Barriers to obtaining diagnostic testing for coronary artery disease among veterans

Objectives. We sought to identify factors associated with appointment nonattendance for diagnostic testing of coronary artery disease among veterans. For patients with possible heart disease, appointment nonattendance may seriously compromise short- and long-term outcomes. Understanding factors associated with nonattendance may help improve care while reducing inefficiency in service delivery.Methods. We surveyed patients who attended (n = 240) or did not attend (n = 139) a scheduled cardiac appointment at a midwestern Veterans Administration medical center. Multivariable regression models were used to assess factors associated with nonattendance.Results. Younger age, lower income, unemployment, and longer wait times for appointments were predictive of nonattendance. Nonattenders reported fewer cardiac symptoms and were more likely to attribute their symptoms to something other than heart disease. Nonattendance was also associated with a coping style characterized by avoidance of aversive information. Logistical issues, fear of diagnostic procedures, disbelief that one had heart disease, and medical mistrust were some of the reasons given for missed appointments.Conclusions. Appointment nonattendance among veterans scheduled for cardiology evaluation was associated with several important cognitive factors. These factors should be considered when one is designing clinical systems to reduce patient nonattendance.Missed appointments, or nonattendance, by patients are common in most health systems. Estimates of nonattendance vary, but in general, rates range from 5% to 30% for primary care.^1–8 Given that patient nonattendance poses problems for both practitioners and patients, extensive research efforts have been focused on identifying factors that are associated with missed appointments. Numerous studies have been undertaken to examine nonattendance and its negative impact on patients and practitioners,^1–28 but almost none have focused on patients who miss appointments for testing or treatment for coronary artery disease (CAD). One study that examined nonattendance among patients with various chronic diseases found that patients with CAD were more likely to miss appointments compared with patients with other chronic diseases.²³ For patients with possible heart disease, a failure to follow through with appointments may seriously compromise patient short- and long-term outcomes.^27,28 Identifying the reasons underlying nonattendance among this at-risk population is therefore important.We examined whether patient features (e.g., sociodemographic characteristics, health status, attitudes, symptom perceptions, health-seeking behaviors, and coping styles) and clinic features (e.g., length of time between scheduling and time to appointment) were associated with nonattendance among patients scheduled for CAD diagnostic testing whose primary care was received through the Veterans Administration (VA). The study was guided by the existing literature on patient nonattendance as well as by attribution theory.On the basis of past research, we posited that individuals of lower socioeconomic status, those who were younger, and those who had better general health would be more likely to miss appointments. As well, the authors of several studies have reported that patients with stronger social support networks are more likely to keep appointments.^29,30 We also hypothesized that nonattendance would be more common among patients who reported less trust in health care³¹ and differences in health-seeking behaviors. Specifically, individuals who avoided information would be less likely to keep appointments or adhere to screening regimens.^32,33 This same finding has also been reported in individuals who prefer a more active role in health care decisionmaking.³⁴ We therefore measured treatment decisionmaking preferences.Attribution theory suggests several additional variables possibly related to nonattendance based on individuals'' fundamental assumptions and beliefs about their world, which form the basis of their attitudes and interpretations of events.^35,36 This theory is suggestive that individuals who are more fatalistic in their attitudes, meaning that they believe events in life are predetermined, will be more likely to be nonattenders. This hypothesis is consistent with previous findings that fatalistic individuals were less likely to be screened for cancer.³⁷ On the basis of similar reasoning, we also hypothesized that individuals who expressed higher religiosity would be more likely to be nonattenders.³⁸ Finally, we measured individuals'' experiences of CAD symptoms and their beliefs about the causes of those symptoms. Specifically, nonattendance was expected to be more frequent among patients who experienced fewer symptoms, less severe symptoms, or those who attributed their symptoms to something other than CAD.     

