Author + information
- Received August 3, 1996
- Revision received March 31, 1997
- Accepted April 17, 1997
- Published online August 1, 1997.
- ↵*Dr. Kenneth G. Lehmann, Section of Cardiology (111C), Veterans Affairs Puget Sound Health Care System, 1660 South Columbian Way, Seattle, Washington 98108.
Objectives. This study investigated the efficacy of four different methods of arterial puncture site management during recovery from invasive cardiac procedures. The primary goals were less patient discomfort and improved clinical outcome.
Background. The increasing use of outpatient catheterization, large interventional devices and potent periprocedural anticoagulation regimens has made the reduction of groin complications a high priority. Despite these trends, there are no randomized trials comparing commonly used techniques in treating the catheter entry site for the first few hours after the procedure.
Methods. Four-hundred consecutive patients undergoing catheterization laboratory procedures were randomly assigned to one of four dressing techniques applied after achieving hemostasis: a sandbag placed over the site; a pressure dressing constructed from surgical gauze and elastic tape; a commercially available compression device; and no use of compressive dressing. Of these 400 patients, 171 would have been eligible for outpatient procedures in the absence of geographic constraints. The dressings were removed, and ambulation was encouraged 5 h after sheath removal. Uniform initial compression times, patient instructions, nursing follow-up and a structured interview and physical examination at 24 h were used.
Results. The level of patient discomfort before and after dressing removal, as well as site tenderness at 24-h follow-up, was statistically similar in all four groups. Hematomas (typically small) and areas of ecchymosis were observed in 58 and 122 patients, respectively, but both their frequency and size were equally represented in each group. Important adverse events were confined to bleeding, rated as mild in 5.8%, moderate in 0.8% and severe in 0.6% of patients. Again, all four groups were statistically similar. Comparable findings were observed in the subgroup of patients eligible for outpatient procedures.
Conclusions. Despite an increase in inconvenience and expense, none of the three compression techniques that were investigated improved patient satisfaction or outcome. Therefore, the routine use of compression dressings after invasive cardiac procedures cannot be recommended.
Cardiac catheterization and related procedures continue to increase in frequency of use, safety and performance in an outpatient setting. The majority of diagnostic catheterizations can be performed in the modern era without sedation, and recovery of most organ systems is nearly immediate. However, the universal requirement of arterial access substantially prolongs the requisite period of immobility and monitoring and remains the single most important impediment to early ambulation and discharge. Moreover, a significant portion of the costs associated with these procedures is expended during this recovery period.
A number of attempts have been made to minimize the impact and complications of this obligate arteriotomy. Over the past decade the size of diagnostic and therapeutic catheters has steadily decreased, allowing a smaller rent in the arterial wall [1–5]. Another advocated approach is to gain arterial access through a site less stressed by ambulation, such as the brachial, radial or axillary arteries [6–10]. Widespread application has been slow to catch on, though, as these upper extremity vessels are less conveniently accessed in most catheterization laboratories and can limit the size of the devices used. Collagen plugs and modified surgical closure devices are undergoing premarket and aftermarket testing in an attempt to speed initial hemostasis [11–16], but whether this accelerated hemostasis will translate to earlier recovery remains unclear.
Nearly all previous efforts at promoting early ambulation have focused on techniques related to sheath insertion or withdrawal. In contrast, scant attention has been paid to speeding ambulation, reducing complications and limiting pain and inconvenience through techniques that exert the most influence after the patient leaves the catheterization laboratory. The current randomized study seeks to discover whether any of four potential methods of managing the arteriotomy site for the first few hours after catheterization will prove superior. This comparison is facilitated by close monitoring both before and after ambulation and by the use of a structured interview and physical evaluation 24 h after the procedure.
1.1 Patient group.
Subjects were selected from consecutive patients undergoing invasive arterial procedures in the cardiac catheterization laboratory. The following exclusion criteria were applied: 1) arterial access at a site other than the right or left femoral artery; 2) emergent procedure; 3) arterial puncture at the same femoral site within the previous 18 h; 4) continued use or effect of warfarin, heparin, thrombolytic agent, ticlopidine, abciximab or other nonaspirin anticoagulant agent at the time of the procedure; 5) vascular perforation, thrombosis or hematoma formation occurring during the procedure; 6) timing of the procedure that would preclude staff availability over the following 24 h; and 7) unwillingness or inability to provide written, informed consent. Overall, 400 patients were enrolled. Three were terminated before follow-up because of inappropriate enrollment (hematoma present before sheath removal). In addition, follow-up data were incomplete in 38 patients (often because of unanticipated early discharge); these individuals were included in all analyses except for outcomes. Power calculations were done to ensure an adequate sample size for the detection of clinically meaningful differences between groups. Based on a two-tailed alpha value of 0.95, the power (1 − β) of the study to detect a one-point difference in pain scale was 0.95; a 2.5-min difference in time to hemostasis was 0.97; a 20% difference in the proportion of patients with rebleeding was 0.99; and a 30% difference in the proportion of patients reporting site tenderness was 0.99.
