You cannot manage what you cannot measure. What you cannot measure, you cannot improve. These management adages are particularly resonant when it come to electronic health records (EHRs) and medical data.
You cannot manage what you cannot measure. What you cannot measure, you cannot improve. These management adages are particularly resonant when it come to electronic health records (EHRs) and medical data.
When the EHR mandates were passed down in the American Recovery and Reinvestment Act (ARRA) of 2009, the idea was that moving patient records to an electronic format would improve clinical efficiency and treatment outcomes, thereby lowering medical costs. While the jury is still out on efficiency, EMR software is being used to collect massive amounts of data that will, in time, improve treatment outcomes.
Previous to EHR adoption, the only way to aggregate large amounts of clinical data was to do so manually. Published clinical trials were the best way to discover new treatment options, but trials are limited in that they only record the data that the administering physician deemed important or appropriate. In addition to data limitations, it takes an average of 17 years—yes, 17 years— for clinical trial research to be incorporated into everyday practices, according to the Agency for Healthcare Research and Quality (AHRQ). EHRs can collect more data, and disseminate it faster than any clinical trial.
While EHR interoperability remains low, in the not-too-distant future, EHRs should be able to export large sets of anonymized patient data, allowing clinicians to discover patterns in treatments, symptoms, demographic information, and more. Physicians will be able to review their patients’ records against large datasets to establish better baselines and averages. This will also help better plan treatments. For example, an oncologist could predict his or her patient’s reaction to a certain treatment based on the reactions of other patients who share similar symptoms, genetics, etc.
This type of data is already being utilized, albeit in a limited capacity, in clinical decision support functionality. Clinical decision support software (CDSS) can review a physician’s diagnosis against an individual patient’s historical record. Also, CDSS can review a patient’s medication history and return data on the efficacy of past and current medications. That data can be used to make medication and dosage recommendations.
While the information is limited to a single patient and EHR vendor at present, improvements in interoperability should allow CDSS to draw from larger datasets. This would help further reduce the possibility of adverse reactions to treatments and medications.
In short, better measurement of health data will help physicians better manage patient health, and thereby improve treatment outcomes. Of course, the old statistical adage “garbage in, garbage out” still applies here. Conclusions drawn from inaccurate or incomplete datasets can be dangerous. Thankfully, the higher specificity of ICD-10 diagnosis codes should improve the quality of data, and the conclusions drawn from said data.
