The Influence of Diabetes Mellitus on the Practice and Outcomes of Percutaneous Coronary Intervention in the Community: A Report

Diabetic patients with coronary artery disease (CAD) have poorer long-term prognosis with more revascularization procedures than non-diabetic patients regardless of the initial treatment strategy.1,2 The role of percutaneous versus surgical revascularization as the initial treatment strategy for diabetic patients remains controversial. The Bypass Angioplasty Revascularization Investigation (BARI) trial found that diabetic patients with Multivessel coronary artery disease randomized to an initial strategy of PCTA versus CABG surgery had significantly higher mortality rates (34.7% versus 19.1%) after five years.3 However, other studies have found similar 1-year and 5-year mortality rates for diabetic patients treated with initial percutaneous coronary interventions (PCI) or CABG surgery.4–7 The evidence also is mixed as to whether or not diabetes mellitus is an independent factor associated with short-term mortality. The literature is conflicted on this point, as studies have shown both higher and lower mortality rates among diabetic patients undergoing PTCA versus CABG surgery.8–11 As a result, for selected diabetic patients PCI continues to be the treatment strategy of choice. Short-term (in-hospital or 30-day) mortality rates among diabetic patients undergoing PCI is very comparable to the rates observed among non-diabetic patients undergoing a PCI treatment strategy.3,12,13 However, long-term outcomes, particularly restenosis rates and additional revascularization rates, are significantly higher for diabetic patients than non-diabetic patients undergoing PCI.9,10 Recent evidence indicates that diabetic patients receiving intracoronary stents have significantly better long-term (1–4 years) outcomes than diabetic patients undergoing PCI without stenting.14–17 Nevertheless, even with stents, diabetic patients have poorer long-term outcomes than non-diabetic patients.16,17 The objective of this study is to provide information on diabetic patients undergoing contemporary PCI in the community setting. First, this study will compare clinical demographics, procedure indications, lesion characteristics, and acute clinical outcomes for diabetic patients versus non-diabetic patients obtained for a series of consecutive community based PCI procedures. Second, this study will report long-term outcomes for diabetic patients undergoing PCI in the community setting, including repeat revascularization rates, evidence of disease progression, and other adverse clinical events. Methods Patient population — inclusion/exclusion. Local institutional review board approval was obtained prior to data collection. Data were prospectively collected and analyzed on all consecutive patients undergoing a PCI procedure on native coronary arteries in the four community based sites (see appendix). Patient enrollment commenced July 1, 1999 and concluded September 30, 2000. A total of 391 (10.9%) patients receiving PCI procedures during the study period were excluded from this study. Patients in whom the index procedure included treatment of a coronary artery bypass graft (n = 321), atherectomy of any vessel (n = 17), or missing information on patients’ diabetes status (n = 53) were excluded. No other patients were excluded. Data collection. Data were obtained from two electronic data collection protocols routinely used by HCA hospitals as part of their quality assurance and outcomes monitoring initiative. Baseline demographic, clinical, angiographic, device use, procedural and in-lab outcome information were recorded prospectively using an ACC-NCDR approved electronic data collection tool. All fields are defined with the approved data dictionary and the software contains internal logic to assure that all appropriate clinical information is completed for each procedure performed. Procedural lesion data include device use and type, lesion length, diameter, and location, and other angiographic characteristics, and success of each coronary lesion attempted. In-lab complications are defined and recorded according to ACC-NCDR standards. In-hospital procedures, complications, length of stay, and cost information were obtained from the HCA Casemix Administrative database. This database includes demographic, procedural, and diagnostic information on all patients hospitalized in any HCA hospital. Detailed information concerning both the CCL data collection tool and the HCA Casemix Administrative