Timed Up and Go Test

Timed Up and Go Test in Falls, Frailty and Economic Evaluation

The Timed Up and Go Test, usually shortened to TUG, times how long a person takes to get out of a chair, walk a short distance, turn round, come back and sit down again. Health economists meet the TUG in two places. Falls and frailty services use it to decide who receives a more detailed assessment or an intervention, so its accuracy shapes the cost per person correctly identified in any pathway model. Trials of exercise, rehabilitation and falls prevention also report it as a mobility outcome, although a time in seconds carries no preference weight and cannot enter a QALY calculation without a further link. A 2014 meta-analysis in community-dwelling older adults found a pooled sensitivity of only 0.31 for future falls at a cut-off of 13.5 seconds, and the 2025 NICE falls guideline (NG249) concluded that tests of this kind do not predict a person's risk of falling.

How the test is carried out and timed

Podsiadlo and Richardson described the TUG in 1991 as a timed modification of the earlier Get-Up and Go test. The person is observed and timed while rising from an arm chair, walking 3 metres, turning, walking back and sitting down again. The score is the time taken, so a higher score means slower and poorer mobility.

The test needs a chair, a measured 3-metre line and a stopwatch. Podsiadlo and Richardson presented it as quick to administer and needing no special equipment or training, which keeps its cost per test low in any screening pathway.

What the original validation study found

The 1991 study recruited 60 patients referred to a geriatric day hospital, with a mean age of 79.5 years. Podsiadlo and Richardson reported that the time score was reliable between raters and within raters, and that it correlated with several established measures of balance and function.

The correlations are negative because a longer time means poorer mobility while higher scores on the comparison measures mean better function. The authors also reported that the time score appeared to predict whether a patient could go outside alone safely, and concluded that the TUG is a reliable and valid test for quantifying functional mobility. The validation sample was small and drawn from one clinical setting, which matters when the test is later applied in fitter community populations.

The TUG and the prediction of falls

Much of the later literature asks whether the TUG can pick out older people who will fall. The answer depends heavily on study design, and the strongest prospective evidence is the least encouraging.

Shumway-Cook, Brauer and Woollacott compared 15 older adults with no history of falls and 15 who had fallen two or more times in the previous six months, timing the TUG on its own and with a cognitive or manual task added. They reported sensitivity and specificity of 87% for identifying older adults prone to falls. A case-control design of this kind compares people already known to differ, so its accuracy figures tend to be higher than those seen when a test is used to predict future events.

Barry and colleagues reviewed prospective evidence in community-dwelling older adults and published a meta-analysis in 2014. Using a cut-off of 13.5 seconds to classify people as high risk, they found a pooled specificity of 0.74 (95% confidence interval 0.52 to 0.88) and a pooled sensitivity of 0.31 (95% confidence interval 0.13 to 0.57). The test was more useful for ruling falls risk in than for ruling it out, the TUG score was not a significant predictor of falls in their logistic regression (odds ratio 1.01, 95% confidence interval 1.00 to 1.02, p = 0.05), and the authors concluded that the TUG should not be used in isolation to identify individuals at high risk of falls in this setting.

NICE falls guidance and the TUG

NICE's guideline on assessing and preventing falls in older people and in people 50 and over at higher risk (NG249, 2025) updates and replaces the earlier guideline CG161. It covers everyone aged 65 and over and people aged 50 to 64 who are at higher risk of falls.

The guideline's rationale names the TUG directly. It records that in community settings a number of studies assessing gait and balance, such as the TUG, were identified, and that although the committee agreed these tests are helpful in observing gait and balance problems, they do not predict a person's risk of falling. NG249 recommends against using falls risk prediction tools to predict a person's risk of falling.

The committee still saw a role for case-finding. It acknowledged that identifying people who have had a previous injury from a fall or multiple falls, are living with frailty, or have gait or balance problems is useful in deciding who may need a more detailed assessment. For an economic model of a falls pathway, the TUG therefore sits more naturally as a description of a gait or balance problem than as a risk score that sets eligibility on its own.

The TUG in identifying frailty

Frailty services use the TUG as a quick screen. The British Geriatrics Society guidance Fit for Frailty recommended gait speed, the TUG and the PRISMA-7 questionnaire for identifying frailty in routine settings, with a TUG cut-off of 10 seconds to get up from a chair, walk 3 metres, turn round and sit down.

The same guidance reports that the PRISMA-7 and the TUG at the 10-second cut-off have very good sensitivity but only moderate specificity for frailty, and that a brief clinical assessment helps to exclude false positives, giving the example of a fit older person whose isolated knee arthritis slows walking. A cut-off that catches most frail people may still miss most future fallers, because the two targets differ and the threshold that suits one does not suit the other.

Where the TUG enters an economic evaluation

The TUG enters economic evaluations as a screening step, an eligibility rule or a trial outcome, and each role needs different handling. As a screening step, its sensitivity and specificity decide how many people receive an intervention, how many of them would have fallen anyway and how many future fallers are missed, so the cost per QALY of a targeted programme depends heavily on test accuracy as well as on the effect of the intervention.

