Most clinical trials involve collecting data in some way or another. It’s essential that you know the basics of clinical trial data collection so you can adequately report on your trial’s results.
In this post, we’ll explain how data is collected in clinical trials and the different ways in which it is analyzed. We’ll also explain what you can do if you miss data collection deadlines and how to handle data discrepancies.
How Data is Collected in Clinical Trials
Data collection in clinical trials happens on a few different levels. Data is collected from the trial subjects, sites involved in the trial, and other parties associated with the trial. Below, we’ll explain these kinds of data collection and give some examples to help you understand how they work.
Clinical Trial Subjects
In most cases, the only data collected from clinical trial subjects is when they come in for their regular appointments and treatment. When a subject visits the clinic, they may be asked to fill out a questionnaire or tally sheet that gives specific information about how they’re doing with their treatment.
Sites Involved in the Trial
Data is usually gathered from sites involved in the trial regularly. Sites involved in the trial may be asked to send specific information about their site, such as how many patients are being treated there, what kinds of treatment they’re receiving, and how subjects are doing with their treatments.
Third Parties Associated with Trial
When data is collected from third parties associated with the trial, they may gather information about how many patients are being treated at each site and what kinds of treatments they’re receiving.
5 Ways to Analyze Data in Clinical Trials
Data collected from clinical trials is usually analyzed by the CRO or statistician who helped design the trial. There are several different types of statistical analyses used to evaluate data in clinical trials, but most of them boil down to these kinds:
1. Descriptive Statistics
Descriptive statistics describe a set of observations, such as the number of patients, average treatment response, and counts of identified adverse events.
2. Statistical Tests
Statistical tests conduct a hypothesis test to make inferences about population parameters from a sample dataset. They may include confidence intervals for the true mean or proportions based on a sample or compare two populations using the differences in their sample means.
3. Hypothesis Testing
Hypothesis testing is used to make inferences about population parameters. For example, hypothesis tests may be used to determine whether there’s a difference between groups concerning treatment outcomes or adverse events.
4. Regression Analyses
Regression analyses are used to study the relationship between variables. For example, they may be used to see how well a subject’s changes in certain lab tests correlate with their corresponding changes in health conditions.
5. Measure of Association
This kind of analysis is used to determine whether two variables are related or not, such as whether a subject’s age at the time of treatment is related to specific health conditions.
How to Prevent and Treat Missing Data in Clinical Trials
Missing data are common in clinical trials, but it’s important to take steps to avoid missing data as much as possible. If you have some missing observations on your dataset, there are several ways to handle this situation that don’t compromise the quality or integrity of your results.
Complete Case Analysis
Complete case analysis is the simplest method for dealing with missing data—it excludes observations that have been recorded as missing from all analyses.
Available Case Analysis
The available case analysis includes only those observations where all variables are available, which leaves out a large number of subjects.
Imputation of Values
This technique assumes values were missing because of random chance. It substitutes the mean of the available observations for each missing value, leading to bias if there’s any overall trend in the data.
Multiple Imputation
Multiple imputation is a more accurate technique that uses statistical methods to create multiple copies of your dataset with different imputed values for each missing point. Then, your analysis is conducted separately over each dataset to make overall inferences from the combined output of all datasets.
Last Observation Carried Forward (LOCF)
This method excludes observations recorded as missing and replaces them with the last non-missing value for that subject before treatment stopped. This approach should not be used if the last data point is after a subject was no longer on the study.
Data in Clinical Trials Wrap Up
Data collection, analysis, and reporting is an integral part of clinical trials. As you can see, there’s a lot to do when it comes to clinical trial data collection and analysis. Ensure you know all of the requirements around these processes before you start your next trial, or you could face adverse consequences down the road.