Where Better Delirium Care Begins

For more than a decade, Hartford HealthCare has been approaching delirium as measurable and preventable. In 2011, the health system implemented universal delirium screening at its flagship site, Hartford Hospital, an 867-bed teaching hospital. Using the screening data, Hartford then built a registry of hundreds of thousands of patients, which it used to quantify the cost of delirium — both to patients and the system itself.

From there, the system added interventions in layers. These included bundled order sets, an EMR-based delirium risk prediction rule, and safe prescribing defaults with clinical decision support for Potentially Inappropriate Medications (PIMs).

Since the program’s inception, the results have shown a positive impact on patient outcomes, as well as on hospital operations and financial performance. Among patients who developed delirium, average length of stay fell from around 16 days to 10 days over the first four years of the program. By year eight, delirium-attributable hospital days had fallen by nearly 40 percent. The hospital also reported a reduction in falls, sitter costs, pressure injuries, ICU restraint use, and 30-day readmissions¹.

Since 2023, Hartford HealthCare has been adapting these interventions to the unique environment of the emergency department (ED). This case study walks through the evolution of Hartford’s approach: what it took, what it returned, and what health systems elsewhere can learn from it.

About Hartford HealthCare

Hartford HealthCare serves around 27,000 patients daily, with a workforce of approximately 48,000 people. Based in Connecticut, it spans nine acute care hospitals, palliative services, and more. At this scale, delirium presented a considerable challenge.

Fortunately, the system already had a history of investing in geriatric care. In 2003, Hartford Hospital became a NICHE (Nurses Improving Care for Healthsystem Elders) site with an active geriatric nursing program. This would serve as a critical component of the work that followed.

In 2011, the system began implementing evidence-based interventions focused on delirium at Hartford Hospital. As the work evolved, it became increasingly clear that age-friendly care was driving better patient outcomes and operational efficiency.

Delirium is a Complex Problem

Delirium represents both a clinical emergency and an operational challenge. One of the most common and costly conditions encountered in hospitals, it affects up to 2.6 million people in the U.S. each year, at an estimated annual cost of $164 billion². It is deeply distressing for both patients and caregivers, yet without systematic screening, as many as 80% of cases may be missed.

Older adults are especially vulnerable to developing delirium during emergency and inpatient care, with underlying cognitive impairment as the leading risk factor. Patients with dementia face a 20–50% chance of developing delirium in the hospital³— and that fuels worse outcomes. Ultimately, delirium leads to longer stays, functional decline, readmission, and significantly higher mortality for older patients⁴.

To address this highly complex problem, Hartford Hospital started with an extensive screening program and an approach they could build on.

An Iterative Approach Driving Success

When Hartford Hospital launched their evidence-based model in 2011, the first step was universal inpatient screening and intervention. From there, they built on their findings. By 2024, the hospital was using screening data to predict the risk of delirium. Around the same time, Hartford introduced a new method of identifying PIMs. Here’s a closer look at the work.

As a starting point, Hartford Hospital implemented Project ADAPT (Actions to Enhance Delirium Assessment, Prevention and Treatment). The program was designed to screen all patients for delirium — across all inpatient settings. A positive screen would automatically trigger interventions for evaluation and management. Beyond this, the care pathway was straightforward: prevent cases from developing, treat those that do, and manage cases that cannot be resolved.

Patients were screened using Single Question in Delirium (SQiD) and attentional testing. Admitted patients were screened three times a day for delirium by the nursing staff using either the Confusion Assessment Method (CAM) or CAM-ICU. Screening was built into nursing documentation as a hard stop and integrated into rounds and handoffs. Completion ran at about 99 percent.

By screening all patients, the data revealed a clear picture of the prevalence and cost of delirium to both patients and the system. The resulting registry enabled Hartford Healthcare to understand the full impact of delirium across their hospital. What they found was alarming.

