- Email: [email protected]
Randomized comparison of hemostasis techniques after invasive cardiovascular procedures
Randomized comparison of hemostasis techniques after invasive cardiovascular procedures Kenneth G. Lehmann, MD, Samantha J. Heath-Lange, BS, and Scott T. Ferris, MS Seattle, Wash
Background The arterial access required during most invasive vascular procedures provides a common source of complications and morbidity. This problem has been made worse by recent trends in earlier ambulation and more aggressive antihemostatic drug regimens. Despite these trends, no randomized trials have been reported comparing the 3 most commonly used techniques in achieving hemostasis at the arterial puncture site.
Methods A cohort of 400 patients undergoing catheterization laboratory procedures were randomly assigned to 1 of 3 groups of arterial compression: manual compression, mechanical clamp, and pneumatic compression device. Standard requirements of the trial included uniformity in initial compression times, patient instructions, nursing follow-up, and timing of ambulation as well as a structured interview and physical examination at 24 hours.
Results Prolonged compression was required in 13% of the manual group, 20% of the clamp group, and 35% of the pneumatic group (P < .0001). In-lab bleeding was more common in the pneumatic group (3%, 4%, and 16%, respectively, P < .0001), as was the need for an alternate compression technique (1%, 1%, and 27%, P < .0001). The groups also differed in respect to mean hematoma size (3.9 cm2, 7.8 cm2, and 19.8 cm2, P = .036) and level of discomfort during compression (1.9, 2.2, and 3.1 on a 1- to 10-point scale, P < .0001). Comparable findings were observed in the subgroup of patients eligible for outpatient procedures.
Conclusions Use of the pneumatic compression device leads to longer compression times, greater discomfort, more bleeding, and larger hematomas. Differences between manual compression and the mechanical clamp were more subtle but tend to favor use of the manual technique. (Am Heart J 1999;138:1118-25.)
See related Editorial on page 1014. Since its inception in 1950, cardiac catheterization has grown to encompass more than 1.5 million procedures per year in the United States alone.1 Although the safety and efficiency of this and related procedures continue to improve, the universal requirement of arterial access prolongs the requisite period of monitoring. This factor remains the single most important impediment to early discharge. Several technical modifications have been advocated to help minimize the negative impact of femoral artery puncture. Over the past decade, the size of diagnostic and therapeutic catheters has steadily decreased.2-6 Although more technically challenging than femoral artery cannulation, arterial access can be obtained through a site less stressed by ambulation, such as the From the University of Washington School of Medicine and the Veterans Affairs Puget Sound Health Care System. Supported by grants from the Research Service of the Department of Veterans Affairs. Reprint requests: Kenneth G. Lehmann, MD, Section of Cardiology (111C), VA Puget Sound Health Care System, 1660 South Columbian Way, Seattle, WA 98108. 0002-8703/99/$8.00 + 0 4/1/101219
brachial, radial, or axillary arteries.7-11 Collagen plugs and percutaneous surgical closure devices have been developed to speed hemostasis, but their expense, uncertain efficacy, and continued requirement for at least some external arterial compression have made their use controversial.12-17 Even these potential improvements, however, have been partially offset by recent trends that place greater stress on the healing artery. These include outpatient procedures, early ambulation, larger sheath sizes required by certain procedures such as directional atherectomy, and the increased use of potent antihemostatic agents such as clopidogrel and abciximab.18 This study prospectively explored the relative safety and efficacy of 3 different but commonly used techniques of achieving hemostasis after arterial sheath removal. This comparison was facilitated by randomized treatment assignment, close monitoring before and after ambulation, and the use of structured interview and physical evaluation 24 hours after the procedure.
Methods Patient population Patients were selected from consecutive individuals undergoing invasive arterial procedures in the cardiac catheterization laboratory. The following exclusion criteria were applied:
American Heart Journal Volume 138, Number 6, Part 1
Figure 1
Lehmann, Heath-Lange, and Ferris 1119
Figure 2
Mechanical devices used in this trial.
(1) arterial access at a site other than the right or left femoral artery; (2) emergent procedure; (3) arterial puncture at the same site within the previous 18 hours; (4) continued use or effect of warfarin, heparin, thrombolytic agent, ticlopidine, abciximab, or other nonaspirin anticoagulant 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 during the following 24 hours; and (7) unwillingness or inability to provide written informed consent. A total of 400 patients were enrolled. Three were excluded shortly after enrollment because of evidence of hematoma formation before sheath removal. In addition, incomplete follow-up data in 38 patients (often from unanticipated early discharge) resulted in exclusion of these individuals from assessment of delayed outcomes. Power calculations were undertaken to ensure adequate sample size for the detection of clinically meaningful differences between groups. Based on a 2-tailed α of .05, the power (1 – β) of the study to detect a 20% difference in the proportion of patients having a hematoma develop was 0.95, a 25% difference in the proportion of patients requiring prolonged arterial compression was 0.98, and a 1-point difference in pain scale was 0.99.
Study protocol The overall study was designed as a prospective randomized trial of arteriotomy site management based on a 3 × 4 factorial design. The second phase investigated 4 different methods of maintaining hemostasis after the patient leaves the catheterization laboratory (sandbag, pressure dressing, compression belt, and no compression device/dressing) and has been reported elsewhere.19 The first phase, representing the subject of this study, consisted of randomized assignment to 1 of 3 techniques used to achieve hemostasis initially. The first group involved manual hold: fingertip pressure applied directly over the femoral artery, positioned both proximal and distal to the puncture site. A mechanical clamp was used in the second group: a C-shaped mechanical compression clamp (Figure 1) designed for femoral artery compression (Compressar, Instromedix, Inc, Hillsboro, Ore). The base of the clamp was positioned under the hip, with the clamp arm over the puncture site. Attached to the end of the clamp arm was a 5-cm diameter sterile disposable pressure pad consisting of a hard
Mean excess compression time required for hemostasis beyond target time of 13 minutes applied to all patients. Error bars denote standard error.
plastic disc containing a 1-cm V-shaped notch on one side. With the notch oriented caudally, the disc was positioned just proximal to the insertion site so that the sheath rested within the notch. The sheath was removed as the disc was lowered to achieve complete arterial compression. The third group used a pneumatic compression device, a femoral compression arch (Figure 1) with an attached transparent inflatable dome (FemoStop USCI, Bard, Inc, Billerica, Mass). The plastic arch was placed over the patient and held in position by a polyester belt wrapped under the patient’s hips. The downwardly directed dome was then inflated over the femoral artery, compressing the puncture site as the sheath was removed. An inflation bulb and gauge (similar to a sphygmomanometer) permitted precise control and monitoring of the pressure inside the dome. To standardize compression times, complete femoral artery compression was applied in all 3 groups for 5 minutes, followed by a gradual release of pressure over the ensuing 8 minutes. Therefore each patient received a minimum of 13 minutes of compression, with further compression applied only if full hemostasis had not been achieved at that point. The 2 mechanical devices were used in accordance with the manufacturer’s directions. In-service training of catheterization laboratory personnel was provided by each manufacturer. Randomization to the FemoStop group was delayed until all study personnel had received extensive experience with the new device, resulting in a smaller number of patients enrolled in that group. The absence of an ongoing learning curve is suggested by the similar results obtained during both halves (early and late) of the study. Randomization was accomplished by a computer-generated random number sequence and a sealed envelope system of assignment. The 4 types of dressings used after the patient left the laboratory (the second phase of the study) were equally represented in each group.
