While medical algorithms can be valuable decision support tools, calculating them by hand can bite you. With hand calculations, the clinician could be risking the patient’s life.
Data is the lifeblood of an algorithm. Manual data entry takes time and poses the risk of data entry errors. Sometimes the data available in a clinical setting such as body weight is not as accurate as data in the electronic health record (EHR). Recording the result of a calculation into the medical record is subject to transcription error. These problems can be avoided by interfacing directly to data sources. As healthcare providers we must improve or minimize any process that poses risks to patient safety.
Many data elements in health care are reported in units, some of which may be unfamiliar to the user. Unit conversions can be challenging, especially when more than one measure is involved, as in handling concentrations. If the units are unfamiliar to the user, then the user may not recognize a mistake since there is no ready point of reference.
While healthcare providers tend to be intelligent, some are mathematically challenged. This is not helped when operations such as logarithms are reported in an ambiguous fashion (does “log” mean natural or base 10?). It is sometimes easy for errors to occur by one or more orders of magnitude, resulting in problems such as the 10-fold dosing error in children. Rounding of numbers can have unintended consequences causing sometimes irreversible harm for the patient and the clinician.
Sometimes people hope to avoid these problems by using a nomogram or other visual tool. They can be helpful, but only when they are used in a precise manner. Failure to use a straight edge or to properly read a value from along a line can introduce errors that add up when there are multiple steps.
The user’s environment can introduce all sorts of opportunities for error. Distractions, interruptions, misunderstandings, fatigue, and hurry can cause a person to make mistakes even with the simplest of algorithms. Any problem may be compounded if the user accepts the results without checking or making sure that the result makes sense.
So why would anyone want to calculate something by hand when something better is available? Algorithm automation allows a clinician both to use limited resources more efficiently and to avoid a myriad of pitfalls that can have serious consequences for patients. Results can be more accurate and readily available. The clinician can then use time more wisely, making sure that the result makes sense and he or she is using it appropriately. Even with automation, mistakes are still made, but the overall goal should be to eliminate every possible source of error so that the clinician is in the best position possible to safely provide healthcare to patients.
