Unclear structures, unclear responsibilities, outdated safety regulations, and worn-out work procedures: Hundreds of different errors in laboratory evaluations can be traced back to a few, albeit overarching, issues.
This is the conclusion of a study that looked at the entire clinical testing system. This includes the doctor ordering the test. Includes shipping and storage. This also includes laboratory training. You also need to consider the company that manufactures and sells the equipment.
Research shows that this complex system is made up of countless individual parts and could rival the complexity of the aviation industry. This had a similar problem, but can be effectively solved by following systems theory principles and taking complexity into account.
As a result, while the absolute number of fatalities due to accidents in the airline industry has decreased in recent decades, the number of airline passengers has increased every year.
solution is available
Equally dangerous are inaccurate test values, samples analyzed too late, or contaminated samples. The error can be traced back to six issues, most of which are avoidable.
These include decentralization, communication and coordination deficiencies, too much emphasis on safety-related rules, and adherence to outdated standards.
Additionally, there is a lack of understanding of the overall system complexity and awareness of the existing risks of incorrect diagnosis.
At least the first four issues could be quickly resolved with standardized letters, uniform designations, and increased training of staff in current standards. On the other hand, the last two issues imply deep intervention into the structure of the inspection system.
Instead, the study says, the prevailing belief is that doctors can prevent most misdiagnoses by recognizing their mistakes at the right time. This shifts responsibility for systemic problems onto individuals.
Although this is a study of the structure of a US laboratory, this approach seems very promising. Eliminating the patterns behind the problem, rather than tackling it locally and in a minimalist way, seems like a recipe that could work elsewhere.
