A recent study published in the BMJ shows that electronic medical health systems must be used appropriately and with caution.
Electronic medical health systems have been used in clinical health practices quite frequently. While their advantages include condensing data, enabling data sharing between health practitioners, and improving patient outflow, they also have a few disadvantages. Research on the impact of electronics on the quality of care has shown that the increased use of electronic systems decreases the doctor-patient relationship by hindering patient interaction. There has also been a record of improper note taking by some health professionals due to a reduced knowledge of the electronic system, creating issues with information distribution when a patient is referred to another specialist.
Electronic health records are useful for research studies
Apart from clinical use, electronic health records have proved to be useful in conducting patient research based on clinical findings in the electronic records. They can also be used to find an ideal sample size, assessing the incidence of a disease along with its relevant prevalence, as well as finding information on patients whose medical conditions may help with an ongoing research study.
However, recent observations have noted a few errors commonly made when taking data from an electronic medical health system for research. One example is identifying the billing code used for that visit to the clinic. This could have bias by being influenced by reimbursement policies rather than proper documentation about the original reason for the visit. When checking for lab reports of the patient, the biochemistry profile would remain totally unknown unless the doctor specifically asked for a lab report during that visit. If the lab reports were brought in by the patient by an external source, there might be a lack of documentation for those results.
Research studies taking data from electronic health record may have a big room for error
Keeping this in mind, researchers Denis Agniel, Isaac Kohane, Griffin Weber decided to see how much bias was present while using electronic health records. They conducted an observational study in a retrospective manner. They acquired their data from two main hospitals in Boston, Massachusetts, USA, that were equipped with inpatient wards, emergency care, and ambulatory care. They found that during the years 2005 to 2006,  there were 669,452 patients treated at the two hospitals during this year. They published their results in the BMJ.
The results of their study showed that if care is not taken while accounting for the appropriate data, most tests that receive data solely from electronic health records could have a big room for error. This implies that since data retrieved from electronic medical health systems is crucial in establishing the prevalence of a disease, the method and criteria chosen for that topic needs to be tightly regulated and followed by all healthcare professionals.
Written by Dr. Apollina Sharma, MBBS, GradDip EXMD
Reference: Agniel, D., Kohane, I. S., & Weber, G. M. (2018). Biases in electronic health record data due to processes within the healthcare system: retrospective observational study. BMJ, 361, k1479.