American Heart Journal December 1999
1120 Lehmann, Heath-Lange, and Ferris
Table I. Baseline characteristics
No. of patients Clinical variables Age (y) Male sex Body surface area (m2) Historic variables Prior MI Hypertension Diabetes Smoking Hypercholesterolemia Peripheral vascular disease Hemodynamic/angiographic variables Heart rate (min–1) Systolic pressure (mm Hg) Pulse pressure (mm Hg) Ejection fraction (%) No. of vessels with CAD None 1 2 3 Mean No. vessels with CAD
Manual
Clamp
Pneumatic
152
143
102
61 ± 11 148 (97) 2.05 ± 0.20
63 ± 11 141 (99) 2.02 ± 0.16
60 ± 11 100 (98) 2.06 ± 0.23
57 (38) 81 (53) 40 (26) 84 (55) 56 (37) 28 (18)
58 (41) 76 (53) 34 (24) 72 (50) 46 (32) 52 (36)
70 ± 12 128 ± 25 59 ± 19 61 ± 16
70 ± 13 128 ± 20 58 ± 19 55 ± 15
70 ± 13 129 ± 23 59 ± 20 58 ± 16
14 (27) 21 (14) 45 (30) 45 (30) 1.62 ± 1.17
46 (32) 18 (13) 29 (20) 50 (35) 1.58 ± 1.26
26 (26) 20 (20) 18 (18) 38 (37) 1.67 ± 1.22
40 (40) 55 (54) 25 (25) 50 (49) 39 (38) 28 (28)*
Values are n (%) or mean ± SD. CAD, Coronary artery disease (>50% stenosis); MI, myocardial infarction. * P < .003 overall.
The protocol was approved by the Human Studies Committee of the University of Washington.
Clinical procedures Vascular access sheaths were used in every patient. Patients undergoing diagnostic procedures were typically treated with 2000 U intravenous heparin immediately after sheath insertion. For interventional procedures, a heparin bolus ranging 10,000 to 15,000 U was administered, followed by repeated boluses as needed to maintain an activated clotting time of 250 to 350 seconds. No patient received protamine sulfate for reversal of anticoagulation. The sheaths were removed immediately after diagnostic procedures. For interventional procedures, sheath removal occurred after heparin discontinuation when the activated clotting time decreased below 150 seconds (generally 4 to 6 hours after the procedure). No platelet glycoprotein IIb/IIIa receptor antagonists were used. Immediately after achieving hemostasis, each arterial access site was carefully inspected for evidence of hematoma formation or other vascular problems. Uniform verbal and written instructions of immobility were provided to each patient on transfer from the catheterization laboratory. Written nursing orders included frequent vital sign checks and bed rest for 5 hours with the head of the bed elevated up to 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. All dressings were removed 5 hours after sheath removal, and all medically able patients were strongly encouraged to ambulate at a level commensurate with their normal activities at home. Approximately 24 hours after sheath removal, all
patients underwent a predetermined standardized interview and were examined by a single blinded observer. Documentation of a hematoma required the palpation of a discrete area of focal induration ≥1 cm in diameter at the puncture site. The size of a hematoma or ecchymosis was computed as an elliptical surface area with orthogonal linear measurements as major and minor axes. Bleeding from the access site was defined as mild if was associated with minimal (<10 mL) blood loss and did not require further therapy, as moderate if estimated blood loss was >10 mL or repeat compression was required, or severe if a major intervention (such as a transfusion) was needed. Arteriovenous fistula, pseudoaneurysm, and arterial occlusion were recorded as major adverse events if surgical correction was required within 1 year of the original procedure.
Data analysis Plus-minus values represent mean ± SD. Differences in categoric variables were compared with the χ2 test, with 1-way analysis of variance used for continuous variables. A P value of .05 was accepted as the limit of statistical significance. All analyses were performed on an intention-to-treat basis.
Results Baseline and procedural variables Table I describes the clinical characteristics of the study population (n = 397). Overall, 39% had a history of myocardial infarction. Atherosclerotic risk factors were common, including hypertension in 53%, diabetes in 25%, a history of current or past smoking in 52%, and
American Heart Journal Volume 138, Number 6, Part 1
Lehmann, Heath-Lange, and Ferris 1121
Table II. Procedure-related parameters
Type of procedure Diagnostic Interventional Postprocedural monitoring Intensive care unit Cardiology ward Sheath size 5F, 6F 7F 8F 9F, 10F Sheath removal By laboratory technician By laboratory or ICU nurse By physician
Manual
Clamp
Pneumatic
143 (94) 9 (6)
127 (89) 16 (11)
94 (92) 8 (8)
48 (32) 104 (68)
45 (32) 98 (69)
36 (35) 66 (65)
1 (1) 136 (90) 8 (5) 7 (5)
4 (3) 119 (83) 7 (5) 13 (9)
4 (4) 85 (83) 6 (6) 7 (7)
127 (84) 24 (16) 1 (1)
111 (78) 27 (19) 5 (4)
87 (85) 15 (15) 0 (0)
Values are n (%) or mean ± SD. All differences were statistically nonsignificant. ICU, Intensive care unit.
Table III. Outcomes and adverse events
Time to hemostasis (min) Prolonged compression required (>13 min) Hematoma In-lab At follow-up Size (cm2) Ecchymosis At follow-up Size (cm2) Bleeding In-lab At follow-up Bleeding severity at follow-up Mild Moderate Severe Other adverse events Alternate compression technique required Arterial occlusion Vascular surgery Transfusion
Manual
Clamp
Pneumatic
P value
13.9 ± 3.5 19 (13)
14.5 ± 4.5 28 (20)
15.6 ± 4.81 36 (35)
.006 <.0001
15 (10) 23 (16) 3.9 ± 3.6
16 (11) 24 (19) 7.8 ± 10.0
13 (13) 11 (12) 19.8 ± 35.0
NS NS .036
53 (38) 20.7 ± 32.8
43 (34) 30.2 ± 133.9
26 (29) 48.8 ± 104.8
NS NS
5 (4) 7 (6)
16 (16) 11 (12)
<.0001 NS NS
9 (82) 1 (9) 1 (9)
6 (86) 1 (14) 0 (0)
9 (82) 1 (9) 1 (9)
1 (1) 0 (0) 0 (0) 1 (1)
1 (1) 0 (0) 0 (0) 0 (0)
26 (27) 0 (0) 0 (0) 0 (0)
4 (3) 11 (8)
<.0001 NS NS NS
Values are n (%) or mean ± SD. NS, Not statistically significant.
