The value of laser Doppler velocimetry and transcutaneous oxygen tension determination in predicting healing of ischemic forefoot ulcerations and amputations in diabetic and nondiabetic patients

The value of laser Doppler velocimetry and transcutaneous oxygen tension determination in predicting healing of ischemic forefoot ulcerations and amputations in diabetic and nondiabetic patients R i c h a r d G. Karanfilian, M.D., T h o m a s G. Lynch, M.D., Victor T. Zirul, B.S., F r a n k T. Padberg, M.D., Zafar Jamil, M.D., and R o b e r t W. H o b s o n II, M.D.,

Newark and East Orange, N.J. The ability to predict successful healing of ulcerations and amputations of the ischemic forefoot continues to be a major clinical challenge, particularly in diabetic patients whose systolic Doppler ankle pressures are often artifactually elevated. We have used the techniques of laser Doppler velocimetry (LD) and transcutaneous oxygen tension monitoring (tcPo2) to quantitatively measure skin blood flow in the distal foot. Fifty-nine limbs were studied (48 patients), of which 37 (63%) were in diabetic and 22 (37%) in nondiabetic patients. All patients were admitted with ischemic ulcerations or gangrenous changes of the forefoot or digit. Twenty transmetatarsal or digital amputations were performed; the remainder o f the lesions were dtbrided and allowed to heal by secondary intention or were covered by a skin graft. Before operation, the systolic pressure (expressed in millimeters o f mercury, mean -+ SEM) was measured by Doppler technique at the ankle, and the ankle/arm index calculated (n = 59 limbs). The tcPo2 (also expressed in millimeters of mercury, mean -+ SEM) was measured from the dorsal foot (n = 56). The baseline skin blood flow velocity (SBFV) and pulse wave amplitude (PWA) were measured with the LD (expressed in miUivolts, mean -+ SEM) on the plantar aspect of the foot (n -- 53 limbs). Criteria for successful healing included a tcPo2 of more than 10 mm Hg, the combination of an LD-SBFV of more than 40 mV and an LD-PWA of more than 4 mV, and an ankle systolic pressure of more than 30 mm Hg. With these criteria, the outcome was predicted correctly in 53 of 56 limbs (95%) by tcPo2, in 46 of 53 limbs (87%) with LD, and in 31 of 59 limbs (52%) with Doppler ankle systolic pressures. We conclude that the estimation of skin blood flow by tcPo2 and LD is significantly (p ~ 0.05) better than Doppler ankle pressure measurements in predicting the healing of forefoot ulcerations and amputations in diabetic and nondiabetic patients. (J VASC SURG 1986; 4:511-6.)

The management o f ischemic ulcerations involving the digits and forefoot continues to be a major clinical challenge. It is often difficult to predict whether this lesion, or a subsequent amputation o f the forefoot, will heal successfully, particularly in the

From the Section of Vascular Surgery,,Department of Surgery, University of Medicine and Dentistry of New Jersey,New Jersey Medical School, Newark, and Veterans Administration Medical Center, East Orange, N.J. Presented at the Tenth AnnualMeeting of the PeripheralVascular Surgery Club, Baltimore,Md., June 5, 1985. Reprint requests: Robert W. Hobson II, M.D., Department of Surgery, Room D-202, Boston UniversityMedical Center, 75 E. Newton St., Boston, MA 02118.

diabetic patient. Since skin blood flow is a major determinant in healing, it follows that methods directed at measuring cutaneous blood flow may be useful in evaluating the potential for healing. We evaluated skin blood flow in both diabetic and nondiabetic patients with transcutaneous oxygen tension (tcPQ) and laser Doppler velocimetry (LD) in an effort to predict whether ischemic ulcerations and amputations o f the forefoot would heal. The measurement o f tcPo2 was first used for continuous monitoring o f newborn infants in neonatal intensive care units and only recently has been applied in patients with peripheral vascular occlusive disease. LD provides a direct measurement o f blood flow velociw 511