database has been described previously.18–21 Definition of diabetes mellitus. In this study a patient was defined as diabetic if, at the time of the initial procedure, they had a history of diabetes mellitus documented in their CCL medical record. This was independent of glycemic treatment. Definitions of clinical outcomes and adverse events. Angiographically successful revascularization of a lesion was defined as an absolute 20% reduction in lesion stenosis with final stenosis Study follow up. Follow-up utilization and long-term clinical adverse events were collected using two sources of information, as has been described previously.22 First, the two HCA data sets were used to prospectively determine health care services consumed by each patient at their initial study facility for the one-year period following the patient’s initial PCI procedure. Follow-up utilization data was examined to determine if a patient had had a diagnostic catheterization, an additional PCI or CABG procedure, an AMI hospitalization, or died during any additional hospitalization or outpatient CCL visit to their original hospital during the follow-up period. Second, 947 of the patients enrolled in this study completed a telephone survey 1 year following their initial CCL encounter. This survey asked respondents to indicate if they had had an additional diagnostic catheterization procedure, any additional PCI or CABG procedures, or a hospitalization associated with an acute MI at any health care facility during the year following their initial PCI procedure. A statistical analysis was conducted to examine if survey respondents were significantly different than non-respondents in terms of their coronary risk factors and the procedure indications at their initial CCL encounter. The analysis found only three statistically significant (p Statistical analysis. Discrete data are reported as the percent observed with selected characteristic, while continuous data are presented as the mean value ± standard deviation. Differences in the observed proportion of discrete variables were compared by Chi-squared analysis or Fisher’s exact test. Comparisons between the mean values of continuous variables were made with paired Student t-test. A multi-variable logistic regression analysis was estimated for each long-term adverse outcome to determine if the presence of diabetes mellitus increased the odds of an adverse event after controlling for gender and age. All analysis was performed with SAS Version 8.2 statistical software (SAS Institute, Carey, North Carolina). Results. Clinical characteristics and demographics. Table 1 presents the baseline clinical characteristics and demographics for the study patients by whether the patient was in the diabetic or non-diabetic group. Approximately 23.5% of the patients had diabetes. There is no statistical difference in the average age between the two study groups. However, diabetic patients are more likely to be female (40.4% versus 36.6%; p = 0.048). In addition, there is a significant difference in the racial make-up between the two study groups with diabetic patients being less likely to be Caucasian (86.0% versus 91.7%), and more likely to be Black (6.5% versus 3.2%) or Hispanic (4.5% versus 2.3%). Diabetic patients enrolled in this study were significantly more likely to have had previous revascularization, previous PCI (17.8% versus 11.7%; p = 0.036) and previous CABG (30.3% versus 26.3%; p Procedural indications, clinical presentation, and procedure information. Table 2 indicates that angina is the most common indication for all patients to undergo the index PCI procedure. Non-diabetics were more likely to present with a primary diagnosis of acute MI (24.4% versus 17.3%) or for treating a residual lesion following a MI (18.5% versus 14.1%). On the other hand, diabetic patients were more likely to present for treatment for a residual lesion following CABG (5.6% versus 3.3%) or following PCI (3.9% versus 2.5%). Of those patients with angina, diabetic patients were more likely than non-diabetic patients to have unstable angina (62.9% versus 57.3%) and more likely to have severe angina (Class III and IV, 62.9% vs. 57.3%). Table 2 also indicates that the index revascularization procedure was more likely to be performed as an elective procedure for diabetic patients than non-diabetic patients (90.0% versus 85.3%). Diabetic patients were more likely to have multiple lesions treated (29.8% versus 24.0%) and less likely to receive a coronary stent (88.9% versus 90.8%) during their