Oxygen tension assessment: an overlooked tool for prediction of delayed healing in a clinical setting

Abstract

Successful wound healing requires adequate transcutaneous oxygen tension (tcpO₂). TcpO₂ may not commonly be incorporated in clinical assessments because of variable measurement response at different sensory temperatures. This study aims to assess the relationship between changes in tcpO₂, measured under basal (39°C) and stimulated (44°C) conditions and healing rate of chronic wounds over 4 weeks, to determine whether tcpO₂ measurement can predict delayed wound healing. TcpO₂ (Radiometer TCM400) measurements at sensor temperatures 39 and 44°C were recorded (twice, 4 weeks apart) adjacent to the ulcer site, and at a mirror image site on the contralateral leg. Ulcer outline was traced on clear acetate and perimeter and area measured (Visitrak™, Smith and Nephew). TcpO₂ measured at 44 and 39°C adjacent to all 13 wounds were lower compared to the contralateral site, significant at 44°C (P = 0·008). Significant correlation (r ² = 0·8) occurred between wound healing rate and increased tcpO₂ at 44°C over 4 weeks. Importantly, the ratio of 39/44°C tcpO₂, measured at the initial appointment, appeared to predict normal or delayed healing rate. TcpO₂ may provide clinicians with information regarding anticipated healing ability of wounds at the initial appointment, and hence identify wounds requiring early implementation of adjuvant therapies to accelerate healing.

INTRODUCTION

The Australian population is ageing and wounds are a common problem in the older population. In 1995, ∼5% of older people reported skin ulceration in the two previous weeks (1). Ulceration is age dependent, with the prevalence of chronic wounds increasing with increasing age (2). In addition, as a result of the increasing age and sedentary lifestyle of the Australian population, diabetes is increasing in prevalence (3). As a result, foot complications, such as foot ulcerations, are also increasing. These wounds are difficult to heal and are notoriously recurrent. The current direct cost of wounds in Australia is suspected to be in the hundreds of millions of dollars. Current wound management in the community is not always optimal but is supported by specialist clinics throughout metropolitan Melbourne that provide appropriate assessment, diagnosis and management of chronic wounds. The Wound Management Service at Heidelberg Repatriation Hospital, Austin Health is one of those clinics.

Currently, patients are assessed for adequacy of arterial supply, presence of sensation and venous return to diagnose the aetiology of the wound. Wound diagnosis guides management and selection of evidence‐based treatment(s) (4). The literature suggests clinicians monitor wounds for 4 weeks, and if healing has not progressed significantly, then adjuvant or other therapies may be required (5). Hence, it usually takes 4 weeks to inform clinicians if alternative therapies are necessary. Research has shown that the longer the delay in treatment, the more difficult a wound is to heal 6, 7. In addition, the delay wastes resources and increases costs of care. Therefore, it would be helpful to have a diagnostic tool that could provide us with appropriate information regarding ability of the wound to heal at a very early stage, to allow for the implementation of appropriate strategies at an earlier time. We believe that oxygen tension measurement may be a diagnostic approach that could be used in the clinic to provide clinicians with information regarding the ability of wounds to heal.

The initial clinical application of transcutaneous oxygen monitors (TCM) was to measure neonatal arterial oxygen tension. Currently, advocates promote oxygen tension measurement around wounds (8), in skin of people with venous oedema (9), in ischaemic limbs to assess limb amputation level (10), prediction of healing diabetes foot ulcers (11) and, more recently, to predict need for revascularisation to heal ischaemic diabetes foot ulcers (12) and to predict ischaemic ulcer healing (13). Confidence in use of skin oxygen tension monitors to monitor wound healing has been reduced in the past because of the variation in responses at different sensory temperatures 14, 15, 16, 17. In their 2000 guidelines, the TransAtlantic Inter‐Society Consensus (TASC) (18) proposed a value of between 30 and 50 mmHg to identify those patients with ischaemic wounds in whom revascularisation is warranted. In June 2007, a new version of TASC was edited, and in this edition only a single value of 30 mmHg was proposed for transcutaneous oxygen tension (tcpO₂) to indicate the value at which revascularisation is warranted to prevent amputation (19).

