Diabetes mellitus is a common disease, affecting approximately 3.7 million people in the UK (6.6% of the UK population in 2015), and its incidence is increasing at the rate of 100 000 per year. Approximately 10% of patients have type 1 and 90% have type 2 diabetes (Diabetes UK, 2018). Foot ulceration occurs frequently in patients with diabetes, particularly type 2 diabetes, with the prevalence ranging from 4% to 10% (Wu et al, 2015). The lifetime risk of developing a diabetic foot ulcer (DFU) is between 5% and 7%, although some research suggests the risk is as high as 15% (Reiber, 2001; International Working Group on the Diabetic Foot, 2003; NHS Digital and Diabetes UK, 2017).
DFUs often develop after minor trauma, such as stubbing a toe, or soft tissue injury when trimming nails. They are frequently indolent, requiring intensive treatment by podiatrists, often over a long period of time. DFUs cause considerable discomfort and inconvenience to the patient and their treatment costs the NHS an estimated £935 million per year (Kerr, 2017). It is therefore important to establish whether there is an underlying, treatable disease process, such as ischaemia, that will prevent conservative treatment from working.
Patients with diabetes are particularly prone to foot ulceration due to a variable combination of neuropathy and vascular insufficiency. Neuropathy causes numbness of the toes or feet, leaving the patient unaware of skin injury. Vascular insufficiency results in poor blood flow, causing delayed healing or a failure to heal at all. In severe cases of diabetic vascular disease, which are often associated with smoking, high blood pressure and high cholesterol, skin breakdown may occur without trauma. Personal experience has shown that in some patients with diabetes, the treatment of vascular disease has resulted in resolution of numbness in their feet, suggesting that peripheral nerve ischaemia is responsible for non-functioning of the peripheral nerves, rather than a true peripheral neuropathy. From a clinical perspective, this makes assessment of the underlying cause of ulceration difficult, as numbness may itself be caused by vascular disease. Measurement of tissue oxygenation should help to resolve this difficulty. Transcutaneous oxygen (tcpO2) monitoring measures the partial pressure of arterial oxygen within the tissues and could therefore potentially identify ischaemia in patients with DFUs.
Aim and objective
This study aimed to determine whether tcpO2 measurement can identify ischaemia and, therefore, undiagnosed critical limb ischaemia (CLI), in patients presenting with DFUs who are assessed clinically as having no significant vascular disease. This pilot study aimed to establish whether tcpO2 measurement:
- Is acceptable to patients in a routine clinical context
- Improves patient assessment by identifying an underlying cause, preventing lengthy and futile conservative management
- Enables earlier referral for investigation and treatment of underlying ischaemia, resulting in better clinical outcomes
- Justifies subsequent study in a larger population.
Method
Participants
Patients referred to NHS Lanarkshire podiatry outpatient clinics for conservative treatment were eligible to participate in the trial if they had ulceration, discolouration or necrosis of the foot, heel or toes, their ulcers were not due to vascular insufficiency, and they had no history of peripheral limb ischaemia. Some patients had already been assessed in a combined diabetic and vascular outpatient clinic. Only patients who were able to give informed consent were included. Patients were excluded from the trial if they had a preexisting clinical diagnosis of CLI or if they had, or were likely to have, an allergy to the cups for the sensors.
At the end of the initial recruitment period, it was decided to recruit additional patients with DFUs, as well as patients with diabetes but no foot ulceration as a control group. Patients in the control group were recruited from the diabetic outpatient clinic.
Transcutaneous oxygen monitoring
A TCM400 tcpO2 monitor with six-channel input was loaned by Radiometer for a period of 6 weeks to perform this study. Radiometer also trained our study technicians, two podiatrists and a staff nurse, to carry out the measurements.
Small adhesive cups were attached to the skin, a few drops of a coupling fluid were placed into the cups and the tcpO2 sensor attached to the cups with a screw connection. The overall appearance and size of the cups were similar to an ECG electrode. The sensors became warm, reaching 44˚C, and it took approximately 15 minutes for tissue oxygenation and the sensor to equilibrate to determine an accurate value. As the sensors are effectively in contact with patients’ skin, they were sterilised by immersion in electrolysed water for a period of 30 seconds between each patient (Dancer et al, 2015).
Process
Patients were given a participant information sheet on attending the outpatient clinic in advance of signing a consent form at their next visit. All patient data were anonymised by allocating a study number to each patient once they had consented. A study data form was then completed for each patient by the study technician recording:
- Which foot was to be measured
- Ulcer status (healing, static or progressing)
- Ulcer duration
- The referring clinician’s clinical assessment about ischaemia and/or peripheral neuropathy as the underlying cause of the ulcer.
