The motivation behind this article was to share my personal experience and interaction with many wonderful friends and colleagues; past, current and future. Culled from many professional meetings and discussions, this is what I ultimately came up with.Biostatisticians and Statistical Programmers are like a successfully married couple. Much like two compassionate and diligent parents, they are individually and collectively raising upstanding children…only in this case, they are delivering high quality analyses after analyses. Hopefully this explains my odd article title. Basically, no matter how serious the personal or professional conflicts that exist for them both, in the end, they manage to achieve their goals with flying colors. Divorce? Not even an option. Couples’ therapy? C’mon now.

Roles and Responsibilities (Lower and Upper bounds): This is very important, but also very easy to implement. In order to be truly successful, both the Biostatisticians and Statistical Programmers need to collectively work at drawing a line in defining as much as possible the upper and lower bounds for any task, especially where overlapping or collective efforts are required. These collective activities could include defining Specifications, TLG’s (Tables Listings and Graphs) Shells, Titles, Footnotes and other items that exist in delivering analyses.

Specifications Review (High Priority): Sometimes, the creation and review of specifications are not given much importance at many places. One of the reasons for this could be the majority of errors found when the final reports are generated or delivered which lead back to errors or updates required in specifications. Then imagine adding the outsourcing (CRO’s and FSP’s in the US or outside the US) component to the mix.

Consider the study programmers program based on specifications. In actuality, they have followed and complete the assignment by following all processes, but the report is not what is expected. Who is to blame? Where is the gap? The outcome from this single activity leads to missed deadlines, incomplete reports, re-work, additional resources, negative financial impacts, dependencies and other negative cascading effects.

This will take place before the Lead Biostatistician and Study Lead Programmer release specifications to study programmers for programming to begin. The allocation of a dedicated 200% effort in design and a wide-spread review of specifications from both parties will result in the best possible results. The key point here is having successful implementation of the process as defined by the company in order to achieve the desired results. This review process can be done independently, through a meeting setup, reviewing non-standard items, putting extra focus on critical/complex algorithms, etc. This is a onetime effort which will reap healthy rewards.

Workload: Individuals need to work closely with their managers to evaluate the workload periodically and speak-up or push back instead of taking on everything themselves. Picture a sponge: at the top and in the front, the sponge is looking to absorb everything. Unfortunately, it could be leaking from the bottom or back. This is where the problems originate. Failing to take into account the bottom and back; areas which are not easily seen by the naked eye. I also want to make sure people are flexible in putting in the necessary extra hours that are sometimes due to agency commitments but should not be continuous or routine as over time, it might have a negative effect on individual health and performance.

This might be one major component or factor involved. It is important to point out that all tasks are performed at a high rate of speed in order to accomplish the tasks at hand. QUALITY soon becomes compromised when one’s focus shifts or drifts to QUANTITY, therefore the results in reduced time or allocation to individual tasks than normally expected. Once the addition of competing timelines enters the mix, and as always, unexpected additional ad hoc tasks, accelerated/changes to timelines – you could be lost and chaos would no doubt erupt.

All companies have Work-Life Balance initiatives, but should we have a Work-Work Balance initiative, too?

Coordinate and Help Each Other: As indicated above, it is important to continue to follow the upper and lower bounds defined, but both should be flexible due to the dynamic environment we live and breathe in by lending a hand during times of need. Therefore, a slight deviation from roles and responsibilities is possible and should be a onetime option and never become routine.

Below are a few examples of what I mentioned above:

• Programmers (Source and Validation), review and be well versed with all study documents such as Protocol, SAP-Statistical Analysis Plan, Specifications, TFL’s, etc. while working on any task. Ensure the information is consistent across the board; do not just go by one specification document required to get the job done. There could be conflicting information which would need immediate attention. Information which would never be thought of during the planning stage.
• Maintain centralized documents. Try not to create multiple versions of the same document, i.e. in different locations, local desktop copies, etc. This is an open invitation for issues which will go un-noticed or until the very end of the process. Sometimes, a very small change could have a ripple effect on multiple dependent documents.
• Ad hoc and exploratory requests take up the majority of the time. Specifications can be given orally, instead of written, by Biostatistician due to other priorities. In this case, might I suggest that the Programmer step forward in creating the specifications and have the Statistician (still on the hook!) to review and approve until which time does not release for programming.
• There could be a few complex analyses, so might I again make a suggestion by having a detailed 3^rd review done in addition to the Programmers source and validation. This gives the Biostatistician an opportunity to become a programmer and the Programmer an opportunity to validate or correct your understanding of complex statistical procedures. This could very well result in Biostatisticians becoming Programmers and vice-versa which is good for professional growth for the both of them; wearing each other’s departmental hat.

Communicate Working Styles: I would highly recommend the Biostatistician and the Statistical Programmer to have an open discussion in order to try to understand each other’s working styles and expectations before getting started to work for the first time, at the very least. By having these work discussions in the very beginning, future problems can be avoided altogether. Simply by having this open dialogue you could be making new friends and positive employees. Additionally, you will achieve a very important professional goal of being able to fit into any work environment and adapt to change much quicker and easier.

My simple take would be to express one’s self and resolve instead of keeping matters to oneself. By penning this article, I am hoping to extinguish the phrase: “It’s Hard to work with a Biostatistician or Programmer”.

Thank you all for reading and apologies if anything said was interpreted differently than the way they were intended. My goal here is to create a successful partnership between Biostatisticians and Programmers, simply by adding the right content of spices which will be accepted by all taste buds and easy to digest. This can be applied to any two teams, projects, individuals in any environment and the results and outcome is not quantifiable.

Finally, the most important point to be gleaned from the above exercise is attempting to maintain high quality deliverables which should always be the ultimate goal for all parties involved effectively.