- Email: [email protected]
A New Automated Toe Blood Pressure Monitor for Assessment of Limb Ischemia
Eur J Vasc Endovasc Surg 24, 304±308 (2002) doi:10.1053/ejvs.2002.1664, available online at http://www.idealibrary.com on
A New Automated Toe Blood Pressure Monitor for Assessment of Limb Ischemia E. Wahlberg1 and R. Gush2 1
Department of Vascular Surgery, Karolinska Hospital, Stockholm, Sweden and 2 Moor Instruments Ltd, Axminster, Devon, U.K.
Objective: toe blood pressure (TBP) is an important method to assess peripheral arterial disease especially in patients with diabetes, but remains difficult to measure. We have developed a simple portable device for TBP measurements. Methods and results: first, TBP was determined in 40 ischemic legs with both laser Doppler and photoplethysmography for perfusion monitoring, to assess if laser Doppler can be used for measurements. The median values recorded were identical, but slightly higher values were obtained with laser Doppler (p 0.03). Secondly, a computer based algorithm for automatic TBP readings with laser Doppler was compared to manual assessment in 28 legs of 20 patients. The median values differed 3 mmHg (p 0.10). Finally the applicability of the new device was tested in eight legs of six patients by two nurses. Conclusion: laser Doppler is appropriate for perfusion monitoring during TBP measurements and automatic pressure readings seem accurate. The automatic portable device is simple to use and can probably determine TBP.
Introduction Chronic critical limb ischemia (CLI) in patients with and without diabetes is a major cause of morbidity with an incidence of 50±100 per 100 000 inhabitants in Western countries.1 In United States alone more than 100 000 amputations are performed each year, a majority of those are performed in patients with CLI and diabetes.2 Diagnosis and accurate assessment of severity of disease is extremely important to identify patients who need vascular intervention,3 and to select the optimal treatment strategy.4 Distal blood pressure assessment is an important part of this management process, particularly in patients with diabetes because neuropathy masks pain as a symptom of severe ischemia. The pressure at distal locations, usually obtained in vascular laboratories, is directly correlated to the severity of disease. In patients with diabetes the measurement of toe blood pressure (TBP)7 overcomes the problems associated with erroneously high ankle pressures.8,9 The recent TransAtlantic Inter-Society Consensus (TASC) document emphasizes the use of TBP when screening and evaluating patients with diabetes.10
Please address all correspondence to: E. Wahlberg, Department of Vascular Surgery, Karolinska Hospital, N2:00, SE-171 76 Stockholm, Sweden.
The purpose of the present study was to describe and evaluate a new, simple portable device to measure TBP. Patients and Methods This prospective study is composed by three parts: (a) Evaluation of laser Doppler (LD) as flow detection method when measuring TBP; (b) evaluation of a computerized algorithm for automatic determination of flow return during cuff deflation, and (c) assessment of the applicability of a new automated instrument (TPM ± toe pressure monitor) when measuring TBP in patients. Consecutive patients referred to the vascular surgical clinic for evaluation of peripheral arterial disease (PAD) were recruited for parts A and B. The patients included were referred by general practitioners to be screened in the vascular laboratory for PAD. For part C patients in the ward were asked to participate (Table 1). Methods Photoplethysmography (PPG) This device (Electromedicin AB, Sweden) analyzes the quantity of reflected light from the skin circulation
1078±5884/02/040304 05 $35.00/0 # 2002 Elsevier Science Ltd. All rights reserved.
