Chronic venous insufficiency (CVI) commonly affects the lower limbs secondary to valvular incompetence, venous obstruction, or post-thrombotic phenomenon. While CVI is not a life-threatening disease, it can result in long-term morbidity with consequent financial and social implications. Due to a dearth of research on this issue, it is difficult to assess the prevalence of leg ulcers in India, which has a population of over 1.2 billion people [1]. In India, the estimated prevalence of leg ulceration was 4.5/1,000 population, with CVI being a significant contributor; male sex and intermediate socioeconomic status were found to be risk factors [2,3].

Adequate tissue perfusion and oxygenation are prerequisites for wound healing. Factors hampering tissue oxygenation predispose patients to ulcer formation and persistence. In CVI, the microcirculation is impaired, resulting in skin changes and ulcer formation. There are few studies with conflicting results documenting the correlation between CVI and tissue oxygen [4-7], with none in the Indian population in the English literature. Tissue oxygenation can be estimated by measuring transcutaneous partial oxygen pressure (TcPO₂).

Measurement of oxygen tension in the tissue is a good tool for predicting wound healing post-amputation in arterial diseases [8]. In this study, TcPO₂ in advanced CVI (stages C4 to C6) was measured and compared to the TcPO₂ in the unaffected limb. Measurement of tissue oxygenation (hypoxia) in CVI using TcPO₂ can help to better understand the disease process, prognostication, and compare the effectiveness of various treatment modalities. We hypothesized that tissue oxygenation, as represented by TcPO₂, will progressively decrease from C4 to C6 according to the CVI severity.

This prospective cross-sectional study was conducted at our teaching hospital between July 2014 and August 2016. The study was approved by the Institutional Review Board of the Christian Medical College (IRB no. 8994) and written informed consent was obtained from the patients. Based on the formula (1.96)²×pq/d² for observational cross-sectional prospective study, the required sample size was calculated to 96 (“p,” i.e., prevalence, was taken to be 50%; “q” = p-1; and “d” is the standard error [10%]) [9].

Patients presenting to the vascular surgery department with unilateral CVI, C4 to C6, were screened and included in the study. All patients underwent duplex ultrasonography. Twenty-two patients were excluded from the study after screening because of bilateral limb disease, peripheral arterial disease, vasculitis, active limb infection, or pedal edema due to other causes (lymphedema, renal, or hepatic failure). C3 disease was also excluded because edema is known to alter the TcPO₂ value [10]. The contralateral limb did not show any clinical symptoms or signs of CVI.

After obtaining informed consent, a detailed clinical examination was performed to confirm the diagnosis of CVI, stage the disease according to the Clinical-Etiology-Anatomy-Pathophysiology classification [11], and rule out other causes of limb ulcers. A single TcPO₂ measurement was taken in the supine and dependent (sitting) positions at the site of maximum skin change since it would reflect the lowest level of oxygenation corresponding to the site of maximal microangiopathic changes or the worst disease. In C6 disease, TcPO₂ was measured close to the ulcer edge, while in C5 disease, it was measured close to the border of the scarred tissue. The TcPO₂ value at the corresponding site of the opposite limb served as an internal control for that particular patient, thus matching patient-related confounding factors. Measurements were obtained in the vascular lab at an ambient temperature of 22°C to 24°C. The electrode was attached with an adhesive ring to the site at a temperature of 43°C. The physiological stabilization time of a patient is 15 to 20 minutes for TcPO₂ reading. During this time, the electrode slowly heats the skin and dilates the arteries. The stable value after 20 minutes was recorded as the measurement for the analysis.

Data were analyzed using IBM SPSS statistics (version 21.0; IBM Corp., Armonk, NY, USA). Descriptive statistics, paired t-test, independent samples t-test, and Bonferroni analysis were performed. Statistical significance was set at P<0.05.

Of the 96 patients, 59.4% were above the age of 40 years (mean age, 44.7±11.7 years; range, 25-73 years). The study included 85 males and 11 females. Left-sided disease was present in 51 (60.0%) males and 7 (63.6%) females (Table 1).

Table 1 . Descriptive statistics of patients included in study.

Characteristic	Male (n=85)	Female (n=11)	Total (n=96)
Stage
C4	20 (23.5)	3 (27.3)	23 (24.0)
C5	18 (21.2)	1 (9.1)	19 (19.8)
C6	47 (55.3)	7 (63.6)	54 (56.3)
Side
Right	34 (40.0)	4 (36.4)	38 (39.6)
Left	51 (60.0)	7 (63.6)	58 (60.4)
Age (y)
≤30	11 (12.9)	2 (18.2)	13 (13.5)
31-45	36 (42.4)	1 (9.1)	37 (38.5)
46-60	31 (36.5)	6 (54.6)	37 (38.5)
>60	7 (8.2)	2 (18.2)	9 (9.4)

Values are presented as number (%)..

