Varicose veins as a risk factor for venous thromboembolism in arthroplasty patients: Meta-analysis

Introduction

Approximately one third of the world’s population have varicose veins, with patients often reporting swelling and discomfort as the precipitating factors for seeking intervention.^1–5 Leg swelling is also a common complaint of a patient with lower limb degenerative joint disease,^6,7 and elective lower limb arthroplasty is considered one of the most common elective surgical procedures carried out worldwide for severe joint disease. ⁸

Several risk factors for venous thromboembolism (VTE) have been established in recent years and there has been much debate surrounding varicose veins as an independent risk factor for VTE. It is believed that a weak association exists between varicose veins and VTE, however, there is a lack of objective evidence. In contrast to varicose veins as a risk factor, there is a well-established and strong association between VTE occurrence and patients undergoing major lower limb surgery. ⁹ This raises the question, do concurrent varicose veins increase the risk of post-operative VTE in patients undergoing major lower limb arthroplasty. Therefore, the aim of this meta-analysis was to determine if the presence of varicose veins impacts on VTE risk in patients undergoing lower limb arthroplasty.

Methods

Results

The initial search revealed 51 articles in Medline and 127 articles in Embase database and two studies in the Cochrane database. The pool of studies identified in the preliminary searches (180 studies) were filtered by the authors for duplicates. 129 articles were then screened against the inclusion and exclusion criteria described under the methods section. 110 studies were reviewed and excluded based on whether or not they were relevant to the aim of the meta-analysis. 19 studies fit the inclusion criteria and underwent careful review. Seven articles fulfilled the inclusion criteria, with an additional four articles identified from reference lists. 11 studies were finally included for analysis, published from 1981 to 2018 shown on the PRISMA flow chart (Figure 1).

OPEN IN VIEWER

Figure 1.

PRISMA flow chart from initial searches to eligible articles for analysis.

Table 1 Summarizes the eligible 11 studies, consisting of nine cohort studies, one case-control study and one randomized control trial. Data on 265,194 patients was collectively obtained from the included studies. 2,188 patients had pre-existing varicose veins (0.83%). 1,835 patients were diagnosed with VTE post-arthroplasty, 122 of which had concurrent varicose veins (0.046%). The Risk of Bias assessment tool was used to assess for bias. Overall there is moderate to high levels of bias detected across the included studies.

OPEN IN VIEWER

Table 1.

Study characteristics.

Author	Year	Type of study	No. of patients	Varicose veins	VTE overall	VTE and varicose veins	Odds ratio
Russell et al. 10	2018	Cohort	402	18	82	10	1.49 (0.54, 4.08)
Fuji et al. 11	2017	Cohort	36,947	98	191	2	3.3 (0.39, 12.26)
Quarto et al. 12	2016	Case-control	44	44	NS	NS	NS
Wu et al. 13	2014	Cohort	114,026	761	518	14	4.19 (2.45, 7.17)
Dua et al. 14	2012	Cohort	109,223	459	766	8	2.53 (1.25, 5.10)
Markovic-Denic et al. 15	2012	Cohort	499	139	13	NS	3.13 (1.03, 9.48)
Samama 16	2007	Cohort	1080	368	20	3	0.71 (0.12, 2.99)
Leizorovicz et al. 17	2005	Cohort	2420	104	28	4	3.81 (0.94, 11.39)
Wang et al. 18	2000	Cohort	102	15	68	11	1.45 (0.38, 6.75)
Lowe et al. 19	1999	Cohort	375	131	120	49	1.48 (0.92, 2.37)
Ishak et al. 20	1981	RCT	76	51	29	21	1.49 (0.49, 4.74)

Meta-analysis: the effect of varicose veins on VTE risk after major limb arthroplasty

A meta-analysis of 10 included studies was performed to examine if there is an association between varicose veins and VTE risk at the time of major lower limb joint replacement. The odds ratio for each study is included in Table 1. There is moderate heterogeneity across the studies, I² = 55.2%, therefore, random effect model was used for the meta-analysis of odds ratios. The pooled odds ratio is 2.37, 95% CI 1.54–3.63, (p < 0.001). This indicates that having varicose veins at the time of arthroplasty increases the risk of post-operative VTE, depicted in the forest plot (Figure 2).

OPEN IN VIEWER

Figure 2.

Forest plot of odds ratio's, The graph illustrates the VTE risk associated with pre-existing varicose veins in patients who are undergoing lower limb arthroplasty.