Automated Toe Blood Pressure
305
Table 1. Characteristics for the patients investigated in the three parts (A±C) of the study. A ± comparing Laser Doppler and photoplethysmography as perfusion method during manual toe blood pressure measurements (TBP), B ± comparing values during manual assessment of TBP to an algorithm for automatic TBP determination, and C ± testing the new device on patients. Study part n (patients)/n (legs) Age Sex (male/female)
years N/N
Diabetes (type I/II) (ins/oral/diet) No symptoms or signs Claudication/rest pain Ulcers or gangrene Brachial blood pressure Ankle blood pressure Ankle brachial index
N (%) N/N N/N/N N N/N N mmHg mmHg
A
B
C
40/40 70 (43±85) 26/14
20/28 75 (54±89) 12/8
6/8 77 (58±80) 3/3
15 (38) 1/14 8/5/2 6 16/7 11 156 (120±210) 92 (0±300) 0.59 (0±1.87)
10 (50) 0/10 5/5/0 ± 8/8 12 148 (115±180) 72 (0±180) 0.40 (0±1.2)
4 (67) 0/4 4/0/0 ± 5/0 3 157 (120±180) 82 (40±170) 0.51 (0.26±1.07)
Median (range).
during illumination with an infrared light-emitting diode. The return of pulsatility during cuff deflation is used to determine TBP. LD A diode delivers laser light to the skin. Photons that encounter moving blood cells become Doppler shifted. The Doppler shifted and unshifted light is reflected back to the detector and the beat frequency spectrum is processed to provide output proportional to blood flow. In skin, the laser penetration depth is considered to be 0.5±1 mm but varies greatly depending on the emitted wavelength and the local skin architecture. For part A of the study a commercially available system (Perimed 1d, Perimed AB, Jarfalla, Sweden) equipped with a standard probe was used. In the TPM that was used in parts B and C of the study a specially designed LD instrument was used (Moor Instruments Ltd, Axminster, U.K.) TPM The TPM incorporates a LD system to monitor perfusion, a transducer to monitor cuff pressure, a pump and air bleed system, a computerized algorithm for determination of LD flux during cuff deflation and reperfusion and a processor to control and monitor the system (Fig. 1). Procedures (A) LD/PPG comparison A 2.0 cm cuff was placed at the base of the hallux and flow detector probes were attached to the pulp. TBP was measured in sequence by recording the LD or PPG tracings in random order, simultaneous with recordings of cuff pressure on a chart. The cuff was inflated by hand as rapidly as possible and
Fig. 1. The combined device.
deflated slowly at 5 mmHg per 2±3 s. TBP was obtained manually from the charts by determining the pressure in the cuff when the perfusion or pulsatile flow reappear. TBP was measured twice, 3 min apart, and the average value for both occasions used for analyses. (B) LD TPM: comparison of manual and automatic assessments TBP was measured using TPM, the device being connected to a PC continuously recording LD flux and cuff pressure. TBP values were obtained from the recorded data manually from LD/pressure traces as previously described. These values (TBPman) were compared to the TBP values obtained from the algorithm (TBPauto). (C) Application of LD TPM The six patients rested supine in their beds for 15 min prior to, and between measurements, with the feet Eur J Vasc Endovasc Surg Vol 24, October 2002
E. Wahlberg and R. Gush
306
kept under a blanket. Each patient was measured ten times during 2 h by two nurses, alternating. Before each measurement of TBP, skin temperature at the probe site was recorded using a thermistor (Exacon1, Copenhagen, Denmark). The TBP values were read on the display by the nurses performing the procedure and filed on to a data sheet together with skin temperatures. In two patients both legs were measured, and the procedure was performed on two different days. Statistics Median, percentiles and range were used to describe the TBP data and patient characteristics. To compare TBP measured two different ways we used the method suggested by Bland and Altman.11,12 This analysis compares the difference between two methods that measure the same parameter to the average value. Wilcoxon signed rank test was used to compare the values in parts A and B. In part C, we used the coefficient of variation (CV), calculated as the interquartile range divided with median, to describe reliability of the data. The association between skin temperature and TBP was analyzed using Pearson's regression test. p less than 5% was used to denote statistical significance. Results (A) LD/PPG comparison The median value of TBP measured with LD was 62 mmHg (range 10±170) and with PPG it was also 62 mmHg (range 0±180). These values were significantly different (p 0.028) between the methods. The difference observed in each individual patient with the methods is shown in Figure 2(a).
Fig. 2. Differences in toe blood pressures, (a) subtracting PPG from LD values (n 40), and (b) subtracting manual from automatic readings of pressure (n 28). Each difference is plotted to the median toe blood pressure (of the two measurements with LD and PPG (a) and manual and automatic (b)).