Optimal Bayesian probability levels for hospital report cards

There is a growing trend towards the production of “hospital report-cards” in which hospitals with higher than acceptable mortality rates are identified. Several commentators have advocated for the use of Bayesian hierarchical models in provider profiling. These methods are frequently based upon the posterior probability that a hospital’s mortality rate exceeds a specific benchmark. However, the minimum probability level required for classifying a hospital as having higher than acceptable mortality has never been formally justified. We developed Bayes Rules for determining optimal probability levels so as to minimize mean posterior costs associated with false classifications under specific loss functions. Using Monte Carlo simulation methods we then determined the ability of posterior tail probabilities of unacceptable performance to accurately identify hospitals with higher than acceptable mortality.     

The reliability of probability analysis in the prediction of coronary artery disease in two hospitals

To assess the accuracy of the Bayesian computer program CADENZA for the prediction of coronary artery disease, the authors examined the probabilities generated by the application of this program to the clinical and noninvasive test results of 303 patients in a private referral center and 199 patients in a veterans' hospital. These probabilities were compared with those produced by applying a six-variable discriminant function derived by logistic regression at the private referral center. Two statistical approaches were employed in evaluating the relative performances of the Bayesian program and the discriminant function. The first of these involved the sorting of patients in both test groups into ascending deciles of probability and comparing expected probability with observed angiographic disease prevalence in each decile. The second involved the calculation and comparison of a standardized reliability measure. The latter was significantly lower for the discriminant function both at the private hospital (0.200 for the discriminant function versus -17.5 +/- 1.96 for the Bayesian program) and at the veterans' hospital (-0.8 +/- 1.96 for the discriminant function versus -11.3 for Bayesian program). This suggests that the discriminant function is significantly superior to the Bayesian algorithm CADENZA for predicting coronary artery disease probabilities in subjects who have relatively high pretest disease probabilities.     

Comparison of hospital and neighborhood controls in a study of coronary artery disease

Case-control studies of risk factors for coronary artery disease (CAD) have almost invariably employed hospital controls, with minimal or no coronary artery stenosis. Although there is an important advantage in knowing the CAD status of controls, such groups are subject to bias related to hospitalization. To evaluate the generalizability of results obtained from studies using hospital controls, we compared risk factors in 342 hospital controls free of angiographic evidence for CAD, 168 neighborhood controls without symptoms of CAD, and 450 CAD patients. Coronary artery disease in cases and hospital controls was established arteriographically. No significant differences were found between the male control groups for total and low density lipoprotein (LDL) cholesterol, LDL apo-B, pack-years of smoking, body mass index, proportion with hypertension, diabetes and family history of coronary heart disease. Compared with neighborhood controls, male hospital controls had significantly lower high density lipoprotein (HDL) cholesterol, higher triglycerides and uric acid and scored higher on the Framingham Type A behavior pattern scale. Among women, the hospital control group had significantly lower LDL cholesterol and fewer pack-years of smoking, and a greater prevalence of hypertension than the neighborhood group. A greater proportion of both male and female hospital controls had left ventricular hypertrophy, and there were more current smokers among the neighborhood controls in both sexes. Age adjustment did not change these comparisons. While very few neighborhood controls were treated with beta-blockers, 32.7% of male and 41.4% of female hospital controls were so medicated. Control for beta-blocker use eliminated the difference in HDL cholesterol and triglycerides between the two male control groups.(ABSTRACT TRUNCATED AT 250 WORDS)     

A probabilistic model for predicting the probability of no-show in hospital appointments

The number of no-shows has a significant impact on the revenue, cost and resource utilization for almost all healthcare systems. In this study we develop a hybrid probabilistic model based on logistic regression and empirical Bayesian inference to predict the probability of no-shows in real time using both general patient social and demographic information and individual clinical appointments attendance records. The model also considers the effect of appointment date and clinic type. The effectiveness of the proposed approach is validated based on a patient dataset from a VA medical center. Such an accurate prediction model can be used to enable a precise selective overbooking strategy to reduce the negative effect of no-shows and to fill appointment slots while maintaining short wait times.     

