Testing a Risk Predictor Tool: Can It Identify Patients at Higher Risk for Vascular Groin Complication Post Cardiac Cath or PCI?

*Zebuline Koran, RN, APN, DNP, is currently employed by Wheaton Franciscan Healthcare, Racine, Wisconsin.

Mike Fronczak can be reached at mfronczak@nch.org.

About the Northwest Community Hospital Cath Lab

The Northwest Community Hospital Cath Lab conducted 1283 combined cardiac cath and PCI procedures in 2013, with a staff size of 8 registered nurses, 6 radiologic technologists, a secretary, and a nurse manager.

Certain patient attributes can help predict which individuals have a higher risk for developing a groin complication following a cardiac cath or PCI. While clinicians have acknowledged these factors and have made modifications in their practice to reduce or prevent complications, the risk for a groin complication has not yet been eliminated.

Northwest Community Hospital looked more closely into risk prediction post cath or PCI by evaluating a tool created to predict groin complications.

According to the American College of Cardiology National Data Registry (ACC-NCDR), in 2010, approximately 1.3 million diagnostic cardiac catheterizations and approximately 627,000 percutaneous coronary interventions (PCIs) were performed in the United States.¹ There are recognized risk factors that have been shown to identify a population of patients who are at higher risk for vascular complications following a cardiac cath or PCI. These risk factors include:  age >70, female, lower body mass index (BMI), hypertension, heart failure, renal failure, diabetes, peripheral vascular disease, anticoagulant agents, and a venous sheath.^2-4 Most acute complications occur during the first 2 hours and can occur up through the next 24 hours following a catheterization.⁵  

During the spring and summer of 2010, an increased number of groin complications were noted at the Northwest Community Hospital, a 496-bed hospital located in the northwest suburbs of Chicago. The cardiac catheterization lab specifically noted an increase in hematomas and pseudoaneurysms. The rise in complications happened to coincide with the trial of a new vascular closure device, and uncertainty existed as to whether the increased number of groin complications was a result of the new device or due to post procedural groin management. It was decided that our lab should review post procedural groin management more closely. A literature search soon uncovered a “Cardiac Cath Lab Risk Predictor Tool” (CCRPT), created to assess and calculate the risk score for an individual undergoing a cardiac cath or PCI.⁶ 

In December 2010, the Northwest Community Hospital cath lab used the CCRPT in a pilot study to determine if the tool could be used to help identify patients at higher risk for a groin complication. This pilot study took all comers, which included patients with either a vascular closure device or manual compression to achieve hemostasis. At the end of the pilot study, questions arose regarding what collected data meant for improving outcomes. From this pilot study and with the help of our nursing research department, we developed a new and more structured study in order to determine if the CCRPT is a valid tool for predicting groin complications. A description of the new study follows.  

About the CCRPT

Sedlacek and Newsom developed a “Cardiac Catheterization Risk Predictor Tool” (CCRPT) in 2010, based on review of the literature and a retrospective study, noting that a CCRPT is a valuable tool that can help predict patients who may have a greater risk for developing vascular complications.⁶ The goal of the predictor tool is to identify patients who are at risk of developing bleeding complications, which may decrease “attributable risk” and mortality.⁶  

The CCRPT lists 13 risk factors and assigns a number relevant to each risk factor. A score of 8.5 or higher places a patient at higher risk for potential groin complication.⁶ 

Vascular complications following a cardiac cath or PCI procedure include: 

• Hematoma -— any swelling or palpable mass;

• Small: 1-5 cm in diameter
• Large: >5 cm in diameter

•  Bleeding occurrence: >3.0 g/dl drop baseline in Hgb, or;
• Retroperitoneal bleed;
• Pseudoaneurysm;
• Arterio-venous (AV) fistula;
• Femoral artery dissection;
• Thrombosis: occlusion of femoral vessel.^2,4,7

It should be noted that oozing is not included, because it is difficult to quantify. Ecchymosis is also not included; however, it is a reminder to the patient of their procedural experience. The benchmark for incidence of all complications, including major adverse events such as stroke, death, renal failure, and vascular complications, is no more than 1% for a cardiac cath and 3% for PCI, as established by the ACC.²

Cath lab staff and physicians are aware of the risk factors, but no universal standard of care exists for the patient when manual pressure is used for sheath removal post procedure. Many hospitals use similar protocols that are “passed or handed down” within the cardiac cath lab community.^2,4 Each facility develops their “best practice” protocols for sheath removal, patient management, and assessment following sheath removal.⁷ Most in the cath lab community agree that a consistent approach to groin management following a cardiac cath or PCI procedure is necessary, and that staff need to develop safe protocols of care for these patients that are research- and evidence-based.⁴ However, most practice guidelines for care are based on expert opinion and are not evidence-based.² Researchers have not been able to document relationships between sheath removal techniques, closure devices, antiplatelet medications, patient characteristics, and complication rates in post-PCI patients. Still, based on evidence in the literature, a population of patients does exist that requires a much longer initial compression time post sheath removal, a longer time on bed rest, and more frequent groin assessment following a cardiac cath or PCI procedure.² The CCRPT (Figure 1) attempts to identify this population of patients.

