Search PubMed⌕ Search

Biomedical subjects

F E Harrell

Publications and source records attributed to F E Harrell.

At least 55 records · Page 3Linked to original sources
PubMed1995

Accurate prognostications of death. Opportunities and challenges for clinicians.

Linking survival time to an array of prognostic variables through a powerful statistical model can provide reliable, valid and potentially useful information for patient care. We present a summary of the recently developed SUPPORT model [Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment] for estimating survival time of seriously ill adult inpatients and illustrate the possible clinical use of such a model. This model is then translated into counseling. Clinicians are positioned to evaluate the relevance and validity of any model and to understand their persistent shortcomings.

Adult↗
PubMed1994

Defining sepsis.

Explore the source record for details and available documents.

APACHE↗
PubMed1994

Accuracy of the clinical examination in detecting hypoxemia in infants with respiratory illness.

OBJECTIVES: To evaluate the diagnostic accuracy of the clinical examination in detecting hypoxemia in infants with lower respiratory tract illness. DESIGN: Cross-sectional study. SETTING: Three university pediatric outpatient departments and one private pediatric practice. PATIENTS: Healthy infants less than 1 year of age seen between December and March 1989 and 1990, with symptoms suggesting acute lower respiratory tract illness. MAIN OUTCOME MEASURES: The test characteristics of 27 elements of the clinical examination, as well as the accuracy of the overall examination and the components of the examination in detecting oxygen saturation < 95% measured by pulse oximetry. Reliability of clinical examination findings. RESULTS: None of the 27 clinical findings had sensitivities that would make them useful diagnostic tests for hypoxemia. By combining all the clinical findings, however, we found good diagnostic accuracy (area under the receiver operator characteristic curve 0.90). Three groups of clinical findings--social interactiveness, respiratory effort, and physical appearance--accounted for much of the diagnostic accuracy of the examination. Auscultatory findings contributed little. In these three groups, five clinical findings accounted for almost all the accuracy: attentiveness, consolability, respiratory effort, color, and movement. Together, these findings also had good accuracy (area under the receiver operator characteristic curve 0.95). CONCLUSIONS: A small number of clinical observations may be mostly responsible for the diagnostic value of the clinical examination of infants with symptoms of LRI. Concentrating on a limited group of findings appears to enhance the accuracy of the examination in detecting hypoxemia.

Cross-Sectional Studies↗
PubMed1994

Daily prognostic estimates for critically ill adults in intensive care units: results from a prospective, multicenter, inception cohort analysis.

OBJECTIVE: To develop daily prognostic estimates for individual patients treated in adult intensive care units (ICU). DESIGN: Prospective, multicenter, inception cohort analysis. SETTING: Forty-two ICUs at 40 U.S. hospitals with > 200 beds including 20 ICUs in tertiary care centers with major teaching activities. PATIENTS: A consecutive sample of 17,440 ICU admissions. MEASUREMENTS AND MAIN RESULTS: A series of multivariate equations were developed using the patient's primary reason for ICU admission, age, chronic health status, treatment before ICU admission, admission Acute Physiology Score, current day Acute Physiology Score, and change between the current and previous day's Acute Physiology Score. The equations were used to create daily risk predictions and cross-validated within the 17,440-patient sample. The single most important factor determining daily risk of hospital death during each of the initial 7 days of ICU care was the current day's Acute Physiology Score of the Acute Physiology and Chronic Health Evaluation (APACHE) III score. The admission Acute Physiology Score and change from previous to current day's Acute Physiology Score were also important, as were ICU admission diagnosis, age, chronic health status, and treatment before ICU admission. Equations incorporating these risk factors had receiver operating characteristics areas ranging from 0.9 on the first ICU day to 0.84 for patients remaining in the ICU for 7 days. The percent of cases with cross-validated predicted risks over 90% increased from 2.3% (n = 406) of cases on day 1 to 9% of all patients remaining in the ICU on ICU day 7 (n = 218). The 1,033 patients who had a daily risk estimate of > 90% during any of their initial 7 ICU days had a 90% mortality rate and represented 47% of all ICU deaths and 31% of the total number of hospital deaths. CONCLUSIONS: Equations using initial and repeated physiologic measurements provide a high degree of explanatory power for subsequent hospital mortality rate. These daily prognostic estimates deserve evaluation for their potential role in improving the process and outcome from clinical decision-making.