Optimal cutoff value of high-density lipoprotein cholesterol for predicting coronary artery disease in Taiwanese population

The aim of this study was to determine if an optimal cutoff value for high-density lipoprotein cholesterol (HDL-C) can be obtained for predicting the risk of coronary artery disease (CAD) in Taiwanese population. We conducted a hospital-based case-control study. Patients identified by cardiac catheterization as having at least 70% stenosis of one major coronary artery and without diabetes were assigned to the case group (n = 184). The control group (n = 516) was composed of healthy individuals with normal blood biochemical values. The multiple logistic regression analysis was used to evaluate linear association between low-density lipoprotein cholesterol (LDL-C), HDL-C, or LDL-C/HDL-C ratio and CAD while adjusting for confounders. Furthermore, receiver operating characteristic curve analyses were constructed. Individuals with an HDL-C value less than or equal to 60 mg/dL had the significantly highest odds ratio (7.72; 95% confidence interval, 2.70-22.07) after adjusting for LDL-C, LDL-C/HDL-C ratio, and other potential confounders. The areas under the curves were 0.85 and 0.61 for HDL-C and LDL-C, respectively. The optimal cutoff value of HDL-C for predicting the presence of CAD was 46 mg/dL. Sensitivity and specificity using this cutoff value were 71.74% and 81.40%, respectively. Our findings suggest that subjects with lower levels of HDL-C have a much higher risk of CAD than those with higher levels of LDL-C. The optimal cutoff value for HDL-C in predicting the risk of CAD is considered as 46 mg/dL in the Taiwanese population.     

Polygenic risk scores outperform machine learning methods in predicting coronary artery disease status

Coronary artery disease (CAD) is the leading global cause of mortality and has substantial heritability with a polygenic architecture. Recent approaches of risk prediction were based on polygenic risk scores (PRS) not taking possible nonlinear effects into account and restricted in that they focused on genetic loci associated with CAD, only. We benchmarked PRS, (penalized) logistic regression, naïve Bayes (NB), random forests (RF), support vector machines (SVM), and gradient boosting (GB) on a data set of 7,736 CAD cases and 6,774 controls from Germany to identify the algorithms for most accurate classification of CAD status. The final models were tested on an independent data set from Germany (527 CAD cases and 473 controls). We found PRS to be the best algorithm, yielding an area under the receiver operating curve (AUC) of 0.92 (95% CI [0.90, 0.95], 50,633 loci) in the German test data. NB and SVM (AUC ~ 0.81) performed better than RF and GB (AUC ~ 0.75). We conclude that using PRS to predict CAD is superior to machine learning methods.     

Diagnostic testing for coronary artery disease in a large population

Extensive public health programs are often proposed without a full appreciation of their effects on the target population. There is often a problem of confusing medical care that may be of benefit to an individual with care that may be of benefit to a population: care that may be highly beneficial for a selected patient may be substantially less effective for an unselected population. Exposing a large, asymptomatic population to diagnostic screening for coronary artery disease has cost and risk ramifications far beyond those of discovering people who might benefit from treatment of previously unsuspected disease. A program to screen 20 million people, and treat the most severely affected with coronary artery bypass graft, would cost nearly +9.2 billion. About +21,000 is spent to find each person with disease, while the cost is more than +169,000 for each person surviving surgical therapy. The cost per year of life extended is over +43,000. In excess of 8,000 persons have major complications and nearly 2,000 die from diagnostic testing and therapy. About half of major complications and deaths occur in persons without disease. This model integrating epidemiologic, economic, and decision analytic methods is presented to illustrate the potential use of myriad techniques in addressing population-based medical care and policy options; it does not provide the definitive answer, however, to the complex problem.     

