Reducing Bedrest Following Arterial Puncture for Coronary Interventional Procedures — Impact on Vascular Complications: The BAC

Prolonged bedrest following percutaneous coronary intervention (PCI) has been rationalized to minimize the risk of vascular complications following femoral arterial puncture. This may not be a benign practice and can result in considerable patient discomfort (primarily back pain) requiring additional nursing resources and medication use. Anecdotal testimonials indicate that for many patients, bedrest following sheath removal is the most uncomfortable part of the PCI hospitalization experience. Furthermore, hospital stay may also be prolonged, adding to the ever present financial burden of most institutions. The desirability of reducing bedrest times is reflected in the growing market for direct arterial closure devices that promote early ambulation, while increasing overall costs.

Current clinical practice regarding bedrest times varies considerably between institutions. A number of methods for determining optimum bedrest times, including matching the bedrest time to the sheath size (i.e., a 7 French sheath = 7 hours of bedrest), institutional practice based on ritual, drug-driven bedrest times (i.e., abciximab use), and those based on research protocols. A survey of Canadian interventional centers1 showed a range of bedrest requirements from 4–24 hours, with 50% of centers requiring 6 hours or less.

A number of studies have examined bedrest times following cardiac catheterization^2–9 and PCI.^1,10–13 While a few studies tested bedrest times as short as 2 hours following cardiac catheterization,^4,8,9 there were no studies examining bedrest times less than 6 hours following PCI when this trial was designed. The purpose of the current trial was to test the hypothesis that bedrest requirements can be reduced from 6 hours to as low as 2 hours following hemostasis of the femoral puncture site without any increase in vascular complications.

METHODS

Patients. Patients presenting for PCI to the Victoria campus of the London Health Sciences Centre, a tertiary teaching hospital in Southwestern Ontario, Canada, were considered for study participation if a femoral arterial approach was planned and if they were able to read and understand English to give informed consent. Primary exclusion criteria included: abnormal baseline coagulation values (prior to the administration of anticoagulants); significant anemia (hemoglobin < 100 g/dl); puncture through a synthetic femoral graft; and severe aortic insufficiency. Secondary exclusion criteria included: uncontrolled hypertension (systolic blood pressure > 180 mmHg or diastolic blood pressure > 110 mmHg) within 2 hours prior to sheath removal; hemodynamic instability requiring inotropic support at the time of sheath removal; insertion of an intra-aortic balloon pump and performance of a diagnostic cardiac catheterization instead of PCI. All patients gave written informed consent. The study was approved by The Review Board for Health Sciences Research Involving Human Subjects at the University of Western Ontario.

Study protocol. During PCI, all patients received intra-arterial heparin to achieve a target activated clotting time (ACT) of 300 seconds, unless intravenous abciximab (ReoPro®, Eli Lilly and Company, Indianapolis, Indiana) was given, lowering the target ACT to 200 seconds. Heparin was not reversed following the procedure. All patients received pre-procedural acetylsalicylic acid (ASA) 325 mg once a day and ticlopidine (Ticlid, Hoff-man-LaRoche, Nutley, New Jersey) 250 mg twice a day in preparation for potential coronary stent implantation.

Following completion of PCI, bedside monitoring of anticoagulation was done by ACT to control for variable clearance of heparin in individual patients and to provide a common baseline prior to sheath removal. The first study-based ACT was performed one hour after the last dose of heparin, then hourly until the reading was less than or equal to 150 seconds. Sheaths were removed as soon as possible following attainment of the target ACT.

Sheath removal was performed only by experienced practitioners to minimize any complications related to operator inexperience. The time of hemostasis (when arterial pressure has been completely removed without bleeding) triggered randomization. An opaque, sealed envelope, indicating the group to which the patient was randomized, was opened and the bedrest protocol initiated (Table 1). Pressure dressings were not recommended. A five pound sandbag could be used to remind the patient not to move the affected leg during the supine portion of bedrest.

Vital signs, pedal pulses and puncture site were checked prior to sheath removal and every 5–10 minutes after sheath removal until hemostasis was achieved, every 30 minutes for the first 2 hours after hemostasis, then hourly until completion of bedrest. Patients were then ambulated for 5 minutes under the direct supervision of a Registered Nurse. The patient then returned to bed and the puncture site and vital signs were assessed. If no complications were evident, the patient could ambulate in the room for the first 2 hours, as tolerated. Vital signs, pedal pulses and puncture site were checked every 15 minutes for the first hour after ambulation, then every hour for 6 hours, then as per ward routine. The final assessment for the presence of any complications was made at discharge, generally the day following PCI.