Adult↗
PubMed1994

Continuing evolution of therapy for coronary artery disease. Initial results from the era of coronary angioplasty.

BACKGROUND: Survival after coronary artery bypass graft surgery (CABG) and medical therapy in patients with coronary artery disease (CAD) has been studied in both randomized trials and observational treatment comparisons. Over the past decade, the use of coronary angioplasty (PTCA) has increased dramatically, without guidance from either randomized trials or prospective observational comparisons. The purpose of this study was to describe the survival experience of a large prospective cohort of CAD patients treated with medicine, PTCA, or CABG. METHODS AND RESULTS: The study was designed as a prospective nonrandomized treatment comparison in the setting of an academic medical center (tertiary care). Subjects were 9263 patients with symptomatic CAD referred for cardiac catheterization (1984 through 1990). Patients with prior PTCA or CABG, valvular or congenital disease, nonischemic cardiomyopathy, or significant (> or = 75%) left main disease were excluded. Baseline clinical, laboratory, and catheterization data were collected prospectively in the Duke Cardiovascular Disease Databank. All patients were contacted at 6 months, 1 year, and annually thereafter (follow-up 97% complete). Cardiovascular death was the primary end point. Of this cohort, 2788 patients were treated with PTCA (2626 within 60 days) and 3422 with CABG (3080 within 60 days). Repeat or crossover revascularization procedures were counted as part of the initial treatment strategy. Kaplan-Meier survival curves (both unadjusted and adjusted for all known imbalances in baseline prognostic factors) were used to examine absolute survival differences, and treatment pair hazard ratios from the Cox model were used to summarize average relative survival benefits. For the latter, a 13-level CAD prognostic index was used to examine the relation between survival and revascularization as a function of CAD severity. The effects of revascularization on survival depended on the extent of CAD. For the least severe forms of CAD (ie, one-vessel disease), there were no survival advantages out to 5 years for revascularization over medical therapy. For intermediate levels of CAD (ie, two-vessel disease), revascularization was associated with higher survival rates than medical therapy. For less severe forms of two-vessel disease, PTCA had a small advantage over CABG, whereas for the most severe form of two-vessel disease (with a critical lesion of the proximal left anterior descending artery), CABG was superior. For the most severe forms of CAD (ie, three-vessel disease), CABG provided a consistent survival advantage over medicine. PTCA appeared prognostically equivalent to medicine in these patients, but the number of PTCA patients in this subgroup was low. CONCLUSIONS: In this first large-scale, prospective observational treatment comparison of PTCA, CABG, and medicine, we confirmed the previously reported survival advantages for CABG over medical therapy for three-vessel disease and severe two-vessel disease. For less severe CAD, the primary treatment choices are between medicine and PTCA. In these patients, there is a trend for a relative survival advantage with PTCA, although absolute survival differences were modest. In this setting, treatment decisions should be based not only on survival differences but also on symptom relief, quality of life outcomes, and patient preferences.

Angioplasty, Balloon, Coronary↗
PubMed1993

The clinical evaluation of new drugs for sepsis. A prospective study design based on survival analysis.

