Sex-Based Assessment of Patient Presentation, Lesion Characteristics, and Treatment Modalities in Patients Undergoing Peripheral Vascular Intervention
Background—Limited evidence suggests that women and men might be treated differently for peripheral arterial disease. This analysis evaluated sex-based differences in disease presentation and its effect on treatment modality among patients who underwent endovascular treatment for peripheral arterial disease.
Methods and Results—Using national registry data from the Vascular Quality Initiative between 2010 and 2013, we examined patient, limb, and artery characteristics by sex through descriptive statistics. We studied 26 750 procedures performed in 23 820 patients to treat 30 545 limbs and 44 804 arteries. Women presented at an older age (69 versus 67 years; P<0.001) and were less often current or former smokers (72% versus 85%; P<0.001). Transatlantic Inter-Society Consensus classification was similar among men and women (Transatlantic Inter-Society Consensus C or D: 37% in men versus 37% in women; P=0.81), as was mean occlusion length (4.5 cm in men versus 4.6 cm in women; P=0.04), even after accounting for lesion location. Women more frequently underwent treatment for rest pain (11% in men versus 16% in women; P<0.001) versus claudication (59% in men versus 53% in women; P<0.001) or tissue loss (28% in men versus 27% in women; P=0.75). Treatment modality did not differ by sex but was associated with disease severity (P for trend <0.001) and lesion location (P for trend <0.001).
Conclusions—Women undergo peripheral endovascular intervention for peripheral arterial disease at an older age with critical limb ischemia. Treatment modalities do not vary by sex but are determined by disease severity and site. Although there exist sex differences in presentation, these differences do not lead to differential treatment for women with peripheral arterial disease.
WHAT IS KNOWN
Studies suggest that women diagnosed with peripheral arterial disease are treated at an older age, present with advanced multilevel disease, and have poorer overall health.
Yet these investigations have limited descriptions of clinical indications, treatment modalities, and disease distribution by sex in addition to being underpowered to detect sex differences.
WHAT THE STUDY ADDS
Using a large, national vascular registry of which 42% are women, this study:
Confirms that women present at an older age with more advanced disease;
Finds no sex differences in disease characteristics, including lesion length, occlusion length, Transatlantic Inter-Society Consensus score, and the type of artery treated;
Identifies no sex differences in treatment type instead, treatment modality driven by Transatlantic Inter-Society Consensus score and artery treated.
Elderly patients (age ≥65 years) constitute the fastest growing segment of the US population, and women outnumber men in this demographic group.1–3 As the population ages, the prevalence of peripheral arterial disease (PAD) will create substantial financial and resource burdens on the healthcare system.4 Moreover, according to the 2010 United States Census, the prevalence of PAD is higher in women than men.2 Yet PAD remains under-recognized and undertreated among women.2,5 Previous studies suggest that compared with men, women are treated at an older age, present with advanced multilevel disease,1 have poorer overall health status, including more severe symptoms and lower quality of life, and are at increased risk for morbidity and mortality.6
Solitary investigations examined several aspects of sex bias in therapy, but these were not the primary objective of investigations and were likely underpowered to determine sex-based differences.6,7 Furthermore, women remain an under-represented minority in most PAD studies, and many investigations have limited descriptions of clinical indications, treatment modalities, and disease distribution by sex.2,4,7 These factors were highlighted by the American Heart Association’s call to evaluate sex-based differences in the presentation of PAD with a goal to identify the potential variation in treatment approaches, benefits and harms among men compared with women.8
To better understand sex-based treatments for PAD, we sought to investigate differences in patient presentation, disease distribution, and treatment modalities among men and women undergoing peripheral endovascular intervention (PVI) for PAD. We used the Society for Vascular Surgery Vascular Quality Initiative data set to analyze these characteristics among men and women.
Established in 2011, the Vascular Quality Initiative prospectively collects data on >120 patient and procedure-level variables for commonly performed major vascular procedures, such as PVI, at >350 participating centers nationwide.9 For this project, we queried the Vascular Quality Initiative PVI data set and abstracted all relevant patient and procedure variables. The data and analytic methods are available to other researchers pending approval by the Research Advisory Committee at Vascular Quality Initiative.