Journal of VASCULAR SURGERY

512 Karanfilian et al.

400-

300-

a

b

¢

rnV 2 0 0 I000 Fig. 1. Laser Doppler tracing demonstrates the pulse wave amplitude (a), vasomotor waves (b), and skin blood flow velocity (c). in the superficial arterioles and capillaries of the skin. We have used both techniques to quantitatively evaluate cutaneous blood flow in the forefoot and have compared these results with the corresponding systolic Doppler ankle pressure in an effort to assess the potential for healing. MATERIAL AND METHODS Fifty-nine limbs in 48 adult male patients (mean age, 57 years), who were admitted with ischemic ulcerations or gangrenous changes of the digits or forefoot, were examined. Thirty-seven (63%) of these limbs were in diabetic patients and 22 (37%) limbs were in nondiabetic patients. Twenty transmetatarsal or digital amputations were performed, whereas the remainder of the lesions were d4brided and allowed to heal by secondary intention or were covered by split-thickness skin grafts. Systolic Doppler ankle pressures were obtained in all patients and the ankle/arm index calculated by dividing the ankle systolic pressure by the brachial systolic pressure. If the ankle/arm index was greater than 1.0 (n = 14), the ankle pressure was considered artifactually elevated, and the value was excluded when the mean ankle pressures were calculated. LD (model LD5000, Medpacific Corp., Seattle, Wash.) was used to measure skin blood flow velocity from either the plantar aspect of the great toe or the plantar aspect of the foot overlying the first metatarsal head (n = 56 limbs). The helium-neon laser Doppler velocimeter uses a monochromatic light source with a wavelength of 632.8 nm. Conducted to the skin via fiber optics, the light penetrates to a maximum depth of 1.5 mm, after which the reflected light is conducted back to a photodiode via a second set of fiber optics. Doppler-shifted light, reflected from moving red blood cells within the surface capillaries

of the skin, is processed and expressed in millivolts. The Doppler shift is directly proportional to the vclocity of blood flow and represents an average of all velocity readings in a skin area of 1.5 mmL LD tracings were evaluated according to previously established criteria. ~ Pulse waves (Fig. 1, a) reflect velocity changes in arteriolar and capillary blood flow that correspond to the cardiac cycle. Vasomotor waves (Fig. 1, b) are the result of rhythmic variations in blood flow velocity secondary to sympathetic inncrvation of the arteriolar bed and occur approximately four to six times per minute. The deviation of the entire tracing from the baseline (Fig. 1, c) represents the skin blood flow velocity. In evaluating skin blood flow in patients with peripheral vascular disease, wc assessed the pulse wave amplitude (LD-PWA) and skin blood flow velocity (LDSBFV). The tcPo2 was measured in millimeters of mercury from the dorsal foot midway between the ankle and toes (model SMI-361, Litton Medical Electronics, Elk Grove, Ill.). The tcPo2 monitor uses a modified polarographic electrode consisting of a low oxygen-consuming platinum cathode, a silver anode, electrolyte, and an oxygen-permeable membrane. The probe also incorporates a heating element and thermistor for control of the skin temperature. When the probe is heated, it "arterializes" the capillary; bed by local vasodilation. This also increases the temperature of the blood under the electrode, shifting the oxyhemoglobin dissociation curve to the right, thereby increasing the partial pressure of oxygen. Oxygen diffusion is further enhanced by the partial liquification of subcutaneous fat below the stratum corneum of the skin. The instrument was calibrated against a known standard before each use and the probe was left in place for 20 minutes before a final reading was obtained. During the measurement the probe temperature was maintained at 45 ° C. All lesions were evaluated clinically for at least 30 days. Amputations were considered to have healed if the wound edges were closed and viable without evidence of infection. D4bridement or local revision was performed if an amputation failed initially. If the amputation failed a second time, the wound was classifted as nonhcaling. Forefoot and digital ulcerations were d4brided and allowed to heal by secondary intention or were covered by split-thickness skin grafts. If after 30 days there was no evidence of healthy granulation tissue, wound contraction, or a viable graft, the ulcer was classified as nonhealing. Although evidence of healing was usually apparent within the

