Proof of life: What we do to patients in the name of safety

The flawed reality of patient observations: The 3 AM flashlight

How does it feel to be woken up in the middle of the night? Are you discombobulated? Does it interrupt your sleep? Is it hard to have a productive morning? Now multiply it. Not once, but thirty-two times over the course of a night. And not for one night, but every night for 5 days straight, at what is arguably the lowest point of your life, and it is mandated and done by people who are doing it to help you.

Now, let's set the stage in an acute care setting. Observations are designed to protect patient safety: every hour of every day of a patient's stay, they are observed on average every 15 minutes.

Welcome to an evening in an acute psychiatric hospital. A nurse must open the door to a patient's room unannounced and conduct the observation, shining a flashlight on them (since the room is dark) to confirm "proof of life". The impact, 32 times over the course of 8 hours of sleep, every 15 minutes, you are checked on to ensure safety. But what is the hidden cost of that lack of sleep? The 3 am flashlight in your eyes, checking if you are safe without any regard to privacy.

Sadly, this is not an anecdote. Researchers placed actigraphy on patients in UK acute admission units and found exactly what you would expect: marked sleep fragmentation and night-time awakenings, with light levels in the room spiking in step with the observations.

The clinical cost of sleep deprivation: How safety protocols increase suicide risk

Poor sleep is not simply an unpleasant side effect of being kept safe. Insomnia is an independent predictor of suicidal ideation, roughly doubling the odds across a meta-analysis of forty-two longitudinal studies, and, crucially, that link survives when you control for depression. It is not that these patients were already depressed and slept badly. Sleep loss carries its own risks.

And it has been measured in exactly the setting we operate in. The one I am arguing needs much more attention. In a study of fifteen hundred adults admitted to inpatient psychiatric care, those whose sleep was most disturbed across their stay left with significantly higher suicide risk at discharge than those who slept.

Yes, I hope that sits with you, as it has with me.

The very routine we built to prevent suicide may be raising the risk of the thing it exists to prevent. And I am not the first to say so. In 2019, a consultant psychiatrist at the Maudsley argued in BJPsych Bulletin that intermittent nurse observations at night serve no purpose and cause sleep deprivation. He noted that suicide at night is rare, roughly two per ten thousand admissions, and that a patient determined to harm themselves simply waits until the door closes.

This is exactly the argument I made in my first post. In fact, a psychiatrist made it seven years before I did, and made it better. So, when are we going to challenge ourselves to think differently?

Mitigating risk without inpatient checks: Lessons from the SleepWell initiative

I can already hear the objection, because I hear it in every hospital I have walked through. "Fine, Todd, but if I stop checking and someone dies on my unit, that is on me." It is a very fair and very real objection, and it deserves better than a reassurance. It deserves evidence, and there is.

An NHS Trust in the UK ran the experiment. Under a program called SleepWell, they suspended the standing policy of hourly overnight checks and gave patients a protected sleep period wherever the clinical team judged it safe to do so. About half of the patients on the pilot units received one, and staff supported the initiative.

Guess what.

Sleeping-pill use went down. And there were no deaths or significant events attributable to the change. The trust decided to extend it. What they challenged was the blanket assumption that a fixed-interval check, applied to every patient regardless of need, is what safety looks like. Fortunately, the unit did not fall apart, patients slept, and nobody died.

Yes, it was an experiment that required courage and trust and showed real results. I am not suggesting it solves the problem.

Beyond virtual sitters: Next-generation contactless patient monitoring technology

This leads to the question of a new approach today. How do we create an environment that improves safety, reduces risk, and supports dignity? In other words, what replaces the flashlight?

The obvious answer is the wrong one; watching them more has not solved anything. A virtual sitter puts a human being on a live video feed of a patient in their bedroom, through the night. That is more eyes on the patient. It is not more dignified and arguably not safer. The watchers get fatigued; they have to watch more patients to "scale" and create "ROI".

Contactless is a different category. Our platform uses infrared to read breathing rate, pulse, and movement with an FDA-cleared medical device. It does not produce a watchable picture of a patient. Nobody is streaming the room. It observes without disturbing and alerts staff to risk as it emerges, rather than 15 minutes later. The system watches for risk without watching the person, which is precisely the distinction that has been missing from this debate. Privacy is not the price we pay for safety. Done properly, it is what the technology is designed to protect.

Patient safety vs. patient recovery: The final question for behavioral health leaders

Finally, I would like to ask you the question I keep asking myself. If you were admitted to an acute psychiatric hospital tomorrow, what would the goal be? To be woken thirty-two times a night in the name of safety? Or to get better?

We have accepted sleep deprivation as the price of safety. But what we have not done is ask whether that price actually buys the safety. In fact, the evidence that it does is remarkably thin, and the evidence that the price is high keeps growing. The technology to observe someone without disturbing them exists today. The only real question left is why we are still opening the door.

‍Author: Todd Haedrich, CEO, LIO