a history of hypercholesterolemia in 37%. Peripheral vascular disease, defined as either a history of claudication, prior vascular surgery, or arterial bruits on exam, was present in 27% of patients. Despite randomization, patients assigned to the manual hold group had a somewhat lower prevalence of peripheral vascular disease (18%). Importantly, when the data were reanalyzed after excluding patients in all 3 groups with peripheral vascular disease, all statistically significant differences in outcomes between groups remained, suggesting that
intergroup differences in the frequency of peripheral vascular disease at baseline did not meaningfully contribute to the results obtained. Hemodynamic and angiographic parameters were equally distributed between groups. Arterial pressure is known to influence the likelihood of vascular complications after cardiac catheterization.20 Centrally measured systolic and pulse pressures averaged 128 mm Hg and 59 mm Hg, respectively, and were nearly identical in each group. Significant angiographic coronary artery
American Heart Journal December 1999
1122 Lehmann, Heath-Lange, and Ferris
Figure 3
Measured size of hematomas observed at 24-hour follow-up examination, grouped by hemostasis technique. Mean values are indicated by open circles. Several points are superimposed because of patients with identical hematoma areas.
disease was noted in 71% of patients overall; most patients lacking coronary disease 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 attainment of hemostasis more difficult; this condition was present in 5 patients (3 in the manual group, 1 in the clamp group, and 1 in the pneumatic group). Of the 397 procedures performed, 92% were diagnostic and 8% were interventional (Table II). After the procedure, 32% of patients returned to an intensive care unit bed, and 68% to the cardiology ward. The arterial sheath diameter used ranged from 5F to 10F, with 7F the most frequently used size.
Outcomes As part of the study protocol, a set arterial compression time of 13 minutes after sheath removal was established a priori. After that point, further compression was only used if continued bleeding was present. As shown in Table III, prolonged compression was required in 13% of the manual group, 20% of the clamp group, and 35% of the pneumatic group (P < .0001). This resulted in a significant difference between groups in mean compression times (Figure 2). Overall, hematomas exceeding 1 cm2 surface area were noted in 11% of patients while in the catheterization laboratory and increased to 16% by 24 hours after the procedure. Although the frequency of hematoma
Figure 4
Cumulative distribution of degree of discomfort reported by each patient during compression. Rightward and downward shift of curve indicates greater discomfort so that pneumatic group reported most discomfort and manual group least discomfort. A 1- to 10-point scale was used, with 1 indicating minimal discomfort.
formation was statistically similar in the 3 groups, the mean size (Figure 3) was smallest in the manual group, intermediate in the clamp group, and largest in the pneumatic group (3.9 ± 3.6, 7.8 ± 10.0, and 19.8 ± 35.0 cm2, P = .036). A similar but nonsignificant pattern was noted in the measured area of ecchymosis at 24 hours. Bleeding in the catheterization laboratory was substantially more common in the pneumatic group (16%) than in the other 2 groups (3% and 4%, P < .0001), but during follow-up similar rates of rebleeding were observed. Severe bleeding was uncommon, occurring in only 2 patients. One patient (1%) each in the manual and clamp groups did not achieve hemostasis with their assigned technique and required crossover to a different method of compression. In sharp contrast, 27% of pneumatic patients did not achieve hemostasis with this technique and required switching to a manual hold (P < .0001). More serious adverse events were quite rare, consisting of a transfusion in one patient. Table IV shows patient discomfort, tenderness, and activity in the 3 treatment groups. As depicted in Figure 4, the mean level of discomfort (on a 1- to 10point scale) during compression was 1.9 ± 1.9 in the manual hold group, 2.2 ± 2.0 in the mechanical clamp group, and 3.1 ± 2.1 in the pneumatic device group (P < .0001). After stopping compression, the mean patient discomfort level diminished somewhat and thereafter was similar in each group (P = not sig-
American Heart Journal Volume 138, Number 6, Part 1
Lehmann, Heath-Lange, and Ferris 1123
Table IV. Patient discomfort and activity Manual Mean level of discomfort During compression After compression but before dressing removal After dressing removal but before ambulation After ambulation Site tenderness at follow-up None Mild Moderate Severe Level of ambulation None Mild Average Strenuous
Clamp
Pneumatic
1.9 ± 1.9 1.9 ± 1.8 1.6 ± 1.3 1.8 ± 1.8
2.2 ± 2.0 1.5 ± 1.1 1.4 ± 0.9 1.6 ± 1.0
3.1 ± 2.1 1.7 ± 1.6 1.5 ± 1.2 1.8 ± 1.6
100 (71) 29 (21) 10 (7) 2 (1)
87 (68) 36 (28) 3 (2) 2 (2)
60 (67) 24 (27) 5 (6) 1 (1)
27 (19) 87 (62) 25 (18) 2 (1)
23 (18) 78 (61) 20 (16) 7 (6)
22 (24) 41 (46) 26 (29) 1 (1)
P value
<.0001 NS NS NS NS
.025
Values are n (%) or mean ± SD. NS, Not statistically significant.
nificant). A physical examination at 24 hours demonstrated site tenderness in a minority of patients (31%). When present, it was mostly mild (77%) and was equally distributed among the 3 groups (P = not significant). All patients who were medically able were encouraged to ambulate as they would at home starting 5 hours after sheath removal. Of the 287 patients (80%) who did, their level of activity at 24 hours was self-described as mild in 72%, average (normal daily activity) in 25%, and strenuous in 3%. Patients assigned to the pneumatic device group were slightly less likely to ambulate (P = .025), perhaps because of the higher in-lab bleeding rate observed with this group.
Discussion Overall findings When considered together, these findings indicate several clear-cut disadvantages with the use of the commercially available FemoStop compression system. The group randomly assigned to this pneumatic device required a longer compression time (P = .006), encountered a higher in-lab bleeding rate (P < .0001), required more frequent crossover to an alternate compression technique (P < .0001), and had greater discomfort during its use (P < .0001). The differences between manual and clamp compression were more subtle, but all trends observed appear to favor the manual technique. Most notable were the differences in requirement for prolonged compression (13% for manual, 20% for clamp, P = .097) and the increased frequency of large hematomas observed at 24 hours (0% for manual, 4% for clamp when including hematomas ≥15 cm2, P = .055).
Relation to prior reports Given the large number of femoral artery cannulations performed worldwide every day, there is a surprising paucity of data in the literature on optimal management strategies after sheath removal. In fact, our study represents the only randomized trial available that compares the 3 most commonly used techniques for obtaining hemostasis at the puncture site. The mechanical clamp is used extensively in many medical centers. In a brief report, Semler21 described a prospective but nonrandomized comparison of manual compression versus mechanical clamp. The findings of this study were limited to a shorter mean compression time (19.9 minutes vs 33.5 minutes) and lower frequency of hematoma formation (2% vs 6%) for the clamp technique. However, no attempt was made to set a target for compression time, no data were reported on discomfort or bleeding, and no evaluation was performed after the patient left the laboratory. In contrast to these findings, Simon22 reported a longer mean compression time and no difference in complication rates with the clamp technique in a 200-patient retrospective chart review. The FemoStop pneumatic compression device is newer and less commonly used. In one retrospective report by Sanuki et al,23 no hematomas or rebleeding were seen in 74 patients.
Relevance to outpatient procedures During the enrollment period of this study, all catheterization procedures undertaken in our laboratory were performed on an in-patient basis mostly because of the 4-state catchment area served by the laboratory. This provided a unique opportunity to monitor many patients closely who would be eligible for out-
American Heart Journal December 1999
1124 Lehmann, Heath-Lange, and Ferris
patient catheterization in the absence of geographic constraints. Patients were continually available for evaluation for at least 24 hours 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 hours after sheath removal at a level commensurate with their anticipated activity at home had they been discharged. In addition to the primary analyses, we assessed outcomes in the subgroup of enrollees who could be considered candidates for outpatient procedures (no unstable angina or interventional procedures).24 The results obtained from these 171 individuals were similar to the overall study group. Thus our findings appear to be applicable to the out-patient setting as well.
