Rural Community Hospital Care NSTEMI/UA: Implications for Triage and Treatment

Douglas Trenkle1* and Cheng Peng2 1American Board of Internal Medicine, American Board of Clinical Lipidology, maine, USA 2Department of Mathematics and Statistics, University of Southern Maine, Portland, ME 04104, USA *Corresponding author: Douglas Trenkle, American Board of Internal Medicine, American Board of Clinical Lipidology, 651 Morgan Bay Road, Surry, maine 04684, USA, Tel: 207-461-6724, E-mail: douglas.trenkle@gmail.com

The time interval currently recommended for early intervention for intermediate and high TIMI (Thrombolysis in Myocardial Infarction) risk score patients with NSTEMI/UA (non-ST elevated MI/unstable angina) is 4 to 24. Physician uncertainty regarding the decision for intervention in patients with NSTEMI/UA disposition is reflected by data from Can Rapid risk stratification of Unstable Angina patients Suppress Adverse outcomes with Early implementation (CRUSADE) where less than half of 19,238 NSTEMI patients were transferred for revascularization, and less than 20% were sent within forty eight hours of presentation [17]. CRUSADE also showed a paradoxical risk: transfer relationship, with 41% of low risk patients and 12.5% of high risk patients transferred in the early period. Our goal was to assess six month readmission and mortality of transferred vs. non transferred patients with NSTEMI/UA who presented for initial treatment at their community hospital.

Methods
Acting as an independent review board (IRB), the Hancock County Aging Study approved this study. Electronic medical records (EMR) of all NSTEMI/UA patients, (ICD9 codes 411.1 and 410.7) treated from January 1, 2009 through December 31, 2009 were identified. Variables for the cohort were coded and retrospectively evaluated for cardiovascular risk. TIMI risk score for morbidity and mortality has widespread acceptance with predictive power for ACS. Patients' risk score was retrospectively determined examining the patient's status at the time of admission to the community hospital. NSTEMI/UA was defined in those patients who presented to the emergency room with symptoms of coronary insufficiency including back/chest pain or pressure, nausea, diaphoresis, or shortness of breath. These symptoms were considered acute coronary insufficiency and NSTEMI when they were associated with positive troponin (T >0.03 ng/ml) and those troponin values varied according to the pattern of myocardial necrosis, with or without abnormal ST depression and T wave inversion. EKG interpretations were performed by experienced and credentialed board certified internists, with confirmatory review of all EKGs by the author/principle investigator. Patients with STEMI (ST elevation myocardial infarction) ACS were excluded from the review, utilizing the standard electrocardiographic criteria for STEMI. Vital signs were defined on entry to the emergency room. Since cholesterol is an acute phase reactant, total cholesterol, HDL cholesterol, LDL cholesterol and triglycerides were defined by the last value in the record prior to admission with calculated ratios of total cholesterol/HDL cholesterol, non-HDL cholesterol, and triglyceride/HDL cholesterol. Other lab values determined at the time of admission include serial troponins, glucose, and creatinine. The original risk measures of TIMI and NCEP ATP III have three and five levels respectively. For statistical purposes the categorical covariants table for readmission and death combined Low and Intermediate categories to define a new TIMI variable, and also combined NCEP ATP III categories with a new model for NCEP ATP III.
Since the key outcome variables are categorical, we used binary logistic regression models to assess the association between the outcome variable and other risk covariates. These measures of association reported in the result section are odds ratio and the cstatistic. We also choose the p-value calculated from the Wald chisquared test statistic to justify the significance of variables in the logistic regression models. Both parametric and non-parametric analysis of variance (ANOVA) procedures were used to assess the association between some categorical variables and continuous variables. Two p-values (based on parametric normal test and nonparametric Kruskal-Wallis ANOVA) are reported. The linear regression model was used to identify the potential significant association between TC/HDL ratio and other variables. All analyses were carried out using SAS® 9.2 (SAS Institute Inc).

Results
Seventy three patients were admitted with NSTEMI/UA to a community hospital from January 1, 2009 through December 31, 2009 (Table 1). Thirty seven patients (51.4%) were transported twenty six miles for urgent revascularization, and thirty two received PCI (Table  4). Less than 17% of transferred patients with NSTEMI/UA received PCI (percutaneous intervention) in under 36 hours. Twenty three (31.8%) of the entire cohort were over 80 years of age, and ten greater than 90 years. The mean age was 71.5 years, and the mean age at time of death was 85.2 years. Overall mortality was 18.0% in 6 months, 8.3% due to cardiac disease. Statistically significant associations with mortality occurred with age (p=0.0037), and systolic (p=0.0008) and diastolic (p=0.001) blood pressures ( Table 2). Readmission had near statistical validity with TC/HDL ratio (p=0.0644). There was no association of TIMI or modified TIMI risk score with death or readmission (Table 3). Statistically significant reduced mortality was associated with transfer for catheterization (p=0.0005) (

Discussion
Our findings suggest that there is a variability in care and nonadherence to guidelines for patients admitted to a community hospital with non-ST elevated MI/unstable angina. There were delays in both transfer and timing of percutaneous intervention. These findings were mirrored by the multi-center CRUSADE study of 19,238 patients. Percutaneous intervention in our study did reduce mortality at six months, but had no impact on readmission rates. In addition our medication review noted that glycoprotein IIb/IIIa therapy was chosen for only 4.1% of all patients (3 of 73). Currently eptifibatide is approved by the FDA for non-ST elevated MI/unstable angina, whether or not percutaneous intervention is performed.
Physicians strive to comply with guidelines and appropriate utilization of services, yet are challenged by different therapies and recommendations for optimal treatments of non-ST elevated MI/UA. In a community of 1000 residents, eight of nine patients who require hospitalization are admitted to community hospitals, and not teaching institutions [18]. Knowing the importance of collaboration and a team approach in providing care, we suggest a prospective study using a model similar to that of Project Leonardo [19]. This multi centered study utilized nurse care managers and current software to improve patient care. Adapting this model to acute coronary syndromes in community hospitals would also predictably serve to improve outcomes.