To support the new TASC guidelines, more recent studies have shown improved results of using tcpO₂, with a predictive value of <34 mmHg to identify which patients with diabetes and ischaemic foot ulcers require revascularisation to heal and prevent amputation (12). While a predictive value of tcpO₂ levels 34 < 40mmHg was suggested by authors as less pressing for surgical intervention, although a considerable probability of amputation was identified even at this level (20). These authors also suggest that revascularisation in cases where tcpO₂ levels are greater than 40 mmHg should be dependent on the severity of tissue loss and possible morbidity caused by the procedure, therein requiring further judgement prior to making the decision.

These cases focus on ischaemic wounds, with our without diabetes‐related sensory neuropathy, and very little has been undertaken in evaluating the ability of the TCM to predict healing of other wound types.

WHAT DOES TCM ACTUALLY MEASURE?

When TCM sensors are heated to physiological levels (e.g. 37°C), skin oxygen tension is increased in older patients (14). Upon heating the sensors to above 42°C, oxygen tension reduces in older persons (17), in people with diabetes (16) and in those with diabetic peripheral neuropathy (15). This variation can be explained by the age‐ and disease‐associated changes that oxygen tension measurement can identify, which depend on the temperature of the TCM sensors. TCM sensors, when applied at 44°C, can reflect the function of sensory nerves (21). The importance of these nerves in tissue repair determines the clinical significance of this measurement. Reduced sensory nerve function significantly delays healing, and those who have reduced oxygen tension levels, measured at sensor temperature of 44°C, are likely to have reduced sensory nerve function (21). Hence, measuring oxygen tension at these temperatures may identify slow healing patients early and allow adjuvant therapies to be commenced earlier to accelerate healing.

This project assessed the differences in oxygen tension at the skin adjacent to various wound types and correlated this with healing rates over a 4‐week period. Different wound types and locations, ages, comorbidities, size, duration of ulceration and depth of ulceration were documented. The aim of the study was to assess whether healing rate is associated with changes in tcpO₂. The establishment of such a relationship would provide a useful tool to assess the ability of wounds to heal, and whether early use of adjuvant therapies was required to accelerate healing. The method of measurement provides this information quickly, within 30–40 minutes/patient.

Participants

Ethics approval was obtained from the Austin Hospital Human Research Ethics Committee. Patients attending the Heidelberg Repatriation Hospital Wound Management Service were invited to participate in the study and those who consented to the research were included.

Those with clinically infected wounds and with grade 3 or 4 pressure wounds were excluded.

Methods

At the beginning of the study, basic demographic data were collected, including age, comorbidities and medications. Peripheral sensation was measured using a 5·07 Semmes–Weinstein monofilament (22). Wound‐specific factors were also collected such as location of the wound, duration of the wound, whether size has increased over the last 4 weeks and wound diagnosis.

At weeks 0 and 4 the following factors were measured:

• 1Oxygen tension at 39 and 44°C at the wound edge.
• 2Wound dimensions:
•  i.wound size
•  ii.wound depth.

Other treatment

Patients' wounds were treated using standard evidence‐based practice. This included appropriate compression and pressure redistribution using felt, orthotics, walkers or casts depending on the requirements of the patient and the impact of such treatments on their quality of life. Wound dressings were chosen according to stage of healing, level of exudate, amount of necrotic tissue and slough and the degree of bacterial bioburden. Visits to the Wound Management Clinic were undertaken as dictated by the wound and the wound management regime, but generally occurred at a minimum of every 4 weeks. At intervening times, management was undertaken in the community. Healing rates were determined at an appointment 4 weeks after the initial visit.

Baseline and 4‐week measurements

Oxygen tension.