1.2 Study protocol.
The overall study consisted of a prospective, randomized trial of arteriotomy site management based on a three-by-four factorial design. The first phase consisted of randomized assignment to one of three techniques used to achieve initial hemostasis (manual hold, clamp and inflatable compression device). In the second phase, representing the subject of this report, we studied four different methods of maintaining hemostasis outside of the catheterization laboratory, referred to as “dressing technique.” Randomization was accomplished using a computer-generated random number sequence and a sealed envelope system of assignment. The four groups are described below, with treatment initiated immediately after achieving adequate hemostasis after sheath removal: 1) “sandbag”—a 4.5-kg (10-lb), 32-cm × 12-cm × 8-cm sandbag placed directly over the site. 2) “Pressure dressing”—six 10 × 10-cm, 12-ply, sterile gauze sponges folded once both vertically and horizontally (final dimensions = 5 × 5 cm) positioned directly over the site. This was held in place by 30-cm strips of 8-cm wide elastic adhesive bandage (Elastoplast, Belerdorf, Inc.). Two strips were stretched from the iliac crest to the ipsilateral inner thigh, and a third strip was placed perpendicularly from the symphysis pubis to the midline of the lateral thigh. The dressing was applied with the hip flexed and externally rotated ∼20° to increase pressure on the arteriotomy site on leg straightening [17, 18]. 3) “Compression device”—a commercially available product (HOLD device, Pressure Products, Inc.) that consists of an 8-cm diameter hemispheric polystyrene disk held under tension against the arteriotomy site by a belt system extending around both the waist and the upper thigh. 4) “None”—the site was not treated with any compression dressing. All dressings were scheduled for removal 5 h after sheath removal, and ambulation was started shortly thereafter. The protocol was approved by the Human Studies Committee of the University of Washington, Seattle.
1.3 Study procedures.
Vascular access sheaths were used in every patient, ranging in size from 5F to 10F. Patients undergoing diagnostic procedures were typically treated with 2,000 U of intravenous heparin immediately after sheath insertion. A heparin bolus dose of 10,000 to 15,000 U was used for interventional procedures and was followed by repeated boluses as needed to maintain an activated clotting time of 250 to 350 s. No patient received protamine sulfate for reversal of anticoagulation. The sheaths were removed immediately after the diagnostic procedures. For interventional procedures, sheath removal occurred 4 to 20 h after completion, and after discontinuation of heparin and documentation of an activated clotting time <150 s. Before application of the randomly selected dressing, each arterial access site was carefully inspected for evidence of hematoma formation or other vascular problems, and measurements of size were made as appropriate. All patients received a small, sterile adhesive strip (Band-aid) to cover the puncture site before placement of the compression dressing.
Uniform verbal and written instructions of immobility were given to each patient at the time of transfer from the catheterization laboratory. Written nursing orders included frequent vital sign checks and bed rest for 5 h with the head of the bed elevated <30°. As part of the study, the nursing team caring for the patient made a special effort to document bleeding, discomfort and deviations from the prescribed activity plan. Five hours after sheath removal all dressings were removed, and all medically able patients were strongly encouraged to ambulate at a level commensurate with their normal activities at home. Approximately 24 h after sheath removal, all patients were interviewed and examined by a single observer. Discomfort was rated on a 1 to 10 scale, where 1 = barely noticeable discomfort and 10 = maximal degree of pain the patient has experienced or could imagine. A hematoma was recorded as present if focal induration >1 cm in diameter was palpated at the puncture site, with the size computed as an elliptical surface area using orthogonal measurements as major and minor axes. The size of any ecchymotic area was similarly measured and computed. Arterial occlusion, arteriovenous fistula and pseudoaneurysm formation were recorded as a major adverse event if surgical correction was required within 1 year of the original procedure.
1.4 Data analysis.
All analyses were performed on an intention-to-treat basis. Data are presented as mean value ± SD. Differences in categoric variables were compared using the chi-square test, and differences in continuous variables were compared using the Kruskal-Wallace analysis of variance. A probability value of 0.05 was accepted as the limit of statistical significance.
2.1 Baseline characteristics.
The clinical characteristics for the study group (n = 397) are presented in Table 1. Most were men with a mean age of 61 years. Previous myocardial infarction was present in 39% of patients. Risk factors for atherosclerosis were common, including hypertension in 53%, diabetes in 25%, a history of current or past smoking in 52% and a history of hypercholesterolemia in 36% of patients. The prevalence of peripheral vascular disease (bruits, claudication and/or previous vascular surgery) was 27%. Importantly, these characteristics were evenly divided between the four study groups.