Volume 4 Number 5 November 1986

Noninvasive assessment of the ischemic forefoot

513

Table I. Transcutaneous oxygen tension (tcPo2), LD-SBFV, LD-PWA, Doppler ankle pressure, and ankle/arm index for diabetic and nondiabetic patients Diabetic

Nondiabetic

Healed tcPo2 (mm Hg) LD-SBFV (mV) LD-PWA (mV) Ankle pressure) (mm Hg) Ankle/arm index~

30 98 14 81 0.58

Nonhealed

_+ 4.0 _+ 13.0 +_ 3.0 +_ 9.0 + 0.07

7 50 4 71 0.49

Healed

_+ 2.5* _+ 8.0* +_ 0.5* +_ 11.0 _+ 0.07

42 88 8 75 0.53

Nonhealed

_+ 3.5 _+ 15.0 -+ 1.4 +_ 6.5 +_ 0.05

2 37 2 41 0.31

_+ 1.6" _+ 2.0* _+ 0.3* _+ 21.0 _+ 0.16

LD-SBFV = skin blood flow velociw measured by laser Doppler velocimetry; LD-PWA = pulse wave amplitude measured bv laser Doppler vclocimetrv. NOTE: All values arc expressed as mean _+ standard error of the mean. *Significant dif}~rence between healed and nonhealed groups (p < 0.05). tFourteen measurements (24%) excluded because of artifactual elevations in pressure.

Table II. Predictive accuracy of tcPo2 All patients Healed Actual Predicted Healing Nonhealing

Nonhealed

31

25

31 (100%) 0

3 (12%) 22 (88%)

Diabetics Total

Healed

56

16

18

16 (100%) 0

3 (17%) 15 (83%)

Accuracy = 53/56 (95%)

Nonhealed

Nondiabetics Total

Healed

34

15

7

15 (100%) 0

0 7 (100%)

Accuracy = 31/34 (91%)

first 30 days, ultimate resolution required up to 4 months in some instances. All measured and calculated data were expressed as a mean value _+ standard error of the mean for all patients and for diabetic and nondiabetic patients. Apparent differences were analyzed for statistical significance with Student's t test for unpaired data. With a best-fit technique, the data for each method were analyzed to establish those criteria at which each would yield the greatest accuracy for predicting healing. The sensitivity of a technique was defined as the ability of that technique to reflect adequate blood flow when it was in fact present. The specificity was defined as the ability to reflect the absence of flow when no flow was present. Ultimate wound healing was the standard against which each of the techniques were compared. Apparent differences in sensitivity, specificity, or accura~ were analyzed for statistical significance with chi-square analysis or Fisher's exact test. RESULTS Of the 59 ischemic limbs studied, 32 ultimately healed (54%) and 27 (46%) failed to heal. Seventeen of 37 (46%) limbs in diabetic patients healed,

Nonhealed

Total 22

Accuracy = 22/22 (100%)

whereas 15 of 22 (68%) limbs in nondiabetic patients eventually healed. O f the 20 transmetatarsal and digital amputations performed, 18 (90%) healed, whereas two had to be revised at a higher level. Of the 39 ulcerative lesions, 14 (36%) eventually healed. The tcPo2, LD-SBFV, LD-PWA, Doppler ankle pressures, and ankle/arm indices for diabetic and nondiabetic patients are presented in Table I. The tcPo2, LD-SBFV, and LD-PWA values are significantly (p < 0.05) greater in those limbs that ultimately healed. In contrast, there is no significant (p < 0.05) difference in the Doppler ankle pressures or the ankle/arm indices. The criteria for each technique that are consistent with healing are as follows: tcPo2 greater than 10 mm Hg, LD-SBFV greater than 40 mV, LD-PWA greater than 4 mV, and ankle systolic pressure greater than 30 mm Hg. Both criteria for LD (LD-SBFV and LD-PWA) must be met. The results of tcPoz measurements are reviewed in Table II. For all patients the test had a 95% accuracy, with a sensitivity of 100% and a specificity of 88%. The three falsely positive studies were in diabetic patients. The results of LD and Doppler ankle pressure

514

Journal of VASCULAR SURGERY

Karanfilian et al.

