Christopher Colombus, Kevin and Perry, men from Mars and women from Venus, Ronald Reagan, Friends Reunited, garages and golf clubs were, surprisingly, all of mentioned at the ICR forum on ‘A Marriage Made in Heaven or Grounds for Divorce: Exploring the interface between Project Management, Data Management and Statistics.’
This forum took the form of a debate on the motion “
The data management and statistics functions are well managed by project management”, between two project managers (Stuart Redding MICR CSci and Daniel Nasmyth-Miller RICR) who (unsurprisingly!) proposed the motion and two data managers (Tony Rees and Tracy Roe) who (equally unsurprisingly!) opposed it. Following the formal debate, there was plenty of time for discussion between the speakers and audience.
However, back to the beginning…
We started with a show of hands to determine the make-up of the audience and their current views on the motion. This showed the following:
- The overwhelming majority of the audience were project managers (29 out of 38);
- Four people had dual responsibility for project management and data management;
- There was only one data manager present. Had we scared off the others?
- Four came into the ‘Other’ category (just in case any statisticians reading wonder why the numbers do not add up!);
- We were fairly evenly split, ie, 11 for the motion, 13 against and 14 abstaining.
Next, a disclaimer was issued: “
The views expressed are deliberately provocative, are not necessarily the opinions of the presenters or their employers and may be deceptive, misleading and just plain wrong. On the other hand, we all know they exist in the industry!” And with the formalities completed, we were off!
Conducting an orchestra
Stuart Redding started by defining project management as the discipline of planning, organising and managing resource to bring about the successful completion of specific project goals and objectives. This has to be achieved within the constraints of scope, quality, time and budget.
He likened the project manager to the conductor of an orchestra: they have to ‘switch on’ other people to perform a task at the correct time, but they don’t actually perform any of the tasks themselves.
Stuart claimed that the scope is defined at the start of the project and all team members know what is included. The scope is also reflected in the budget and other similar documents, including the responsibilities matrix that are distributed to all team members and are signed off by data management so they know what tasks they are supposed to be doing, the project plan and the biometrics plan.
Quality metrics are received from the client (in a CRO environment) and are all defined up-front so everyone knows the requirements. The study timelines and budget (including the scope of work) are also defined up-front.
Project Managers are continuously planning, and data management and statistics are just parts of the team that the project manager has to manage. All elements of a project are uncontrollable, but data management is more controllable than some other elements. For example, if problems with investigator sites mean that recruitment takes longer than expected, then there is little that the project manager can do about that, but investing more resource in data management can help to catch up with the timelines at the end of the study.
Project managers have to communicate on a regular basis with up to 15 individuals and/or departments. This communication is held on an ongoing basis throughout the project. Data managers need to make sure that they attend all meetings, not just the first one, so that they know well in advance what will be needed in terms of resources in order to achieve their deliverables.
Some of my best friends are data managers…
Following on from Stuart’s presentation, Danny Nasmyth-Miller told us that he really likes project management: he finds it enjoyable, interesting and entertaining! But more than that, some of his best friends work in data management and statistics and it was a data manager who taught him about project management. They taught him that data managers have to be involved right from the start of the project and not just at the end. However, many data managers do not want to be involved at the beginning of the project even when given the opportunity.
He gave the following example of a project: “
When Columbus started out he did not know where he was going. When he got there, he didn’t know where he was. When he got back, he didn’t know where he had been. And he did it all on other people’s money.”
By definition, projects are unique and temporary and involve lots of interacting groups that change throughout their life-cycle ie, the project team at the start of the project is different from the team at the middle, which is different from the team at the end of the project. It is the job of the project manager to set clear objectives for the team.
Finally he posed the question “
Are data management and statistics the Kevin and Perry (as per Harry Enfield) of clinical trials?” Are they saying “
It’s not fair. Nobody loves us.”? His provocative closing statement was that they should stop moaning from the sidelines and get on the pitch…
Project managers are from Mars…
Tony Rees and Tracy Roe started their joint presentation with another disclaimer: “
Any resemblance to any real life situations, people, companies, job function, Clinical Project Managers or any resemblance to any person alive or otherwise is purely coincidental…. but highly convenient for our argument against the motion!”