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. Second, unlike current practice in many laboratories, the most common size of arterial sheath used in this study was 7F. Although this could potentially affect the absolute level of discomfort and adverse events reported in this trial, it is unlikely to affect the relative differences that would be anticipated between groups when smaller sheath sizes are used. Hence our findings should retain their applicability to the out-patient as well as in-patient setting. Third, this study included a mixture of both diagnostic and interventional procedures. The number of interventional patients assigned to each group, however, was statistically similar, and the differences in outcomes found between groups remained after their exclusion. Finally, the nature of the study precluded blinding of treatment strategy for either patient or treating physician. Most outcomes, though, were evaluated without knowledge of the assigned technique.
Clinical implications In addition to the differences in this trial, the choice of compression technique involves economic considerations and use of medical personnel. Currently, the US retail price of the Compressar clamp system is $595 plus $5 per patient for the disposable disk. The retail price of the FemoStop pneumatic compression system is $125 plus $69 per patient for the single-use components. In contrast, the manual hold system requires no additional hardware or expense. The required setup time for each technique differs as well, with manual compression requiring the shortest setup in our laboratory and the FemoStop technique requiring the longest. Unlike the manual hold, the clamp and pneumatic technique do not automatically require the continuous pres-
ence of medical personnel throughout the compression period. However, in our opinion, the risks associated with each technique mandate the full-time attention of a technician or nurse during setup and use, making the personnel requirements of each method roughly equal. Based on the current trial, the use of the FemoStop pneumatic compression device results in longer compression times, greater discomfort, more bleeding, more frequent failures, and higher costs than the other 2 techniques, without any obvious advantages to justify its use. The relative merits between manual compression and the mechanical clamp are less obvious. However, trends in compression times, hematoma size, the amount of discomfort, economic factors, and setup time all favor the manual technique. Therefore manual compression after sheath removal may be the preferred technique for achieving hemostasis after femoral artery cannulation. We thank Susan Gilbert, RN, MSN, Ruben Flores, RT, Daryl Jones, CPT, Donna Kline, RN, and Cary Shepherd, CVT, for expert technical assistance as well as Michael Stadius, MD, for editorial comments.
References 1. Zimmerman HA, Scott RW, Becker ND. Catheterization of the left side of the heart in man. Circulation 1950;1:357-9. 2. Talley JD, Wilkins C, Ciccone J, et al. Comparison of 7 and 8 French guiding catheters for elective PTCA: results of a prospective randomized trial. Cathet Cardiovasc Diagn 1994;33:205-11. 3. Kern MJ, Cohen M, Talley JD, et al. Early ambulation after 5 French diagnostic cardiac catheterization: results of a multicenter trial. J Am Coll Cardiol 1990;15:1475-83. 4. Pande AK, Meier B, Urban P, et al. Coronary angiography with four French catheters. Am J Cardiol 1992;70:1085-6. 5. Urban P, Meier B, Haine E, et al. Coronary stenting through 6 French guiding catheters. Cathet Cardiovasc Diagn 1993;28:263-6. 6. Mehan V, Meier B, Urban P, et al. Coronary angioplasty through 4 French diagnostic catheters. Cathet Cardiovasc Diagn 1993; 29:267-72. 7. Kiemeneij F, Laarman GJ, de Melker E. Transradial artery coronary angioplasty. Am Heart J 1995;129:1-7. 8. Lotan C, Hasin Y, Mosseri M, et al. Transradial approach for coronary angiography and angioplasty. Am J Cardiol 1995;76:164-7. 9. Bush CA, VanFossen DB, Kolibash AJ Jr, et al. Cardiac catheterization and coronary angiography using 5 French preformed (Judkins) catheters from the percutaneous right brachial approach: a comparative analysis with the femoral approach. Cathet Cardiovasc Diagn 1993;29:267-72. 10. Kiemeneij F, Laarman GJ. Percutaneous transradial artery approach for coronary Palmaz-Schatz stent implantation. Am Heart J 1994; 128:167-74. 11. Heuser RR, Mehta SS, Strumpf RK, et al. Intracoronary stent implantation via the brachial approach: a technique to reduce vascular bleeding complications. Cathet Cardiovasc Diagn 1992;25:300-3. 12. Bartorelli AL, Sganzerla P, Fabbiocchi F, et al. Prompt and safe femoral hemostasis with a collagen device after intracoronary implantation of Palmaz-Schatz stents. Am Heart J 1995;130:26-32.
American Heart Journal Volume 138, Number 6, Part 1
13. Slaughter P, Chetty R, Flintoft VF, et al. A single center randomized trial assessing use of a vascular hemostasis device vs conventional manual compression following PTCA: what are the potential resource savings? Cathet Cardiovasc Diagn 1995;34:210-4. 14. Kussmaul WG, Buchbinder M, Whitlow PL, et al. Rapid arterial hemostasis and decreased access site complications after cardiac catheterization and angioplasty: results of a randomized trial of a novel hemostatic device. J Am Coll Cardiol 1995;25:1685-92. 15. Ernst SM, Tjonjoegin RM, Schrader R, et al. Immediate sealing of arterial puncture sites after cardiac catheterization and coronary angioplasty using a biodegradable collagen plug: results of an international registry. J Am Coll Cardiol 1993;21:851-5. 16. Schrader R, Steinbacher S, Burger W, et al. Collagen application for sealing of arterial puncture sites in comparison to pressure dressing: a randomized trial. Cathet Cardiovasc Diagn 1992;27:298-302. 17. Sanborn TA, Gibbs HH, Brinker J, et al. A multicenter randomized trial comparing a percutaneous collagen hemostasis device with conventional manual compression after diagnostic angiography and angioplasty. J Am Coll Cardiol 1993;22:1273-9.
Lehmann, Heath-Lange, and Ferris 1125
18. The Epic Investigators. Use of a monoclonal antibody directed against the platelet glycoprotein IIb/IIIa receptor in high-risk coronary angioplasty. N Engl J Med 1994;330:956-61. 19. Lehmann KG, Heath-Lange SJ, Ferris ST. Maintenance of hemostasis following invasive cardiac procedures. J Am Coll Cardiol 1997; 30:444-51. 20. Waksman R, King SB, Douglas JS, et al. Predictors of groin complications after balloon and new-device coronary intervention. Am J Cardiol 1995;75:886-9. 21. Semler HJ. Transfemoral catheterization: mechanical versus manual control of bleeding. Radiology 1985;154:234-5. 22. Simon AW. Use of a mechanical pressure device for hemostasis following cardiac catheterization. Am J Crit Care 1994;3:62-4. 23. Sanuki E, Kunieda H, Fujioka K, et al. Clinical evaluation of FemoStop in arrest of hemorrhage of the femoral artery. Nippon Igaku Hoshasen Gakkai Zasshi 1995;55:7-14. 24. Clark DA, Moscovich MD, Vetrovec GW, et al. Guidelines for the performance of outpatient catheterization and angiographic procedures. Cathet Cardiovasc Diagn 1992;27:5-7.