Table III. Predictive accuracy of laser Doppler velocimetry All patients

Actual Predicted Healing Nonhealing

Diabetics

Nondiabetics

Healed

Nonhealed

Total

Healed

Nonhealed

Total

Healed

29

24

53

16

17

33

13

23 (79%) 1 (4%) 6 (21%) 23 (96%) Accuracy = 46/53 (87%)

14 (87%) 1 (6%) 2 (13%) 16 (94%) Accuracy = 30/33 (91%)

Nonhealed

Total

7

9 (69%)

0

4 (31%)

7 (100%)

20

Accuracy = 16/20 (80%)

Table IV. Predictive accuracy of Doppler ankle pressures All patients

Actual Predicted Healing Nonhealing

Diabetics

Healed

Nonhealed

Total

Healed

32

27

59

17

24 (75%) 20 (74%) 8 (25%) 7 (26%) Accuracy = 31/59 (52%)

Nonhealed

Nondiabetics Total

Healed

37

15

20

11 (65%) 17 (85%) 6 (35%) 3 (15%) Accuracy = 14/37 (38%)

measurements are presented in Tables III and IV, respectively. Both LD and tcPo2 have an accuracy that is significantly (p < 0.05) greater than that of Doppler ankle pressures. Although LD has a specificity of 96%, it has a sensitivity of only 79%, which is significantly (p < 0.05) less than the sensitivity of tcPo2. Doppler ankle pressures have a sensitivity of 75% but a specificity that is significantly (p < 0.05) less than that of LD or tcPo2. In addition, the specificity and accuracy of Doppler ankle pressures is significantly decreased in diabetic patients. The sensitivity, specificity, and accuracy of each technique are summarized in Table V. DISCUSSION

Digital or transmetatarsal amputation (TMA) is a valuable adjunct in the treatment of gangrenous lesions of the digits and forefoot, particularly in diabetic patients. 2-r However, it is difficult to predict ultimate healing, and authors have reported failure rates in excess of 50%, with even higher failure rates in younger patients and those patients with diabetes. 8 Berardi and Keonin 9 reported 16 transphalangeal amputations and TMAs, all of which eventually failed 4 months to 5 years after operation. Porter, Bauer, and Taylor 1° reported ultimate success in 70% of phalangectomies and TMAs, although only 66% of the phalangectomies and 42% of the TMAs healed primarily. In 1976 Jamieson and Hill H concluded that a trial of distal amputation was the only

Nonhealed

Total

7

22

13 (87%) 3 (43%) 2 (13%) 4 (57%) Accuracv = 17/22 (77%)

definitive test to indicate whether healing would Occur.

The value of a laboratory test in predicting healing is based on its sensitivity, or ability to reflect adequate blood flow when it is present. A false-positive study, which predicts healing in an extremity that ultimately fails, results in amputation or the revision of an amputation to a higher level. However, a false-negative study, which predicts failure to heal a lesion or the failure of an amputation that eventually heals, potentially results in an unnecessary amputation or amputation at a higher level than is actually necessary.