Oxygen tension (tcpO₂) at sensor temperatures at 39 and 44°C at the wound edge was measured using transcuaneous monitor (TCM 400, Radiometer Pacific, Copenhagen, Denmark). Oxygen tension was also measured on a mirror image site on the unaffected limb for comparison. The measuring sites were carefully cleaned with alcohol swab. The fixation rings were fixed to the skin with adhesive rings at the peri wound site and a mirror image site on the contralateral limb, both avoiding any large veins, hair, skin defects, bone or tendons. A reference electrode was applied to the chest wall in the left second intercostal space, in the mid‐clavicular line. The electrodes were then calibrated by the TCM and heated to 39°C. Drops of contact liquid, supplied by the manufacturer, were applied inside the fixation ring and the TCM electrode screwed in place. The resultant values were recorded once the reading was stable, usually taking approximately 15 minutes. The electrodes were then returned to the TCM, and recalibrated to 44°C and re‐applied to the fixation rings – further drops of contact fluid were applied if required.

Wound healing.

Wound healing was measured as percentage reduction in area per 4 weeks. This monitors healing rate without being affected by the initial size of the ulcer 23, 24, 25. A tracing of the ulcer outline was taken on a clear piece of acetate and the perimeter and area was traced using a digital planimeter. Three planimeter readings were taken, with the average used for each of the perimeter and area.

Other factors that may affect healing and diabetes control.

Medications, smoking history, alcohol use, level of schooling, ulcer duration, ulcer size and amount of arterial supply to the limb [using ankle‐brachial pressure indices (ABPI) or vascular scans] were also recorded.

Statistics

Fifteen patients were initially included in the study. Two patients did not attend the second session (4 weeks after the initial visit). Statistic analysis included correlations and analysis of variance (ANOVA).

RESULTS

Mean ulcer size at baseline was 7·4 cm² (SEM ± 3·3), mean wound duration was 34·7 weeks (SEM ± 6·4) and mean patient age at the start of the study was 74·7 years (SEM ± 2·9). The study population included 57% males. At 4 weeks, 23% of wounds fully healed.

TcpO₂ measured at both 39 and 44°C adjacent to wounds (6·7 ± 1·1) and (24·9 ± 4·5), respectively, were lower compared to the contralateral site (22·2 ± 2·8) and (45·6 ± 5·4), respectively, reaching significance for the 44°C measurement (P = 0·008) as shown in Table 1. The relationship between the rate of healing and percentage change of oxygen tension from baseline at 39 and 44°C are shown in 1, 2, respectively. Calculations of the rate of healing and percentage change in baseline of tcpO₂ at 44°C (measured as levels at week 4‐week 0/week 0) were performed for 11 participants. As shown in Figure 2, there was a significant correlation (r ² = 0·8) between wound healing rate and the increase in tcpO₂ at 44°C over a 4‐week period.

Table 1.

Data comparing tcpO₂ measured at sensor temperatures of 39°C (baseline) and 44°C (stimulated) adjacent to the wound and the mirror image on the contralateral limb

	Wound (39°C)	Control (39°C)	Wound (44°C)	Control (44°C)
Mean ± SEM	6·7 ± 1·1	22·2 ± 2·8	24·9 ± 4·5	45·6 ± 5·4
One way ANOVA	F (1,24) = 23·7,P > 0·05		F (1,24) = 8·11,P < 0·05 (0·008)

ANOVA, analysis of variance; tcpO₂, transcutaneous oxygen tension.

Figure 1.

Rate of healing versus percentage change from baseline at sensor temperature of 39°C (r ² = −0·24).

Figure 2.

Rate of healing versus percentage change from baseline at sensor temperature of 44°C (r ² = −0·8).