A Shapiro–Wilk test and visual inspection of histograms, normal Q-Q plots, and box plots showed that the TcPO₂ values of the affected and non-affected limbs were normally distributed when measured in the dependent position (skewness<0.35), but had significant skewness when measured in the supine position (affected limb, 0.729; non-affected limb, 0.814). However, parametric tests were employed in all cases, as they fulfilled the other required conditions for parametric tests.

In the supine position, irrespective of the stage of advanced CVI, the mean TcPO₂ in the affected limb (23.3 mmHg) was significantly lower than that in the non-affected limb (32.1 mmHg; P<0.01). In the dependent position, the TcPO₂ in the affected limb (40.5 mmHg) was significantly lower than that in the non-affected limb (50.7 mmHg; P<0.01; Table 2).

Table 2 . Comparison of supine and dependent mean TcPO₂ values in unaffected and affected limbs.

Limb	Position	TcPO2 (mmHg)	Paired t-test
Normal	Supine (I)	32.05±11.16
	Dependent (J)	50.74±11.89
	Difference (J–I)	18.69±6.83	P<0.01
Diseased	Supine (I)	23.28±11.52
	Dependent (J)	40.50±12.22
	Difference (J–I)	17.12±6.68	P<0.01

Values are presented as mean±standard deviation..

Subgroup comparison according to the affected legs of various stages versus the unaffected leg (Table 3) showed a statistically significant difference between the unaffected leg and legs with stage C5 to C6 disease (P<0.01). The difference in TcPO₂ was statistically non-significant for C4 disease. However, the values were lower in legs with C5 or C6 disease, both in the supine and dependent positions (Table 3). In the supine position, a statistically significant decrease in mean TcPO₂ values was observed between C4 and C5, but not between C5 and C6. This decrease was not significant among stages C4 to C6 in the dependent position. The range of TcPO₂ values is listed in Table 4.

Table 3 . Comparison of mean TcPO₂ in different stages of CVI and their significance in different positions (unit: mmHg).

Position	CVI stage (I) (mean TcPO2)	CVI stage (J) (mean TcPO2)	Mean TcPO2 difference (I–J)	P-value
Supine	C0 (32.05)	C4 (26.57)	5.48	0.187
		C5 (16.63)	15.42	<0.001*
		C6 (24.22)	7.83	<0.001*
	C4 (26.57)	C5 (16.63)	9.94	0.022*
		C6 (24.22)	2.34	>0.999
	C5 (16.63)	C6 (24.22)	–7.59	0.058
Dependent	C0 (50.74)	C4 (44.83)	5.91	0.182
		C5 (35.47)	15.27	<0.001*
		C6 (40.43)	10.31	<0.001*
	C4 (44.83)	C5 (35.47)	9.36	0.063
		C6 (40.43)	4.40	0.791
	C5 (35.47)	C6 (40.43)	–4.96	0.681

CVI, chronic venous insufficiency..

*Statistically significant..

Table 4 . Range of TcPO₂ in clinical stages 4, 5, 6 and normal limb (unit: mmHg).

Clinical stage	Position
	Supine	Dependent
Stage 4 TcPO2 (n=23)
Mean±SD	26.57±13.59	44.83±13.05
Median (range)	28.00 (6–65)	49.00 (22–75)
SEM	2.97	2.85
Stage 5 TcPO2 (n=19)
Mean±SD	16.63±8.80	35.47±9.95
Median (range)	18.00 (4–32)	35.00 (15–52)
SEM	2.14	2.41
Stage 6 TcPO2 (n=54)
Mean±SD	24.22±10.43	40.43±12.00
Median (range)	24.50 (1–54)	43.00 (4–60)
SEM	1.61	1.85
Normal limb TcPO2 (n=96)
Mean±SD	32.05±11.16	50.74±11.89
Median (range)	30.00 (5–74)	50.00 (20–85)
SEM	1.25	1.33

SD, standard deviation; SEM, standard error of the mean..

The TcPO₂ had fair accuracy with an area under the receiver operating characteristic curve >0.7 for both supine (Fig. 1) and dependent positions (Fig. 2) in detecting advanced CVI, when compared to the unaffected limb. A TcPO₂ value of 40 mmHg in the supine position had a sensitivity of 95% and specificity of 20% for predicting the diseased state. In the dependent position, a value of 56 mmHg had a sensitivity of approximately 91%, with a specificity of 25%. A cut-off value could not be recommended because the increase in sensitivity was at the expense of specificity and a high false-positive rate. There was a positive correlation between supine and dependent TcPO₂ values in both affected (0.839) and non-affected (0.818) limbs (Fig. 3).

Figure 1. Receiver operating characteristic (ROC) of supine TcPO₂ in predicting disease state. Area under the curve=0.727, standard error=0.036, P<0.001.

Figure 2. Receiver operating characteristic (ROC) of dependent TcPO₂ in predicting disease state. Area under the curve=0.733, standard error=0.036, P<0.001.

Figure 3. Scatter plot of supine and dependent TcPO₂ values in affected and unaffected legs. Control R² linear = 0.669. Case R² linear = 0.703.