Vascular complications. The primary outcome events were puncture site complications, which were defined as major (requiring blood transfusion, surgical repair, ultrasound-guided compression or prolongation of hospital stay) or minor (not requiring specific therapy other than site compression). Vascular problems may include external bleeding, hematoma formation, retroperitoneal bleeding, compromise of circulation to the affected limb, pseudoaneurysm formation and formation of an arterio-venous fistula. Hematoma formation was quantified in centimeters using a disposable ruler. Timing of the occurrence of a vascular complication in relation to sheath removal (prior to sheath removal; after sheath removal, but prior to hemostasis) randomization (after hemostasis) and ambulation (during bedrest or ambulation) was noted.

Statistical analysis. The hypothesis of this trial was one of clinical equivalence, i.e., that there would be no difference in vascular complications between the control and experimental groups. The original sample size calculation was estimated using the method of Day,15 assuming an overall rate of vascular events of 4% and a 95% confidence interval (CI) width of ± 4%. After half the patients had been enrolled, an interim analysis showed that the rate of vascular events was < 1% overall and that a difference of 4% or less was ensured; at that point, the trial was terminated. Data analysis was performed using SPSS version 7.5.1. Baseline data were summarized using descriptive statistics. Continuous variables were compared using analysis of variance. Categorical variables are reported as proportions and the differences between groups compared using the Chi-square or Fisher’s exact test, as appropriate. A two-sided p-value of 0.05 was assumed to indicate statistical significance.

RESULTS

From March 3, 1997 to October 1, 1998, a total of 354 subjects were recruited for the study. Of these, there were 299 who completed the full protocol. In addition, there were 46 who were excluded because they required only diagnostic cardiac catheterization; nine were excluded for other secondary exclusion criteria [significant hypertension (n = 3); hemodynamic instability (n = 2); peri-procedural neurological event (n = 1); and physician preference (n = 3)]. Subjects were allocated as follows: ninety-nine to each of the 2 and 4 hour experimental groups and 101 to the 6 hour control group. Baseline demographics are shown in Table 2. The groups were well matched for important baseline variables.

The majority of patients had an acceptable ACT, allowing for sheath removal within 4 hours of the last dose of heparin. Sheath removal was performed by one of five physicians (41.4%) or one nurse practitioner (58.6%). A compression clamp was used in all cases. The median total clamp time was 45 minutes with no difference between groups (p = 0.39). Sandbags were used 97.3% of the time and pressure dressings were used in only 0.7% of patients. Vascular complications are summarized in Table 3.

Hematoma formation and external bleeding occurring prior to hemostasis do not reflect differences in bedrest because randomization was triggered by initial hemostasis at the puncture site. Hematoma formation was most common prior to sheath removal (30.8%), with a smaller number occurring from sheath removal to hemostasis (8.7%). Of the 14 hematomas occurring after randomization (4.7%), nine occurred during bedrest and 5 with ambulation; only 2 required repeat arterial compression. The overall hematoma rate was 44.1%, with the majority less than 5 x 5 cm in diameter. There were no significant differences between groups. External bleeding prior to sheath removal was defined as bleeding sufficient to soak at least two 4 x 4 squares of gauze. This occurred in 22.4% of patients; no one required a blood transfusion. External bleeding following randomization occurred in 3% of patients and resolved with further arterial compression. Again, there were no differences between groups.

Only one patient, a 38-year-old obese female who received abciximab, developed a pseudoaneurysm 24 hours after sheath removal. This patient had a large hematoma prior to sheath removal, which extended with rupture of the pseudoaneurysm requiring urgent surgical repair and prolonging her hospital stay. Another elderly female patient had her length of stay prolonged by 1 day due to a hematoma that limited ambulation. An ultrasound examination ruled out a pseudoaneurysm or other vascular complication and she was discharged the next day. 

A subgroup analysis (Table 4) was performed on the 14.4% of patients who received the intravenous glycoprotein (GP) IIb/IIIa receptor inhibitor abciximab. There were no significant differences in either major or minor vascular complications between patients receiving and not receiving abciximab.

DISCUSSION

The main finding of our study is that 2 hours of bedrest following sheath removal after PCI is safe and not associated with an increased risk of vascular complications compared with 4 or 6 hours. Patients anecdotally verbalized increased comfort with earlier movement, both in reduced time to turn over and to ambulate.

The intravenous GP IIb/IIIa inhibitor abciximab was not approved for use in our hospital until about 9 months into the study, which is reflected in the relatively low (14.4%) usage. Contemporary recommendations for prolonged bedrest with abciximab (6–8 hours following discontinuation of the infusion or about 20 hours of bedrest for most patients) seem excessive in light of our findings.