Evaluating the CCRPT score at Northwest Community Hospital

Our study sought to: 

1. Identify whether a relationship exists between the calculated score of the CCRPT and the development of vascular complications following sheath removal by manual compression. 
2. Validate the results of the previous (original) study on the effectiveness of the CCRPT for identifying patients for groin complications.

The study results would also determine whether specific patient factors within the risk prediction tool are able to predict a higher risk for a vascular complication. The hope to reduce the number of vascular groin complications, increase the timely recognition of complications, decrease length of stay and associated costs, and optimize patient outcomes. Confirming the accuracy of the CCRPT would permit us to identify higher risk patients in advance and thus modify our current post procedure protocol for these particular patients, increasing manual compression time during sheath removal and the frequency of assessments during the recovery phase. 

Study objective

The objective of the study was to determine if there was a vascular complication following the procedure and compare that to the cumulative score of the CCRPT. Only patients that had a sheath removed using manual pressure to achieve hemostasis were used for data collection. 

Study methods

Once the patient was scheduled and consented for the cardiac cath or PCI, the nurses and x-ray technologists in the CCL staff reviewed the patient’s medical record and determined which risk factor(s) applied on the CCRPT. A cumulative score was calculated and written on the bottom of the table. Certain variables were weighted on a scale of 1.5, 2, or 3, then the full amount totaled for a final score.  Variables (score following each) included:

• Patient age (1.5, 2, or 3); 
• Gender (female increased risk to 2); 
• Body surface area (smaller body surface area increased risk to 2); 
• Presence of heart failure (1.5); 
• Blood disorders (1.5);
• Peripheral vascular disease (1.5); 
• Chronic obstructive pulmonary disease (1.5); 
• Renal disease (or creatinine greater than or equal to 2mg/dl) (2); 
• Whether the procedure was emergent (2);
• Number of lesions treated (more than 2 lesions is weighted at 2);
• Use of clopidogrel (pre or intra procedure) (1.5);
• Use of glycoprotein IIb/IIIA inhibitors (pre or intra procedure) (2), and whether the patient was treated for a myocardial infarction or cardiogenic shock (2).

Additional information that was collected and listed on the page included the procedure type and duration, groin complication: yes or no, and type of complication: hematoma, bleeding, arteriovenous (AV) fistula, pseudoaneurysm, retroperitoneal bleed, artery dissection, loss of distal pulse, or “other”. 

Once hemostasis was obtained, the patient was observed for development and type of vascular complication. The patient was transferred to an appropriate unit where bed rest and assessments would continue. The staff nurses on the unit continued to monitor the groin site and vital signs according to physician’s orders. The physician ordered bed rest and vital sign monitoring as a common order set every 15 min x 4, 30 min x 2, 1 hour x 4 and 4 hours x 2. Groin site monitoring was continued until the patient was discharged, up to a 5-day period post procedure. Nurses documented their observations in the electronic medical record (EMR) and on the CCRPT sheet that followed the patient to the floor throughout their hospital stay. Several point people were utilized on the nursing units to help with notification of vascular complications. The CCRPT checklist contributed to a fundamentally consistent evaluation by staff, since we counted any hematoma or bleeding event that occurred. 

Results and findings

There were 179 manual compression patients with complete data. The mean CCRPT score was 6.18, with a standard deviation of 3.48. The internal consistency of items, as measured by Cronbach alpha, was 0.37; this is very low, but reflected the fact that the CCRPT scale included many low-incident indicators, and it was unlikely that any patients had all indicators (the maximum score was 15, obtained by one patient) (Figure 1). 

The mean CCRPT score for the 24 patients who developed groin complications was 8.48 with a standard deviation of 3.16, which is consistent with the mean in the original study.⁶ The 155 patients with no groin complications had a mean CCRPT score of 5.82, with a standard deviation of 3.40. This was higher than the mean of 4.5 in the original study for the non-complication group.⁶ A t-test comparing the two groups gave t (177) = -3.60, P < 0.001. This was consistent with the original study, which found P < 0.003.⁶ The mean scores of the two groups were significantly different, with the groin complication patients higher. Figure 2 shows that the groin complication patients had CCRPT scores that were almost as well distributed as the no post procedural groin complication patients.