index procedure than non-diabetic patients. Finally, while not reported in Table 2, this study found no significant differences (p Lesion characteristics. Table 3 describes selected lesion characteristics for the 4,092 lesions treated in the index procedures. Table 3 indicates that the distribution of lesion type and the average reference vessel diameter are significantly different between the two groups. In general, diabetic patients were more likely than non-diabetic patients to have the PCI procedure performed on a type A or type B1 lesion, while non-diabetic patients were more likely than diabetic patients to have the PCI procedure performed on a type C lesion. A greater proportion of diabetic patients had small (Acute and long-term outcomes. Despite the differences in clinical characteristics, demographics, procedural indication, clinical presentation, and lesion characteristics between diabetic and non-diabetic patients, Table 4 (Part A) indicates that there were no significant differences in any of the acute clinical outcomes (success or adverse events) between diabetic and non-diabetic patients. Overall, diabetic patients experience slightly lower in-hospital mortality rate (1.08 versus 1.29; p = 0.849), while slightly less than 1.1% in both groups undergo CABG surgery following PCI during their initial hospitalization. Nevertheless, diabetic patients had significantly longer hospital lengths of stay (3.8 days versus 3.4 days) than non-diabetic patients. The 1-year follow-up results reported in Table 4 (Part B) indicate that diabetic patients have more CAD progression than non-diabetic patients during the 1-year follow-up period. A comparison between the two groups indicates that there was a trend toward diabetic patients having more target lesion revascularization (TLR) with PCI at one year (13.6% versus 8.9%; p = 0.07) and target vessel revascularization (TVR) by PCI at one year (17.6% versus 12.7%; p = 0.058) than non-diabetic patients. Diabetic patients were significantly more likely than non-diabetic patients to have a second PCI procedure performed in any vessel (31.7% versus 23.0%; p = 0.012) or to undergo CABG surgery (10.1% versus 4.5%; p = 0.003) within one year of their initial PCI procedure. As a result, diabetic patients were significantly more likely than non-diabetic patients to have at least one additional revascularization procedure (38.2% versus 25.0%; p Estimated odds ratio. Table 5 reports the adjusted odds ratio (estimated using logistic regression analysis) for the variable indicating a patient’s diabetic status. The logistic regression models controlled for the patient gender and whether or not the patient was age 65 or older, in addition to the diabetic status of the patient. Overall the model confirms the long-term univariate findings in Table 4; that diabetic patients are more likely than non-diabetic patients to have each of the adverse events during the one-year follow up period. As in Table 4, the estimated odds ratio was not significantly (p Discussion. Patients with diabetes mellitus and coronary artery disease experience significantly more long-term adverse outcomes than non-diabetic patients with CAD independent of the initial revascularization strategy. Nevertheless, diabetic patients will continue to need revascularization procedures for relief of angina pectoris, and improvement in functional health status and quality of life. There is substantial evidence in the literature to indicate that diabetes mellitus patients can be successfully treated using an initial treatment strategy of PCI.1,4–7 This paper reports and compares clinical characteristics, demographics, procedure indications, lesion characteristics, and acute and long-term outcomes of diabetic and non-diabetic patients obtained from a prospectively collected data set of consecutive community-based PCI procedures. Several important conclusions can be drawn from this study about the demographic and clinical differences between diabetic and non-diabetic patients undergoing PCI treatment strategy. First, diabetic patients undergoing PCI are significantly more likely to be female and from a minority race than non-diabetic patients. In particular, diabetic patients are nearly twice as likely to be Black or Hispanic than non-diabetic patients. Second, at the time of their index encounter coronary risk factors and procedure indications suggest that diabetic patients have more advanced and extensive CAD than the typical non-diabetic patient. For example, diabetic patients are more