Comparison to other studies. At the time this trial was designed, there were no published trials examining bedrest times of less than 6 hours among the PCI population. Koch and colleagues^12,13 reported on two studies, the first comparing 4 and 12 hours of bedrest following PTCA in 907 patients and in a subsequent study of 300 patients assessing ambulation at 2 hours. In both studies, sheath removal (6 French) was performed immediately following the procedure and the dose of heparin was low (only 5,000 IU). There were no differences between the two groups and vascular complications were low. In addition, Keeling and colleagues¹¹ published a small trial involving 51 patients that supports 4 hours of bedrest following PCI.

Study limitations. The trial was designed to keep the bedrest time blinded until hemostasis at the puncture site had been achieved. At this time, due to the nature of the intervention, the trial became unblinded, possibly allowing for outcome ascertainment bias. Our study was relatively small, and as such, may be underpowered. Notwithstanding, the upper bound of our observed event rate (2/299, 0.67% with exact binomial, 95% CI = 0.08–2.4%) is in keeping with the 2.6% rate of serious vascular complications following “conventional” PCI as reported by Muller.¹⁴ Our findings with abciximab should be interpreted with caution given the vagaries of subgroup analyses and that patients were not randomized to abciximab. Finally, because this study was performed at only one center, the results may not be generalizable to other centers.

CONCLUSION

Patients undergoing femoral arterial puncture for PCI can safely ambulate as early as 2 hours after hemostasis is achieved. Importantly, no additional costs are required to implement this change in practice in comparison to the use of direct arterial closure devices. The findings from this trial may represent an opportunity for cost savings, since the use of direct arterial closure devices to allow for early ambulation can add significantly to overall procedural costs. The length of stay for patients in this trial (on average about 24 hours for elective patients) was unchanged from routine care post-PCI to allow for adequate follow-up. It is possible that patients ambulated earlier post-PCI may be able to be discharged sooner; this is an avenue for future investigation. Patients receiving glycoprotein IIb/IIIa inhibitors as part of their PCI may also be able to ambulate as early as 2 hours after hemostasis. Further study among this ever-increasing population is warranted.

Acknowledgments. We would like to acknowledge the contributions of the following individuals: K. Sridhar, MD, P. Teefy, MD, N. Huq, MD, E. Grech, MD, J. White, RN, and the Registered Nurses in the Cardiac Cath Lab, Coronary Care Unit, Telemetry Unit and Pre-admission Clinic at the Victoria Campus of the London Health Sciences Centre, London, Ontario, Canada.

1. Peet G, McGrath M, Brun JH, Hilton JD. Femoral arterial sheath removal post-PTCA: A cross Canada survey. Can J Cardiovasc Nurs 1995;6:13–19. 2. Barkman A, Lunse C. The effect of early ambulation on patient comfort and delayed bleeding after cardiac catheterization. Heart Lung 1994;23:112–117. 3. Keeling A, Taylor V, Nordt L, et al. Reducing time in bed after cardiac catheterization (TIBS II). Am J Crit Care 1996;5:277–281. 4. 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;15:1475–1483. 5. Lim R, Anderson H, Walters M, et al. Femoral complications and bedrest duration after coronary arteriography. Am J Cardiol 1997;80:222–223. 6. Lunse-Pooler C, Barkman A, Fenwich Bock B. Effects of modified positioning and mobilization on back pain and delayed bleeding in patients who had received heparin and undergone angiography: A pilot study. Heart Lung 1996;25:117–123. 7. Steffenino G, Dellavalle A, Ribichini F, et al. Ambulation three hours after elective cardiac catheterization through the femoral artery. Heart 1996;75:477–480. 8. Talley D. Further streamlining cited with 5 French diagnostic catheters. PTCA News for Practitioners 1995;10:3. 9. Vlasic W, Almond D. Research-based practice: Reducing bedrest following cardiac catheterization. Can J Cardiovasc Nurs 1999;10:19–22. 10. Fowlow B, Price P, Fung T. Ambulation after sheath removal: A comparison of 6 and 8 hours of bedrest after sheath removal in patients following a PTCA procedure. Heart Lung 1995;24:28–37. 11. Keeling AW, Fisher CA, Haugh KH, et. al. Reducing time in bed after percutaneous coronary angioplasty (TIBS III). Am J Crit Care 2000;9:185–187. 12. Kock KT, Pick JJ, de Winter RJ, et al. Early ambulation after coronary angioplasty and stenting with six French guiding catheters and low-dose heparin. Am J Cardiol 1997;80:1084–1086. 13. Kock KT, Pick JJ, de Winter RJ, et al. Two hour ambulation after coronary angioplasty and stenting with 6 Fr guiding catheters and low dose heparin. Heart 1999;81:53–56. 14. Muller D, Shami K, Ellis S, Topol E. Peripheral vascular complications after conventional and complex percutaneous coronary interventional procedures. Am J Cardiol 1992;69:63–68. 15. Day S. Clinical trial numbers and confidence intervals of pre-specified size. Lancet 1988;12:1427.