An evaluation of the use of transcutaneous oxygen pressure measurement in the non-invasive vascular laboratory : with special reference to selection of amputation level.
Transcutaneous oxygen pressure measurement (TCp02) using a miniaturised Clarke electrode and a heating thermistor was developed independently by Huch et al and Eberhardt et al in 1972. After its initial use to non invasively monitor arterial partial pressure (Pa02) in neonates it was proposed as a useful test of skin blood flow and possibly amputation wound healing level selection in patients with peripheral vascular disease. Unfortunately a wide range of predictive values emerged with some authors reporting amputations healing when the TCp02 value was 0 mmHg. The investigation, while still considered useful, has not gained widespread support. This study investigates the use of TCp02, establishes a value for the use of the TCp02 Index to predict amputation wound healing potential and examines the hypothesis that the use of the TcpO Index to select amputation level can reduce patient morbidity and mortality. The literature is reviewed and a series of studies evaluating TCp02 use, undertaken in the Durban Metropolitan Vascular Service Non-Invasive Laboratories, are presented. TCp02 measurements were performed in a standardised manner with the subject supine breathing room air. Measurements were taken at fixed sites, on the mid dorsum of the foot (Foot), 10 cm distal to the tibial tuberosity and 2 cm lateral to the anterior tibial margin (BKA), 10 cm proximal to the patella in the midline (AKA) and on the chest in the mid-clavicular line. A TCp02 Index, the limb to chest ratio was defined. TCp02 data derived from control subjects asymptomatic of peripheral vascular disease were shown to be similar to age matched pooled data derived from the literature. In patients with peripheral vascular disease, absolute TCp02 and the TCp02 Index were shown to fall from proximal to distal sites and again were no different to pooled data derived from the literature. Based on presenting symptoms, the fall in TCp02 and the TCp02 Index was significant from proximal to distal sites. The reduction in absolute TCp02 and the TCp02 was also related to the most distal pulse present. TCp02 values were found to be no different in patients with peripheral vascular disease with or without diabetes. When comparing TCp02 and the TCp02 Index with Doppler pressure measurements at the Popliteal artery and at the foot, and the Doppler ankle brachial index (ABI), Doppler derived data were significantly higher in diabetic patients than in non-diabetic patients. No differences were noted in TCp02 data. TCp02 was compared with the 133Xe radio-isotope skin washout test. The best correlation was (r = 0.46) was obtained with a logarithmic curve y = 10.862Ln(x) + 38.751. TCp02 was compared with antibiotic concentrations (Cefoxitin) in muscle obtained from the site of amputation and the Cefoxitin Index, the ratio of muscle antibiotic concentration to plasma concentration, as an indication of the relationship of skin TCp02 to muscle blood flow. A significant correlation was shown between the Cefoxitin Index and TCp02 (r = 0.67, p = 0.035) and the TCp02 Index (r = 0.64, P = 0.045), suggesting that skin oxygen delivery may reflect muscle antibiotic delivery and hence blood flow. TCp02 and the TCp02 Index were compared with heated and unheated laser Doppler fluxmetry (LDF) in 35 patients undergoing amputation wound healing assessment. Significant correlations were shown between heated LDF, heated LDF Index and the TCp02 Index (r = 0.63 and r = 0.69, P < 0.0001). TCp02 Index values of 0.5 and 0.55 showed an accuracy of 96.2 % in predicting amputation outcome while LDF values of 3, 4 and 5 arbitrary units gave an accuracy of 88.5 %. Using receiver operator curves, a TCp02 Index of 0.55 was shown to be the best test. Over the years 1987 and 1988, TCp02 data were gathered on 193 patients undergoing lower limb amputation for peripheral vascular disease. Information on the outcome of the amputation was available for 152 amputations. Circumstances which might result in a reduced pre-operative TCp02 reading were identified and criteria were set for the use of TCp02 to predict amputation wound healing potential. 122 amputations which met the defined entry criteria were available for evaluation. A TCp02 Index of 0.50 gave a definitive predictive value below which no amputation healed. Similarly no amputation with an absolute TCp02 of less than 27 mmHg healed. Receiver operator characteristic curves showed the TCp02 Index to be a better test than absolute TCp02. A TCp02 Index of 0.55 was shown to have the best sensitivity of96.7 %, with a specificity of79.8 % and an accuracy of 90.2 %. When introduced to clinical practice, correct use of the TCp02 Index of 0.55 resulted in a reduction in amputation revision rate from 40.3 % in 1987, to 8.2 % in 1990. Initially some surgeons felt that the TCp02 Index predicted amputation wound failure at distal sites at which healing could be expected on clinical criteria, and chose amputate at sites with a TCp02 Index value less than 0.55. These amputations failed to heal. As surgeons gained confidence in the test, they chose to follow the TCp02 data more often and the percentage of amputations performed at sites predicted by the TCp02 Index to fail , fell from 35.5 % in 1987 to 6.6 % in 1990. Over a 15 year period at King Edward VIII Hospital, the amputation revision rate has fallen from an average of 32.7 % in the first five years when Tcp02 data were not available to the surgeon, to 21.4 % and 22.9 % in the two subsequent 5 year periods when Tcp02 data were available. The mortality rates were unchanged. The decline in revision rates was less than expected and relates to the fact that approximately only 42 % of patients requiring amputation undergo the test. This is because it is time consuming and available only during weekday office hours. These studies have confirmed the usefulness of Tcp02 measurement in the non-invasive vascular laboratory. The index is shown to be superior to absolute Tcp02 as a predictive test of amputation wound healing. The introduction of several criteria to define when Tcp02 use is appropriate has refined the investigation and made it clinically useful in our setting. A Tcp02 Index of 0.55 in the appropriate patient is a useful test to predict amputation wound healing and its use has resulted in reduced patient morbidity and mortality, confirming the hypothesis tested.