The inability of Doppler ankle pressures to predict the healing of forefoot amputations was rcportcd by Mehta et al. I2 They observed that TMAs healed in 75% of patients with ankle pressures of less than 40 mm Hg, whereas 80% of those with ankle pressures between 40 and 60 mm Hg and 76% with pressures greater than 60 mm Hg healed. As in the present study they observed no significant difference between the Doppler ankle pressures of those patients whose amputations healed and those whose amputations did not. They concluded that a low ankle pressure does not contraindicate a TMA, while a higher pressure does not guarantee success. Similar difficulties in interpretation of Doppler ankle systolic pressures were documented by Carter, ~3,~4who recommended use of toe pressure measurements. He reported only one amputation among

Volumc 4 Number 5 November 1986

Noninvasive assessment of the ischemic Jbrefoot

515

Table V. Results of transcutaneous oxygen tension, laser Doppler velocimet~, and Doppler ankle pressures in evaluating ischemic forefoot ulcerations and amputations

Transcutaneous Po2 Laser Doppler velocimetry Doppler ankle pressures

Sensitivi~

Spccifici~.

Accuracy

100% 79% 75%

88% 96% 26%

95% 87% 52%

21 limbs with an ankle pressure exceeding 30 mm Hg in nondiabetic patients and 55 mm Hg in diabetic patients. However, wound healing also occurred in 25 of 38 limbs (65%) with toe pressures less than these levels. Similar deficiencies were documented in the series reported by Holstein et al. ''~ in which healing was observed in 34% of patients with toe pressures less than 30 mm Hg. As a result of these data, evaluation of skin blood flow measurements by LD and tcPo2 monitoring seemed appropriate in an effort to improve on these results. LD has been used clinically to assess skin blood flow since 1977, after reports by Holloway and Watkins 16were published. It has been used to assess tissue trauma as a result of bullet wounds ~7 or injections, ~s to evaluate intestinal blood supply, 19-21 and to evaluate the potential viability of skin flapsfl 2,23Its use in patients with peripheral vascular disease was described by our group in 1984. I In that report we were successful in differentiating between patients with limb-threatening ischemia and those without vascular disease. Use of the tcPo2 monitor to evaluate skin blood flow in patients with peripheral vascular disease was originally described by Tonnesen 24 in 1978 and later by Matsen et al. 2s Clync et al. 26 observed decreased tcPo2 values in patients with claudication and rest pain, whereas Eickhoff and Engel127 demonstrated that tcPo2 was significantly decreased in patients with symptomatic ischemia and increased after successful revascularization. They suggested use of tcPo2 monitoring in the immediate postoperative period to assess graft patency. Use of tcPo2 to determine the ultimate level of amputation was reported by White et al.,28 who observed that a tcPo2 greater than 50 mm Hg was indicative of probable healing, whereas values less than 40 mm Hg indicated probable failure. Values between 40 and 50 mm Hg indicated a tenuous situation in which amputations probably would not heal in the presence of infection. Franzeck et al. 29 reported on 69 patients with peripheral vascular disease in whom the calf tcPo2 was significantly decreased. Their data indicated that the tcPo2 incorrectly predicted the outcome in 3 of 35 amputations,

all three being successful despite a tcPo2 less than 10 mm Hg. They suggested that a tcPo2 less than 5 mm Hg indicated a poor prognosis for healing, whereas a tcPo2 greater than 20 mm Hg indicated a good potential for healing. This same group later reported 119 amputations, 39 of which involved the forefoot. 3° They measured the tcPo2 before and 10 minutes after 100% oxygen was administered and concluded that a baseline tcPo2 greater than 10 mm Hg, or an increase of more than 10 mm Hg during oxygen inhalation was consistent with a successful outcome. They reported a 95% accuracy rate with below-kalee (BK) amputations but only a 71% accuracy rate with digital amputations and TMAs. The latter improved to 76% with oxygen inhalation. Of seven TMAs evaluated, five healed (mean tcPo2, 29 _+ 24 mm Hg) and two failed (mean tcPo2, 15 + 8 mm Hg). The authors concluded that the technique was more accurate in predicting the outcome of BK amputations than it was for digital and forefoot amputations. On the basis of our present data, however, we would disagree with this conclusion. Our accuracy of 95% for forefoot amputations is equal to that reported by Harward et al? ° for BK amputations and superior to their data for forefoot amputations. In addition, an increase in accuracy from 71% to 76% during oxygen administration has no statistical significance, and thus there is no evidence that oxygen inhalation increases the accuracy of the technique. The decreased sensitivity of LD compared with that of tcPo2 may be due in part to local vasoconstriction as a result of variations in ambient temperature. Although currently available tcPo2 monitors have a thermistor incorporated into the probe to control skin temperature, no heating elements are currently incorporated into the LD probe. Recently Matsen et al. 3~ compared the tcPo2 monitor and LD at identical sites on the feet of normal volunteers, with the skin heated and unheated. They reported that tcPo2 and LD reflect changes in the arteriovenous gradient (and thus skin blood flow) over areas of warmed skin but not unwarmed skin, suggesting that unless the skin is warmed, LD and tcPo2 may be of limited value. This may have contributed to the significantly lower sensitivitw of LD compared with