2 min to obtain TBP. The device was put on a small trolley and placed close to the patient to be measured. The variability in TBP values obtained are presented in Table 2. The median CV for all eight legs was 22% (18±25), and similar for the two nurses (median 16 and 18%, respectively). TBP values were associated to skin temperature (r 0.58, p 5 0.001).
(B) Automatic/manual comparison When examining the accuracy of the algorithm, values were similar (p 0.104) when measured by TBPauto, 48 mmHg (range 12±164), compared to TBPman, 51 mmHg (range 15±161). The difference between the methods in each individual patient is shown in Figure 2(b). For both comparisons the values for patients with and without diabetes were similar. (C) Application in patients The TPM was considered easy to use by the two nurses utilizing the system. It took them on average Eur J Vasc Endovasc Surg Vol 24, October 2002
Discussion This study confirms that it is possible to develop a device (TPM) for simple automated TBP measurements. The TPM is not dependent on pulsatile flow and no training to interpret the perfusion signal is needed. Accordingly, our nurses testing the system had no problems making the measurements and considered it easy to use. This is in line with previous attempts.13 Our observation of slightly higher values using LD compared to PPG is consistent with earlier findings.14±16 While PPG is based on opacity changes, LD assesses
Automated Toe Blood Pressure
307
Table 2. Toe blood pressures recorded by two nurses in eight legs of six patients using the new device. In patients 4 and 5 both legs were measured. Each leg is measured five times by each nurse during two hours. Values given are median, range and coefficient of variation (CV). Patient #
1 2 3 4 5 6
Ankle pressure mmHg
140 40 80 170 160 55 65 85
Skin temperature Median (range) C
34.6 (2.9) 31.6 (1.5) 27.3 (3.2) 24.6 (2.4) 27.5 (4.2) 26.7 (4.6) 30.9 (3.1) 28.1 (2.1)
TBP Total (n 10)
Nurse A (n 5)
Nurse B (n 5)
Median (range) mmHg
CV %
Median (range) mmHg
CV %
Median (range) mmHg
CV %
104 (63) 20 (15) 39 (27) 27 (22) 46 (32) 37 (21) 45 (26) 61 (37)
23 23 25 24 24 18 19 20
101 (55) 17 (15) 29 (14) 29 (12) 55 (15) 36 (9) 40 (13) 71 (19)
25 29 18 17 11 13 13 12
129 (41) 24 (7) 45 (22) 24 (22) 38 (18) 39 (17) 52 (16) 50 (20)
18 13 19 31 21 18 12 15
movement of blood. This difference may be of importance in the low pressure range when pulsatile flow is absent. Under such circumstances, with PPG in the AC mode, the TBP is recorded as zero. Accordingly, in this study higher values with LD were mainly observed among pressures lower than 20 mmHg. Similar results have been reported by de Graaff. This LD advantage and the excellent correlation in the higher pressure range convinced us to use LD in the TPM. The most important feature of the new TPM device is the automated algorithm. There was only a small difference between manual and automatic readings.16 However, the TPM needs to be evaluated further before it can be used to accurately assess small changes in TBP in severely ischemic legs ± as needed in treatment trials. In the pressure range of 30±60 mmHg, where TBP can be of value for selecting patients with diabetes for treatment, the algorithm has performed well. TBP varies greatly, especially in patients with mild disease and almost normal pressures.6 A number of factors probably contribute to this: skin and ambient temperature;17 cuff width;18 the patient's level of stress and examiner skill.19 The overall variability of just above 20% of the mean TBP values obtained in our patients is close to what has been previously reported.17 Accordingly, when the circumstances during measurements are stable and standardized, the variability associated with TBP is seldom linked to the TBP method chosen. The suggestion of Bowers is probably valid: to screen for severe ischemia by observation of TBP values below 40 mmHg recorded on two separate occasions.7 This concept will be tested by using TPM in a cohort of patients with diabetes having foot ulcers referred to our clinic. The cost of the TPM is not established at this time since it is still under development, but it is likely that it will be in the range of 1500±2000 Euro.