What the Data Revealed

• 90% of inpatient geriatric consults were for delirium.
• Approximately 19% of screened inpatients were CAM-positive at some point during their stay, but the rate varied enormously by service¹.
  • 3% after elective joint replacement
  • 20% after fragility fracture or colorectal surgery
  • 36% in trauma
  • 40% after coronary artery bypass grafting
• Delirium has serious consequences.
  • Patients who developed delirium stayed an average of 12 days, compared with four days for those who did not¹.
  • 30 percent went home compared with 70 percent
  • 10 percent died in the hospital compared with fewer than 1 percent
  • 43 percent were discharged to a skilled nursing facility compared with 17 percent

Associated Costs of Delirium at Hartford Hospital

ADAPT Interventions — Low Cost, High Impact

ADAPT included interventions to prevent delirium where possible, treat emerging delirium, and manage delirium that had already set in. It should be noted that this pathway prioritized non-pharmacological interventions. These involved low-cost, patient-centered strategies such as promoting mobility, sleep hygiene, hearing and vision aids, avoiding NPO, maintaining hydration, reducing ambient noise, and avoiding restraints. Patients and families also received education and participated in care.

In addition, patients 65 and older received a geriatric medicine consult. Younger patients or those with major psychiatric diagnoses were referred to psychiatry. Daily discussions during progression rounds allowed staff to keep each other informed on each patient’s progress. These conversations centered on one important question: Was the patient improving?

Added Low-Cost Delirium Management Support

Hartford Hospital layered the low-cost supports throughout the care pathway to bolster delirium management. These included delirium information in the patient handbook, encouraging families to participate in care, a personalized “Hartford HealthCare Cares About Me” poster at the bedside, and a sensory modulation room.

They also introduced a structured volunteer program to assist with socialization, nutrition, and mobilization where appropriate. Meal Mates was available for nutrition support, and an activity cart for social and mental engagement was available.

For patients who screened positive for delirium at any point in their stay, these interventions had a positive impact. And the hospital saw benefits, too.

Project PREDICT (2024)

An EMR-Driven Risk Stratification Tool

Leveraging data from the ADAPT registry, the team at Hartford Hospital discovered that 68 percent of patients who had screened CAM-positive at any point during their stay did so within the first 48 hours of admission. By the time geriatrics was consulted, or mitigating strategies were in place, it was already too late. Determined to get ahead of that first episode of delirium, the team implemented Project PREDICT (Predictive Real-Time Evaluation of Delirium in Clinical Therapy).

They adapted an existing EMR risk prediction model — the Infant Drop Risk Tool — and created a tool for delirium risk prediction. The new tool pulled a defined set of eight evidence-based risk factors from the EMR every 15 minutes, scoring each patient’s delirium risk with no manual entry. It then automatically flagged those at high risk, providing clear advisories for nurses and physicians.

At the threshold score, the nurse received an advisory to add the delirium care plan, which populated flowsheet rows for sensory items, family at bedside, mobility, sleep-wake cycles, and patient and family education. The physician received an advisory that linked directly to the ICU or non-ICU delirium order set, with the option to decline and give a reason. A delirium risk banner stayed on the chart for the rest of the encounter.

Project SCREEN (2023)

Geriatric Prescribing Panels — Making the Safer Dose the Default

Alongside Project ADAPT and PREDICT, the Hartford team developed an interprofessional workgroup including nursing, clinical informatics, geriatrics, and pharmacy. They identified the most overprescribed PIMs and, upon reviewing the existing prescribing panels, discovered that even the lowest doses and frequencies were often too high for older adults.

The interdisciplinary workgroup determined that a geriatric-specific prescribing panel for high-risk medications should be built based on literature review and a model out of the University of Chicago which had shown success.

Starting with 9 commonly ordered age-sensitive medications, the team created a set of geriatric-appropriate prescribing panels in their EMR (in this case, EPIC) that would automatically populate for patients aged 65 and older. They removed all inappropriate doses and added clinical decision support for high-risk medications directly on the prescribing panels.

The intervention launched without the need for staff education, and the results were immediately positive. Hartford HealthCare has since expanded SCREEN to include 38 additional medications, while adding age-sensitive medication details to EMR and 4Ms views. The approach was also tailored to critical care. Here, ICU clinicians were routed to a separate order set with the appropriate dosing options.

