Statistik-Visual
13 September 2017
Jacqueline Berlin

A wizard at maths with passion and expertise

“I only believe in statistics that I doctored myself.” This well-known quotation encapsulates the popular conviction that statistics can be used to prove anything. Whether it actually originates from Winston Churchill, however, is open to debate. Statistics are essential to a pharmaceutical company like Boehringer Ingelheim. Without statistics, there would be no market approvals. Before the authorities can approve a drug for use in a specific market, the manufacturer must demonstrate that it is both safe and effective. To do this, it conducts a number of tests and studies. This is where statisticians come into play: They have to prove that the results are representative and can be applied to a very wide range of applications. In principle, they apply results from random sampling to the body of patients as a whole. Daniela Fischer is one of our biostatisticians. A guest article by Jutta Cook.

As a statistician, you need a good head for figures. Daniela Fischer, 28, is not only a wizard at maths she also loves discussing her work. So what does a biostatistician do exactly? “Quite a lot of talking, actually,” Fischer says with a laugh. This is because the bulk of her work involves translation. “The medical team is looking for the answer to a question, and comes to us with a study proposal. Our task is to see whether we have the right methods and identify what the results actually mean. It’s important to communicate with great care to avoid misunderstandings.”

 

Each statistical method has its own prerequisites. Fischer must therefore first check which methods are most appropriate for the underlying question. “Only then can I evaluate whether the value that I calculate is really telling me what I want to know,” says Fischer. A certain element of standardization helps in this regard, even though such an approach may initially appear paradoxical. After all, clinical studies generally examine innovative new therapeutic approaches. “Nonetheless, using similar criteria provides comparability and can save a lot of work,” explains Fischer.

 

These evaluation criteria must be defined before the start of the study. Working together with physicians and data managers, Fischer develops questionnaires that enable the investigators to ask patients the right questions during the studies and deliver the required measuring points for entry into the system. She also determines the minimum number of patients required to draw a statistically relevant conclusion. “Calculating the sample size isn’t easy. Often, we need to conduct research of the relevant literature and try out various statistical methods,” she explains. “And again, we have to engage in close communication with the physicians because certain assumptions need to be set.”

 

Programming knowledge

 

Biostatisticians also need to know how to program. Colleagues from the Statistical Programming department are responsible for programming the tables and graphics used in the study reports. “They work in close collaboration with us and use our specifications to develop the required programs. However, they are not usually on board during the initial calculation of the sample size,” explains Fischer.

 

Programming, biology, medicine, mathematics – biostatisticians need knowledge of all these areas in order to carry out plausibility and quality checks during the course of the study. “Statistics software can do quite a lot and provides a number of predefined methods. However, we still need to check whether the results tally with what we want to know.” In line with the principles for good statistical practice, most programs are developed on a dual independent basis and are subject to strict quality controls. Among other things, this ensures that the design meets the requirements of the ethical review performed prior to each study.

 

Teamwork

 

Biostatistics bears a great deal of responsibility. A clinical study can cost 800 million euros. Sometimes, answers to questions that have not even been posed arise on a post-hoc or situational basis. “As with surprises that occur within the data itself, we need to take a very close look at exactly why this result occurred. This involves working very closely with colleagues from TMCP and the medical team.” And this is one of the things that Fischer enjoys most about her job: “I have a lot of contact with other people and often work as part of a team. It’s a great feeling when a study that I’ve helped to plan is successful.” Fischer particularly enjoys receiving positive results from cancer studies, which run over an extended period and require some specific analysis methods. “In these cases, we are improving survival rates. If we do everything right, patients benefit from an increased life expectancy,” says Fischer.

 

All data generated during the study is collated in a constantly growing study database. “Together with the statistical programmers, we prepare the statistical outputs – the tables and graphics – during the course of the study. This involves a great deal of front-loading, but means we don’t have to worry about the result layouts at the end.”

 

Providing proof

 

Until the study has been completed, the data is handled in blinded form. “Only once the study is over do we upload the treatment assignment. In other words, what data originates from patients in the group receiving the placebo or the comparator? Once this information is available, it is not possible to make any further changes to the database. Then comes the exciting part: We perform the specified analyses and then wait curiosly for the results,” explains Fischer.

 

Once all calculations and evaluations are complete, the results are entered in the study report. Around 100 pages of this document provide an explanation of the study and of the main analyses. A further 300 to 400 pages with tables and graphics explain the data used as a basis, and annexes of up to 1,000 pages provide summaries of the individual patient data. If the new drug is approved, the report is published together with the protocol and analysis plan.

 

With passion and expertise

 

How does someone become a biostatistician – and why? “It actually happened by accident,” says Fischer. She had always enjoyed math. So when she chanced upon a newspaper report about mathematical biometrics, the combination of math, biology, medicine and computer science had an immediate appeal.

 

Fischer’s first contact with Boehringer Ingelheim came during a plant visit. The company regularly invites students to visit and gain an impression of it as an attractive employer. When Fischer later attended a university career fair and learned that BI was offering a student work placement, she applied immediately. The master’s thesis that she went on to complete at the company provided an opportunity to show just what she could do: she compared two designs for dose finding – one conventional and one new. The result was so convincing that her boss offered her a permanent position – a role that she now performs with passion and expertise.

"If we do everything right, patients benefit from an increased life expectancy!, says Daniela Fischer. "
Jacqueline Berlin
Daniela Fischer_Boehringer Ingelheim
Before the authorities can approve a drug for use in a specific market, the manufacturer must demonstrate that it is both safe and effective. To do this, it conducts a number of tests and studies. This is where statisticians come into play: They have to prove that the results are representative and can be applied to a very wide range of applications. In principle, they apply results from random sampling to the body of patients as a whole. Daniela Fischer is one of our biostatisticians.

0 comment(s) for 'A wizard at maths with passion and expertise'

Add new comment

CAPTCHA
This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.
"We simply cannot dedicate one hundred percent of our energy to our family and our career – nor should we have to, says Claudia Lehmann. "
Jacqueline Berlin, Talent Acquisition Specialist
"Swimming against the tide can be very worthwhile!, says Dr. Elke Simon. "
Jacqueline Berlin, Talent Acquisition Specialist
"My efforts are helping to ensure that patients and healthcare providers learn about the therapies we offer!, says Johanna Flesch. "
Sarah Pfeffer
"Making More Health has deepened my understanding of charity and the right way to create social value that is, ‘teaching a man to fish rather than giving him a fish’, says Cynthia Chen. "
Jacqueline Berlin, Talent Acquisition Specialist
"If we do everything right, patients benefit from an increased life expectancy!, says Daniela Fischer. "
Jacqueline Berlin, Talent Acquisition Specialist
Find 500+ open Positions at Boehringer Ingelheim Jobs