Background The arterial access required during most invasive vascular procedures provides a common source of complications and morbidity. This problem has been made worse by recent trends in earlier ambulation and more aggressive antihemostatic drug regimens. Despite these trends, no randomized trials have been reported comparing the 3 most commonly used techniques in achieving hemostasis at the arterial puncture site.
Methods A cohort of 400 patients undergoing catheterization laboratory procedures were randomly assigned to 1 of 3 groups of arterial compression: manual compression, mechanical clamp, and pneumatic compression device. Standard requirements of the trial included uniformity in initial compression times, patient instructions, nursing follow-up, and timing of ambulation as well as a structured interview and physical examination at 24 hours.
Results Prolonged compression was required in 13% of the manual group, 20% of the clamp group, and 35% of the pneumatic group (P < .0001). In-lab bleeding was more common in the pneumatic group (3%, 4%, and 16%, respectively, P < .0001), as was the need for an alternate compression technique (1%, 1%, and 27%, P < .0001). The groups also differed in respect to mean hematoma size (3.9 cm2, 7.8 cm2, and 19.8 cm2, P = .036) and level of discomfort during compression (1.9, 2.2, and 3.1 on a 1- to 10-point scale, P < .0001). Comparable findings were observed in the subgroup of patients eligible for outpatient procedures.
Conclusions Use of the pneumatic compression device leads to longer compression times, greater discomfort, more bleeding, and larger hematomas. Differences between manual compression and the mechanical clamp were more subtle but tend to favor use of the manual technique. (Am Heart J 1999;138:1118-25.)
See related Editorial on page 1014. Since its inception in 1950, cardiac catheterization has grown to encompass more than 1.5 million procedures per year in the United States alone.1 Although the safety and efficiency of this and related procedures continue to improve, the universal requirement of arterial access prolongs the requisite period of monitoring. This factor remains the single most important impediment to early discharge. Several technical modifications have been advocated to help minimize the negative impact of femoral artery puncture. Over the past decade, the size of diagnostic and therapeutic catheters has steadily decreased.2-6 Although more technically challenging than femoral artery cannulation, arterial access can be obtained through a site less stressed by ambulation, such as the From the University of Washington School of Medicine and the Veterans Affairs Puget Sound Health Care System. Supported by grants from the Research Service of the Department of Veterans Affairs. Reprint requests: Kenneth G. Lehmann, MD, Section of Cardiology (111C), VA Puget Sound Health Care System, 1660 South Columbian Way, Seattle, WA 98108. 0002-8703/99/$8.00 + 0 4/1/101219
brachial, radial, or axillary arteries.7-11 Collagen plugs and percutaneous surgical closure devices have been developed to speed hemostasis, but their expense, uncertain efficacy, and continued requirement for at least some external arterial compression have made their use controversial.12-17 Even these potential improvements, however, have been partially offset by recent trends that place greater stress on the healing artery. These include outpatient procedures, early ambulation, larger sheath sizes required by certain procedures such as directional atherectomy, and the increased use of potent antihemostatic agents such as clopidogrel and abciximab.18 This study prospectively explored the relative safety and efficacy of 3 different but commonly used techniques of achieving hemostasis after arterial sheath removal. This comparison was facilitated by randomized treatment assignment, close monitoring before and after ambulation, and the use of structured interview and physical evaluation 24 hours after the procedure.
Methods Patient population Patients were selected from consecutive individuals undergoing invasive arterial procedures in the cardiac catheterization laboratory. The following exclusion criteria were applied:
American Heart Journal Volume 138, Number 6, Part 1
Figure 1
Lehmann, Heath-Lange, and Ferris 1119
Figure 2
Mechanical devices used in this trial.
(1) arterial access at a site other than the right or left femoral artery; (2) emergent procedure; (3) arterial puncture at the same site within the previous 18 hours; (4) continued use or effect of warfarin, heparin, thrombolytic agent, ticlopidine, abciximab, or other nonaspirin anticoagulant 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 during the following 24 hours; and (7) unwillingness or inability to provide written informed consent. A total of 400 patients were enrolled. Three were excluded shortly after enrollment because of evidence of hematoma formation before sheath removal. In addition, incomplete follow-up data in 38 patients (often from unanticipated early discharge) resulted in exclusion of these individuals from assessment of delayed outcomes. Power calculations were undertaken to ensure adequate sample size for the detection of clinically meaningful differences between groups. Based on a 2-tailed α of .05, the power (1 – β) of the study to detect a 20% difference in the proportion of patients having a hematoma develop was 0.95, a 25% difference in the proportion of patients requiring prolonged arterial compression was 0.98, and a 1-point difference in pain scale was 0.99.
Study protocol The overall study was designed as a prospective randomized trial of arteriotomy site management based on a 3 × 4 factorial design. The second phase investigated 4 different methods of maintaining hemostasis after the patient leaves the catheterization laboratory (sandbag, pressure dressing, compression belt, and no compression device/dressing) and has been reported elsewhere.19 The first phase, representing the subject of this study, consisted of randomized assignment to 1 of 3 techniques used to achieve hemostasis initially. The first group involved manual hold: fingertip pressure applied directly over the femoral artery, positioned both proximal and distal to the puncture site. A mechanical clamp was used in the second group: a C-shaped mechanical compression clamp (Figure 1) designed for femoral artery compression (Compressar, Instromedix, Inc, Hillsboro, Ore). The base of the clamp was positioned under the hip, with the clamp arm over the puncture site. Attached to the end of the clamp arm was a 5-cm diameter sterile disposable pressure pad consisting of a hard
Mean excess compression time required for hemostasis beyond target time of 13 minutes applied to all patients. Error bars denote standard error.
plastic disc containing a 1-cm V-shaped notch on one side. With the notch oriented caudally, the disc was positioned just proximal to the insertion site so that the sheath rested within the notch. The sheath was removed as the disc was lowered to achieve complete arterial compression. The third group used a pneumatic compression device, a femoral compression arch (Figure 1) with an attached transparent inflatable dome (FemoStop USCI, Bard, Inc, Billerica, Mass). The plastic arch was placed over the patient and held in position by a polyester belt wrapped under the patient’s hips. The downwardly directed dome was then inflated over the femoral artery, compressing the puncture site as the sheath was removed. An inflation bulb and gauge (similar to a sphygmomanometer) permitted precise control and monitoring of the pressure inside the dome. To standardize compression times, complete femoral artery compression was applied in all 3 groups for 5 minutes, followed by a gradual release of pressure over the ensuing 8 minutes. Therefore each patient received a minimum of 13 minutes of compression, with further compression applied only if full hemostasis had not been achieved at that point. The 2 mechanical devices were used in accordance with the manufacturer’s directions. In-service training of catheterization laboratory personnel was provided by each manufacturer. Randomization to the FemoStop group was delayed until all study personnel had received extensive experience with the new device, resulting in a smaller number of patients enrolled in that group. The absence of an ongoing learning curve is suggested by the similar results obtained during both halves (early and late) of the study. Randomization was accomplished by a computer-generated random number sequence and a sealed envelope system of assignment. The 4 types of dressings used after the patient left the laboratory (the second phase of the study) were equally represented in each group.