Information about medication, diabetes history and other medical conditions was obtained from the patient’s notes.
TcpO2 readings were taken by the study technician in a warm, comfortable environment with the patient lying supine. All measurements were taken with reference to the user’s manual. The tcpO2 probes were attached to the skin as directed in the manual. Six simultaneous readings were acquired by probes placed as follows:
- Lower chest/upper abdomen, as a baseline for that patient
- Thigh, 10 cm above the knee
- Dorsum of the calf, approximately one-third of the way up the calf from the ankle to the knee, being the most distal part of the flap required to close a below-knee amputation
- Dorsum of the foot medially
- Dorsum of the foot laterally
- As near as possible to the area of ulceration, on healthy skin.
For the tcpO2 readings to be reproducible and comparable, they were performed using a standard routine. This routine involved calibration of the equipment, the patient avoiding caffeine and smoking prior to readings and lying supine in a room with an ambient temperature of 21–23˚C. Equilibration between the tissue oxygen and the sensor required approximately 15 minutes, before stable readings are acquired.
Afterwards, patients were asked to complete a brief study evaluation questionnaire. They were asked ‘How much would you say that you were inconvenienced by your participation in this research study?’ and ‘How much discomfort or itching did you experience due to the application of the monitor sensors?’ There were five possible responses to each question:
- No inconvenience/discomfort
- Minor inconvenience/discomfort
- Some inconvenience/discomfort
- Major inconvenience/discomfort
- Not sure.
Participants were asked to tick the box that most closely related to their experience. They were also invited to add additional comments, if they wished.
Patients were reassessed clinically at 12 months and placed in to one of four categories: healed, improving, static or deteriorating.
Results
Forty-six patients with a DFU and 26 patients with diabetes but without a foot ulcer were entered into the trial. In the ulcerated group there were 39 male patients (85%) with an average age of 64 (range 39–87 years) and seven female patients (15%) with an average age of 49 (range 36–66 years). In the non-ulcerated control group there were 14 male patients (54%) with an average age of 65 (range 33–84 years) and 12 female patients (46%) with an average age of 63 (range 49–74 years). Table 1 summarises the clinical characteristics of the participants. Some information was not recorded, accounting for the discrepancies in the percentages. Interestingly, the prevalence of hypertension and hyperlipidaemia was higher in the patients without ulcers.
The tcpO2 results for patients with ulceration are given in Table 2. Of the 37 patients who were re-assessed after 12 months, 16 patients’ ulcers (34.7%) had healed and eight (17.5%) were showing signs of improvement, three (6.6%) were static and the remaining 10 (21.7%) were showing signs of deterioration. The other nine patients (19.5%) with DFUs died during the study period.
Table 3 gives the tcpO2 results for patients in the control group, who did not have a diabetic foot ulcer. There is no striking difference between the readings for the non-ulcerated group and the ulcerated group. This is, perhaps, not unexpected as the majority of the patients with ulcers did not have underlying ischaemia.
Four patients (15.4%) in the non-ulcerated group died during the follow-up period; a similar proportion to those patients with ulceration. One of these patients (patient 1005) had ischaemic tcpO2 readings and another (patient 1021) had a high tcpO2 reading.Two patients (7.7%) were referred to podiatry for foot ulceration. One of these patients (patient 1001) had ischaemic tcpO2 readings.
A similar proportion of individuals in both groups had a tcpO2 of ≤30 mmHg: eleven out of 46 (23.9%) in the group with diabetic foot ulcers and five out of 26 in the control group (19.3%). There was also no real difference in the proportions of patients in each group with high distal readings (26.0% in the group with foot ulceration versus 34.6% in the control group). There were a number of high peripheral readings, which have been highlighted in pink, in both groups.
Patient evaluation survey
All patients completed the evaluation questionnaire. Only one patient commented upon experiencing mild discomfort. All other responses indicated that tcpO2 measurement caused no inconvenience or discomfort. There were no negative comments; where offered, respondents remarked that the experience was comfortable, relaxed, and even pleasurable. The routine use of this investigation was found to be acceptable by participants.
Discussion
In this pilot study, there was remarkable similarity in the clinical make up of the two cohorts of patients. Surprisingly, perhaps, there was a higher prevalence of hypertension and hyperlipidaemia in the patients without DFUs. The group of patients with ulcers were thought clinically to have neuropathic rather than ischaemic DFUs and the aim of the study was to try to identify those patients with sub-clinical ischaemia.