Table 1 compares prediction of risk to actual rate of groin complication. Using the CCRPT score of 8.5, considered high risk by the original study, against the groin complication indicator, shows an increased number of errors in risk assessment rather than accurate predictions. Fifty-five (55) patients were identified as high risk, yet only 14 (26%) had groin complications, with 74% developing no groin complications. This is a high rate of false positives. Of 24 patients who had groin complications, 10 (42%) were classified as low risk per the CCRPT, meaning that these patients would be overlooked and would be at risk for a groin complication.

The total CCRPT score appeared to be influenced the most by the following variables: emergent procedures; MI or cardiogenic shock; and pre- or intra-procedural use of glycoprotein IIb/IIIa inhibitors. However, these were not necessarily the variables that distinguished a groin complication patient. Separate t-tests were run against each CCRPT variable. The variables with significant differences, at P < 0.10, were: 

• Age;
• Lower extremity vascular disease;
• Renal failure; 
• Pre- or intra-procedural use of glycoprotein IIa/IIIb inhibitors. 

Discussion 

The CCRPT is not accurate enough for decisions about post-procedural care to prevent potential groin complications. In both studies, the mean CCRPT score of the patients that developed a groin complication was close to 8.5. However, in our study, the instrument had too many false negatives (10/179) and far too many false positives (41/179). False positive patients would be given additional post procedural care that was not warranted, reducing staff availability for other patients or procedures. Additionally, the false negative patients would receive only nominal post procedure care.

The sample size in Sedlacek and Newsom’s retrospective study consisted of 20 patients in each group; those that did develop complications and those that did not. Their conclusion is based on the mean score of 8.5 for patients who developed a groin complication. Also worth noting, 12 of the 20 patients in Sedlacek and Newsom’s study who developed bleeding complications scored below the mean of 8.5. Sedlacek and Newsom did state that their sample size was small and indicated that the CCRPT needed further testing. 

Limitations

Some of the limitations of our study included the following:

1. Sheaths are pulled in multiple locations in the hospital post procedure. Although there is a general technique, not everyone applies pressure with the same manner or duration.
2. Different departments may use different approaches to post procedure care, i.e., using a sandbag.
3. There was no differentiation for a myocardial infarction (MI). The different classifications are ST-elevation MI (STEMI) or non-ST elevation MI (NSTEMI). This variable may be over- or under-used.
4. There was no account for the use of warfarin or any of the newer pharmacology used for atrial fibrillation.
5. Use of a venous sheath, intra-aortic balloon pump, or sheath size was not considered.

Conclusion

There may be a need for predicting patients at risk of groin complication with a risk predictor tool, and so the CCRPT may be a useful step in that direction. However, our findings indicate that a common-sense approach should continue to be applied to each individual patient post cardiac cath and/or PCI. Future possibilities may exist for additional research to develop an effective risk predictor tool that may help in the creation of a universal standard of care following cardiac procedures.

Acknowledgements

Gloria Reidinger and Zeb Koran for their assistance in research fellowship.

Jeanne Richmond and Julie Hunter for helping with the logistics in the study and John McConnell for providing the statistics for the research project.

NCH cath lab staff, ICU, and CSU nurses for documentation and care post procedure.

References

1. American College of Cardiology-National Cardiovascular Data Registry (NCDR). Retrieved from https://www.ncdr.com/webncdr/cathpci/.
2. Dumont CJ, Keeling AW, Bourguignon C, Sarembock IJ, Turner M. Predictors of vascular complications post diagnostic cardiac catheterization and percutaneous coronary interventions. Dimens Crit Care Nurs. 2006 May-Jun; 25(3): 137-142.
3. Gonzales L, Fields W, McGinty J, Gallo AM. Quality improvement in the catheterization laboratory: redesigning patient flow for improved outcomes. Crit Care Nurse. 2010 Apr;30(2): 25-32. doi: 10.4037/ccn2010832.
4. Sulzbach-Hoke LM, Ratcliffe SJ, Kimmel SE, Kolansky DM, Polomano R. Predictors of complications following sheath removal with percutaneous coronary intervention. J Cardiovasc Nurs. 2010 May-Jun; 25(3): E1-8. doi: 10.1097/JCN.0b013e3181c83f4b.
5. Harper J. Post-diagnostic cardiac catheterization: development and evaluation of an evidence-based standard of care. J Nurses Staff Dev. 2007 Nov-Dec; 23(6): 271-276.
6. Sedlacek MA, Newsome J. Identification of vascular bleeding complications after cardiac catheterization through development and implementation of a cardiac catheterization risk predictor tool. Dimens Crit Care Nurs. 2010 May-Jun; 29(3): 145-152. doi: 10.1097/DCC.0b013e3181d24e31.
7. Dressler D, Dressler K. Caring for patients with femoral sheaths: after percutaneous coronary intervention, sheath removal and site monitoring are the nurse’s responsibility AJN, American Journal of Nursing. 2006 May; 106(5): 64A-64H.