In conclusion, a new portable device based on LD for perfusion monitoring, featuring an algorithm for automatic TBP measurements is introduced. This device can be used to quickly measure TBP in patients in clinics and wards. Its accuracy need to be evaluated but it may provide a future screening tool able to identify diabetic patients with foot ulcers in need of invasive treatment. Acknowledgements The authors gratefully acknowledge Kerstin Ahlgren and Nanette Alberti for their assistance in data acquisition.
References 1 The Vascular Surgical Society of Great Britain and Ireland: critical limb ischaemia: management and outcome. Eur J Vasc Endovasc Surg 1995; 10: 108±113. 2 The global lower extremity amputation study group: Epidemiology of lower extremity amputation in centres in Europe, North America and East Asia. Br J Surg 1999; 87: 328±337. 3 Ramsey SD, Newton K, Blough D et al. Incidence, outcomes, and cost of foot ulcers in patients with diabetes. Diabetes Care 1999; 22: 382±387. 4 Pell JP, Fowkes FGR, Lee AJ. Indications for arterial reconstruction and major amputation in the management of chronic critical lower limb ischaemia. Eur J Vasc Endovasc Surg 1997; 13: 315±321. 5 Ramsey DE, Manke DA, Sumner DS. Toe blood pressure: a valuable adjunct to ankle pressure measurement for assessing peripheral arterial disease. J Cardiovasc Surg 1983; 24: 43±48. 6 Carter SA. Ankle and toe systolic pressures comparison of value and limitations in arterial occlusive disease. Int Ang 1992; 11: 289±297. 7 Bowers BL, Valentine J, Myers SI, Chervu A, Clagett P. The natural history of patients with claudication with toe pressures of 40 mmHg or less. J Vasc Surg 1993; 18: 506±511. 8 Emanuele MA, Buchanan BJ, Abraria C. Elevated leg systolic pressures and arterial calcification in Diabetic occlusive disease. Diabetes Care 1981; 4: 289±292. 9 Goss DE, DeTrafford JC, Roberts VC, Flynn MD, Edmonds ME, Watkins PJ. Raised ankle/brachial pressure index in insulin treated diabetic patients. Diabetic Medicine 1989; 6: 576±578. Eur J Vasc Endovasc Surg Vol 24, October 2002
308
E. Wahlberg and R. Gush
10 TASC working group: Management of peripheral arterial disease (PAD). Eur J Vasc Endovasc Surg 2000; 19(Supplement A). 11 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurements. Lancet 1986; i: 307±310. 12 Bland JM, Altman DG. Comparing methods of measurement: why plotting difference against standard method is misleading. Lancet 1995; 346: 1085±1087. 13 Fronek A. Blazek V, Curran. Toe pressure determination by audioplethysmography. J Vasc Surg 1994; 20: 267±270. 14 Andersson S, Linderholm H, RinnstroÈm O, Burlin L. A laser Doppler technique for measuring distal blood pressure: a comparison with conventional strain gauge technique. Clinical Physiology 1986; 6: 329±335. 15 Beinder E, Hoffman U, Franzeck UK, Huch A, Huch R, Bollinger A. Laser Doppler technique for the measurement of digital and segmental systolic blood pressure. VASA 1992; 21: 15±21.
Eur J Vasc Endovasc Surg Vol 24, October 2002
16 de Graaff JC, Ubbink D Th, Legemete DA, de Haan RJ, Jacobs MJHM. The usefulness of laser Doppler in the measurement of toe blood pressure. J Vasc Surg 2000; 32: 1172±1179. 17 Carter SA, Tate RB. The effect of body heating and cooling on ankle and toe systolic pressures in arterial disease. J Vasc Surg 1992; 16: 148±153. 18 Gundersen J. Segmental measurements of systolic blood pressure in the extremities including the thumb and the great toe. Acta Chir Scand 1972; Supplement. 19 Fowkes FGR, Housley E, McIntyre CCA, Prescott RJ, Ruckley RJ. Variability of ankle and brachial systolic pressures in the measurement of atherosclerotic peripheral arterial disease. J Epidem Com Health 1988; 42: 128±133.
