I’m back in the medical ICU, heading into seven straight nights on call. The last time I was on this schedule was in December—and I wrote about it then too: ICU Noise and Sleep Disruption. At the time, my daughter was six months old and waking up 2–3 times a night. Between her and the ICU, I was living in a constant state of interrupted sleep.
Now, she’s sleeping through the night (thankfully), but the ICU still isn’t. And the more time I spend here, the more I realize how the ICU environment itself—despite all its life-saving interventions—isn’t always built for healing.
In case you missed the first article: ICU patients experience fragmented, non-restorative sleep, with up to 50% of their sleep happening during the day. When I was waking up with my daughter, sure, I was tired—but I could function. For ICU patients, that kind of disruption can be dangerous. It worsens physical and cognitive recovery and increases the risk of delirium, which is as common as it is underappreciated.
Delirium affects up to 60% of surgical ICU patients. It’s largely preventable, but when it hits, the consequences are serious: increased mortality, longer ICU stays, cognitive issues that can look like dementia, and $148 billion in annual costs to the healthcare system.
Six months since writing about this… and not much has changed.
We still prioritize tasks over rest. Vital signs, labs, baths—they’re done at night for our convenience, not the patient’s recovery.
We still under-recognize delirium. Especially hypoactive delirium, which often goes unnoticed until it’s too late.
We still haven’t standardized basic environmental improvements. Things like bundled sleep hygiene, noise control, or circadian-friendly lighting are backed by data, but rarely implemented at scale.
The problem is real, it’s solvable, and yet here we are—still waking our sickest patients for things that could wait until morning.
Insights
Root Cause Analysis: 5 Whys
The 5 Whys process in root cause analysis involves repeatedly asking "Why?" five times to drill down into the root cause of a problem by exploring the cause-and-effect relationships underlying the issue.
My root cause analysis of the problem remains the same.
The problem: The ICU, designed to save lives, paradoxically disrupts recovery by causing persistent sleep fragmentation.
Why do ICU patients experience disrupted sleep? They are constantly exposed to noise, bright lights, and frequent interruptions for medical care (e.g., vital checks, medication administration, lab draws).
Why are there frequent interruptions and environmental disturbances in the ICU? The ICU prioritizes continuous monitoring and care delivery to manage critically ill patients, which involves constant use of alarms, 24/7 workflows, and essential tasks that can’t always be delayed.
Why are alarms and workflows not optimized to minimize patient disruptions? The ICU environment and processes are historically designed for clinical efficiency rather than patient-centered care, such as sleep promotion and minimizing sensory overload.
Why hasn’t the ICU environment evolved to better balance clinical needs and patient recovery? There’s a lack of standardization in environmental design, workflows, and interventions aimed at improving sleep.
Why is sleep disruption in critically ill patients not treated as a higher priority? Awareness of the long-term harm caused by sleep disruption—such as delirium, cognitive impairments, and increased mortality—is still growing, and addressing it requires cultural, operational, and infrastructure changes that are resource-intensive.
Impact Analysis
Impact analysis is the assessment of the potential consequences and effects that changes in one part of a system may have on other parts of the system or the whole.
Patient: Sleep disruption in the ICU significantly impairs patient recovery, leading to delirium, a condition linked to increased mortality, prolonged ICU and hospital stays, and long-term cognitive impairments resembling dementia. Beyond the ICU, these patients often face lasting physical, emotional, and psychological burdens that diminish their quality of life. The inability to achieve restorative sleep worsens physical recovery, exacerbates mental health challenges, and increases dependency on rehabilitation services post-discharge.
Clinician or Provider: The chaotic ICU environment complicates the delivery of effective, patient-centered care. Persistent noise contributes to alarm fatigue, reducing clinicians’ ability to respond efficiently to critical events. Additionally, managing the downstream effects of delirium, such as prolonged recovery and the need for sedatives or physical restraints, adds to provider stress, increases workload, and contributes to burnout.
System: ICU sleep disruption drives up costs through longer lengths of stay, higher resource utilization, and increased readmission rates. As I stated above, delirium alone is estimated to cost the U.S. healthcare system $148 billion annually. Further, patients with prolonged recovery often require post-ICU rehabilitation services, such as cognitive and physical therapy, placing additional strain on already stretched healthcare resources. By neglecting sleep disruption, the system unintentionally perpetuates inefficiencies, higher expenditures, and poorer long-term outcomes.
Solutions
Redeye Protocols for the ICU: Borrow a play from the airline industry. On long-haul redeyes, airlines front-load all the announcements, meals, and service at the beginning of the flight—then dim the lights and minimize disruptions until just before landing. We should do the same in the ICU. Bundle the evening care tasks (vitals, meds, labs) early in the night, then commit to a protected “sleep window” where we limit unnecessary noise and activity. Lights off. Voices down. We could even give out earplugs, eye masks, and white noise machines, since they’ve been shown to improve sleep quality and reduce delirium.
AI-Powered Smart Monitoring and Predictive Alarms: Leverage AI to analyze patient data continuously and predict clinical events, reducing the need for frequent checks that disrupt sleep. For example, AI can predict stability trends to help cluster care and medication administration. Machine learning algorithms can also optimize alarm thresholds, ensuring alarms sound only when truly necessary while maintaining patient safety.
Circadian Rhythm Restoration: Implement protocols to promote natural light exposure during the day (e.g., opening blinds, increasing daytime activity) and reduce artificial light at night. Use smart lighting systems that automatically adjust to mimic natural circadian rhythms, helping restore normal sleep-wake cycles.