OBJECTIVE: To develop a survival model and severity assessment method to estimate the 28-day mortality risk for patients with sepsis syndrome entering phase 2 and 3 drug evaluations. DESIGN: Retrospective analysis of intensive care unit admissions with sepsis syndrome by means of log-normal regression to identify risk factors for 28-day mortality. Prospective application of the model to patients with gram-negative infection meeting sepsis syndrome criteria from separate data collection (validation group). PATIENTS: A total of 58,737 intensive care unit admissions at 107 hospitals in the United States and Western Europe screened to yield 1195 patients meeting entry criteria for the sepsis syndrome study for the original model; 295 hospitalized patients with gram-negative infection meeting criteria for sepsis syndrome for validation. MAIN OUTCOME MEASURES: Survival time and mortality at 28 days after fulfillment of the sepsis syndrome criteria. RESULTS: Acute physiologic abnormalities were the most important prognostic factors influencing outcome (82% of total chi 2). Specific disease resulting in intensive care unit admission and the time the patient was in the hospital and intensive care unit before qualification were also independent risks, as were age and a clinical history of cirrhosis. The model's overall classification accuracy was a Somers' Dyx of .52 (rank correlation between predicted risk and 28-day mortality) (receiver operating characteristic area, 0.76), with equal accuracy (Dyx = .59; receiver operating characteristic area, 0.80) in the independent group of patients. CONCLUSIONS: We created an accurate independent estimate for 28-day mortality risk for patients with sepsis syndrome (severe sepsis). This estimate could improve the evaluation of new drugs by investigating whether the drug's benefit varies by patient risk and then determining the amount of benefit for individual patients.

Adolescent↗
PubMed1993

Value of the history and physical in identifying patients at increased risk for coronary artery disease.

OBJECTIVE: To determine whether information from the physician's initial evaluation of patients with suspected coronary artery disease predicts coronary anatomy at catheterization and 3-year survival. DESIGN: Prospective validation of regression model estimates in an outpatient cohort. SETTING: University medical center. PATIENTS: A total of 1030 consecutive outpatients referred for noninvasive testing for suspected coronary artery disease; 168 of these patients subsequently underwent catheterization within 90 days. MEASUREMENTS: Information from the initial history, physical examination, electrocardiogram, and chest radiograph was used to predict coronary anatomy (the likelihood of any significant coronary disease, severe disease [left main or three-vessel], and significant left main disease) among 168 catheterized patients and to estimate 3-year survival among all patients. These estimates were compared with those based on treadmill testing. Cardiovascular testing charges were calculated for all patients. RESULTS: Predicted coronary anatomy and survival closely corresponded to actual findings. Compared with the treadmill exercise test, initial evaluation was slightly better able to distinguish patients with or without any coronary disease and was similar in the ability to identify patients at increased risk for dying or with anatomically severe disease. Based on arbitrary definitions, 37% to 66% of patients were at low risk and responsible for 31% to 56% of the charges for cardiovascular testing. CONCLUSIONS: The physician's initial evaluation, despite the subjective nature of much of the information gathered, can be used to identify patients likely to benefit from further testing. The development of strategies for cost-conscious quality care must begin with the history, physical examination, and simple laboratory testing.

Aged↗
PubMed1993

Comparison of analytic models for estimating the effect of clinical factors on the cost of coronary artery bypass graft surgery.

The cost of treating disease depends on patient characteristics, but standard tools for analyzing the clinical predictors of cost have deficiencies. To explore whether survival analysis techniques might overcome some of these deficiencies in the analysis of cost data, we compared ordinary least square (OLS) linear regression (with and without transformation of the data) and binary logistic regression with two survival models: the Cox proportional hazards model and a parametric model assuming a Weibull distribution. Each model was applied to data from 155 patients undergoing coronary artery bypass grafting. We examined the effects of age, sex, ejection fraction, unstable angina, and number of diseased vessels on univariable and multivariable predictions of costs. The significant univariable predictors of cost were consistent in all models: ejection fraction was significant in all five models, and age and number of diseased vessels were each significant in all but the OLS model, while sex and angina type were significant in none of the models. The significant multivariable predictors of cost, however, differed according to model: ejection fraction was a significant multivariable predictor of cost in all five models, age was significant in three models, and number of diseased vessels was significant in one model. All five models were also used to predict the costs for an average patient undergoing surgery. The Cox model provided the most accurate predictions of mean cost, median cost, and the proportion of patients with high cost. This study shows: (1) lower ejection fraction and older age are independent clinical predictors of increased cost of CABG, and (2) the Cox proportional hazards model shows considerable promise for the analysis of the impact of clinical factors upon cost.

Aged↗
PubMed1992

High and low strength nonsynchronized shocks given during canine ventricular tachycardia.