The ratio of 39/44°C tcpO₂ measured at the initial appointment was also calculated. As shown in Figure 3, the ratio of 39/44°C tcpO₂ might be a useful predictor of a normal or a delayed rate of healing. The data showed that a ratio of 39/44°C tcpO₂ below 0·4 is a predictor of delayed healing. This predictor was applicable to all patients tested except for one patient with an ABPI value greater than 1·3, suggesting medial arterial calcification of the ankle arterial walls and thereby incompressible vessels.

Figure 3.

Relationship between 39/44°C tcpO₂ measurements at initial visit and rate of healing over 4 weeks.

Of note, 50% (3/6) of patients whose wounds did not progress to healing had diabetes, whereas 30% (2/7) of patients with wounds that had significant healing had diabetes and both patients exhibited a loss of protective sensation.

Correlations between the wound dimensions, duration and the ratio of tcpO₂ at 39/44 sensor temperatures as well as the healing rate are shown in 2, 3. There was no correlation between the ratio of tcpO₂ at 39/44 sensor temperatures and other suggested prognostic indicators for wound healing. There was, however, a moderate positive correlation (r ² = 0·6) between the ratio of tcpO₂ at 39/44 sensor temperatures and the rate of healing as well as between ulcer duration and wound depth.

Table 2.

Correlations on rate of healing at 4 weeks, wound area at baseline and wound duration to ratio of tcpO₂ at ratio 39/44 tcpO₂ sensor temperatures ^*

Ratio 39/44°C measurements	Rate of healing	Ulcer duration (weeks)	Wound area (cm) at week 0	Wound depth (mm) at week 0
0·4	0·48	5	5·8
0·2	−0·09	60	3·5	4
0·1	−0·2	39	48	3
0·2	0·06	28	18·4	1
0·1	−1		0·3	1
0·4	−0·42	36	5·5	1
0·4	1	24	0·5	0·5
0·5	0·49	6	3·5	1
0·4	0·69	52	6·1	3
0·4	1	65	2·5	0·5
0·5	1		2·3	3
0·8	0·36	24	4·4	0·5
0·1	0·05	66	1·9	3

^*There was no correlation between the ratio of tcpO₂ at 39/44 sensor temperatures and other suggested prognostic indicators for wound healing. There was, however, a moderate positive correlation (r ² = 0·6) between the ratio of tcpO₂ at 39/44 sensor temperatures and the rate of healing as well as between ulcer duration and wound depth.

Table 3.

Correlations between the prognostic factors shown in Table 2

	Column 1	Column 2	Column 3	Column 4	Column 5
Column 1	1
Column 2	0·562234	1
Column 3	−0·46404	−0·10033	1
Column 4	−0·36949	−0·24285	−0·0391285	1
Column 5	−0·4172	−0·07827	0·5649152	0·245318	1

DISCUSSION AND CONCLUSIONS

International groups have used the oxygen tension monitor as a tool to dictate amputation site for those patients require lower extremity amputation (26). In addition, a consensus group has stated that tcpO₂ obtained while breathing normobaric air can assist in identifying which patients will not heal spontaneously (27). The TASC consensus group have recommended using values of <30 mmHg to indicate which patients with critical limb ischaemia and wounds require revascularisation to prevent amputation (20).

It should be noted that in our study where nine patients with chronic venous ulcers and four patients with mixed venous and arterial ulcers were tested, tcpO₂ measured at 44°C adjacent to wounds (24·9 ± 4·5 mmHg) were lower compared to the contralateral site (45·6 ± 5·4 mmHg) and indeed both values are relatively low in view of the above discussion on tcpO₂ measurements in critical limb ischaemia. This could be related to the underlying pathology in our patient group where venous insufficiency resulting in bilateral oedema could contribute to the low tcpO₂ values reported in this study. However, Nemeth et al. (28) argued that oedema may not constitute a barrier to oxygen diffusion through the skin and does not account for the low tcpO₂ values in the ulcerated oedematous limb. On the other hand, Boyko et al. (29) suggested that depending on measurement site, pedal oedema on the dorsal foot and leg could be a factor that contributes to reduce lower limb tcpO₂ measurements.