The transcutaneous oxygen content is supposed to reflect the microcirculatory state of the underlying tissue, both in health and disease. Its content or partial pressure should decrease with severe disease in CVI, with values being lowest in stage C6, with or without trophic changes. A higher TcPO₂ would indicate a better state of health in the tissue. The null hypothesis would state that TcPO₂ does not reflect the microcirculation the CVI.

The prevalence of CVI is higher in females, as seen in many population studies [9,12]. However, this study of 96 Indian patients included 85 males (88.5%) and 11 (11.5%) females, and the prevalence of CVI was lower in females, which may reflect a selection bias. The prevalence of CVI increases with age [13]. In this study, 59.4% of patients were above the age of 40 years. TcPO₂ had fair accuracy in detecting advanced stage CVI (C4 to C6) when compared to the unaffected limb. The values were significantly lower in the C5/C6 disease, both in the supine and dependent positions. Segregated by stage, TcPO₂ was significantly lower in C5/C6 than in the unaffected limb. C6 had a statistically non-significantly higher TcPO₂ than that of C5. There was a significant decrease in TcPO₂ from C4 to C5. This is expected because scarred or healed lesions are known to have low oxygen tension.

In a study by Franzeck et al. [14], TcPO₂ was measured and correlated with video microscopy to examine the microvasculature. The mean TcPO₂ of control group was 56.8 mmHg, and that of affected group was 47.7 mmHg. The difference was not statistically significant (P>0.05). In this study, TcPO₂ was measured at the site of maximal change, which reflects poor microcirculation internally.

Barnikol and Pötzschke [15] reported a comparative study of the mean value of four TcPO₂ measurements in patients with C6 CVI; the mean TcPO₂ value was significantly lower in the C6 disease group than in the control group (17.9 vs. 63.0 mmHg, respectively). In this study, the mean TcPO₂ was also significantly lower in the C6 disease group than in the control group (24.2 vs. 32.1 mmHg, respectively). The greater difference in the mean values in the study by Barnikol and Pötzschke [15] was probably due to four measurements taken around the ulcer, which is more reliable than a single measurement. However, even a single measurement at the site of greatest skin change was statistically significant, as shown in this study. Recently, TcPO₂ was also studied in post-thrombotic syndrome (PTS) by Cuen-Ojeda et al. [16], who found low TcPO₂ levels in severe cases of PTS.

A cutoff value of 40 mmHg in the supine position had 95% sensitivity and 20% specificity; the same cutoff in the dependent position had a 50% sensitivity and 80% specificity. A particular cutoff value could not be recommended, as the increase in sensitivity came at the expense of specificity and a high false-positive rate. The study also found that the TcPO₂ was significantly lower in the supine position than in the dependent position, thus demonstrating the effect of limb position on TcPO₂ values.

There are several limitations to this study. The number of patients in each stage of advanced CVI was not equal, with the majority of patients in the C6 stage. A single peri-articular TcPO₂ measurement may not provide representative information concerning the oxygen status of the entire wound due to oxygen inhomogeneity. Multiple same-limb control measurements were not performed. However, this was primarily done to determine whether TcPO₂ correlates with the severity of CVI. In addition, repeated measurements would take time and add expense due to the need for new electrodes, which is difficult in resource-limited settings. There could be a potential selection bias, as the location for TcPO₂ measurement was based on the subjective assessment of area of maximum skin change. Peripheral arterial disease was ruled out in all patients clinically with the presence of palpable pulses and normal ankle-brachial pressure index; however, an arterial duplex ultrasound was not performed in all patients.

In a limb affected by CVI, the TcPO₂ was significantly lower in the C5/C6 stage than in the unaffected limb, with a definite effect of limb positioning on TcPO₂ values. Since TcPO₂ was lowest in the C5 stage, its role or clinical relevance in this stage is doubtful, as scarred tissue will obviously have low oxygen tension. The statistically significantly lower value than that in the unaffected limb and the decrease in TcPO₂ from C4 to C6 (though not statistically significant) provides a direction for studies focusing on single TcPO₂ value as a prognostic tool in relation to interventions and as a marker for C6 disease. An increase in TcPO₂ following intervention could predict the efficacy of the intervention. Future studies using TcPO₂ to compare interventions or objectively assess improvement after an intervention can be conducted.

TcPO₂ is a non-invasive test that can be used with a single value in an outpatient setting to assess clinical progression (worsening or improvement) of CVI. This is the first study in an Indian population focusing on the relevance of TcPO₂ in advanced CVI for prognostication.

We would like to thank Dr. Grace Rebecca for help in statistical analyses and Dr. Aubin Mathew Varghese in preparation of this manuscript.

This study was supported by internal fluid research grant (No. 8994).

The authors have nothing to disclose.

Concept and design: AAK, SA. Analysis and interpretation: LJ, AAK, SA. Data collection: LJ, AAK. Writing the article: LJ, AAK. Critical revision of the article: LJ, AAK, VS, PP, DS, ES, SA, PG. Final approval of the article: all authors. Statistical analysis: LJ, AAK. Obtained funding: LJ, AAK. Overall responsibility: AAK.