Tools for Fall-Risk Assessment
Almost every fall-risk research paper begins by underscoring how frequent falling occurs – more than 1 in 4 individuals over the age of 65 fall every year, a fifth of which lead to injury and too often death.
It’s surprising, then, that there have been a great many approaches proposed to evaluate a person’s fall-risk – some more effective than others. The spectrum spans from simple, no-tech questionnaires to high-tech computerized recording and analysis of functional performance tests.
Soft Assessment Tools
There are many factors in play when it comes to fall-risk: muscle weakness, visual impairment, cognitive impairment, medications, foot pain, foot neuropathy, and more.
Some of these can be captured in the form of questionnaires (e.g., Johns Hopkins Fall Risk Assessment Tool, Morse Fall Scale). In this kind of fall-risk assessment tool, each question typically has a score associated with each of its multiple choice answers, and the cumulative score is checked against a predefined scale. It’s an easy and inexpensive way to begin implementing a fall-risk strategy in a health care centre or even as a tool to help your elderly parents acknowledge that they may be at risk of future falls.
But such tools aren’t necessarily perfect. In one study, the simple question, “Have you fallen in the last 12 months” was found to be more predictive of future falls than four questionnaire-only based assessments.
Low-Tech Functional Performance Tools
The next level of fall-risk assessment is to have the patient undergo functional testing, under the watchful eye of a health practitioner armed with a stopwatch and a meter stick. The goal of such functional testing is to see how effectively the patient performs movements that are generally related to their daily activities.
The simplest and perhaps most popular of these tests is the Timed Up and Go (TUG) test. The patient sits in a chair, is told to begin, rises from the chair, walks 3m, turns around, walks back to the chair, turns and sits back down. The time taken between the instruction to begin until they are seated again can be easily captured with a stopwatch and compared against a predefined scale.
Other tests such as the Berg Balance Scale and the Tinnetti Performance Oriented Mobility Assessment go into more detail, each involving a collection of various functional tests (e.g., sitting to standing balance, walking stance, etc.) with instructions on how to score the patient in each.
Like the questionnaires, the cumulative score is used to rate the fall-risk. These approaches are somewhat complementary to the soft assessment tools described above, since they capture functional performance rather than historical/situational information that put individuals at risk for fall.
However, one downside is the potential for inconsistency or subjectivity during the scoring process of the 14 or 16 functional subtests/assessments of the Berg and Tinnetti tests, respectively.
High-Tech Functional Performance Tools
Going high-tech has several super-human advantages that help you see beyond its cost.
Firstly, technical solutions can see or measure quantities that people just can’t eyeball. Multiple scientific articles show how future fallers exhibit significant stride-to-stride variability during gait or increased sway while standing. Yet future fallers are only on the order of 10s of milliseconds and a few millimeters apart from their steadier counterparts in such scenarios, which would be practically undetectable to an observer, but not to computer hardware.
Secondly, some fall-risk research uses complex algorithms (e.g., local dynamic stability) in their predictive metrics which a clinician could never hope to apply without software. While a clinician might look up a single number in a table during an assessment, software can do the calculations a clinician just doesn’t have time for.
In fact, software actually makes it feasible to implement in the field the latest fall-risk research, which is not normally streamlined for medical practice.
For example, Howcroft et al. (2017) found that the difference in center of pressure movement extent during standing with eyes open versus eyes closed is significantly less in fallers than non-fallers. To estimate risk, the center of pressure extent values must first be calculated from a list of x-y positions for each balance recording. Then a ratio must be formed and evaluated with respect to the article’s normal (Gaussian) distribution – overall, a job best done by software.
Software can also automate integration of information from multiple sources to support a better informed fall-risk determination. Verghese et al. (2009) offered a variety of different fall risk metrics based on a patient’s gait such as double support time, gait velocity, and stride length variability. These were each related in a clinically relevant fashion, but to combine them to get a multi-faceted fall-risk prediction would require further statistical knowledge and a sit-down session with a calculator, a notepad, and the patient’s data.
Thirdly, when a multi-faceted approach to fall-risk assessment can be taken, the cause of high-risk predictions can also be narrowed and addressed. For example, if out of a list of different metrics a high-risk prediction were made due to high peak plantar pressures (another of the many factors associated with fall-risk), the clinician would have a starting point from which to reduce that risk.
Indeed a software system may also be able to identify the location of such a peak pressure for relief through prescription of a custom orthotic or other intervention. This is perhaps a strength of the low-tech multi-functional performance tests as well. One specific subtest might reveal a particular patient weakness leading to an associated intervention.
Finally, automation eliminates the reliance on people to reliably rate functional movement, a help to patients and clinicians. Patients get the most consistent assessment possible and clinicians don’t have to worry about having possibly improperly rated an individual on one or more functional subtests.
For example, some of the Berg balance scale ratings require the rater to judge whether an action can be done “confidently”, “safely”, or “needs supervision”. In the Tinneti test, it is easy to imagine situations where a rater would be unsure which of the multiple choices to make (e.g., steady vs. unsteady), possibly adding noise to the result.
In short, it would seem that high-tech solutions have the potential to be easier to use, more thorough, faster, and more consistent than the lower cost, low-tech functional performance tests. Questionnaires have their place – they provide a quick and easy entry to fall-risk assessment and are at least partially complementary to functional performance tests. Fall history, in particular, is a strong predictor of future falls that you just won’t find out without asking.
What fall-risk assessment tools have you used? What care scenario were you in? What pros or cons did you experience? Please let us know in the comments section below.