Accepted 27 March 2002
A New Automated Toe Blood Pressure Monitor for Assessment of Limb Ischemia E. Wahlberg1 and R. Gush2 1
Department of Vascular Surgery, Karolinska Hospital, Stockholm, Sweden and 2 Moor Instruments Ltd, Axminster, Devon, U.K.
Objective: toe blood pressure (TBP) is an important method to assess peripheral arterial disease especially in patients with diabetes, but remains difficult to measure. We have developed a simple portable device for TBP measurements. Methods and results: first, TBP was determined in 40 ischemic legs with both laser Doppler and photoplethysmography for perfusion monitoring, to assess if laser Doppler can be used for measurements. The median values recorded were identical, but slightly higher values were obtained with laser Doppler (p 0.03). Secondly, a computer based algorithm for automatic TBP readings with laser Doppler was compared to manual assessment in 28 legs of 20 patients. The median values differed 3 mmHg (p 0.10). Finally the applicability of the new device was tested in eight legs of six patients by two nurses. Conclusion: laser Doppler is appropriate for perfusion monitoring during TBP measurements and automatic pressure readings seem accurate. The automatic portable device is simple to use and can probably determine TBP.
Introduction Chronic critical limb ischemia (CLI) in patients with and without diabetes is a major cause of morbidity with an incidence of 50±100 per 100 000 inhabitants in Western countries.1 In United States alone more than 100 000 amputations are performed each year, a majority of those are performed in patients with CLI and diabetes.2 Diagnosis and accurate assessment of severity of disease is extremely important to identify patients who need vascular intervention,3 and to select the optimal treatment strategy.4 Distal blood pressure assessment is an important part of this management process, particularly in patients with diabetes because neuropathy masks pain as a symptom of severe ischemia. The pressure at distal locations, usually obtained in vascular laboratories, is directly correlated to the severity of disease. In patients with diabetes the measurement of toe blood pressure (TBP)7 overcomes the problems associated with erroneously high ankle pressures.8,9 The recent TransAtlantic Inter-Society Consensus (TASC) document emphasizes the use of TBP when screening and evaluating patients with diabetes.10
Please address all correspondence to: E. Wahlberg, Department of Vascular Surgery, Karolinska Hospital, N2:00, SE-171 76 Stockholm, Sweden.
The purpose of the present study was to describe and evaluate a new, simple portable device to measure TBP. Patients and Methods This prospective study is composed by three parts: (a) Evaluation of laser Doppler (LD) as flow detection method when measuring TBP; (b) evaluation of a computerized algorithm for automatic determination of flow return during cuff deflation, and (c) assessment of the applicability of a new automated instrument (TPM ± toe pressure monitor) when measuring TBP in patients. Consecutive patients referred to the vascular surgical clinic for evaluation of peripheral arterial disease (PAD) were recruited for parts A and B. The patients included were referred by general practitioners to be screened in the vascular laboratory for PAD. For part C patients in the ward were asked to participate (Table 1). Methods Photoplethysmography (PPG) This device (Electromedicin AB, Sweden) analyzes the quantity of reflected light from the skin circulation
1078±5884/02/040304 05 $35.00/0 # 2002 Elsevier Science Ltd. All rights reserved.
Automated Toe Blood Pressure
305
Table 1. Characteristics for the patients investigated in the three parts (A±C) of the study. A ± comparing Laser Doppler and photoplethysmography as perfusion method during manual toe blood pressure measurements (TBP), B ± comparing values during manual assessment of TBP to an algorithm for automatic TBP determination, and C ± testing the new device on patients. Study part n (patients)/n (legs) Age Sex (male/female)
years N/N
Diabetes (type I/II) (ins/oral/diet) No symptoms or signs Claudication/rest pain Ulcers or gangrene Brachial blood pressure Ankle blood pressure Ankle brachial index
N (%) N/N N/N/N N N/N N mmHg mmHg
A
B
C
40/40 70 (43±85) 26/14
20/28 75 (54±89) 12/8
6/8 77 (58±80) 3/3
15 (38) 1/14 8/5/2 6 16/7 11 156 (120±210) 92 (0±300) 0.59 (0±1.87)
10 (50) 0/10 5/5/0 ± 8/8 12 148 (115±180) 72 (0±180) 0.40 (0±1.2)
4 (67) 0/4 4/0/0 ± 5/0 3 157 (120±180) 82 (40±170) 0.51 (0.26±1.07)
Median (range).