After asking the audience for suggestions, they informed us that in their opinion the best handbook for Project Management is the book “Men are from Mars, Women are from Venus”. Project Managers are the ‘male’ in the relationship, whose method of dealing with a problem is a tendency to retreat into their ‘cave’. They may literally retreat, eg, to the garage or spend time with other project managers in the bar. Data Managers find this hard to understand as they talk about issues in order to find a solution. This is a potential source of conflict. Do Project Managers and Data Managers need ‘marriage guidance’ counselling?
CRAs and most project managers do not understand the process of database lock; they just think that they need to send in the last CRF page and then the data manager rushes over to a large switch on wall called ‘database lock’ and then there is lots of crashing and whirring and the database is locked. It is more complicated than this and data management cannot lock the database if they receive 1000 answered queries the day before the planned lock date!
Continuing the analogy, they said that data management has a long-term relationship with another ‘man’, who understands their passion for data: the statistician. Actually, it could be considered that all three (project manager, data manager and statistician) are ‘in bed together’.
Project managers consider themselves to be superior to data managers and they do not accept that there is an overlap in their roles and activities. They think that they always know best, even if they don’t. As most project managers are ex-CRAs, Tony and Tracey argued, their primary commitment is to clinical rather than to the team as a whole. They therefore do not support and communicate with data management and ignore their suggestions.
The quality of leadership from the project manager determines the difference between a team passionate about what they’re doing compared with one that is only following orders. Delegation is also very important: as Ronald Reagan said, “
surround yourself with the best people you can find, delegate authority and don’t interfere as long as the policy you’ve decide upon is being carried out.”
Project Managers are at the top of the pyramid. They see themselves as someone who:
- Plays by the rules;
- Works exactly as the contract stipulates;
- Is a credible person who honours the contract;
- Works to one goal, which has been agreed with management or the sponsor;
- Makes the final decision.
Data managers see themselves as someone who is more of a team player, who:
- Focuses more on the relationships;
- Suggests changes to ways of working that would benefit the project and then gets rejected;
- Is a reliable person who honours change, structure and planning;
- Is a team player.
However, project managers and data managers actually need to have the same qualities ie,:
- Be a natural leader
- Enjoy challenges
- Take responsibility
- Create a vision
- Grab hold of a problem
- Take accountability
- Always define the structure
Tony and Tracey felt that divorce is inevitable due to the following issues:
- Fragmented teams;
- Poor communication;
- Slow recruitment (but the project manager gets cross with the data manager if they send a data query the day after a monitoring visit, that they did not know about);
- Squeezing of data management timelines;
- Data management need to see the project plan, but it is not always made available to them
The project manager should review the Statistical Analysis Plan and the Data Handling Plan. They should give protocols and other documents to the data management for their review at the beginning of the study; these documents will be improved with data management input. Importantly, project managers should not try to know how to perform the tasks performed by the data manager.
They ended their presentation with a real life example… “
The worst example I ever had was when the PM agreed that the primary variable should no longer be captured at site due to ‘logistical difficulties’; first thing the statistician knew was at the soft lock. The study did not meet its objective!” …and a cartoon, that showed a project manager saying “
I just saved a bunch of queries, I threatened the DM with my golf club!”
Discussion
Following the presentations there was time for discussion between the audience and the presenters. Some of the issues and questions discussed are mentioned below; please note that in many cases we did not come up with a solution!
- It was thought that the project management process generally works better if the data manager and project manager are part of the same organisation as there can be more flexibility, although in theory it should not make a difference.
- How much knowledge does the project manager need to have of all the functions involved in the study?
- An example was given where data management were told by the project manager, not to attend team calls during the start-up period of the study. None of us present thought that this was a good idea!
- Project management does work if you apply the theory, but it is not always applied as per theory. Are SOPs written sufficiently well to help the process or does the process rely on the experience and ability of the project manager to know how they should communicate with data management and the other departments? The use of diagrams to describe processes in SOPs is a big step forward; it helps to visualise how everyone interacts.
- Should we consider using more specialist project managers from outside the industry, with no clinical experience and background: is their knowledge of project management more important?
- Who should have the line management responsibility? Should this be with the project managers or someone independent of the project team? This does not happen in most organisations, but would this improve project management?
Finally, we had another show of hands to see how people’s minds had been changed following the presentations and the resulting discussion. This gave some complicated results, as there was an 8% swing away from the motion after the presentation and an 11% swing towards the motion after the discussion(!), but a total of 45% changed their minds. So maybe we project managers are prepared to listen after all!