The Next Phase

Adapting Interventions to the ED

Despite all the important work that had been done at the inpatient level, Hartford HealthCare had no dedicated individual to lead improvements for older adults across any of its EDs. To complicate matters, data on the patient population was lacking.

In 2023, Dr. Katherine Farmer stepped into the role of Geriatric Champion at Hartford Hospital Emergency Department. Very quickly, she found that around 33% of the patient population was over 65. This revealed a pressing need to extend Hartford’s delirium work into the ED. With this insight, Dr. Farmer began working closely with the geriatrics department. Through their ongoing collaboration, Hartford Hospital saw success with the following geriatric-friendly ED interventions.

Geriatric Ambassador Program

Building a Sustainable ED Volunteer Program

Early on in her tenure, Dr. Farmer spoke with the volunteer coordinators at the hospital (those involved in the ADAPT interventions). She discovered that there was significant interest from local healthcare-stream college students seeking patient-facing roles. In response, she and her team created the Geriatric Ambassador Program to place them alongside ED staff. Volunteers assisted with therapeutic activities, hydration, nutrition, and ADLs. They also provided support with socialization and reorientation tasks – all shown to prevent delirium in older adults.

ED to BED Program

Prioritizing Patients Through Data

An “ED to Bed” program leveraged the delirium prediction tool developed by Project PREDICT. Here, the tool identified the patients who were at the highest risk of developing delirium. It then prioritized them for earlier bed assignment to reduce ED boarding time – a known precipitator of delirium in older patients. While the current data shows that Hartford has not yet been able to decrease boarding times for these patients, the hope is that more system-wide delirium interventions will improve throughput and make more beds available. It starts with prioritizing ED patients based on risk rather than order of entry, which leads to greater equity in care.

Quiet Corners

Creating Better Spaces

Many older adult patients with dementia or delirium in the ED do not require telemetry or isolation, and so they are deprioritized for rooms in the ED. This can have a negative effect on their health outcomes, leading to further confusion or worsening conditions.

At Hartford Hospital, the geriatric ED team sought a solution. They selected a low-traffic back hallway and designated it as a “quiet corner” for older adults without a room. The goal was to remove the deliriogenic elements of the ED as much as possible from the environment. The area has since been overseen by a nurse or PCT trained in geriatric care and equipped to assist with redirecting, ADLs, toileting, feeding, and other delirium-specific care processes. Geriatric Ambassador volunteers are often part of the quiet corner care also.

Geriatric Focus Group

Learning From Staff

Hartford is also looking to draw from the experience of ED staff. These frontline workers face innumerable challenges when caring for vulnerable older adults with cognitive impairment in the ED.

To better understand their pain points and find opportunities for improvement, the ED team at Hartford started a Geriatric Focus Group meeting. The optional meeting convenes in the mornings as the day and night shifts overlap.

The meetings have already yielded insights from staff and patients. In one instance, nurses were being asked to ambulate patients so that they could eat their meals out of bed. But this was nearly impossible, because there were no chairs for patients in the ED rooms once they were up. Based on this simple discovery, the department was able to order more chairs, so nurses could successfully help their at-risk patients move out of bed and maintain function.

More to Come

What started as an intervention for identifying delirium in inpatient settings has grown into something with much broader potential. By building on their successes and applying their learnings, Hartford HealthCare has been able to iterate on their approach, so proven care processes can be adapted to the ED.

Guided by this important work, the ED team at Hartford Hospital is now developing geriatric-specific order sets to provide ED doctors with a one-click menu for commonly ordered labs, nursing orders, and consults for older patients. They are also adapting Project SCREEN, making it symptom-based rather than medication-based for ease of use in the ED.

As they continue to build on their approach, Hartford HealthCare is sharing their knowledge across healthcare systems with support from the BRAIN Project — a collaboration between the Alzheimer’s Association, the West Health Institute, and GEDC.

References

1. IHI: Business Case for Becoming an Age-Friendly Health System
2. Improving Hospital Delirium Screening and Documentation
3. Prevalence, Risk Factors, and Impact of Delirium on Hospitalized Older Adults With Dementia: A Systematic Review and Meta-Analysis
4. Long-term clinical outcomes of delirium after hospital discharge: a systematic review and meta-analysis