Hemodynamic and angiographic variables were also equally distributed between the groups. Central systolic and pulse pressures, representing hemodynamic forces that might influence arteriotomy site rebleeding, averaged 128 and 59 mm Hg, respectively. Overall, 29% of enrollees lacked significant angiographic coronary artery disease, most of whom received invasive evaluation for valvular or other forms of heart disease. Because of its hyperdynamic effect on blood pressure, the presence of aortic regurgitation can make the achievement and maintenance of hemostasis more difficult; this condition was present in five patients (two in the “pressure dressing” group and one each in the other three groups).
2.2 Procedure-related variables.
The distributions of procedure-related variables are shown in Table 2. The arterial sheath diameter ranged from 5F to 10F, with 7F the most frequent size used. The sheath was inserted directly through prosthetic vascular bypass graft material in a small number of patients (7%) . The majority of procedures (92%) were diagnostic, with the remainder interventional. After the procedure, 68% of patients returned to the cardiology ward and 32% to an intensive care unit bed.
As part of the study protocol, a target minimal arterial compression time of 13 min after sheath removal was established. The actual mean time to hemostasis in each group ranged from 14.3 to 14.9 min. Before the dressing could be applied, minor rebleeding was noted in 6.3% of patients, and small hematomas (not present before sheath removal) were observed in 11.1%. Again, there were no statistically significant differences between the groups with regard to any of these variables.
All patients who were medically able were encouraged to ambulate as they would at home starting 5 h after sheath removal. Of the 287 patients (80%) who did, their level of activity at 24 h was self-described as mild in 72%, average (normal daily activity) in 25% and strenuous in 3%. In general, each of the dressing techniques used was well tolerated. The mean level of discomfort (on a scale of 1 to 10) during dressing use was 1.8 ± 1.8 for the “none” group, 1.7 ± 1.6 for the “sandbag” group, 1.7 ± 1.5 for the “pressure dressing” group and 1.8 ± 1.4 for the “compression device” group (p = 0.69). The mean discomfort level in each group remained low after the dressing was removed, both before (p = 0.46) and during (p = 0.80) ambulation (Fig. 1).
A physical examination at 24 h showed site tenderness to be both uncommon (31%) and, when present, mostly mild (77%). Severe tenderness was found in only five patients (three of whom were in the “pressure dressing” group). A new hematoma at the site was recorded in an average of 12% of patients in each group, covering a mean surface area of 6.9 ± 18.4 cm2(Fig. 2). Ecchymoses were more common—found in 37%, 26%, 36% and 38% of the four groups (p = 0.25), as shown in Table 3and Fig. 3.
Adverse events occurred in a total of 26 patients. This consisted of mild rebleeding in 21 (5.8%), moderate rebleeding in 3 (0.8%), and severe rebleeding in 2 (0.6%). Although rebleeding was least common in the “pressure dressing” group, the rates of rebleeding were not statistically different between the groups. Arterial occlusion was not observed, and surgical vascular repair was not required in any patient within 1 year of the catheterization procedure.
Taken together, these findings show no benefit, either subjective or objective, with the use of any of the three compressive dressing techniques.
3.1 Relation to outpatient catheterization.
At the time of the study, all catheterization procedures undertaken in our laboratory were performed on an inpatient basis, due in large part to the four-state catchment area served by the laboratory. This provided a unique opportunity to study and closely monitor many patients who would be eligible for outpatient catheterization in the absence of geographic constraints. Patients were continually available for monitoring and evaluation for at least 24 h after their invasive procedure, maximizing the likelihood that suboptimal results and adverse outcomes would be fully detected and appropriately assessed. As part of this protocol, all participants were encouraged to ambulate 5 h after sheath removal at a level commensurate with their anticipated activity at home, had they been discharged. Overall, 287 patients (80%) did successfully ambulate. Of these, 6 (2%) experienced some bleeding after they started to ambulate, but there were no differences between the groups in the frequency of late bleeding. In addition to the primary analyses presented earlier, we assessed outcomes in the subgroup of enrollees who might be considered for outpatient procedures (no unstable angina or interventional procedures) . The results obtained from these 171 individuals were the same as for the overall study group, except for a slightly lower frequency of hematoma formation with the use of a sandbag (offset by a slightly higher frequency of hematoma formation with sandbag use in nonambulatory patients). Thus, our findings of a lack of improved efficacy with any of the compression techniques studied appear applicable to outpatient procedures as well.