likely to have each of the major cardiac risk factors and already have had a previous revascularization procedure. Third, there are significant differences between diabetic and non-diabetic patients in the number of lesions treated and their reference vessel diameter during the index encounter. Diabetics are more likely to have multiple lesions treated within a single vessel. This supports the presence of more diffuse coronary disease in diabetics. Additionally, diabetics have both smaller reference vessels and more small vessels treated. However, the treated lesions of diabetics are less complex, as diabetics are more likely to have type A or B lesions intervened upon. Our findings indicate that acute clinical outcomes, both in-lab and during the index hospitalization, are relatively rare, and statistically equivalent for both diabetic and non-diabetic patients. These findings are consistent with a growing literature indicating that short-term outcomes between diabetic and non-diabetic patients undergoing PCI procedures are comparable.23 Conversely, in our study, diabetics demonstrated significantly worse long-term outcomes. Both univariate and adjusted logistic regression results establish worse 1-year outcomes for diabetics. Diabetics, experienced more restenosis, as reflected by the higher incidence of repeat target lesion PCI (OR = 1.6; p = 0.06). However, restenosis alone does not account for the markedly worse long-term outcomes amongst diabetics. Diabetics also experienced more coronary artery disease progression. This is evident in the higher incidence of TVR (OR = 1.5; p = 0.07), non-target vessel PCI (OR=1.6; p = 0.013), and CABG (OR = 2.4; p = 0.003) at one year. Furthermore, the incidence of repeat PCI within 30 days, or CABG during the index hospitalization was equivalent in both groups, supporting the notion of similar acute, but poorer long-term outcomes amongst diabetics. Limitations. There are two limitations that warrant discussion. The first limitation is that all the data in this study were self-reported, both by patients and physicians. There were no clinical or independent core lab validations of clinical or angiographic variables. However, there are a detailed set of decision rules built into the data reporting system to ensure procedure and product validation, as well as validation of data during quality reviews. Additionally, this level of data validation is similar to other data sources reported in the literature but not associated with clinical trials.2,8,23 A second limitation is that not all patients enrolled in this study completed the surveyed for long-term utilization of cardiovascular services. However, we conducted a statistical analysis to determine if the sample of individual who completed the follow-up survey were representative of the entire study population. Our analysis indicates that there are no consistent statistically significant differences in terms of coronary risk factors or procedure indications between the follow-up sample and the study population. In addition, we re-estimated the logistic regression models using the one-year follow-up data on the entire study population. This analysis also confirms our major result that diabetic patients were more likely than non-diabetic patients to have adverse events. As a result, we believe that the long-term revascularization and adverse outcome rate reported are representative of community-based practice and the treatment of diabetic patients. Conclusions. This study validates that diabetic patients have similar acute outcomes as non-diabetic patients undergoing PCI in the community setting. Additionally, our study supports the widely held notion that diabetic patients undergoing PCI treatment experience poorer long-term outcomes than non-diabetics patients. Our findings suggest that the poor long-term outcomes observed among diabetic patients is due to both increased restenosis and more native coronary artery disease progression. Overall the results from this study suggest that future efforts for reducing the extent of disease progression in diabetics will continue to require efforts to improve primary and secondary prevention. Acknowledgments. We would like to thank the following: Marsha Colwell, Administrative Director, and Michael Ware, MD, Redmond Regional Medical Center, Rome, Georgia; Gail Grasso, Administrative Director, and Jay Midwall, MD, JFK Medical Center, Atlantis, Florida; Paul Jenson, Chief Operating Officer, and Dwight Reynolds, MD, OU Medical Center, Oklahoma City, Oklahoma.

