- Care plans and order sets should be digitized and implementable so clinicians can focus on hands-on care instead of performing repetitive work.
- Evidence suggests that clinical protocols and order sets (e.g., ICU bundles) result in improved quality and outcomes and lower cost of care.
- Artificial intelligence will take digitized care plans to a new level, but machine learning will never eliminate the need for human insights.
As technology continues to change how we practice medicine, hospitals and health systems should consider digitizing care plans to support providers in clinical decision making that is evidence-based and avoids unnecessary variation.
Too often, providers rely on memory to create orders sets or develop pathways. This approach can lead to overutilization of tests or treatments and sometimes to delay or omission of critical elements. The downstream results are disorganized and inefficient, resulting in lower-quality patient care that can lead to injury, suboptimal outcomes and increased cost.
Digitization of care plans can help organizations achieve the Triple Aim and promote value-based care. The goal is to use technology to create processes of care that focus on achieving high-value outcomes.
At stake: Working smarter, not harder
Providers resist practicing “cookbook medicine” — but even the most famous chefs follow recipes that ensure the quality of the dish is consistent, allowing them to focus more on the presentation. In healthcare, time is wasted on recrafting care plans for new patients in similar situations.
There will always be nuances that require an individual care plan to be adjusted, but providers should not have to perform repetitive work that takes away from critical thinking time. Whether the provider is admitting a patient to the hospital with an exacerbation of a chronic illness or preparing for an elective surgical procedure, care plans and order sets should be digitized and implementable so providers can focus on hands-on care. In addition, building care plans into the electronic health record (EHR) for common conditions and procedures frees up providers to spend more time at the bedside.
The Centers for Medicare & Medicaid Services and other payers are using claims-based analysis to determine whether outcome and utilization patterns are within acceptable norms for certain conditions. In addition, many value-based programs have been developed that either reward or penalize providers based on quality, outcomes and costs. Implementing care plans will help providers be successful in this expanding value-focused paradigm.
Treatments driven by care plans make for better care
The focus on clinical outcomes and healthcare-associated conditions has led to the development of care plans to manage acute admissions from the emergency department (ED) and to avoid injury to patients in the intensive care unit (ICU). Likewise, the field of obstetrics has relied on standardized care plans for many years, with the management of normal labor and postpartum care following logical, evidence-based protocols. In addition, the American Academy of Family Physicians has endorsed care plans and algorithms for common conditions in the ambulatory setting.
Best practice is to incorporate these care plans and order sets in the EHR such that they are easily available for application. Their use should be monitored to gauge clinician adherence to agreed-upon protocols for healthcare utilization in appropriate settings.
Evidence suggests that clinical protocols and order sets result in improved quality and outcomes and a reduction in the cost of care. ICU bundles, sepsis protocols, preoperative checklists and admission order sets for common inpatient conditions are great examples that have been implemented and studied for their effectiveness. Similar results have been seen in the ambulatory care setting with clinical-decision algorithms designed to lead a provider through appropriate testing and choice of treatment to avoid overuse of antibiotics.
Accounting for the rise of artificial intelligence
Digitization of care plans can help providers implement established treatment decisions and algorithms, but the next question involves the role for machine learning to more deeply support healthcare decision making.
In the most basic of applications, artificial intelligence may take the form of monitor alerts in the ICU or drug interaction warnings in the EHR. This form of guidance will and should evolve, with background algorithms suggesting treatment pathways and providing likelihood of potential outcomes and risks.
But machine learning will not replace or override human decision making in clinical care delivery. Too many situations require the learning, practice and experience of bedside caregivers. Even though computers are quick at analyzing data, complex and urgent situations still require humans to make decisions.
Opportunities for additional research and longitudinal registry capture
Research should continue regarding the impact of digitizing care plans on quality, outcomes, costs and provider satisfaction. Allowing clinicians to work smarter and have more time for patient interaction and critical thinking should better fulfill the reasons most entered the profession. There may be an association between use of digitized care plans and reduction in administrative burden, with a corresponding improvement in provider burnout.
Beyond that, capturing the utilization and results of digitized care plans in shared longitudinal registries will open new research and application opportunities that have yet to be imagined.