516

Journal ol VAS(~[~LAF. SURGERY

Karanfilian et al.

that oftcPo2 measurements. The future development of heating elements for LD probes could increase the sensitivity of the technique to parallel that of tcPo> Although the present data represent a retrospective correlation of the potential for wound healing and the results of preoperative testing, it can be concluded that tcPo2 measurements and LD provide an accurate indication of skin blood flow. The high falsenegative rate (21%) associated with LD makes it inferior to tcPo2 measurements, although both tests are superior to Doppler ankle pressures. Furthermore, both examinations were equally useful in diabetic and nondiabetic patients, whereas Doppler ankle pressures yielded only a 38% accuracy in the diabetic population. However, whether these studies will ultimately offer the means with which to accurately predict the healing of ischemic ulcerations and amputations of the forefoot must await a prospective application of these criteria.

15.

16. 17.

18.

19.

20.

21. REFERENCES 1. Karanfilian RG, Lynch TG, Lee BC, Long JB, Hobson RW. The assessment of skin blood flow in peripheral vascular disease by laser Doppler velocimetq< Am Surg 1984; 50: 641-4. 2. McKirtrick LE, McKittrick JB, Risley TS. Transmetatarsal amputation for infection or gangrene in patients with diabetes mellitus. Ann Surg 1949; 130:826-42. 3. Wheelock FC. Transmetatarsal amputations and arterial surgery in diabetic patients. N Engl J Med 1961; 264:316-20. 4. Schwindt CD, Lulloff R& Rogers SC. Transmetatarsal amputations. Orthop Clin North Am 1973; 4:31-42. 5. Haimovici H. Criteria for and results of transmetatarsal amputation for ischemic gangrene. Arch Surg 1955; 70:45-51. 6. Bradham GB, Lee WH, Stallworth JM. Transmetatarsal amputation. Angiology 1960; 11:495-8. 7. Warren R~ Crawford ES, Hardy TB. The transmetatarsal amputation in arterial deficiency of the lower extremi~. Surgery 1952; 31 : 132-40. 8. Eflkney DJ, Lira RC, Schecter WP. Transmctatarsal amputation. Arch Surg 1977; 112:1366-70. 9. Berardi RS, Keonin Y. Amputations in peripheral vascular disease. Am J Surg 1978; 135:231-4. 10. Porter JM, Bauer GM, Taylor LM. Lower-extremity amputations fbr ischemia. Arch Surg 1981; 116:89-92. 11. Jamieson CW, Hill D. Amputation for vascular disease. Br J Surg 1976; 63:683-90. 12. Mehta K, Hobson RW, Jamil Z, Hart L, O'DonneH JA. Fallibiliq, of Doppler ankle pressures in predicting healing of transmetatarsal amputation, I Surg Res 1980; 28:466-470. 13. Cartcr SA. The relationship of distal systolic pressures to healing of skin lesions in limbs with arterial occlusive disease, with special reference to diabetes mellitus. Scand J Clin Lab Invcst 1973; 31(Suppl 128):239-43. 14. Carter SA. Role of pressure measurements in vascular disease.