Optimal use of literature knowledge to improve the Bayesian diagnosis of coronary artery disease

Bayes' theorem with the independence assumption is applied to a test sample of 141 subjects, using two sets of test sensitivities and specificities. The first set is derived by averaging over literature reports on the accuracy of the exercise electrocardiogram, exercise thallium scintigraphy, and carciac fluoroscopy. The second set of indices is derived by applying multivariate regression to the technical, population, and methodologic attributes obtained from the same literature by the use of meta-analysis. The meta-analytically corrected sensitivities and specificities resulted in significant improvement in the discriminatory power of the Bayes model. (Area under ROC curve increased, p = less than 0.01). However, the corrected model was not as accurate as a data-derived logistic regression model of the same test variables. Meta-analysis may be useful for modest improvement in the accuracy of literature-derived Bayesian models for predicting disease probabilities.     

颈动脉粥样硬化与不同类型冠心病的相关性分析

目的 探讨冠心病病人颈动脉粥样硬化程度与冠状动脉粥样硬化(CAAS)程度的关系及相关危险因素;总结不同类型冠心病患者颈动脉粥样硬化(CAS)病理特点.方法 回顾性分析228例经冠状动脉动脉造影确诊为冠心病的病例冠状动脉造影结果、颈动脉超声检测结果及相关危险因素;根据颈动脉是否有粥样硬化分为粥样硬化组及非粥样硬化组;根据WHO诊断标准将病例分为稳定型心绞痛组、不稳定型心绞痛组及心肌梗死组;根据冠状动脉造影结果将病例分为单支病变组(组A)、双支病变组(组B)、三支病变组(组C)及左主干组(组D);对颈动脉粥样硬化(CAS)程度进行积分处理.结果 发现228例冠心病病人中198例均有不同程度颈动脉粥样硬化,发病率86.8%;颈动脉粥样硬化(CAS)与高血压显著相关,与年龄、性别、体重指数、吸烟、嗜酒、高脂血症、高尿酸血症及糖尿病无明显相关;颈动脉粥样硬化(CAS)程度随冠状动脉粥样硬化(CAAS)的程度加重而加重,但仅仅一支病变组总斑块数显著低于左主干组(P＜0.05);稳定型心绞痛组的颈动脉等级积分、Crouse积分均低于不稳定型心绞痛组;稳定型心绞痛组颈动脉等级积分(grading integral)较急性心肌梗死组低,而Crouse积分高于急性心肌梗死组;稳定型心绞痛组的总斑块数、扁平斑数及软斑数均低于不稳定型心绞痛组及急性心肌梗死组,三组的硬斑数差异无统计学意义(P＞0.05).三组均未发现溃疡斑.结论 颈动脉粥样硬化与冠心病有相关性.

Abstract：

Objective To determine the relationship between the aggravation of CAAS and coronary atherosclerosis (CAS) ;and to summarize the pathologic character of CAAS of the patients with various coronary artery disease. Methods Review the result of coronary angiography and carotid artery ultrasonography and the related risk factors of the patients who were diagnosed as CAD through coronary angiography ( CAG). The patients were divided into the scleratheroma group and the non scleratheroma group. The patients were divided into stable angina pectoris( AP) group, unstable angina pectoris(UAP) group and acute myocardial infarction (AMI) group according to the criterion of coronary artery disease of WHO. The patients were divided into group A (coronary artery of single vessel lession) ,group B( coronary artery of double vessel lession ) , group C( coronary artery of triple vessel lession )and group D(left main vessel lession ).The aggravation of CAAS was graded. Results There were 198 patients with various CAAS among 228 patients with coronary artery disease ( 86. 8%). CAAS was much related with hypertension and non-related with age, sex, BMI, smoking, drinking, hyperlipoidemia, hyperuricosuria and diabetes mellitus. The aggravation of CAAS much graver with much graver CAS. But only the plaque number of Group A was more than Group D( P ＜0. 05). The grading integral and Crouse integral of CAAS of AP group was not remarkable less than UAP group. The grading integral of CAAS of AP group was less than AMI group and the Crouse integral of AP group was more. But there was no remarkable difference. The number of all plaque,plaque and plaque of AP group was not remarkable less than UAP group and AMI group. There was the plaque among the three group. Conclusion Carotid artery atherosclerosis (CAAS) is relative to coronary artery disease ( CAD).     

