Thursday, April 12, 2007


Hospitals do all they can to avoid patient falls. Falls can lead to minor cuts and bruises, but they can also cause serious injuries. It is a cruel irony to be injured in a fall when you are being cured in a hospital.

All falls are recorded to evaluate what happened and why. One of our folks was recently looking through our reports and noticed a pattern. Three people had recently suffered falls just as they were about to be discharged. No, not after they left their room and were heading home. But while they were sitting on the edge of their bed, fully clothed, ready to go.

What was happening here? We think that our staff members were receiving a subliminal message: They would see a healthy, dressed person in the room and might not have paid the same degree of attention to the patient as they would have an hour earlier when he or she might have been sitting on the edge of the bed in a hospital gown. Slight dizziness or instability of this person would then lead to the fall.

So, now we have circulated the word to the floors to be alert to this possibility, and we are hoping to see a difference. This takes no major effort, just an extra bit of attention at the right time.

Unexpected problem, good analytical pickup by one of our quality and safety staffers, and a simple solution. Not all safety improvements require a huge effort.

I'm curious whether other hospital folks out there have seen this particular phenomenon, too, and, if so, what you did about it. Ditto for other types of safety and quality observations.


Val Jones said...

In our rehab unit, many patients did their best to fall on the day of discharge. The sad truth is that many of them got better care and were happier in the hospital than in their lonely apartments with a disability. Falls can be a cry for help - and the way to prevent them involves a complex, holistic response to their psychosocial situations.

Hylton said...

I can actually add a bit of data: a family member of mine was at BIDMC last year after an accident and fell under the very same circumstances - poised on the side of the bed. She'd been in bed for a few weeks and while not dressed to leave was in considerably better shape than she'd been in previous weeks and no doubt didn't attract/merit the same sort of attention now that she was more lucid.

Anonymous said...

Interesting find. The act of getting dressed, gathering one's things at the end of a hospital stay and waiting for discharge is taxing enough. In addition to alerting staff, patients and family members should be cautioned prior to discharge that a sturdy armchair is a safer waiting place than the edge of the bed.

Anonymous said...

Hi Paul
Unrelated question here... I am currently reading The Culprit and the Cure by Steven G. Aldana MD. In it he makes some interesting claims about how docs are trained to treat disease and not on prevention.

In your experience with BIDMC and HMS do you see that to be the case?

He suggests that the healthcare industry would be better off working like the dental industry (prevent first then treat).

And on a side note, have you read this book? It has a really convincing argument (built on hundreds of published studies) that lifestyle reason for most of America's health issues.

Here is a local review:
"This book can do more to improve well-being, now and for years to come, than the best available medicines."

Walter C. Willett, M.D.
Chair of Nutrition, Harvard School of Public Health

Anonymous said...

As the one doctor assigned to do all the fall evals, I have noticed another couple trends: 1) the fall at the nursing shift change, which at 7-8pm syncs up with 2) the fall when family goes home for the day. I do 95% of my fall evals 7-8pm or 10-11pm and the rest usually 4-5am (the elderly pee in the middle of the night fall).
Personal alarms and bed alarms are 99% ineffective in preventing falls.

Greg P said...

Hospitals are not "doing all they can" to prevent falls. This is in keeping with the statement, "I'm doing the best I can." You can always do more, do better.

There are many people functioning, yet marginally so, before they come in the hospital. So they come in, and they're kept at bedrest, get weaker, and even they don't realize it. In many cases, if the doctor isn't writing for the patient to be gotten up and walked in the hallways during their stay, it doesn't happen.

Christian Sinclair, MD said...

Anonymous (doctor who does the fall evals)

Could you give a citation on the ineffectiveness of alarms? Or is it anecdotal evidence? The argument over the effectiveness of bed alarms recently came up in a staff meeting and I would like to bring that back to them.



Anonymous said...

Purely anecdotal, but based on the experience of our group of docs. I won't go into details here, but working night shift I really see how long it takes for staff to respond and how patients can get the alarms off. This is a hot topic at our hospital as well. Especially after two patients died after hanging themselves accidentally on the posey at bedside.
Speculating here, how helpful would a prospective alarm trial be? Staff, knowing they are being studied, will be much more responsive to the alarms.

Anonymous said...