Most importantly, the inclusion of the contralateral intact skin measurement in our study is an important reference reflecting the overall systemic effect of the underlying pathology. In addition, we have evaluated outcomes within 2–4 weeks, which is consistent with Keast et al. (30), who showed that the percentage decrease in ulcer area [measured with the Visitrak system (Smith and Nephew, Melbourne, Australia)] during that period was a predictor of healing at 12–24 weeks.

Based on the outcomes of this exploratory study, we believe that this machine could provide useful information regarding the potential of different wound types to heal, therefore reducing delays in implementation of appropriate wound therapies and leading to accelerated healing, reduced costs and improved patient quality of life. Indeed, the current results suggest that the oxygen tension monitor may be a useful diagnostic tool to incorporate into an initial clinical wound assessment.

As stated previously, TCM sensor measurements, when sensors are heated to 39°C, reflect basal microvascular blood flow, while its application at 44°C results in measurements that better reflect the function of sensory nerves (21). This is attributed to the fact that 44°C is a temperature sufficient to activate C fibres to release sensory neuropeptides that contribute to increase microvascular blood flow among other functions that promote tissue healing 31, 32. The clinical significance of emphasising the differences between what both measurements reflect is related to the importance of these nerves in tissue repair. Reduced sensory nerve function significantly delays healing (33), and reduced oxygen tension levels when measured at sensor temperature of 44°C is correlated with reduced sensory nerve function (21). Hence, measuring oxygen tension at these temperatures may identify slow healing patients early and adjuvant/alternative therapies can be commenced earlier to accelerate healing.

The current data supports our previous work 21, 34, 35, 36 and endorses the notion that adequate C fibre function is important for skin integrity. While the data showed that tcpO₂ measured at both 44 and 39°C adjacent to wounds were lower compared to the contralateral site, it is only those measurements that reflect C fibre function that reached significance (P = 0·008). The results suggest that while a decline in both basal microvascular blood flow and C fibre function contributes to wound initiation, a decline in C fibre function has a greater effect on wound healing.

Furthermore, over a 4‐week period, the data showed an unexplained very weak negative correlation between wound healing rate and the increase in tcpO₂ at 39°C while there was a significant correlation (r ² = 0·8) between wound healing rate and the increase in tcpO₂ at 44°C. The results suggest that it is the improvement in C fibre function as reflected by the improved 44°C tcpO₂ measurements that positively impacted on wound healing.

The above results might not seem to be novel, as it would be anticipated that a decline in basal microvascular blood flow and C fibre function (as reflected by basal and stimulated tcpO₂ measurements) would contribute to wound initiation and their improvement would help the healing process. It should be noted, however, that we have provided new evidence to support the notion that changes in C fibre function do have more significant effects on wound initiation and subsequent healing.

The above proposition does not necessarily mean that tcpO₂ measurements at 44°C should be the only measurements to be used in clinical settings. Indeed, taking both 39 and 44°C tcpO₂ measurements adds additional clinical and prognostic information.

This study showed that it is the ratio of 39/44°C tcpO₂, measured at the initial appointment which might be a useful predictor for a normal or a delayed rate of healing in chronic venous ulcers. This was a clinically significant result. The data showed that a ratio of 39/44°C tcpO₂ below 0·4 is a predictor of delayed healing. This predictor was applicable to all patients tested except for one patient with an ABPI value greater than 1·3 suggesting calcification of the walls of the arteries and incompressible vessels. This incompressibility has been suggested to be a result of diabetes or renal failure rather than peripheral vascular disease (37); however; in this case the patient was not suffering from diabetes. Further information regarding medical history is unavailable; however, the presence of incompressibility may have been because of systemic issues that may have affected the tcpO₂ results. Further investigation of possible links between incompressibility, microcirculation and wound healing is warranted.

