Background: Loss of protective plantar foot sensation is the major
cause of diabetic foot ulcerations and ultimate limb loss. Identification of
patients without protective sensation can reduce the risk of unrecognized foot
injury. The current recommended screening protocol requires 10-g monofilament
testing of ten foot sites with use of a forced-choice paradigm. The objective
of the present study was to determine whether testing of fewer than ten sites
could provide accuracy comparable with that obtained by testing all ten
sites.
Methods: A cross-sectional comparative study of plantar sensory
levels in diabetic subjects with and without plantar ulceration was conducted
in a tertiary-care teaching hospital setting. We examined forty-seven diabetic
subjects with a history of foot ulceration and forty-five diabetic subjects
with no history of foot ulceration. Plantar sensory threshold values at five
sites on the sole of each foot were measured with a quasi-continuous range of
applied forces, and receiver operating characteristic analysis techniques were
applied.
Results: Screening on the basis of only the maximum force threshold
for the left and right first metatarsal head sites provided comparable or
better performance at high levels of sensitivity than did either the mean or
the maximum force threshold across all ten sites. A sensory threshold of 4.5 g
for both the left and right first metatarsal head sites predicted the risk of
ulceration with a sensitivity of 100% and a specificity of 67%.
Conclusions: Testing of diabetic patients for protective sensation
may be simplified to testing under both first metatarsal heads with a 4.5-g
monofilament. If a patient cannot sense the application of a 4.5-g
monofilament under either first metatarsal head, he or she probably has lost
protective sensation and should be considered to be at risk for undetected
injury.
Level of Evidence: Diagnostic study, Level II-1
(development of diagnostic criteria on the basis of consecutive patients [with
universally applied reference "gold" standard]). See Instructions
to Authors for a complete description of levels of evidence.