Establishing the Cohort
Patients who underwent PVI from January 1, 2010, to December 31, 2013, were eligible for inclusion in our analysis. There were 35 199 eligible PVI procedures performed in 28 173 patients. We excluded procedures with missing information on sex, indication, side treated, and artery treated (n=2469; 7%). We also excluded interventions on a previously treated artery (n=2405; 7%), aneurysmal disease pathology (n=339; 1%), and acute limb ischemia as indication for treatment (n=3113; 9%). After applying these criteria, we had 26 873 procedures performed in 23 940 patients, 30 668 limbs, and 44 927 arteries (Figure 1). For analysis, we further excluded cases in which only the aorta was treated (n=123) because this is a different subgroup of pathology within our population of peripheral vascular disease. The final cohort for analysis included 26 750 procedures performed in 23 820 patients, 30 545 limbs, and 44 804 arteries.
The primary exposure was sex (men versus women). As shown in Figure 1, we analyzed our cohort in 3 different ways: by patient, by limb, and by artery treated. At the patient level, we studied sex-based differences in demographics and comorbidities. At the limb level, we examined the history of prior procedures, indication for treatment (claudication versus rest pain versus tissue loss), and unilateral versus bilateral treatment. Finally, at the artery level, we studied sex differences in type and number of arteries treated, lesion characteristics, and treatment modality used for each artery treated. Tibial arteries (anterior tibial, posterior tibial, peroneal, and tibioperoneal trunk) were grouped together for analysis given their individual low frequency of treatment.
Descriptive univariate statistics were conducted for all characteristics, stratified by sex. We also present absolute standardized differences (d) between men and women, which is a measure of the effect size (difference in means or proportions divided by the SD). We used d>0.1 as an threshold for imbalance, or significant difference, between the groups.10 Statistical significance was determined using the χ2 test of independence and Student t test for continuous variables. Because the focus of this article is in the description of current practice, we did not use a formal method to account for multiple testing and present the uncorrected Pvalues. Although we refer to P<0.05 as a threshold for statistical significance, we focused on absolute difference of frequencies to identify significant and clinically meaningful findings. Linear trend analysis for categorical characteristics was conducted using simple logistic regression. All statistical analyses were performed using Stata 13.1 software (College Station, TX). Our study was designated as institutional review board exempt by the Center for the Protection of Human Subjects at the Geisel School of Medicine at Dartmouth.
Women comprised 42% of our cohort. The mean age of patients was 68 years. Unilateral limb treatment was performed in 84% of patients, and right and left limbs were treated with equal frequency. The aorta was primarily treated in conjunction with bilateral limb therapy (60% of all aortic treatments). The proportions of interventions at each arterial level were as follows, listed in descending order of frequency: superficial femoral artery (29%), common iliac artery (22%), popliteal artery (14%), external iliac artery (13%), anterior tibial artery (6%), tibioperoneal trunk (4%), posterior tibial artery (4%), peroneal artery (4%), common femoral artery (4%), profunda femoris artery (1%), and aorta (1%).
Characteristics at Presentation
Most demographic characteristics at time of the index procedure were similar for men and women (Table 1). However, women presented at an older age than men (69 versus 67 years; P<0.001; d=0.23) and were more likely to be black (15% versus 11%; P<0.001; d=0.14). Furthermore, women were less likely to have a history of coronary disease (26% versus 32%; P<0.001; d=0.14), less likely to be ambulatory (77% versus 82% ambulatory; P<0.001; d=0.11), and were more commonly nonsmokers (28% versus 15%; P<0.001; d=0.31). In terms of medications, women were less likely to be on a statin before the procedure (64% versus 70%; P<0.001; d=0.12).
Men and women differed slightly by indication for treatment (Table 1). Most interventions were performed for claudication, namely 59% and 53% of limbs treated in men and women, respectively (P<0.001). Rest pain was more prevalent in women, accounting for 16% of limbs treated among women compared with only 11% among men (P<0.001). Endovascular outflow procedures of the ipsilateral or contralateral limb were the most common previous interventions (15% of the cohort), but there were no meaningful differences in any prior procedure between men and women.
