[The walking-counting test. A simple test for assessing the risk of falling]

Falls and fractures caused by falls are common in the elderly. Interventions to prevent such falls and/or fractures are available. But there are no appropriate tests in general practice for the targeted use of preventative measures. This prospective study was undertaken to validate a simple "walking and counting" test for assessing the risk of falling.

92 patients of a general practice, aged between 70 and 96 years (64 women, 28 men) were included in the study. All patients were given the "walking and counting" test in which they would walk as quickly as possible along a stretch of 4 m, repeated immediately in a 3-step sequence while counting backwards from 100. The walking speed without counting was measured initially (vo), as was the speed during distraction by counting (vD). The difference between the two values, delta v%, measured the percentage change under distraction. In 80 patients it was possible to record the number of falls over the entire subsequent period of 12 months. All data were assessed by relative operating characteristics (ROC) and Bayesian analysis.

40 patients had falls in the follow-up period of 12 months, falls being more frequent the higher the age and if there had been previous falls. While all patients were able to do the test, individual values differed considerably. There was a statistical correlation between the calculated change in walking speed and the risk of falling. Using a threshold value of 20% slowing of the walking speed, the test sensitivity was 95% (0.51-0.81: 95% confidence interval) and the specificity 0.85 (0.71-0.93), with a positive predictive value of 0.77 (0.59-0.84).

The walking and counting test is a simple and informative test, more reliable than other parameters, to predict for an individual patient with danger of falling over a 12-months period. It provides the general practitioner with a tool for demanding appropriately targeted preventative measures.