Enhancing Medication CPOE Safety and Quality by Indications-Based Prescribing

Project Final Report (PDF, 1.13 MB) Disclaimer

Project Details -
Completed

Research shows that incorporation of medication indications-–defined as the notation of a medication’s purpose-–into the prescribing process has the potential to prevent medication errors and patient harm, as well as make prescribing more efficient. However, current computerized prescriber order entry (CPOE) systems do not effectively support adding indications to prescriptions. This project sought to address this limitation by convening six stakeholder panels to achieve consensus on the rationale, needs, requirements, and implications of incorporating medication indication into CPOE. Applying a user-centered design process and incorporating the recommendations from the panels, the researchers designed a prototype that incorporates drug indication into the prescription and ordering workflow. In addition, they interviewed patients and pharmacists to obtain feedback on sample medication lists and bottles with indications. The prototype was compared to two leading commercial CPOE systems. The team hypothesized that the new system would significantly improve ordering speed, error rate, and prescribers’ experience and satisfaction.

The specific aims of the project were as follows:

  • Convene six stakeholder expert panels to achieve consensus and buy-in on the rationale, multi-user needs, operational and interoperability requirements, interface design principles, limitations and barriers, and policy implications of incorporating medication indication into CPOE. 
  • Design and build a working prototype of an indications-enabled CPOE system, informed by the recommendations from the first aim. 
  • Formally test and compare this indications-based prototype to two widely deployed CPOE systems using eight pre-defined use-case clinical scenarios. 

The project convened six 90-minute webinars with more than 300 individuals from 75 organizations to define requirements, discuss implications, and create model features of an innovative CPOE system that incorporated medication indications into the prescription. The topics for the six webinars, along with the PowerPoint slides and recordings are available online at: https://www.ahrq.gov/chain/research-tools/featured-certs/improving-hit-prescribing-safety-year-1.html.

The team used the lessons and key insights learned from the stakeholder panels to design a working prototype for indications-based prescribing. Development of the prototype included expert input, contextual inquiry sessions, participatory design sessions, usability roundtables, and formative usability testing. Usability testing was conducted on the final prototype to compare performance of the prototype to two existing leading commercial vendor electronic health records and found that the prototype CPOE system outperformed the commercial systems for all scenarios in terms of participants’ efficiency, effectiveness, and provider satisfaction. The research team concluded that it is imperative to design CPOE systems to efficiently and effectively incorporate indications into prescriber workflows and optimize ways this can best be accomplished.

A Prototype Computerized Provider Order Entry System Reduced Medication Errors

Key Finding and Impact

A prototype CPOE was developed that allowed providers to record medication indications and showed them the drug of choice for that indication. Providers testing the prototype CPOE correctly placed medication orders 95 percent of the time, compared to 61 percent and 85 percent with two commercial systems.

Mistakes made by patients and providers can lead to medication errors.

Historically, doctors wrote prescriptions in Latin so that patients could not read the reason the medication was prescribed. It was an era of medical practice that believed keeping patients in the dark was beneficial. Medicine has come a long way in terms of patient empowerment. Today, doctors help patients to make informed choices by explaining risk and benefits of medications prescribed. However, if the patient has several prescriptions, they may confuse their medications. Medication errors may also occur during the prescribing process, including prescribing the wrong medication, wrong dose, or the wrong frequency of taking the medication.

Including medical indications in the prescribing process can reduce medication errors.

In medicine, an “indication” is a condition or reason for prescribing a medication or performing a test. Specifying the indication during the prescribing process can help avoid medication errors; however, the computerized prescriber order entry (CPOE) systems used by providers to prescribe medications do not usually include indications, or the process of adding it is not easy or intuitive. If the indication was recorded in the CPOE, pharmacists would have more information about the prescription, and the indication could be printed on the label of the medication bottle to help patients understand their medications.

“It is important for patients to keep their medicines straight – patients stop taking medicines without realizing which one they are taking for which condition. I recently saw a patient who stopped taking a drug she thought was for one problem (depression which was better) but was actually for her diabetes.”
- Dr. Gordon Schiff

Developing a prototype CPOE with indications and CDS.

Dr. Gordon Schiff and his team at the Brigham and Women’s Hospital developed and pilot tested a prototype CPOE system that incorporated indications into the prescribing process. They convened six 90-minute national webinars to gather input from providers on how to develop the prototype. Additionally, they observed providers entering medication orders to understand their process and inform the features that were ultimately included in the prototype.

The prototype gave prescribers a way to record the indication, and it showed them the drug or drugs of choice for that indication. The system included a feature that not only listed recommended medication choices, but allowed users to hover over a drug choice and read the rationale for prescribing the medication. It was also customizable such that the medication was not recommended if the patient was allergic to it.

The prototype CPOE outperformed two widely used CPOE systems.

Clinicians were then asked to test the prototype by entering information from eight test cases into the CPOE, and the results were compared to two leading commercial CPOE systems. Clinicians correctly placed 95 percent of orders using the prototype compared to 61 percent and 85 percent with the commercial systems. Those testing the system also reported greater satisfaction, noting it was easy to use and they would like to use it frequently. Preliminary findings from interviews with pharmacists indicated they almost universally supported including indications in the prescribing process because it would help them counsel patients, ensure medication errors were not made, and increase patients’ understanding of their medications. The prototype developed by Dr. Schiff was superior with respect to prescribing speed, safety, and user satisfaction as compared to two leading commercial CPOE systems.