Number of Arteries Treated
The number of arteries treated per procedure ranged from 1 to 6, with a median number of 1 artery treated per limb. By sex, 54% men versus 51% women had a single artery treated (P<0.001), 31% men versus 34% women had 2 arteries treated (P<0.001), and 15% men versus 15% women had ≥3 arteries treated (P=0.71).
When we assessed the relationship between the number of arteries treated and indication, we noted among unilateral limb treatment (Figure 2A) a minor sex difference in rest pain patients, where 43% of women had ≥2 arteries treated compared with 39% in men (P=0.04). However, men and women with claudication or tissue loss had a similar number of arteries treated. In bilaterally treated limbs (Figure 2B), there was a slight sex difference in number of arteries treated for claudication (1 artery treated: 77% in men versus 74% in women; P=0.01), but no significant difference in those treated for rest pain or tissue loss. Taken in aggregate across different indications for intervention, the number of arteries treated was not associated with sex.
Disease severity, however, was significantly associated with the number of arteries treated. Among all patients who had bilateral treatment, increasing severity from claudication to rest pain to tissue loss lead to a significant increase in multilevel disease treatment (2+ arteries treated; P for trend <0.001) and significant decrease in single artery treatment (P for trend <0.001).
Distribution of Arteries Treated
Men and women had a similar distribution of arteries treated (Figure 3A). Figure 3B shows the proportion of treatments performed in men versus women at each artery level. Except for the aorta, the number of interventions at each artery level was higher in men than women. When calculating rates of treatment at artery levels, women were more likely to receive treatment in the aorta (risk ratio [RR]=2.0; 95% confidence interval [CI], 1.6–2.5), common iliac artery (RR=1.2; 95% CI, 1.1–1.2), superficial femoral artery (RR=1.03; 95% CI, 1.00–1.06), and popliteal artery (RR=1.1; 95% CI, 1.0–1.1). Women were less likely to have treatment in the external iliac artery (RR=0.9; 95% CI, 0.8–0.9) and tibial arteries (RR=0.8; 95% CI, 0.76–0.82). Women and men were equally likely to receive treatment in the common femoral artery (RR=0.97; 95% CI, 0.9–1.1) and profunda femoris artery (RR=0.9; 95% CI, 0.7–1.0).
We also studied sex differences in the distribution of common combinations of arteries treated at the limb level while accounting for unilateral versus bilateral treatment (Table I in the Data Supplement). We found no significant differences by sex.
Lesion characteristics included Transatlantic Inter-Society Consensus (TASC) score, treated length, occlusion length, and percent of lesions that were occluded (Table 2). Except in tibial arteries, TASC A lesions were the most commonly treated lesions, comprising between 30% of superficial femoral artery lesions treated and 56% of profunda femoris artery lesions treated. In the tibial arteries, TASC D lesions were the most common, accounting for 37% of all tibial artery lesions treated. Overall, men and women presented with similar TASC score lesions. Similarly, the mean occlusion length was similar in men and women (4.5 cm in men versus 4.6 cm in women; P=0.04). In general, at each artery level, men and women demonstrated similar lesion characteristics, including TASC score, treatment length, occlusion length, and frequency of occlusion (Table 2).
Treatment Types Delivered
Treatment types studied included percutaneous transluminal angioplasty, self-expanding stent, balloon-expandable stent, stent graft, atherectomy (including laser atherectomy, orbital atherectomy, and excisional atherectomy), and other treatments (including cryoplasty and cutting balloon angioplasty). Across indication and artery treated, men and women received the same treatment modality, with few exceptions (Figure 4). For example, in superficial femoral arteries treated for claudication, 41% were treated with a self-expanding stent in men versus 36% in women (P=0.04). See Table II in the Data Supplement for detailed results.
