How are you with numbers?
It’s important for all medical communicators to master the essentials of medical statistics. Medical writers are not the ones carrying out statistical tests or procedures, but they may be responsible for interpreting and reporting statistics.
Reading, writing, and editing statistical information are essential skills for excellent medical communication.
Medical writers are the ones who need to discover the “story” behind commonly used statistical procedures, tests, and results. This is a critical mission that can improve patient outcomes, advance scientific discoveries, and help the public understand what’s going on behind the numbers.
Statistics for Medical Writers and Editors by Bart J. Harvey, MD, PhD, MEd, is a module in AMWA’s Essential Skills Certificate program. It provides valuable tips for demystifying and understanding the important role of medical statistics in medical communication.
Imagine you are editing a research report, and you encounter the following statistical information:
Length of hospital stay (days): mean, 13.3; median, 6; standard deviation, 12.1
Here are some questions medical writers should be able to answer to understand what these numbers mean:
- What level of measurement is used to describe the length of hospital stay?
- What is mean?
- What is median?
- What are the implications of the large difference between the mean and the median?
- What is standard deviation?
- What additional information does the standard deviation provide?
For some people, numbers can be intimidating. But, as the saying goes, it’s not rocket science.
The Role of Statistics in Medical Research and Communication
Medical research is the method scientists use to determine whether a treatment or test is effective, or how common a health‑related behavior is. Researchers who are designing studies choose a specific study design (for example, a randomized clinical trial or a general population survey) to measure the outcomes.
Research outcomes can determine how many patients experience disease relapses, the efficacy of new treatments, or the percentage of subjects who report they are smokers. This is important information, not just for the patients, but for public health and society.
What Do Medical Communicators Need to Know about Statistics?
Medical communicators should be able to do the following
- Identify the type of data used to measure a variable.
- Summarize categorical and continuous data with appropriate descriptive statistics, and interpret these descriptive statistics.
- Understand 95% confidence intervals and understand how they are calculated.
- Interpret P values and explain the concepts behind hypothesis testing (using the t test as an example).
- Apply these concepts to improve medical communication documents.
Data Are the Building Blocks of Health Research
Patients, health care providers, policymakers, and researchers all make observations and take measurements. These values form a collection of data that can take many forms. Here are a few examples:
- Using a home self‑monitor, a woman with diabetes finds that her serum glucose concentration (blood sugar level) is 132 mg/dL.
- A man with a family history of high blood pressure uses an automated blood pressure cuff at a local pharmacy and learns that his blood pressure is 120/80 mm Hg.
- On her initial visit to a family doctor, a woman reports that she has been pregnant 4 times and has given birth to 3 children.
When these observations and measurements differ from person to person, or from time to time, they are called variables.
Data are generally either categorical or quantitative.
Categorical data fall into specified categories, such as nominal data (occupation, country of birth, language spoken at home) and ordinal data (age groups, amount of pain, level of health).
Quantitative data can be described by numerical quantities. They are based on counts or measurements.
As medical communicators are looking at group data, they use descriptive statistics as the starting point for data analysis.
Here’s a real‑life example: We obtained heart rates from 40 medical writers and editors.
76, 81, 65, 86, 70, 79, 83, 87, 71, 80, 90, 77, 71, 84, 78, 77, 73, 85, 80, 81
82, 74, 81, 90, 79, 82, 82, 74, 86, 80, 83, 77, 81, 81, 78, 83, 79, 85, 82, 80
It’s tricky to make sense of the numbers as they are. That’s why data distributions are important.
Medical communicators use several graphical and numerical statistical techniques to describe and summarize these data. Graphical displays can include
- frequency distributions,
- measures of central tendency, and
- measures of variability (dispersion).
The following is just one example of a frequency distribution that uses the participants’ heart rates data.
From Statistics for Medical Writers and Editors: An Essential Skills Workshop of the American Medical Writers Association, p. 14.
These are just a few examples of the ways that medical writers can use statistics. Without statistics, medical writers are missing a valuable tool for conveying important information to various audiences.
Embracing Medical Statistics
Medical communicators don’t need to become statisticians. However, they should feel comfortable and confident when encountering and reporting statistical methods and results.
Most important, they need to learn to make the most commonly used statistical procedures, tests, and results clear or understandable to their audience.
When all readers grasp the meaning behind the data and the graphics, they can better comprehend the results of medical studies and clinical trials, and a healthier world is possible.