1. Mathew V, Holmes D. Outcomes in diabetics undergoing revascularization: The long and the sort of it. J Am Coll Cardiol 2002;40:424–427. 2. Thourani V, Weintraub W, Stein B, et al. Influence of diabetes mellitus on early and late outcome after coronary artery bypass grafting. Ann Thorac Surg 1999;67:1045–1052. 3. BARI investigators. Influence of diabetes on 5-year mortality and morbidity in a randomized trail comparing CABG and PTCA in patients with multivessel disease. Circulation 1997;96:17616–17619. 4. Detre K, Guo P, Holubkov, et al. Coronary revascularization in diabetic patients: A comparison of the randomized and observational components of the Bypass Angioplasty Revascularization Investigation (BARI). Circulation 1999;99:633–640. 5. Barsness G, Peterson E, Ohman E, et al. Relationship between diabetes mellitus and long-term survival after coronary bypass and angioplasty. Circulation 1997;96:2551–2556. 6. Serruys P, Sousa J, Jatene A, et al. Arterial Revascularization Therapies Study Group. Comparison of coronary artery bypass surgery and stenting for the treatment of multivessel disease. N Engl J Med 2001;344:1117–1124. 7. Niles N, McGrath P, Malenka D, et al., and the Northern New England study group. Survival of patients with diabetes and multivessel coronary artery disease after surgical or percutaneous coronary revascularization: Results of a large regional prospective study. J Am Coll Cardiol 2001;37:1008–1015. 8. Thourani V, Weintraub W, Stein B, et al. Influence of diabetes mellitus on early and late outcomes after coronary artery bypass grafting. Ann Thorac Surg 1999;67:1045–1052. 9. Salomon N, Page U, Okies J, et al. Diabetes mellitus and coronary artery bypass. J Thorac Cardiovasc Surg 1983;85:264–271. 10. Lawrie G, Morris G, Gleser D. Influence of diabetes mellitus on the results of coronary bypass grafting. JAMA 1986;256:2967–2971. 11. Risum O, Abdelnoor M, Svennevig J, et al. Diabetes mellitus and morbidity and mortality risk after coronary artery bypass surgery. Scand J Thorac Cardiovasc Surg 1996;30:71–75. 12. Kleiman N, Lincoff A, Kereiakes D, et al. Diabetes mellitus, glycoprotein IIb/IIIa blockade, and heparin: Evidence for a complex interaction in a multicenter trial. Circulation 1998;97:1912–1920. 13. Stein B, Weintraub W, Gebhart S, et al. Influence of diabetes mellitus on early and late outcome after percutaneous transluminal coronary angioplasty. Circulation 1995;91:979–989. 14. Kornowski R, Mintz G, Kent K, et al. Increased restenosis in diabetes mellitus after coronary interventions is due to exaggerated intimal hyperplasia: A serial intravascular ultrasound study. Circulation 1997;95:1366–1369. 15. Van Belle E, Bauters C, Hubert E, et al. Restenosis rates in diabetic patients: A comparison of coronary stenting and balloon angioplasty in native coronary vessels. Circulation 1997;96:1454–1460. 16. Abizaid A, Kornowski R, Mintz G, et al. The influence of diabetes mellitus on acute and late clinical outcomes following coronary stent implantation. J Am Coll Cardiol 1998;32:584–589. 17. Van Belle E, Perie M, Braune D, et al. Effects of coronary stenting on vessel patency and long-term clinical outcome after percutaneous coronary revascularization in diabetic patients. J Am Coll Cardiol 2002:40:410–417. 18. Heuser R, Houser F, Culler SD, et al. Do coronary stents improve clinical outcomes during the first six months following revascularization? A retrospective study of 6,671 patients comparing coronary stenting and balloon angioplasty. J Intervent Cardiol 2000;12:392–398. 19. Becker ER, Cohen D, Culler SD, et al. Benchmarking CCLoratories: The impact of patient age, gender, and risk factors on variable costs, device costs, and total time and procedural time in 53 catherization laboratories. J Intervent Cardiol 1999;11:533–542. 20. Cohen D, Becker ER, Culler SD, et al. Impact of patient characteristics, complications and facility volume on the cost and time of cardiac catherization and coronary angioplasty in 70 catherization laboratories. Am J Cardiol 2000;86:595–601. 21. Brown PP, Mack MJ, Simon AW, et al. Comparing clinical outcomes in high-volume and low-volume off-pump coronary bypass operation programs. Ann Thorac Surg 2001;72:S1009–S1015. 22. Kugelmass AD, Cohen DJ, Houser F, et al. Practice and outcomes of percutaneous coronary intervention in the community before drug-eluting stents: A report from the HCA database. J Invas Cardiol 2003;14:121–127. 23. Laskey WK, Selzer F, Vlachos HA, for the Dynamic Registry Investigators. Comparison of in-hospital and one-year outcomes in patients with and without diabetes mellitus undergoing percutaneous catheter intervention (From the National Heart, Lung, and Blood Institute Dynamic Registry). Am J Cardiol 2002;90:1062–1067.