during illumination with an infrared light-emitting diode. The return of pulsatility during cuff deflation is used to determine TBP. LD A diode delivers laser light to the skin. Photons that encounter moving blood cells become Doppler shifted. The Doppler shifted and unshifted light is reflected back to the detector and the beat frequency spectrum is processed to provide output proportional to blood flow. In skin, the laser penetration depth is considered to be 0.5±1 mm but varies greatly depending on the emitted wavelength and the local skin architecture. For part A of the study a commercially available system (Perimed 1d, Perimed AB, Jarfalla, Sweden) equipped with a standard probe was used. In the TPM that was used in parts B and C of the study a specially designed LD instrument was used (Moor Instruments Ltd, Axminster, U.K.) TPM The TPM incorporates a LD system to monitor perfusion, a transducer to monitor cuff pressure, a pump and air bleed system, a computerized algorithm for determination of LD flux during cuff deflation and reperfusion and a processor to control and monitor the system (Fig. 1). Procedures (A) LD/PPG comparison A 2.0 cm cuff was placed at the base of the hallux and flow detector probes were attached to the pulp. TBP was measured in sequence by recording the LD or PPG tracings in random order, simultaneous with recordings of cuff pressure on a chart. The cuff was inflated by hand as rapidly as possible and
Fig. 1. The combined device.
deflated slowly at 5 mmHg per 2±3 s. TBP was obtained manually from the charts by determining the pressure in the cuff when the perfusion or pulsatile flow reappear. TBP was measured twice, 3 min apart, and the average value for both occasions used for analyses. (B) LD TPM: comparison of manual and automatic assessments TBP was measured using TPM, the device being connected to a PC continuously recording LD flux and cuff pressure. TBP values were obtained from the recorded data manually from LD/pressure traces as previously described. These values (TBPman) were compared to the TBP values obtained from the algorithm (TBPauto). (C) Application of LD TPM The six patients rested supine in their beds for 15 min prior to, and between measurements, with the feet Eur J Vasc Endovasc Surg Vol 24, October 2002
E. Wahlberg and R. Gush
306
kept under a blanket. Each patient was measured ten times during 2 h by two nurses, alternating. Before each measurement of TBP, skin temperature at the probe site was recorded using a thermistor (Exacon1, Copenhagen, Denmark). The TBP values were read on the display by the nurses performing the procedure and filed on to a data sheet together with skin temperatures. In two patients both legs were measured, and the procedure was performed on two different days. Statistics Median, percentiles and range were used to describe the TBP data and patient characteristics. To compare TBP measured two different ways we used the method suggested by Bland and Altman.11,12 This analysis compares the difference between two methods that measure the same parameter to the average value. Wilcoxon signed rank test was used to compare the values in parts A and B. In part C, we used the coefficient of variation (CV), calculated as the interquartile range divided with median, to describe reliability of the data. The association between skin temperature and TBP was analyzed using Pearson's regression test. p less than 5% was used to denote statistical significance. Results (A) LD/PPG comparison The median value of TBP measured with LD was 62 mmHg (range 10±170) and with PPG it was also 62 mmHg (range 0±180). These values were significantly different (p 0.028) between the methods. The difference observed in each individual patient with the methods is shown in Figure 2(a).
Fig. 2. Differences in toe blood pressures, (a) subtracting PPG from LD values (n 40), and (b) subtracting manual from automatic readings of pressure (n 28). Each difference is plotted to the median toe blood pressure (of the two measurements with LD and PPG (a) and manual and automatic (b)).