Disease severity and artery treated seem strongly associated with treatment type delivered (Figure 4). Descending from the common iliac to the tibial arteries, there is a significant linear trend in the increase of percutaneous transluminal angioplasty (P for trend <0.001) atherectomy (P for trend <0.001) and other treatment modalities (cutting balloon angioplasty or cryoplasty; P for trend <0.001) across all indications. Likewise, descending from the common iliac to tibial arteries, there is a significant linear trend in the decreased use of self-expanding stent (P for trend <0.001), balloon-expandable stent (P for trend <0.001), and stent graft (P for trend <0.001) across all indications.
The treatment of peripheral vascular disease in women has been suggested to present sex-specific challenges in technique and therefore initial approach.7 For example, some analyses found that women hospitalized for PAD are less likely to undergo a procedure compared with men.11 Furthermore, female sex has been associated with worse outcomes after endovascular and open vascular procedures.12,13 Yet differences in clinical and demographic factors have not been able to completely account for this noted sex bias in initial treatment presentation or treatment approach.6 We, therefore, sought to explore the underlying sex-based differences in the clinical presentation and endovascular treatment of patients with PAD using a large, validated quality improvement study cohort to address the lack of power and variation in information on patient, limb, and artery-specific data inherent to the current published literature.
We provide a brief overview of the most relevant sex-based studies in PAD patients in Table 3. As shown, there remains a fair amount of heterogeneity in the conclusions drawn while the limitations of each study make it difficult to draw any global conclusions in meta-analysis. This paucity of data leads to the current lack of consensus in the literature on sex-based differences among patients with PAD. Some of this conflict may be because of the pervasiveness of single-center, small sample studies with limited external validity on which prior literature conclusions are drawn.
Our study was designed to systematically assess sex-based variation in PVI considering these shortcomings with the aim to support, supplant, or expand the current knowledge base. Our findings add to this current literature in several important ways (Table 4). First, regarding the initial clinical patient presentation, we found few meaningful sex-based differences in terms of comorbidities or history of previous interventions. As such, our study reaffirms previously reported literature results. It seems consistent across ours and other authors’ work that women more commonly present at an older age with more advanced disease, such as ischemic rest pain versus claudication, when compared with men. We also noted that men more commonly smoke or previously smoked compared with women. However, although current literature was inconclusive on sex differences in comorbidities such as hypertension, our much larger cohort showed no such differences. It is unclear whether socioeconomic or genetic factors are attributable in this consistent discrepancy in initial presentation.
Second, in terms of disease characterization, few single-center studies4,14 briefly addressed sex differences in lesion characteristics. As listed in Table 4, they suggest that men more commonly have tibial disease while women more commonly have iliac disease although others report no difference in TASC score or lesion length. Yet these analyses did not account for disease severity or disease location, that is, the type of artery that was treated, which may ultimately bias results. Our study was able to account for these confounders and noted no sex differences in disease characteristics on lesion length, occlusion length, or disease distribution as measured by TASC score or the type of artery treated.
Third, sex differences in endovascular treatment modality used in PVI have not been previously addressed in the literature (Table 4). Our meticulous analysis comparing various treatment types by sex found that there were no significant differences in men compared with women. However, we found a clear and persistent trend in the increasing use of percutaneous transluminal angioplasty with a concomitant decline in the use of stents or atherectomy as the treated arteries became more distal. This finding objectively confirms a common notion that smaller caliber vessels are less likely to receive stent placement or atherectomy and are more likely treated with percutaneous transluminal angioplasty only. Thus, rather than sex, the disease location and distribution were the major determinant of treatment type. Future work will examine how these decisions contribute toward treatment outcomes in PVI.
In addition, we identified a significant association between indication for treatment, or disease severity as assessed by rest pain versus tissue loss versus claudication, and the number of arteries treated. Similarly, disease severity was also strongly linked to the treatment modality used. In essence, the worse the disease process, the more aggressive the treatment as described by number of arteries treated and treatment modality used. The interaction of these characteristics has not been previously reported but seems often intuitive in the decision-making process about limb ischemia treatment.
This article serves as a real-world description of practice patterns among 24 000 men and women undergoing PVI across the United States. Although this cohort lacked significant representation from all racial and ethnic groups because most patients were white (83%) and non-Hispanic (96%), this might be the true representation of patients receiving PVI treatment and is in and of itself an important finding in health disparities studies. Furthermore, women represented a substantial (42%) proportion of the cohort, which is an uncommon strength in vascular disease studies.
