figures not a thing of the past, but things everyone can do even during the Frühjahrsmüdigkeit season.

So, when it comes to overcoming Frühjahrsmüdigkeit, I guess we could just drink more tea, coffee or even coke, not that they have been successful in the past. We could try to stimulate our mind by moving to new things like CDISC standards, new programming concepts, new tools which does the same task but using a more complicated process, or to a new company. Finally we could just accept that it is that time of the year, and perhaps it’s the hayfever that is causing tiredness, or perhaps its having to re-hash the same programming code for the umpteenth time that is responsible for the lack of stimulation. Whatever it is, it will just last a short period of time, and there is no need to think up a new company programming philosophy or implement a new programming concept during this time. Should we always try to move forward? or are there times when we should just take a break and say its "Frühjahrsmüdigkeit"!

and not enough of local level support to ensure we have local programmers in the future?

One of the things I noticed during a question and answer session at the PhUSE conference was that
when companies are looking for programmers the list of requirements was always on top of the
assumption that the person can already program. My question is, are we still hiring enough trainees
so that we have programmers in the future?

I guess my worry is that we are increasingly looking for managers when we are searching for programmers?
They must be good at communicating, they must be able to lead, organise and manage timelines. Previously
there were far more technical requirements in the adverts for programmers, but as that is slowly
disappearing, is there a danger that we end up with managers with little programming experience? If they
are then responsible for managing programming groups in other locations, then over time the experienced
programmers there can themselves be managers, and there will be no need for local managers!

It seems that there is no time for development these days. Why develop when you can buy off the shelf as
it were. If we develop new programmers then they will only go somewhere else afterwards. In some

countries this is the way organisations think, very short term thinking in my opinion, and it only works

because of the programmers are taking the initiative themselves to train and develop.

If we ask ourselves whether the organisations we work for are doing enough to develop future programmers,
do we feel happy about our answer? Is our organisation taking a lead to develop good programmers with the
work ethics and skills to represent the organisation? Or are we merely bystanders, assuming that other
organisations will develop, and we can take someone from there with their work ethics and standards,
assuming they can program already!