2 min to obtain TBP. The device was put on a small trolley and placed close to the patient to be measured. The variability in TBP values obtained are presented in Table 2. The median CV for all eight legs was 22% (18±25), and similar for the two nurses (median 16 and 18%, respectively). TBP values were associated to skin temperature (r 0.58, p 5 0.001).
(B) Automatic/manual comparison When examining the accuracy of the algorithm, values were similar (p 0.104) when measured by TBPauto, 48 mmHg (range 12±164), compared to TBPman, 51 mmHg (range 15±161). The difference between the methods in each individual patient is shown in Figure 2(b). For both comparisons the values for patients with and without diabetes were similar. (C) Application in patients The TPM was considered easy to use by the two nurses utilizing the system. It took them on average Eur J Vasc Endovasc Surg Vol 24, October 2002
Discussion This study confirms that it is possible to develop a device (TPM) for simple automated TBP measurements. The TPM is not dependent on pulsatile flow and no training to interpret the perfusion signal is needed. Accordingly, our nurses testing the system had no problems making the measurements and considered it easy to use. This is in line with previous attempts.13 Our observation of slightly higher values using LD compared to PPG is consistent with earlier findings.14±16 While PPG is based on opacity changes, LD assesses
Automated Toe Blood Pressure
307
Table 2. Toe blood pressures recorded by two nurses in eight legs of six patients using the new device. In patients 4 and 5 both legs were measured. Each leg is measured five times by each nurse during two hours. Values given are median, range and coefficient of variation (CV). Patient #
1 2 3 4 5 6
Ankle pressure mmHg
140 40 80 170 160 55 65 85
Skin temperature Median (range) C
34.6 (2.9) 31.6 (1.5) 27.3 (3.2) 24.6 (2.4) 27.5 (4.2) 26.7 (4.6) 30.9 (3.1) 28.1 (2.1)
TBP Total (n 10)
Nurse A (n 5)
Nurse B (n 5)
Median (range) mmHg
CV %
Median (range) mmHg
CV %
Median (range) mmHg
CV %
104 (63) 20 (15) 39 (27) 27 (22) 46 (32) 37 (21) 45 (26) 61 (37)
23 23 25 24 24 18 19 20
101 (55) 17 (15) 29 (14) 29 (12) 55 (15) 36 (9) 40 (13) 71 (19)
25 29 18 17 11 13 13 12
129 (41) 24 (7) 45 (22) 24 (22) 38 (18) 39 (17) 52 (16) 50 (20)
18 13 19 31 21 18 12 15
movement of blood. This difference may be of importance in the low pressure range when pulsatile flow is absent. Under such circumstances, with PPG in the AC mode, the TBP is recorded as zero. Accordingly, in this study higher values with LD were mainly observed among pressures lower than 20 mmHg. Similar results have been reported by de Graaff. This LD advantage and the excellent correlation in the higher pressure range convinced us to use LD in the TPM. The most important feature of the new TPM device is the automated algorithm. There was only a small difference between manual and automatic readings.16 However, the TPM needs to be evaluated further before it can be used to accurately assess small changes in TBP in severely ischemic legs ± as needed in treatment trials. In the pressure range of 30±60 mmHg, where TBP can be of value for selecting patients with diabetes for treatment, the algorithm has performed well. TBP varies greatly, especially in patients with mild disease and almost normal pressures.6 A number of factors probably contribute to this: skin and ambient temperature;17 cuff width;18 the patient's level of stress and examiner skill.19 The overall variability of just above 20% of the mean TBP values obtained in our patients is close to what has been previously reported.17 Accordingly, when the circumstances during measurements are stable and standardized, the variability associated with TBP is seldom linked to the TBP method chosen. The suggestion of Bowers is probably valid: to screen for severe ischemia by observation of TBP values below 40 mmHg recorded on two separate occasions.7 This concept will be tested by using TPM in a cohort of patients with diabetes having foot ulcers referred to our clinic. The cost of the TPM is not established at this time since it is still under development, but it is likely that it will be in the range of 1500±2000 Euro.