Selection of neighborhood controls for a study of coronary artery disease

This report describes the selection process for a neighborhood control group recruited between February 1985 and July 1986 to augment a hospital-based case-control study investigating the relation of traditional and nontraditional risk factors to coronary artery atherosclerosis. A total of 219 cases with angiographically defined coronary artery disease residing within a 60-mi (96-km) radius of Winston-Salem, NC, were assigned to surveyors to be matched. Thirty-seven per of the study population were rural, 92% were white, 58% were male, and 52% were older than 50 years of age. One hundred and eighty-seven controls were age-(+/- five years), sex-, and race-matched pairwise to cases. After locating the residence of the case, the surveyor systematically visited neighboring households to ascertain eligibility status of residents. To achieve a match, a maximum of three visits was made to the neighborhood; up to 25 households were surveyed per visit. Refusal rate was less than 5% of eligible matches. Young white urban males were the easiest to match, while rural females, especially older persons, were the most difficult. Average time to complete an assignment included 129 minutes for travel, 237 minutes for surveying the neighborhood, and 62 minutes for clerical tasks. Average distance traveled was 85 mi (136 km) per case. As expected, the time and distance to complete a case were greater in rural than urban areas. The average cost per case was $122.97.     

Clinical assessment of the probability of coronary artery disease: judgmental bias from personal knowledge.

Probability estimates of angiographic coronary artery disease made by experienced, board-certified staff cardiologists were compared with those of cardiologists in training (fellows). In addition, estimates made before coronary angiography were compared with those made several months later based on written clinical summaries of 15 items of objective clinical and test data. Cardiologists were asked to estimate the probabilities of coronary artery disease, multivessel disease, and triple-vessel or left main disease. The study population consisted of 510 consecutive patients without valvular disease referred for the first time for coronary angiography to three hospitals. Both staff and fellows consistently overestimated the pre-angiographic probability of coronary artery disease. The probabilities estimated from patient summaries were always significantly lower than the pre-angiographic assessments. Only staff cardiologists reliably assessed the probabilities of coronary artery disease during the second assessment (p less than 0.05). Thus, estimates of disease probability based on clinical judgment vary according to the source of information, and these estimates are more accurate when physicians have objective data on hand and do not know the identities of the patients.     

Practical Bayesian analysis of a simple logistic regression: predicting corneal transplants

The Bayesian analysis of a logistic regression model is described using an example of predicting the need for a corneal transplant in keratoconus. Controversy over the use of subjective prior information in Bayesian methods is avoided by a formulation representing negligible prior information. Simple computational procedures are described, and it is argued that the results are more accurate, clearer and make fuller use of the information contained in the data. Analysis of more complex models is considered. In particular, it is argued that classical methods as implemented in the computer package GLIM can be used as approximations to Bayesian methods, particularly at the initial stage of model selection.     

Risk for coronary artery disease and morbid preeclampsia: a commentary

PURPOSE: A predisposition to coronary artery disease (CAD) may put women at risk for preeclampsia. Morbid preeclampsia (early, severe, recurrent, and with neonatal morbidity) represents the subset of preeclampsia of greatest public health concern. METHODS: We review here the published links between preeclampsia and CAD. RESULTS: Many risk factors are common to both CAD and preeclampsia. These include obesity; elevated blood pressure; dyslipidemia; insulin resistance; and hyperglycemia, together termed "Syndrome X"; as well as endothelial dysfunction; hyperuricemia; hyperhomocysteinemia; and abnormalities of inflammation, thrombosis, and angiogenesis. After pregnancy, women with preeclampsia are more likely to experience later life CAD. CONCLUSIONS: Both the association between CAD risk factors and preeclampsia and the association between preeclampsia and later CAD appears to be more pronounced among the subset of women with morbid preeclampsia. Thus, women at elevated risk for CAD may be at particularly high risk for morbid preeclampsia and women with morbid preeclampsia may be those at highest risk for later life CAD.