3.2 Importance of puncture site management.
Since its inception in 1950 , the use of left heart catheterization has expanded greatly in frequency, scope and safety. These improvements are attributable in large part to the tremendous advances in techniques and devices used for this and related procedures. However, over the same period little has changed in the approach to postprocedural arterial puncture site management. To date, no randomized trial has been published that investigates this universal problem. As a result, considerable differences can be found in the recommended approach to this problem, detailed in standard texts [22–25]. The most common recommendation, though, involves puncture site compression with a sandbag or pressure dressing for several hours after the procedure .
This topic is gaining importance for a number of reasons. First, the last decade has witnessed a major shift worldwide from inpatient to outpatient catheterization. These patients must be able to ambulate quickly and safely. Second, early discharge for inpatients undergoing an intervention has been promoted and practiced as a way to control medical expenses. Third, the introduction of certain procedures such as directional atherectomy, intraaortic balloon counterpulsation and percutaneous cardiopulmonary bypass has led to an increase in arterial sheath size used in many patients [26, 27]. Finally, there is a growing trend for the in-laboratory use of potent anticoagulation agents such as ticlopidine and abciximab, not only for coronary stenting but also during routine balloon angioplasty procedures . As a result of these trends, the rates of vascular complications have become a principal focus of several major interventional trials .
3.3 Potential advantages of compression dressings.
Each of the three compression dressings investigated in this trial has unique theoretic advantages. Sandbags are simple to apply and encourage leg immobility by their bulk, but they only exert diffuse (and therefore minimal) pressure directly on the puncture site and tend to easily slip off the groin site. Pressure dressings exert considerable direct pressure and are positionally stable, but they take time to construct, can result in skin damage at points of attachment of the adhesive strip and can hinder visual inspection of the puncture site. The commercially available compression device addresses some of the shortcomings of the pressure dressings, but it has the disadvantages of higher expense and an easily soiled support belt system.
None of these shortcomings is of sufficient magnitude to preclude the use of compression dressings if clinical benefit results. However, based on the results of the current study, none proved statistically superior, in terms of clinical outcome or patient discomfort, to the use of no compression dressing at all.
3.4 Potential study limitations.
Several possible limitations of this trial deserve mention. First, the combination of an intermediate sample size with a low expected frequency of serious adverse events precludes any statistically meaningful comparison of the rate of major vascular complications between the groups. Based on an estimated incidence of 1% [29–35], power studies (using an alpha value of 0.05 and a beta value of 0.10) suggest that 153,000 patients would be needed to reliably detect a 25% intergroup difference in these complications. Obviously this sample size would have represented an unrealistic target for this trial, and will likely be unattainable in future studies in this area. Second, the most common size of arterial sheaths in this study was 7F. Most outpatient catheterizations done currently in our and other centers routinely use 6F or 5F sheaths. Although this could potentially affect the absolutelevel of discomfort and adverse events reported in this study, it is likely that a comparison of the relativedifferences between compression dressing groups will remain valid for smaller sheath sizes as well. Hence, our findings should retain their applicability to the outpatient as well as inpatient setting. Finally, we have attempted to standardize all variables possible, including initial hold times, duration of dressing use, timing and extent of ambulation, patient instructions, evaluation of puncture site before dressing application, quantitation of hematomas and areas of ecchymosis and heparin use. Nevertheless, some unmeasured confounding variable could have skewed our results. The use of random treatment assignment, however, should have minimized this potential error.
3.5 Clinical implications.
All of the three dressing techniques investigated in this randomized trial resulted in impaired visualization of the arteriotomy site, greater use of catheterization laboratory time, higher expense and increased inconvenience to both the physician and patient. These marginal disadvantages could be offset by a decrease in discomfort or a lower complication rate. However, none of the three techniques investigated proved statistically superior to no dressing at all. Therefore, the routine use of arterial puncture site compression techniques for either inpatient or outpatient invasive cardiac procedures cannot be recommended.
We gratefully acknowledge the expert technical assistance provided by Daryl Jones, CPT, Cary Shepherd, CVT, Ruben Flores, RT, Donna Kline, RN, and Susan Gilbert, RN, MSN, as well as the editorial comments provided by James Ritchie, MD.
☆ Financial support was provided by grants from the Research Service of the Department of Veterans Affairs, Washington, D.C.
- Received August 3, 1996.
- Revision received March 31, 1997.
- Accepted April 17, 1997.
- The American College of Cardiology
- Kussmaul WG,
- Buchbinder M,
- Whitlow PL,
- et al.
- Sanborn TA,
- Gibbs HH,
- Brinker J,
- Knopf WD,
- Kosinski EJ,
- Roubin GS
- Ernst SM,
- Tjonjoegin RM,
- Schrader R,
- et al.
- Zimmerman HA,
- Scott RW,
- Becker ND
- Baim DS
- Daily EK,
- Tilkian AG
- Grossman W
- Hill JA,
- Lambert CR,
- Pepine CJ
- Green GS,
- McKinnon CM,
- Rosch J,
- Judkins MP