American Heart Journal December 1999
1120 Lehmann, Heath-Lange, and Ferris
Table I. Baseline characteristics
No. of patients Clinical variables Age (y) Male sex Body surface area (m2) Historic variables Prior MI Hypertension Diabetes Smoking Hypercholesterolemia Peripheral vascular disease Hemodynamic/angiographic variables Heart rate (min–1) Systolic pressure (mm Hg) Pulse pressure (mm Hg) Ejection fraction (%) No. of vessels with CAD None 1 2 3 Mean No. vessels with CAD
Manual
Clamp
Pneumatic
152
143
102
61 ± 11 148 (97) 2.05 ± 0.20
63 ± 11 141 (99) 2.02 ± 0.16
60 ± 11 100 (98) 2.06 ± 0.23
57 (38) 81 (53) 40 (26) 84 (55) 56 (37) 28 (18)
58 (41) 76 (53) 34 (24) 72 (50) 46 (32) 52 (36)
70 ± 12 128 ± 25 59 ± 19 61 ± 16
70 ± 13 128 ± 20 58 ± 19 55 ± 15
70 ± 13 129 ± 23 59 ± 20 58 ± 16
14 (27) 21 (14) 45 (30) 45 (30) 1.62 ± 1.17
46 (32) 18 (13) 29 (20) 50 (35) 1.58 ± 1.26
26 (26) 20 (20) 18 (18) 38 (37) 1.67 ± 1.22
40 (40) 55 (54) 25 (25) 50 (49) 39 (38) 28 (28)*
Values are n (%) or mean ± SD. CAD, Coronary artery disease (>50% stenosis); MI, myocardial infarction. * P < .003 overall.
The protocol was approved by the Human Studies Committee of the University of Washington.
Clinical procedures Vascular access sheaths were used in every patient. Patients undergoing diagnostic procedures were typically treated with 2000 U intravenous heparin immediately after sheath insertion. For interventional procedures, a heparin bolus ranging 10,000 to 15,000 U was administered, followed by repeated boluses as needed to maintain an activated clotting time of 250 to 350 seconds. No patient received protamine sulfate for reversal of anticoagulation. The sheaths were removed immediately after diagnostic procedures. For interventional procedures, sheath removal occurred after heparin discontinuation when the activated clotting time decreased below 150 seconds (generally 4 to 6 hours after the procedure). No platelet glycoprotein IIb/IIIa receptor antagonists were used. Immediately after achieving hemostasis, each arterial access site was carefully inspected for evidence of hematoma formation or other vascular problems. Uniform verbal and written instructions of immobility were provided to each patient on transfer from the catheterization laboratory. Written nursing orders included frequent vital sign checks and bed rest for 5 hours with the head of the bed elevated up to 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. All dressings were removed 5 hours after sheath removal, and all medically able patients were strongly encouraged to ambulate at a level commensurate with their normal activities at home. Approximately 24 hours after sheath removal, all
patients underwent a predetermined standardized interview and were examined by a single blinded observer. Documentation of a hematoma required the palpation of a discrete area of focal induration ≥1 cm in diameter at the puncture site. The size of a hematoma or ecchymosis was computed as an elliptical surface area with orthogonal linear measurements as major and minor axes. Bleeding from the access site was defined as mild if was associated with minimal (<10 mL) blood loss and did not require further therapy, as moderate if estimated blood loss was >10 mL or repeat compression was required, or severe if a major intervention (such as a transfusion) was needed. Arteriovenous fistula, pseudoaneurysm, and arterial occlusion were recorded as major adverse events if surgical correction was required within 1 year of the original procedure.
Data analysis Plus-minus values represent mean ± SD. Differences in categoric variables were compared with the χ2 test, with 1-way analysis of variance used for continuous variables. A P value of .05 was accepted as the limit of statistical significance. All analyses were performed on an intention-to-treat basis.
Results Baseline and procedural variables Table I describes the clinical characteristics of the study population (n = 397). Overall, 39% had a history of myocardial infarction. Atherosclerotic risk factors were common, including hypertension in 53%, diabetes in 25%, a history of current or past smoking in 52%, and
American Heart Journal Volume 138, Number 6, Part 1
Lehmann, Heath-Lange, and Ferris 1121
Table II. Procedure-related parameters
Type of procedure Diagnostic Interventional Postprocedural monitoring Intensive care unit Cardiology ward Sheath size 5F, 6F 7F 8F 9F, 10F Sheath removal By laboratory technician By laboratory or ICU nurse By physician
Manual
Clamp
Pneumatic
143 (94) 9 (6)
127 (89) 16 (11)
94 (92) 8 (8)
48 (32) 104 (68)
45 (32) 98 (69)
36 (35) 66 (65)
1 (1) 136 (90) 8 (5) 7 (5)
4 (3) 119 (83) 7 (5) 13 (9)
4 (4) 85 (83) 6 (6) 7 (7)
127 (84) 24 (16) 1 (1)
111 (78) 27 (19) 5 (4)
87 (85) 15 (15) 0 (0)
Values are n (%) or mean ± SD. All differences were statistically nonsignificant. ICU, Intensive care unit.
Table III. Outcomes and adverse events
Time to hemostasis (min) Prolonged compression required (>13 min) Hematoma In-lab At follow-up Size (cm2) Ecchymosis At follow-up Size (cm2) Bleeding In-lab At follow-up Bleeding severity at follow-up Mild Moderate Severe Other adverse events Alternate compression technique required Arterial occlusion Vascular surgery Transfusion
Manual
Clamp
Pneumatic
P value
13.9 ± 3.5 19 (13)
14.5 ± 4.5 28 (20)
15.6 ± 4.81 36 (35)
.006 <.0001
15 (10) 23 (16) 3.9 ± 3.6
16 (11) 24 (19) 7.8 ± 10.0
13 (13) 11 (12) 19.8 ± 35.0
NS NS .036
53 (38) 20.7 ± 32.8
43 (34) 30.2 ± 133.9
26 (29) 48.8 ± 104.8
NS NS
5 (4) 7 (6)
16 (16) 11 (12)
<.0001 NS NS
9 (82) 1 (9) 1 (9)
6 (86) 1 (14) 0 (0)
9 (82) 1 (9) 1 (9)
1 (1) 0 (0) 0 (0) 1 (1)
1 (1) 0 (0) 0 (0) 0 (0)
26 (27) 0 (0) 0 (0) 0 (0)
4 (3) 11 (8)
<.0001 NS NS NS
Values are n (%) or mean ± SD. NS, Not statistically significant.
a history of hypercholesterolemia in 37%. Peripheral vascular disease, defined as either a history of claudication, prior vascular surgery, or arterial bruits on exam, was present in 27% of patients. Despite randomization, patients assigned to the manual hold group had a somewhat lower prevalence of peripheral vascular disease (18%). Importantly, when the data were reanalyzed after excluding patients in all 3 groups with peripheral vascular disease, all statistically significant differences in outcomes between groups remained, suggesting that
intergroup differences in the frequency of peripheral vascular disease at baseline did not meaningfully contribute to the results obtained. Hemodynamic and angiographic parameters were equally distributed between groups. Arterial pressure is known to influence the likelihood of vascular complications after cardiac catheterization.20 Centrally measured systolic and pulse pressures averaged 128 mm Hg and 59 mm Hg, respectively, and were nearly identical in each group. Significant angiographic coronary artery
American Heart Journal December 1999
1122 Lehmann, Heath-Lange, and Ferris
Figure 3
Measured size of hematomas observed at 24-hour follow-up examination, grouped by hemostasis technique. Mean values are indicated by open circles. Several points are superimposed because of patients with identical hematoma areas.
disease was noted in 71% of patients overall; most patients lacking coronary disease 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 attainment of hemostasis more difficult; this condition was present in 5 patients (3 in the manual group, 1 in the clamp group, and 1 in the pneumatic group). Of the 397 procedures performed, 92% were diagnostic and 8% were interventional (Table II). After the procedure, 32% of patients returned to an intensive care unit bed, and 68% to the cardiology ward. The arterial sheath diameter used ranged from 5F to 10F, with 7F the most frequently used size.