TcpO2 measurement is a non-invasive method of measuring oxygen concentration (pO2 in mmHg) in the subcutaneous tissues, providing an objective measurement of oxygenation and an indication of perfusion. It is a useful method of identifying and assessing CLI and is more accurate than ankle/brachial pressure index (ABPI), particularly in patients with diabetes whose vessels are often calcified and non-compressible. TcpO2 measurement has also been used to predict whether ulcers are likely to heal without revascularisation (Ruangsetakit et al, 2010) and to predict healing after revascularisation procedures (Caselli et al, 2005).
A tcpO2 of ≤30 mmHg is indicative of CLI (Norgenn et al, 2007). There was no difference in the proportion of participants with CLI in either group, according to their tcpO2 readings. Several low readings were from the sensor on the calf, the significance of which is unclear. If only the low readings from the foot sensors are used, then eight patients in the ulcerated group (17%) and two from the control group (8%) had readings suggestive of CLI. Of these eight patients, two patients’ ulcers healed (patients 2 and 46) and one was improving (patient 39) after 12 months of conservative treatment. Two patients’ DFUs remained static over this period (patients 15 and 23) and three deteriorated (patients 1, 19 and 27). The five patients whose ulcers failed to heal (13% of the DFU group) are likely to have had ischaemia as an underlying cause and would have benefitted from vascular investigation and consideration of revascularisation, despite having already been assessed clinically as having no vascular disease. The two patients whose DFUs healed with conservative treatment may have healed more quickly if underlying vascular disease, suggested by the low tcpO2 readings, had been identified and treated.
The high peripheral readings are difficult to explain. A high peripheral tcpO2 may be highly significant in patients with DFUs. One patient consented to have tcpO2 readings taken during a peripheral angioplasty procedure. The patient had diabetes, diffuse disease of the calf vessels and a necrotic great toe. TcpO2 sensors were placed on the lateral and medial aspects of the foot and a sensor at the base of the second and third toes, as close to the necrotic great toe as possible. The tcpO2 readings at the start of the procedure were 42 mmHg at the lateral foot, 13 mmHg at the medial foot and 86 mmHg at the base of the toes. The high reading at the base of the toes was unexpected. On inflating the angioplasty balloon in the posterior tibial artery, the tcpO2 reading over the medial foot dropped to 2 mmHg, suggesting no collateral flow. On releasing the balloon, it started to rise above 20 mmHg, but the tcpO2 at the base of the toes started to fall to below 60 mmHg. Possibly the only explanation is that the last collateral flow that can occur is through the skin vessels, so skin over the base of the toes was receiving a much larger blood supply than it required as blood made its way to supply the necrotic great toe. This would have an impact on the clinical assessment of this patient: good perfusion at the base of the necrotic toe makes a clinical diagnosis of ischaemia unlikely.
On the basis of the pilot study results, tcpO2 measurement seems a potentially good method of identifying ischaemia in patients with DFUs. A larger trial would seem more than justified in view of the considerable cost to the NHS of treating these patients. A number of changes, both technical and procedural, should be considered in such a trial:
- The measurements carried out during the angioplasty procedure suggest ischaemia may be very localised, as although the adherent cup was positioned as close as possible to the necrotic toe, the tcpO2 was unusually high. Smaller adherent cups or some other connection is required
- The adhesive holding the cup to the skin was very strong and caused some damage to the skin, which is already delicate, in a small number of patients. The adhesive has to be water-resistant, as a water-based liquid is used as the coupling agent with the sensor. A nontoxic solvent is required to remove the cups
- To maximise the opportunity to identify ischaemia, the measurements should be taken with the leg elevated by 30˚, as this reduces the tcpO2 in patients with ischaemia, increasing the sensitivity of the test (Ruangsetakit et al, 2010).
If a larger study can demonstrate the efficacy of tcpO2 in identifying ischaemia in patients with DFUs, this could be used to assess all new referrals to ensure ulcers requiring revascularisation are not overlooked. This would avoid the patient embarking on a long, unnecessary treatment programme — often with weekly hospital visits — with little chance of a successful outcome.
Conclusion
The results of this pilot study suggest that tcpO2 measurement may be a suitable method for identifying ischaemia in DFUs. Participants found it to be comfortable and convenient. A larger trial should be considered with the aim of demonstrating the efficacy of tcpO2 in identifying ischaemia in patients with DFUs and preventing the instigation of intensive and costly treatment with little chance of a successful outcome.
Acknowledgement: Thanks to Radiometer for kindly lending the TCM400 monitor and for training the staff to use it. Thanks also to the Podiatry department, Hairmyres Hospital.