Of note, diabetes was present in 50% (3/6) of patients whose wounds did not progress to healing and in only 30% (2/7) of patients with wounds that had significant healing. These two patients with diabetes and good healing also had a loss of protective sensation. Loss of protective sensation indicates a reduction in sensory nerve activity, reflected by reduced tcpO₂ readings at 44°C. However, the ratio of 39/44°C tcpO₂ was still able to identify the wounds that had significant healing in these patients. Therefore, we can propose that the ratio of 39/44°C may also be applicable to patients with diabetes and loss of protective sensation. Further investigation using a larger sample is necessary to confirm these findings.

This result indicates that adequate basal microvascular blood flow (reflected by 39°C tcpO₂ measurements) and adequate blood flow when stimulated via C fibre activation (reflected by 44°C tcpO₂ measurements) is essential to significant healing over a 4‐week period. In other words, improvement in C fibre function alone will not necessarily result in healing if it is not supported by adequate basal microvascular blood flow.

This study therefore provides evidence that tcpO₂ measurements could provide clinicians with information predicting the healing ability of wounds at the initial appointment, and hence identify wounds requiring early implementation of adjuvant or alternative therapies to accelerate healing. Further studies are needed to support the above proposition.

Other studies have examined differentiating between those venous wounds that healed and those that did not heal by the 12th or 20th week of care using various measurements at 2, 4 or 8 weeks of care including progressive percentage change in area, log healing rate and log area ratio (38). However, this requires waiting for these time periods (2–8 weeks) to identify slow healing patients. Moreover, this approach does not provide information regarding what factors are involved in preventing these wounds from healing within the expected time frame.

A recent study showed that a venous ulcer that has an area of less than 10 cm² and is less than 12 months duration at the first visit has a 29% chance of not healing by the 24th week of care, while a wound greater than 10 cm² and greater than 12 months old has a 78% chance of not healing (39). However, like the earlier studies, this prognostic information does not provide much information regarding the factors causing the delay in healing of these wounds and thereby this approach cannot help decide on management that may reverse this delay.

The benefit of using this study's tcpO₂ measurement technique is that it identifies both the skin nutritive capacity at baseline, as well as the ability of the sensory nerves to respond to the wound, and thereby the ability of the person to deliver the requisite microcellular components necessary for the wound to heal in a timely manner, if it can heal at all. This information can be obtained within 30–40 minutes and requires only a simple ratio calculation. We have previously shown that increasing sensory nerve activity by using low‐frequency sensory nerve stimulation (LF‐SNS – Zeinab Khalil – Patent No: PCT/AU2004/001079 Nerve function and tissue healing 21, 35, 40 in addition to compression can accelerate wound healing in those with venous leg ulcers who have reduced sensory nerve function (36). While the link between healing and improvement in sensory nerve function by both compression and use of LF‐SNS requires further study, having the ability to identify deficits in C fibre function in a clinical setting using the TCM may help select those who could benefit from improving C fibre function with the use of LF‐SNS.

This study's results highlight the important role incorporating adjuvant therapies added to standard evidence‐based wound management techniques may have in achieving improved wound healing. We may not be far from identifying patients with sensory nerve deficits using this method, and then initiating adjuvant therapy to improve microvascular blood flow by reducing sympathetic vasoconstrictor tone as well as stimulating sensory nerve function with the use of the LF‐SNS.

There are several limitations of this study. It is best regarded as a proof of concept study, to evaluate the ability of tcpO₂ to identify wounds that are less likely to heal. There were only a small group of participants from one wound centre who had wounds of various durations and of mainly venous aetiology (nine venous and four mixed venous arterial). Replication of this study using a larger population with various wound aetiologies and from a variety of different wound settings is necessary to confirm these results. In particular, data on those with diabetes and foot ulcers, known to have a significant neuropathic component, needs further investigation.

ACKNOWLEDGEMENTS

This study was supported by Rebecca L. Cooper Medical Research Foundation and Radiometer Pacific Pty Ltd. Andrew Jardine performed the measurements for this study.