Outcomes As part of the study protocol, a set arterial compression time of 13 minutes after sheath removal was established a priori. After that point, further compression was only used if continued bleeding was present. As shown in Table III, prolonged compression was required in 13% of the manual group, 20% of the clamp group, and 35% of the pneumatic group (P < .0001). This resulted in a significant difference between groups in mean compression times (Figure 2). Overall, hematomas exceeding 1 cm2 surface area were noted in 11% of patients while in the catheterization laboratory and increased to 16% by 24 hours after the procedure. Although the frequency of hematoma
Figure 4
Cumulative distribution of degree of discomfort reported by each patient during compression. Rightward and downward shift of curve indicates greater discomfort so that pneumatic group reported most discomfort and manual group least discomfort. A 1- to 10-point scale was used, with 1 indicating minimal discomfort.
formation was statistically similar in the 3 groups, the mean size (Figure 3) was smallest in the manual group, intermediate in the clamp group, and largest in the pneumatic group (3.9 ± 3.6, 7.8 ± 10.0, and 19.8 ± 35.0 cm2, P = .036). A similar but nonsignificant pattern was noted in the measured area of ecchymosis at 24 hours. Bleeding in the catheterization laboratory was substantially more common in the pneumatic group (16%) than in the other 2 groups (3% and 4%, P < .0001), but during follow-up similar rates of rebleeding were observed. Severe bleeding was uncommon, occurring in only 2 patients. One patient (1%) each in the manual and clamp groups did not achieve hemostasis with their assigned technique and required crossover to a different method of compression. In sharp contrast, 27% of pneumatic patients did not achieve hemostasis with this technique and required switching to a manual hold (P < .0001). More serious adverse events were quite rare, consisting of a transfusion in one patient. Table IV shows patient discomfort, tenderness, and activity in the 3 treatment groups. As depicted in Figure 4, the mean level of discomfort (on a 1- to 10point scale) during compression was 1.9 ± 1.9 in the manual hold group, 2.2 ± 2.0 in the mechanical clamp group, and 3.1 ± 2.1 in the pneumatic device group (P < .0001). After stopping compression, the mean patient discomfort level diminished somewhat and thereafter was similar in each group (P = not sig-
American Heart Journal Volume 138, Number 6, Part 1
Lehmann, Heath-Lange, and Ferris 1123
Table IV. Patient discomfort and activity Manual Mean level of discomfort During compression After compression but before dressing removal After dressing removal but before ambulation After ambulation Site tenderness at follow-up None Mild Moderate Severe Level of ambulation None Mild Average Strenuous
Clamp
Pneumatic
1.9 ± 1.9 1.9 ± 1.8 1.6 ± 1.3 1.8 ± 1.8
2.2 ± 2.0 1.5 ± 1.1 1.4 ± 0.9 1.6 ± 1.0
3.1 ± 2.1 1.7 ± 1.6 1.5 ± 1.2 1.8 ± 1.6
100 (71) 29 (21) 10 (7) 2 (1)
87 (68) 36 (28) 3 (2) 2 (2)
60 (67) 24 (27) 5 (6) 1 (1)
27 (19) 87 (62) 25 (18) 2 (1)
23 (18) 78 (61) 20 (16) 7 (6)
22 (24) 41 (46) 26 (29) 1 (1)
P value
<.0001 NS NS NS NS
.025
Values are n (%) or mean ± SD. NS, Not statistically significant.
nificant). A physical examination at 24 hours demonstrated site tenderness in a minority of patients (31%). When present, it was mostly mild (77%) and was equally distributed among the 3 groups (P = not significant). All patients who were medically able were encouraged to ambulate as they would at home starting 5 hours after sheath removal. Of the 287 patients (80%) who did, their level of activity at 24 hours was self-described as mild in 72%, average (normal daily activity) in 25%, and strenuous in 3%. Patients assigned to the pneumatic device group were slightly less likely to ambulate (P = .025), perhaps because of the higher in-lab bleeding rate observed with this group.
Discussion Overall findings When considered together, these findings indicate several clear-cut disadvantages with the use of the commercially available FemoStop compression system. The group randomly assigned to this pneumatic device required a longer compression time (P = .006), encountered a higher in-lab bleeding rate (P < .0001), required more frequent crossover to an alternate compression technique (P < .0001), and had greater discomfort during its use (P < .0001). The differences between manual and clamp compression were more subtle, but all trends observed appear to favor the manual technique. Most notable were the differences in requirement for prolonged compression (13% for manual, 20% for clamp, P = .097) and the increased frequency of large hematomas observed at 24 hours (0% for manual, 4% for clamp when including hematomas ≥15 cm2, P = .055).
Relation to prior reports Given the large number of femoral artery cannulations performed worldwide every day, there is a surprising paucity of data in the literature on optimal management strategies after sheath removal. In fact, our study represents the only randomized trial available that compares the 3 most commonly used techniques for obtaining hemostasis at the puncture site. The mechanical clamp is used extensively in many medical centers. In a brief report, Semler21 described a prospective but nonrandomized comparison of manual compression versus mechanical clamp. The findings of this study were limited to a shorter mean compression time (19.9 minutes vs 33.5 minutes) and lower frequency of hematoma formation (2% vs 6%) for the clamp technique. However, no attempt was made to set a target for compression time, no data were reported on discomfort or bleeding, and no evaluation was performed after the patient left the laboratory. In contrast to these findings, Simon22 reported a longer mean compression time and no difference in complication rates with the clamp technique in a 200-patient retrospective chart review. The FemoStop pneumatic compression device is newer and less commonly used. In one retrospective report by Sanuki et al,23 no hematomas or rebleeding were seen in 74 patients.
Relevance to outpatient procedures During the enrollment period of this study, all catheterization procedures undertaken in our laboratory were performed on an in-patient basis mostly because of the 4-state catchment area served by the laboratory. This provided a unique opportunity to monitor many patients closely who would be eligible for out-
American Heart Journal December 1999
1124 Lehmann, Heath-Lange, and Ferris
patient catheterization in the absence of geographic constraints. Patients were continually available for evaluation for at least 24 hours 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 hours after sheath removal at a level commensurate with their anticipated activity at home had they been discharged. In addition to the primary analyses, we assessed outcomes in the subgroup of enrollees who could be considered candidates for outpatient procedures (no unstable angina or interventional procedures).24 The results obtained from these 171 individuals were similar to the overall study group. Thus our findings appear to be applicable to the out-patient setting as well.