In conclusion, a new portable device based on LD for perfusion monitoring, featuring an algorithm for automatic TBP measurements is introduced. This device can be used to quickly measure TBP in patients in clinics and wards. Its accuracy need to be evaluated but it may provide a future screening tool able to identify diabetic patients with foot ulcers in need of invasive treatment. Acknowledgements The authors gratefully acknowledge Kerstin Ahlgren and Nanette Alberti for their assistance in data acquisition.
References 1 The Vascular Surgical Society of Great Britain and Ireland: critical limb ischaemia: management and outcome. Eur J Vasc Endovasc Surg 1995; 10: 108±113. 2 The global lower extremity amputation study group: Epidemiology of lower extremity amputation in centres in Europe, North America and East Asia. Br J Surg 1999; 87: 328±337. 3 Ramsey SD, Newton K, Blough D et al. Incidence, outcomes, and cost of foot ulcers in patients with diabetes. Diabetes Care 1999; 22: 382±387. 4 Pell JP, Fowkes FGR, Lee AJ. Indications for arterial reconstruction and major amputation in the management of chronic critical lower limb ischaemia. Eur J Vasc Endovasc Surg 1997; 13: 315±321. 5 Ramsey DE, Manke DA, Sumner DS. Toe blood pressure: a valuable adjunct to ankle pressure measurement for assessing peripheral arterial disease. J Cardiovasc Surg 1983; 24: 43±48. 6 Carter SA. Ankle and toe systolic pressures comparison of value and limitations in arterial occlusive disease. Int Ang 1992; 11: 289±297. 7 Bowers BL, Valentine J, Myers SI, Chervu A, Clagett P. The natural history of patients with claudication with toe pressures of 40 mmHg or less. J Vasc Surg 1993; 18: 506±511. 8 Emanuele MA, Buchanan BJ, Abraria C. Elevated leg systolic pressures and arterial calcification in Diabetic occlusive disease. Diabetes Care 1981; 4: 289±292. 9 Goss DE, DeTrafford JC, Roberts VC, Flynn MD, Edmonds ME, Watkins PJ. Raised ankle/brachial pressure index in insulin treated diabetic patients. Diabetic Medicine 1989; 6: 576±578. Eur J Vasc Endovasc Surg Vol 24, October 2002
308
E. Wahlberg and R. Gush
10 TASC working group: Management of peripheral arterial disease (PAD). Eur J Vasc Endovasc Surg 2000; 19(Supplement A). 11 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurements. Lancet 1986; i: 307±310. 12 Bland JM, Altman DG. Comparing methods of measurement: why plotting difference against standard method is misleading. Lancet 1995; 346: 1085±1087. 13 Fronek A. Blazek V, Curran. Toe pressure determination by audioplethysmography. J Vasc Surg 1994; 20: 267±270. 14 Andersson S, Linderholm H, RinnstroÈm O, Burlin L. A laser Doppler technique for measuring distal blood pressure: a comparison with conventional strain gauge technique. Clinical Physiology 1986; 6: 329±335. 15 Beinder E, Hoffman U, Franzeck UK, Huch A, Huch R, Bollinger A. Laser Doppler technique for the measurement of digital and segmental systolic blood pressure. VASA 1992; 21: 15±21.
Eur J Vasc Endovasc Surg Vol 24, October 2002
16 de Graaff JC, Ubbink D Th, Legemete DA, de Haan RJ, Jacobs MJHM. The usefulness of laser Doppler in the measurement of toe blood pressure. J Vasc Surg 2000; 32: 1172±1179. 17 Carter SA, Tate RB. The effect of body heating and cooling on ankle and toe systolic pressures in arterial disease. J Vasc Surg 1992; 16: 148±153. 18 Gundersen J. Segmental measurements of systolic blood pressure in the extremities including the thumb and the great toe. Acta Chir Scand 1972; Supplement. 19 Fowkes FGR, Housley E, McIntyre CCA, Prescott RJ, Ruckley RJ. Variability of ankle and brachial systolic pressures in the measurement of atherosclerotic peripheral arterial disease. J Epidem Com Health 1988; 42: 128±133.
Accepted 27 March 2002