Bias

The funnel plot is asymmetrical. Three studies are inside the funnel with the remaining five studies outside of the funnel. This is potentially due to publication bias among the included studies or as a result of small study effect. It is plausible to assume that the small sample size of the selected studies results in the asymmetry in Figure 3 particularly as Eggers test was non-significant, (p = 0.98).

OPEN IN VIEWER

Figure 3.

Funnel plot depicting publication bias among the included studies. Eggers test for asymmetry was non-significant (p = 0.98).

Discussion

Whether varicose veins are an independent risk factor for VTE remains controversial. There is also limited objective evidence relating to varicose veins and an increased risk of post-operative VTE in patients undergoing major lower limb surgery. ⁹ This is a meta-analysis of observational studies, with only one trial, with moderate heterogenicity across the studies. Data on 265,194 patients was analyzed in order to determine if concurrent varicose veins increase the risk of post-operative VTE in patients undergoing major lower limb arthroplasty. Post-operative symptomatic VTE was only identified if patients were symptomatic as an inpatient or represented soon after discharge. There was minimal use of routine venous duplex post-op as part of the study design for the respective studies. Details surrounding varicose veins were lacking. Patients were reported to as having varicose veins, or previous history of varicose vein surgery, at the time of arthroplasty.

Collectively, 2188 patients had varicose veins as per the studies included in this review.^10–22 Overall, the clinical detail surrounding these patients with concurrent venous disease is sparse, with no reference made to the severity of disease, and only one study investigating the effect of previous varicose vein surgery and the effect on VTE risk. The incidence of post-operative VTE in orthopedic patients with varicose veins stated in the relevant studies ranges from 1 to 41%, respectively.^10–22 The tabulated data gives a combined incidence of 0.05% (Table 1). The pooled odds ratio obtained as part of the meta-analysis is 2.37, 95% CI 1.54–3.63, p < 0.001 (Figure 2). This is suggested by all but one of the authors included as part of this study. However, it is important to highlight that although the pooled OR is greater than 1, only two studies have a confidence interval that do not cross the mid-line. In addition to this, there is considerable bias associated with the included studies, of which there is also moderate heterogeneity. The odds ratio should be interrupted with caution and we must not lose sight of the fact that VTE risk is multifactorial with patient specific risk factors all of which should be addressed pre-arthroplasty. All this considered the incidence and risk of post-operative VTE that can be attributed to varicose veins is minor, which is also the opinion of Beksac et al., and Samama.^16,22 The highest thromboembolic risk factor is hip/knee surgical intervention, increasing the VTE risk to 48%–61%. ²³

Given our results, is there evidence to support pre-arthroplasty varicose vein surgery in an attempt to reduce VTE risk. Dua et al. indicated in their study that the VTE in patients who underwent previous varicose vein surgery is less than those who have co-existing varicose veins at time of arthroplasty, concluding that the risk of VTE in patients with previous varicose vein surgery is the same as those with no history of varicose veins. In only one study, Quarto et al. attempted to evaluate prophylactic treatment of the great saphenous vein pre-arthroplasty and had one symptomatic DVT post-arthroplasty, an incidence of 2.3%. The authors make valid arguments that varicose vein surgery is potentially beneficial at reducing VTE risk, however, the results are not statistically significant and more evidence is necessary to come to better conclusions. Lastly, when you compare the overall incidence of VTE from the included studies, 0.05%, to the incidence of DVT post-arthroplasty in patients who underwent prophylactic varicose vein surgery, 2.3%, we are unable to suggest that pre-arthroplasty varicose vein surgery actually reduces VTE risk when it potentially increases the overall risk.

Conclusion

Varicose veins can be considered as a risk factor for post-operative VTE in arthroplasty patients, however they can be considered minor when compared lower limb surgery and other factors. Further evidence is needed to determine if prophylactic varicose vein surgery reduces the risk of VTE in lower limb arthroplasty patients. Finally, the incidence of VTE in patients with varicose veins in this study was 0.05%. Although varicose vein surgery has advanced in recent years, a safe and effective procedure, the incidence of DVT associated is approximately 1%. ²⁴ Compared to the VTE risk of hip or knee surgery, 48–61%, ²³ it remains difficult to justify subjecting asymptomatic patients to a venous treatment for varicose veins in order to reduce a minor thrombotic risk factor. Further studies are necessary in order to determine if patients should have their varicose veins operated on prior to hip or knee replacement.