Methodological aspects of toe blood pressure measurements for evaluation of arterial insuffiency in patients with diabetes

Abstract

Diabetes mellitus (DM) is a growing health problem in the aging population of the western world with prevalence figures around 6-7%, which are increasing fast. Treatment of this disease is complex and expensive for society. One of the most feared DM complications is foot ulcerations. As a causative factor for this complication peripheral arterial disease (PAD) is second only to neuropathy in importance, and screening for PAD is vital to prevent amputation. Toe blood pressure (TBP) measurements are often used for this purpose and this method is also included in the definition of critical leg ischemia. Unfortunately, TBP measurements are affected by a high variability and obtained pressures can be of limited value because of this. Accordingly, the aim of this thesis was to evaluate, standardise and optimise TBP measurements.

Study I: Photopletysmography (PPG) has been the standard for blood flow detection during TBP assessment. Laser Doppler (LD) has some theoretical advantage over PPG due to its higher sensitivity. The aim of this study was to evaluate LD for TBP measurements. It included 36 patients referred for PAD screening, who were examined both with PPG and LD. The correlation between pressures obtained with the two methods was high (r=0.99). Significantly (p< .005) higher values were obtained with LD. In a second part of the study TBP measured with LD were able to categorize 40 patients and control subjects according to the Fontaine classification. In 10 of the patients who had DM, ankle and TBP values were compared, and two of them had falsely elevated ankle blood pressures (ABP).

Study II: This study aimed to evaluate the importance of vessel wall compliance on blood pressure measurements with cuff at the ankle and toe level using the hydrostatic pressure (pole test) as gold standard. Twenty-five legs in 23 patients with DM were examined. Due to persistent signal at maximal leg elevation, it was impossible obtain pole test pressures in all patients, and only 11 at ankle and 19 at toe level could be assessed. The ABP values were higher if measured with cuff than with pole test (p< .01), but no difference was found in the toe. The conclusion was that falsely elevated blood pressure in patients with DM probably is of less importance for TBP than ABP.

Study III: The focus of this study was to examine if the cuff size influences TBP values. Eleven patients with DM without a history of PAD were investigated and compared to six healthy control subjects. In the entire cohort of subjects (patients and controls) TBP values were significantly higher (p< .01) if measured with a narrow as compared to a wider cuff. There was no difference between patients and controls, but this could be due to a limited cohort size. Differences in hallux sizes were one of the determinants of the TBP variability. The conclusion was that the cuff width does influence the obtained TBP values and needs to be considered when using TBP to screen for PAD.

Study IV: The objective of the study was to determine the optimal cuff width for measuring TBP. Twenty PAD patients with and without DM were recruited at two centers. Four cuff sizes, 1.5, 2.0, 2.5 and 3.0 cm were used to measure TBP in the cohort and the values compared with pole test recordings, which were considered to represent the true TBP value. Values were overestimated when 1.5 and 2.0 cm (p<.001) wide cuffs were used, while 2.5 and 3.0 cm wide cuff use produced values more in line with the real TBP. The results from this study suggested that a 2.5 cm wide cuff should be used for all TBP measurements in clinical praxis.

Study V: In this study the general variability of the TBP method was investigated and an automated device’s ability to measure TBP was evaluated. TBP was recorded in 23 legs of 16 diabetic patients and pressures assessed manually from perfusion-pressure graphs and by an automated algorithm. The inter- and intra-observer variability of TBP was less than 10% of the recorded pressure. Automatic assessments of TBP was acceptably accurate for measurement of TBP <45 mm Hg (SD 4.0 mm Hg), but not for higher pressures. Adjustment of the automatic algorithm to manage the problem of biphasic wave patterns reduced the difference compared to visual readings from 5.9 to -2.9 mm Hg (SD from 16.9 to 8.9).

Conclusion Stiff arteries do not affect TBP measurements as much as ABP in patients with DM, but examinations have to be standardised to be a reliable. LD can be used as blood flow detector when measuring TBP and is probably the best alternative when TBP is low. To reflect the hydrostatic pressure the toe cuff width should be 2.5-3 cm or 30% of the toe circumference. Standardized TBP measurements for skin temperature, flow detector and cuff size produce pressures with variability below 10%, which probably is less than for ABP. An automated TBP device is reliable for identification of diabetic patients with foot ulcerations with a risk of having critical limb ischemia.