Medication therapy is extremely commonplace and more complex than might be supposed. From choosing the right medication to ensuring it gets to the patient and the patient uses it to assessing the effectiveness or need to adjust the therapy; hundreds of billions of dollars are affected. A multitude of technology, primarily software, has been brought to bear on these problems. A recent Agency for Healthcare Research and Quality report looks at the evidence regarding use of health information technology to manage medication therapies. (AHRQ Report) The report focused on evidence for improvement of processes and outcomes by use of electronic systems that collect and exchange information about patients, are integrated with existing HIT systems, like EMRs, and provide advice or suggestions to providers or patients.
A total of 428 articles were reviewed, reporting on a variety of study types. The studies evaluated a number of interventions such as use of CDSS, CPOE, eprescribing, barcoding and PDAs. The overall results suggest that care processes such as workflow, knowledge, skills and avoidance of medication errors can be improved by use of medication management technology, particularly in the prescribing and monitoring phases of MM, but there were few studies evaluating actual outcomes, which is understandable given the often indirect nature of the relationship between technology use and ultimate outcomes. Patient and provider acceptance and satisfaction appear high, with some issues regarding workflow changes. Reduction in errors in prescribing or dispensing may help avoid serious mistakes that can lead to major patient complications, or even death.
Some studies did an economic evaluation of the use of HIT for medication management, but no definitive conclusions could be reached from those studies. Prescribing cheaper drugs and avoiding adverse effects would be the primary cost savings, with some potential labor savings, but it is unclear that these offset the cost of buying and operating the technology. In general, the most beneficial use of HIT seems to be when it is a specific application targeted at a very discrete problem. Overall, one is left with the impression that although there are many studies, they don’t necessarily provide a lot of guidance on value, which is a common problem with understanding the costs and benefits of HIT in other areas.