I dug around a bit. Take a look at these sites/articles:
Key quote: "It is important to remember that BADs will not prevent the patients from falling out of bed; they are intended to warn nursing staff that a fall situation is occurring."
Look to pages 24 and 36-37 (references 62, 77-79).
No statistical significance. Was the trend due to the nurses knowing the study was going on?

These are good jumping off points for research. Looking at these, I am not surprised.

RJS said...

When I was in the hospital (BIDMC, actually), I dressed in warmup pants to leave. You know, the slippery, shiny kind. The coefficient of friction on these items of clothing is much lower than with something like jeans, and they're much easier to put on.

Just a strange, off-the-wall thought, and not nearly as insightful as some of the things brought up by the other commenters, but maybe another data point to look at for those who track these things.

Paul Levy said...

Please explain what kind of bed alarms you are referring to, and why they would be helpful in these cases.

Anonymous said...

I remember when I was in the hospital last July at BI my roomate was an old man who had hip surgery. Well after the surgery he got demencha(spelling?) and constantly tried to get out of his bed. The nurses put the alarm on the bed to alert them whenever he tried to get out but when the alarm went off no one ever came to check on him. So basically since I had a broken leg and could not get out of my bed I would constantly have friends and family help this poor old man back into bed.

Kathryn said...

A couple of late comments on the above. There are a few review studies that have been done showing alarms are not effective. According to two systematic reviews of the literature [i.e., Evans D, Hodgkinson B, Lambert L, et al. Falls in Acute Hospitals: A Systematic Review. The Joanna Briggs Institute, 1998, and Oliver D, Hooper A, Seed P. Do hospital fall prevention programs work? A systematic review. JAGS 2000; 42(12):1679-1689], only two randomized control studies (at the time) have been conducted of fall prevention devices (pressure-based bed alarms and identification bracelets), neither of which showed a reduction in falls. (These would be Tideiksaar R, Feiner CF, Maby J. Falls prevention: the efficacy of a bed alarm system in an acute-care setting. Mount Sinai Journal of Medicine 1993; 60:522-527, and Mayo NE, Gloutney L, Levy AR. A randomized trial of identification bracelets to prevent falls among patients in a rehabilitation hospital. Archives of Physical Medicine and Rehabilitation 1994; 74:1302-1308.) A third 1985 study of an alarm device attached above the knee by a fabric band reported a 33-45% decrease among an unknown number of patients observed on two wards (Widder B. A new device to decrease falls. Geriatr Nurs 1985; 6:287-288).

In 2001, a report from the DHHS Agency for Healthcare Research and Quality stated “At this time, there is insufficient evidence regarding the effectiveness of bed alarms in preventing falls in elderly patients to recommend the practice.” (Making Health Care Safer: A Critical Analysis of Patient Safety Practices.) In 2004 the University of Iowa Gerontological Nursing Interventions and Research Center published a falls guideline (University of Iowa Gerontological Nursing Interventions Research Center. Falls Prevention for Older Adults.) that says "Considering the cost associated with purchase of bed alarms, with no evidence of their benefit, it is unclear why these devices continue to be used as a fall prevention intervention." (U of I GNIRC 2004 Guideline: “Fall Prevention for Older Adults”).

I have spent some time researching this issue because we have just submitted the results of a randomized clinical trial showing some very hopeful results (e.g., 82% reduction in fall-related injuries) from a new wireless device just on the market, FallSaver. It was developed by a physician concerned about falls in acute care, but tested in skilled nursing facilities (notably vets who notoriously won't call for assistance with anything). The results of the prototype study were published in JAMDA (Kelly KE, Phillips CL, Cain KC, et al. Evaluation of a nonintrusive monitor to reduce falls in nursing home patients. JAMDA 2002;3:377-382). The results of the new confirmatory study are available at We have several hundred patient-days of use in hospitals with similar results. Sorry if this sounds like an advertisement but we are obviously very enthusiastic about the chance to make a big impact on falls. The inventor is the same MD who invented the CellSaver, by the way.

On timing of falls, there was an interesting article published some years back that showed that statistically falls tend to occur within a few days of admission or discharge. I believe they attributed the early falls to lack of familiarity with new surroundings, and the later falls to overconfidence about renewed physical ability tied to pending discharge. I can't find that reference, however.