Optimizing patient throughput across the acute patient journey is essential if hospitals and health systems are to improve healthcare efficiency and clinical outcomes.The fundamental purpose of optimized throughput is to ensure the patient is in the best care setting possible for them at every point in their patient journey. Spending too little or too much time in an acute care setting can be detrimental to the patient. Take the case of a patient who stays in a hospital bed longer because they can’t get to a skilled nursing or rehab facility. It’s not a good clinical situation for the patient because acute caregivers aren’t set up to deliver the next care treatment plan the patient needs. Consequently, the patient’s rehab is delayed.
Poor patient throughput also results in lower capacity for a hospital as acute care beds remain occupied by patients who should be at another point in their patient journey. This increases the likelihood that a hospital will be forced to deny a bed to a patient with acute care needs. There’s a financial impact to poor throughput as well. When a patient stays an extra day longer than the time prescribed by their Diagnosis Related Group (DRG) classification, they cost hospitals more than $3,000 per day without generating additional revenue.
Further, these lingering patients are significantly impacting revenue opportunities for the hospital from new patients who could access beds if they were available. According to Kaufman Hall’s most recent National Hospital Flash Report, reducing the average patient length of stay by one day could generate nearly $20 million in additional revenue for the average 425-bed hospital.
Avoidable barriers
Several common pain points reduce a hospital’s acute care patient throughput efficiency. These include:
- When physicians fail to submit discharge orders before 11 a.m. (which significantly decreases the likelihood of a hospital freeing up a specific patient’s bed that day)
- Physical therapy and occupational therapy evaluations that are delayed or don’t happen
- When there is confusion over whether a patient is leaving with the appropriate medications or waiting on a prescription coming from the pharmacy
- Transportation delays preventing patients from going home or to another care setting
- Delays in getting prior authorization for a patient going to a skilled nursing facility or rehab facility
While those are among the major pain points, the list above hardly is exhaustive. More to the point, there’s a common thread running through each item: These are self-created barriers, and thus totally avoidable. It is incumbent upon hospitals, therefore, to streamline the process and eliminate any barriers to efficient patient throughput.
Improving throughput with AI
As artificial intelligence (AI) continues to gain traction in healthcare, providers are increasingly adopting it to enhance their operational efficiency. However, most health systems currently focus on applying AI to administrative rather than clinical tasks. This distinction is significant because while clinical applications of AI receive much attention, the more immediate and widespread use of AI is in optimizing hospital workflows and resource allocation. Administrative tasks – such as patient throughput, coordination between departments, and discharge planning – offer tangible opportunities for AI to improve efficiency, reduce bottlenecks, and enhance the overall patient journey. This focus on streamlining non-clinical processes paves the way for better care delivery and resource management across healthcare systems.
AI can be deployed by hospitals to improve efficiency at every stage of the acute patient journey. From triage and admission to the patient’s hospital stay and through discharge, AI streamlines the throughput process while improving coordination between departments, ensuring seamless transitions and efficient use of resources.
A patient needing an acute care bed comes into the hospital in one of three ways – through the emergency department (which is roughly 60% to 70% of the time), as someone who has a scheduled procedure such as surgery (20% to 30%), or as a transfer from another care facility (10% to 20%). And they’re all competing for the same bed.
The problem is that each of these pathways can be very siloed, turning the process into a “first in, first out” model instead of prioritizing beds based on clinical conditions such as which patient’s vitals are deteriorating faster. There’s little to no coordination, even within a hospital, much less a health system.
AI and data provide the tools to look at information not only in real time, but to project out 24, 48, or 72 hours based on analysis of the types of patients who usually come through the hospital doors. In addition, AI can include data that highlights planned discharges and help the caregiver decide on the best facility to transfer the patient. Using data and AI to offer predictive and prescriptive actions can help health systems avoid bottlenecks that form at the beginning, middle, and end of the acute patient journey.
Once a patient is admitted, AI can be used with workflow specific, operational, and EHR data to enable more coordinated care, proactive communication between caregivers, and optimized resource allocation. AI provides the transparency necessary to determine what internal units, such as telemetry, are backlogged or when they may become backlogged based on predictive analytics. This enables hospitals to anticipate and avoid potential patient resource utilization challenges.
Another way AI delivers value in patient throughput is enabling automation to complete a multitude of tasks that are necessary but of relatively low value, such as administrative work.
Automation also can be clinically valuable. One health system had a transfer acceptance rate of 98% for cardiology patients that matched very specific diagnostic criteria. However, it often took up to five hours and numerous phone calls to coordinate the transfer.
After reviewing their admission workflow and electronic health records (EHR) data, the health system changed their protocols to auto-accept patients with that diagnosis. The result was greater efficiency and, more importantly, improved speed to care.
On the discharge side, let’s say we have a patient who has been earmarked for discharge. They are progressing well and going to a post-acute care or skilled nursing facility. Today, we gather some basic information about the skilled nursing facilities in the general area that would be appropriate for the patient. We know some information about the patient themselves – what type of treatment plans they’ll need in this care setting, their type of insurance, etc.
A case management worker will take in all that information, have a conversation with the patient, and then identify five or 10 facilities to which the patient can be discharged. It’s a chaotic process without much clarity. AI allows a faster, more advanced process for provider matching because it can quickly analyze a patient’s EHR data and insurance information, along with historical trends of these providers. This allows hospitals to quickly narrow down potential destinations to one or two providers and get a much faster response time than they do now.
AI and workflow integration
By integrating data and bringing intelligence into their workflows, hospitals and health systems can leverage novel insights into real-time care orchestration which streamlines throughput in a way that clinically benefits patients, improves efficiency, and optimizes revenue.
But driving workflow changes through AI and automation requires a strategy, a commitment, and a platform that can easily be integrated into a hospital or health system’s digital infrastructure. A hospital can compile the greatest data sets and create dashboards of amazing complexity, but even the most advanced technology in the world will deliver no value in acute care environments if it can’t be implemented, understood by users, and actioned upon.
That’s why workflow integration of AI, automation, and data analytics is crucial to adoption and use by clinicians and support staff. A tightly integrated AI platform delivers actionable data presented in an easily consumed form to guide clinical and care orchestration decisions across the acute patient journey.
AI and automation can play a pivotal role in the acute patient journey by optimizing patient logistics into, through, and out of the health system. The technology’s ability to improve resource allocation, streamline patient flow, and enhance monitoring and management leads to faster recovery times and better outcomes. AI also plays a pivotal role in improving coordination between departments, ensuring seamless transitions and efficient use of resources. Finally, the transfer, patient progression, and discharge processes benefit from AI’s predictive capabilities that reduce bottlenecks, deliver situational awareness, and identify actions that improve the acute patient journey.
About Bob Zdon
Bob Zdon joined ABOUT in 2020 as Chief Operating Officer. Zdon is responsible for overseeing the product, marketing, and services delivery capabilities. Before joining ABOUT, Zdon was President of RAZR Marketing and Grayduck Health, a Minneapolis-based marketing and technology agency. Prior to that, he held senior leadership roles at other healthcare- and technology-focused companies, including Healthland, Universal Hospital Services, Hill-Rom, and Allina Health. Zdon earned bachelor’s degrees in electrical engineering and biomedical engineering from the University of Minnesota-Twin Cities.