As a trial outcome, a change in seconds is a measure of mobility and not a utility. A cost-utility analysis of a falls-prevention or rehabilitation programme normally takes its QALYs from a preference-based measure such as the EQ-5D collected in the trial, and models falls and fractures as events with their own costs and utility losses. The TUG can then support the mechanism of effect without carrying the QALY gain, and any attempt to convert TUG change into utility would need a mapping function estimated for that purpose.

Illustrative worked example of screening with the TUG

All numbers in this example are illustrative, apart from the sensitivity of 0.31 and specificity of 0.74, which are the rounded pooled estimates reported by Barry and colleagues. A community service screens 1,000 older people, of whom 30% would fall in the next year without intervention. Everyone who scores above the cut-off is offered an exercise programme costing GBP 500 per person, and each TUG test costs GBP 10 in staff time.

1. Future fallers flagged and missed.

$$TP = N \times \pi \times Se = 1000 \times 0.30 \times 0.31 = 93$$

where $TP$ is the number of future fallers flagged, $N$ is the number screened, $\pi$ is the proportion who would fall and $Se$ is the sensitivity. The number of future fallers missed is 300 − 93 = 207.

2. Non-fallers flagged.

$$FP = N \times (1-\pi) \times (1-Sp) = 1000 \times 0.70 \times 0.26 = 182$$

where $FP$ is the number of people flagged who would not have fallen and $Sp$ is the specificity.

3. Programme cost. The service offers the programme to 93 + 182 = 275 people. The programme costs 275 × 500 = 137,500 and screening costs 1,000 × 10 = 10,000, a total of GBP 147,500. Of the programme spending, 182 × 500 = 91,000, or GBP 91,000, goes to people who would not have fallen in the year.

The TUG-targeted programme reaches 93 of the 300 people who would fall, so 207 future fallers receive nothing. About two thirds of the programme budget goes to people who would not have fallen, although they may still gain strength and balance from the exercise. A model that compared this strategy with offering the programme to people identified by a history of falls, or to everyone, would show how far the cost per QALY depends on the screening rule, which is the practical consequence of the NICE view that such tests do not predict falls.

Boundaries with nearby measures

The TUG is a clinical outcome assessment of the performance type, in which a person carries out a standardised task under observation. It differs from a patient-reported outcome, such as a fear-of-falling questionnaire, which records what the person says about their own health.

The Six-Minute Walk Test measures how far a person can walk in six minutes and is used mainly as an endpoint in heart and lung disease, whereas the TUG times a brief sequence of everyday movements and is used mainly in older people. The Short Physical Performance Battery combines several short tests of lower-limb function into one score. The falls rate is a count of falls per 1,000 occupied bed days used as a hospital safety indicator, an outcome of care and not a test of the individual. Slow TUG times reflect an activity limitation in walking and transfers, and frailty is a wider syndrome that the TUG helps to screen for but does not define.

Limitations and common misreadings

The TUG is easy to collect, and its simplicity invites over-interpretation. Several misreadings recur in service design and in models.

• Treating a cut-off as a risk score. A pooled sensitivity of 0.31 at 13.5 seconds means most future fallers in community samples score below the threshold.
• Carrying case-control accuracy into a model. Figures from comparisons of known fallers and non-fallers overstate the accuracy of prediction in an unselected population.
• Borrowing thresholds across purposes. The 10-second frailty cut-off and the 13.5-second falls cut-off answer different questions and come from different samples.
• Treating seconds as utility. A change in TUG time carries no preference weight, so QALY gains need a preference-based measure or a validated link.

Sources

• Barry E, Galvin R, Keogh C, Horgan F, Fahey T. Is the Timed Up and Go test a useful predictor of risk of falls in community dwelling older adults: a systematic review and meta-analysis. BMC Geriatrics. 2014;14:14. doi:10.1186/1471-2318-14-14
• British Geriatrics Society. Fit for Frailty: consensus best practice guidance for the care of older people living in community and outpatient settings. Part 1 published 2014. https://www.bgs.org.uk/sites/default/files/content/resources/files/2018-05-23/fff2_full.pdf
• National Institute for Health and Care Excellence. Falls: assessment and prevention in older people and in people 50 and over at higher risk (NG249). 2025. https://www.nice.org.uk/guidance/ng249/chapter/Rationale-and-impact
• Podsiadlo D, Richardson S. The timed "Up & Go": a test of basic functional mobility for frail elderly persons. Journal of the American Geriatrics Society. 1991;39(2):142-148. doi:10.1111/j.1532-5415.1991.tb01616.x
• Shumway-Cook A, Brauer S, Woollacott M. Predicting the probability for falls in community-dwelling older adults using the Timed Up & Go Test. Physical Therapy. 2000;80(9):896-903. doi:10.1093/ptj/80.9.896