22.

23.

24.

25.

26. 27.

28.

29.

30.

31.

In: Bernstein EF, ed. Nonim,asivc diagnostic techniques m vascular disease. St. Louis: CV Mosbv Co, 1982:301-10. Holstein P, Noer I, Tonncsen KH, Sager P~ Lassen NA. Distal blood pressure in severc arterial insufficiency. In: Bcrgan JJ, Yao JST, eds. Gangrene and severe ischemia of the lower extremities. Ncw York: Grunc & Stratum, lnc, 1978:95-114. Holloway GA, Watkins DW. Lascr Doppler measuremcnt of cutaneous blood flow. J Invest Dermatol 1977; 69:306-9. Oberg PA, Nilsson GE, Tcnland T, Holmstrom A, Lcwis DH. Use of a new laser Doppler flowmeter tbr measurement of capillary blood flow in skeletal muscle aRer bullet wounding. Acta Chir Scand 1979; Suppl 489:145-50. Hollowav GA. Cutaneous blood flow responses to injection trauma measured by laser Doppler vclocimctry, l Invcst Dermatol 1980; 74:1-4. Shepherd AP, Riedel GL. Continuous measurement of intestinal mucosal blood flow by laser-Doppler velocimetry. Am J Physiol 1982; 242:G668-72. Rotering RH, Dixon JA, Holloway GA, McCloskey DW. A comparison of the HeNe laser and ultrasound Doppler systems in the determination of viability of ischemic canine intestine. Ann Surg 1982; 196:705-8. Feld AD, Fondacaro JD, Holloway GA, Jacobson El). Mcasurcment of mucosal blood flow in the canine intestine with laser Doppler velocimetry. Life Sci 1982; 31 : 1509-17. Jones BM, Mayou BJ. The laser Doppler flowmetcr tbr microvascular monitoring: A preliminary report. Br J Plast Surg 1982; 35:147-9. Larrabee WF, Holloway GA, Trachy RE, Sutton D. Skin flap tension and wound slough: Correlation with laser Doppler velocimctry. Otolaryngol Head Neck Surg 1982; 90:185-7. Tonnesen KH. Transcutancous oxygcn tcnsion in immincnt fbot gangrene. Acta Anaesthesiol Scand 1978; Suppl 68: 107-10. Matscn FA, Wyss CR, Pedegana LR, Krugmire RB, Simmons CW, King RV, Burgess EM. Transcutaneous oxygcn tension measurement in peripheral vascular disease. Surg Gvnccol Obstet 1980; 150:525-8. Clync CA, Ryan J, Webster JH, Chant AD. Oxygen tension on thc skin ofischemic legs. Am J Surg 1982; 143:315-8. Eickhoff JH, Engell HC. Transcutaneous oxygen tcnsion (tcPo,) measurements on the foot in normal subjects and in patients with peripheral arterial disease admitted for vascular surgeq,. Scand J Clin Lab Invest 1981; 41:743-8. White RA, Nolan L, Harley D, Long J, Klcin S, Temper K, Nelson R, Tabrisky J, Shoemaker W. Noninvasivc evaluation of pcripheral vascular diseasc using transcutaneons oxygcn tension. Am J Surg 1982; 144:68-75. Franzeck UK, Talke P, Bernstein EF, Golbranson FL, Fronck A. Transcutaneous Po2 mcasurcments in hcalth and peripheral arterial occlusive disease. Surgery 1982; 91:156-63. Harward TR, Volney J, Golbranson F, Bernstcin EF, Fronck A. O~gen inhalationqnduccd transcutaneous Po2 changcs as a predictor of amputation lcvcl. J VAse SuRe; 1985; 2: 220-7, Matscn FA, Wyss CR, Robertson CL, Oberg PA, Holloway GA. The relationship oftranscutaneous Po2 and laser Dopplcr measurements in a human model of local arterial insufficiency. Surg Gynecol Obstct 1984; 159:418-22.