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. Second, unlike current practice in many laboratories, the most common size of arterial sheath used in this study was 7F. Although this could potentially affect the absolute level of discomfort and adverse events reported in this trial, it is unlikely to affect the relative differences that would be anticipated between groups when smaller sheath sizes are used. Hence our findings should retain their applicability to the out-patient as well as in-patient setting. Third, this study included a mixture of both diagnostic and interventional procedures. The number of interventional patients assigned to each group, however, was statistically similar, and the differences in outcomes found between groups remained after their exclusion. Finally, the nature of the study precluded blinding of treatment strategy for either patient or treating physician. Most outcomes, though, were evaluated without knowledge of the assigned technique.
Clinical implications In addition to the differences in this trial, the choice of compression technique involves economic considerations and use of medical personnel. Currently, the US retail price of the Compressar clamp system is $595 plus $5 per patient for the disposable disk. The retail price of the FemoStop pneumatic compression system is $125 plus $69 per patient for the single-use components. In contrast, the manual hold system requires no additional hardware or expense. The required setup time for each technique differs as well, with manual compression requiring the shortest setup in our laboratory and the FemoStop technique requiring the longest. Unlike the manual hold, the clamp and pneumatic technique do not automatically require the continuous pres-
ence of medical personnel throughout the compression period. However, in our opinion, the risks associated with each technique mandate the full-time attention of a technician or nurse during setup and use, making the personnel requirements of each method roughly equal. Based on the current trial, the use of the FemoStop pneumatic compression device results in longer compression times, greater discomfort, more bleeding, more frequent failures, and higher costs than the other 2 techniques, without any obvious advantages to justify its use. The relative merits between manual compression and the mechanical clamp are less obvious. However, trends in compression times, hematoma size, the amount of discomfort, economic factors, and setup time all favor the manual technique. Therefore manual compression after sheath removal may be the preferred technique for achieving hemostasis after femoral artery cannulation. We thank Susan Gilbert, RN, MSN, Ruben Flores, RT, Daryl Jones, CPT, Donna Kline, RN, and Cary Shepherd, CVT, for expert technical assistance as well as Michael Stadius, MD, for editorial comments.
References 1. Zimmerman HA, Scott RW, Becker ND. Catheterization of the left side of the heart in man. Circulation 1950;1:357-9. 2. Talley JD, Wilkins C, Ciccone J, et al. Comparison of 7 and 8 French guiding catheters for elective PTCA: results of a prospective randomized trial. Cathet Cardiovasc Diagn 1994;33:205-11. 3. Kern MJ, Cohen M, Talley JD, et al. Early ambulation after 5 French diagnostic cardiac catheterization: results of a multicenter trial. J Am Coll Cardiol 1990;15:1475-83. 4. Pande AK, Meier B, Urban P, et al. Coronary angiography with four French catheters. Am J Cardiol 1992;70:1085-6. 5. Urban P, Meier B, Haine E, et al. Coronary stenting through 6 French guiding catheters. Cathet Cardiovasc Diagn 1993;28:263-6. 6. Mehan V, Meier B, Urban P, et al. Coronary angioplasty through 4 French diagnostic catheters. Cathet Cardiovasc Diagn 1993; 29:267-72. 7. Kiemeneij F, Laarman GJ, de Melker E. Transradial artery coronary angioplasty. Am Heart J 1995;129:1-7. 8. Lotan C, Hasin Y, Mosseri M, et al. Transradial approach for coronary angiography and angioplasty. Am J Cardiol 1995;76:164-7. 9. Bush CA, VanFossen DB, Kolibash AJ Jr, et al. Cardiac catheterization and coronary angiography using 5 French preformed (Judkins) catheters from the percutaneous right brachial approach: a comparative analysis with the femoral approach. Cathet Cardiovasc Diagn 1993;29:267-72. 10. Kiemeneij F, Laarman GJ. Percutaneous transradial artery approach for coronary Palmaz-Schatz stent implantation. Am Heart J 1994; 128:167-74. 11. Heuser RR, Mehta SS, Strumpf RK, et al. Intracoronary stent implantation via the brachial approach: a technique to reduce vascular bleeding complications. Cathet Cardiovasc Diagn 1992;25:300-3. 12. Bartorelli AL, Sganzerla P, Fabbiocchi F, et al. Prompt and safe femoral hemostasis with a collagen device after intracoronary implantation of Palmaz-Schatz stents. Am Heart J 1995;130:26-32.
American Heart Journal Volume 138, Number 6, Part 1
13. Slaughter P, Chetty R, Flintoft VF, et al. A single center randomized trial assessing use of a vascular hemostasis device vs conventional manual compression following PTCA: what are the potential resource savings? Cathet Cardiovasc Diagn 1995;34:210-4. 14. Kussmaul WG, Buchbinder M, Whitlow PL, et al. Rapid arterial hemostasis and decreased access site complications after cardiac catheterization and angioplasty: results of a randomized trial of a novel hemostatic device. J Am Coll Cardiol 1995;25:1685-92. 15. Ernst SM, Tjonjoegin RM, Schrader R, et al. Immediate sealing of arterial puncture sites after cardiac catheterization and coronary angioplasty using a biodegradable collagen plug: results of an international registry. J Am Coll Cardiol 1993;21:851-5. 16. Schrader R, Steinbacher S, Burger W, et al. Collagen application for sealing of arterial puncture sites in comparison to pressure dressing: a randomized trial. Cathet Cardiovasc Diagn 1992;27:298-302. 17. Sanborn TA, Gibbs HH, Brinker J, et al. A multicenter randomized trial comparing a percutaneous collagen hemostasis device with conventional manual compression after diagnostic angiography and angioplasty. J Am Coll Cardiol 1993;22:1273-9.
Lehmann, Heath-Lange, and Ferris 1125
18. The Epic Investigators. Use of a monoclonal antibody directed against the platelet glycoprotein IIb/IIIa receptor in high-risk coronary angioplasty. N Engl J Med 1994;330:956-61. 19. Lehmann KG, Heath-Lange SJ, Ferris ST. Maintenance of hemostasis following invasive cardiac procedures. J Am Coll Cardiol 1997; 30:444-51. 20. Waksman R, King SB, Douglas JS, et al. Predictors of groin complications after balloon and new-device coronary intervention. Am J Cardiol 1995;75:886-9. 21. Semler HJ. Transfemoral catheterization: mechanical versus manual control of bleeding. Radiology 1985;154:234-5. 22. Simon AW. Use of a mechanical pressure device for hemostasis following cardiac catheterization. Am J Crit Care 1994;3:62-4. 23. Sanuki E, Kunieda H, Fujioka K, et al. Clinical evaluation of FemoStop in arrest of hemorrhage of the femoral artery. Nippon Igaku Hoshasen Gakkai Zasshi 1995;55:7-14. 24. Clark DA, Moscovich MD, Vetrovec GW, et al. Guidelines for the performance of outpatient catheterization and angiographic procedures. Cathet Cardiovasc Diagn 1992;27:5-7.