Reported frequencies were not adjusted for patient risk factors given the descriptive nature of this project. To address potential confounding, we stratified by indication and artery type treated where appropriate. In addition, because this is an observational cohort study, causality cannot be assumed in proposed associations.
Missing data are limitations inherent to large registries. Any procedures that were missing critical information, such as sex, indication, and artery type treated, were excluded from the outset but comprised a small proportion of the cohort. No variables analyzed were missing >5% of observations except TASC score, which was missing in ≈10% of artery-level observations. To address this limitation, a stratified analysis of procedures by missing versus completed TASC classification was conducted for sex, indication, and artery type treated for which there was no statistically significant differences.
In summary, we identified that women clinically present at an older age with more advanced arterial disease but are less likely to have comorbid coronary artery disease or be actively smoking. These differences could be because of a varied disease pathophysiology or represent limitations to healthcare access, among others, for women compared with men and thus warrant further study. For instance, our future endeavors will analyze the associations between sex disparities in PAD treatment and postprocedural outcomes with emphasis placed on regional variations and socioeconomic status as predictive variables.
Men and women who undergo endovascular treatment for PAD in our national registry are largely equivalent in terms of disease characterization and treatment modality. Sex-based disease characterization in terms of number of arteries treated and artery type treated seems to be driven by indication. Likewise, treatment modality seems associated with indication and artery type treated, rather than sex. Further investigation is required to appreciate the effect of disease severity and treatment modality on patient outcomes after PVI.
We are grateful to Dr Todd Mackenzie for his statistical expertise and support.
Sources of Funding
This work was supported by the US Food and Drug Administration (FDA) Office of Women’s Health. This project was supported, in part, by an appointment to the Oak Ridge Institute for Science and Education (ORISE) Research Participation Program at The Dartmouth Institute for Health Policy and Clinical Practice, FDA, administered by ORISE through an interagency agreement between the US Department of Energy and FDA.
N. Ramkumar, Dr Goodney, and. Dr Sedrakyan were supported by a grant from the Food and Drug Administration (FDA grant U01FD005478-01; Dr Sedrakyan, principal investigator) as part of The Vascular Implant Surveillance and Interventional Outcomes Network. N. Ramkumar and Dr Suckow were supported by an Oak Ridge Institute for Science and Education fellowship from the FDA. The other authors report no conflicts.
The Data Supplement is available at http://circinterventions.ahajournals.org/lookup/suppl/doi:10.1161/CIRCINTERVENTIONS.117.005749/-/DC1.
- Received July 28, 2017.
- Accepted November 29, 2017.
- © 2018 American Heart Association, Inc.
- Lo RC,
- Bensley RP,
- Dahlberg SE,
- Matyal R,
- Hamdan AD,
- Wyers M,
- Chaikof EL,
- Schermerhorn ML
- Grenon SM,
- Cohen BE,
- Smolderen K,
- Vittinghoff E,
- Whooley MA,
- Hiramoto J
- Fowkes FG,
- Rudan D,
- Rudan I,
- Aboyans V,
- Denenberg JO,
- McDermott MM,
- Norman PE,
- Sampson UK,
- Williams LJ,
- Mensah GA,
- Criqui MH
- Hardman RL,
- Jazaeri O,
- Yi J,
- Smith M,
- Gupta R
- Dreyer RP,
- van Zitteren M,
- Beltrame JF,
- Fitridge R,
- Denollet J,
- Vriens PW,
- Spertus JA,
- Smolderen KG
- Hirsch AT,
- Allison MA,
- Gomes AS,
- Corriere MA,
- Duval S,
- Ershow AG,
- Hiatt WR,
- Karas RH,
- Lovell MB,
- McDermott MM,
- Mendes DM,
- Nussmeier NA,
- Treat-Jacobson D
- Brahmbhatt R,
- Brewster LP,
- Shafii S,
- Rajani RR,
- Veeraswamy R,
- Salam A,
- Dodson TF,
- Arya S