Every time medicine is prescribed in a hospital, a computer gives recommendations of the medicine and its dose for the patient. Health professionals update these patient records on the computer. The new research finds that these electronic systems are flawed. Programming errors or poor design and are less to do with the health workers using the system and more to do with IT glitches.
Let's see how often these technology errors occur and what they mean for patient safety.
The research team reviewed more than 35,000 medication orders at a major metropolitan hospital to understand how frequently technology-related errors occur. The focus was on errors made when medicines were prescribed or ordered via a computer based system. The research showed that one in three medication errors were facilitated due to the faulty design or functionality of the electronic medication system.
It was also examined how technology errors changed over time by reviewing rates of errors at three time points: in the first 12 weeks of using the system, and at one and four years after it was implemented. These errors are expected to become less frequent over time as health professionals become more familiar with the systems. It was also observed in this study that the rate of technology errors was the same four years after the system went in as it was in the first year of use.
There are multiple reasons for the occurrence of errors. For example, prescribers face a long list of possible dose options for a medication on the computer and accidentally choose the wrong one. This can lead to a wrong than the one intended specifically for the patient's disease.
It was found that high-risk medications were frequently associated with technology errors. These included oxycodone, fentanyl and insulin, all of which can have serious adverse effects if prescribed incorrectly.
Earlier this year there was an error in South Australia’s electronic medical record system. This miscalculated the due date for more than 1,700 pregnant women, possibly prompting premature inductions of labor.
For instance, there was a case in the United States where a nurse obtained the medicine from a computer-controlled dispensing cabinet (known as an automated dispensing cabinet), which is used to store, dispense and track medicines. Due to the poor design, the cabinet allowed the nurse to search for a medicine by entering just two letters. This led to the nurse selecting and administering the wrong drug to the patient, causing cardiac arrest. The nurse faced a criminal trial due to this IT error.
This research study has identified specific examples of poor system design and even those errors that have been highlighted by the ones working in the system. The have produced a series of safety bulletins for the health system that describe and address these glitches. These include a drop-down menu that allows prescribing of a medicine via injection into the spine. This particular medicine would be fatal if administered this way. Another instance is where an in-built calculator rounds up or down the doses for medication according to set rules. But this could lead to incorrect doses in very young or lower-weight children.
With increasing digitization in our hospitals and health services, the risk of technology errors increases even before we consider the potential for error in artificial intelligence used in our health systems. This study has included recommendations to optimize the systems for each example. Organizations could then use these specific examples to test their systems and take action accordingly:
- Systems need to be continually monitored and updated, to make them easier and safer to use and to prevent issues from becoming catastrophic.
- Health IT managers and developers need to understand errors and recognize when system design is suboptimal.
- Since clinicians are often the first to notice issues, there should also be mechanisms to investigate and address their concerns promptly, supported by systematic data on technology-related errors.
This study is not to call for a return to paper-based records. But until the computer-based systems are made safe, there won't be full benefits gained from the enormous potential digital systems could deliver in health care.
Reference:
1. Peter J Gates et al., “How effective are electronic medication systems in reducing medication error rates and associated harm among hospital inpatients? A systematic review and meta-analysis,” Journal of the American Medical Informatics Association 28, no. 1 (November 8, 2020): 167–76, https://doi.org/10.1093/jamia/ocaa230.
2. Magdalena Z Raban et al., “Longitudinal study of the manifestations and mechanisms of technology-related prescribing errors in pediatrics,” Journal of the American Medical Informatics Association, September 11, 2024, https://doi.org/10.1093/jamia/ocae218.
(Rehash/Dr. Shreya Dave/MSM)