UNLABELLED: Cardioversion shocks given during ventricular tachycardia may cause ventricular fibrillation or acceleration of ventricular tachycardia, or arrest the tachycardia. A recently proposed theory may explain why the former two phenomena may occur. Briefly, this theory states that potential gradient shock fields of a critical strength delivered to tissue with a critical degree of refractoriness will cause circulating wave fronts of ventricular activation ("rotors") manifest as ventricular arrhythmia. We tested this theory by delivering nonsynchronized shocks 50% higher than defibrillation threshold or 50% lower than defibrillation threshold during 275 episodes of ventricular tachycardia in eight dogs with 5- to 7-day-old myocardial infarcts. Shocks stronger than the defibrillation threshold are likely to create shock fields in the ventricles everywhere stronger than this critical value, and therefore would not generate rotors. Shocks less strong than the defibrillation threshold may create shock fields within the ventricles that include the critical value, and therefore cause rotors if given when critically refractory tissue is present. Nonsynchronized shocks were used to increase the likelihood of encountering tissue with a critical degree of refractoriness. Ventricular fibrillation or acceleration of ventricular tachycardia occurred following 83 of 138 (60%) low strength shocks and following 20 of 137 (14.6%) high strength shocks. The pooled odds ratio for induction of ventricular fibrillation or accelerated ventricular tachycardia after low strength shocks as compared to high strength shocks was 8.9. CONCLUSION: when given during ventricular tachycardia, low strength shocks are much more likely to cause ventricular fibrillation or accelerated ventricular tachycardia than are high strength shocks (P less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)

Animals↗
PubMed1992

Observational comparison of event-free survival with medical and surgical therapy in patients with coronary artery disease. 20 years of follow-up.

BACKGROUND: The purpose of this study was to describe the long-term event-free survival patterns of patients with significant coronary artery disease treated medically versus patterns of those treated surgically and to evaluate the factors associated with improved event-free survival. METHODS AND RESULTS: We studied the results of 5,824 patients undergoing medical and surgical therapy for ischemic heart disease from 1969 to 1984, with follow-up to 1991. Events considered for this evaluation were nonfatal myocardial infarction or cardiovascular death. The Cox proportional hazards model was used to determine factors differentially affecting surgical event-free survival. The survival benefits previously reported for bypass surgery in this population were largely preserved when event-free survival was examined. The two factors associated with significant event-free survival benefits for surgically treated patients were more severe coronary artery disease and a more recent surgery data. Patients with more severe coronary obstruction had a greater relative improvement with surgery in event-free survival than did patients with less severe anatomic disease. Event-free survival with surgery progressively improved over the period of the study and, by 1984, was significantly better than medical therapy for most patient subgroups. Patients with poor prognosis because of risk factors such as older age, severe angina, or left ventricular dysfunction had a risk reduction with surgery proportional to their overall risk under medical therapy. CONCLUSIONS: Higher-risk patients with more severe disease (due to either coronary disease or other risk factors and age) should be considered for coronary revascularization because it is in these patients that coronary artery bypass graft surgery has the greatest impact in reducing future cardiovascular events.

Cardiac Catheterization↗
PubMed1991

Prognostic value of a treadmill exercise score in outpatients with suspected coronary artery disease.

BACKGROUND: The treadmill exercise test identifies patients with different degrees of risk of death from cardiovascular events. We devised a prognostic score, based on the results of treadmill exercise testing, that accurately predicts outcome among inpatients referred for cardiac catheterization. This study was designed to determine whether this score could also accurately predict prognosis in unselected outpatients. METHODS: We prospectively studied 613 consecutive outpatients with suspected coronary disease who were referred for exercise testing between 1983 and 1985. Follow-up was 98 percent complete at four years. The treadmill score was calculated as follows: duration of exercise in minutes--(5 x the maximal ST-segment deviation during or after exercise, in millimeters)--(4 x the treadmill angina index). The numerical treadmill angina index was 0 for no angina, 1 for nonlimiting angina, and 2 for exercise-limiting angina. Treadmill scores ranged from -25 (indicating the highest risk) to +15 (indicating the lowest risk). RESULTS: Predicted outcomes for the outpatients, based on their treadmill scores, agreed closely with the observed outcomes. The score accurately separated patients who subsequently died from those who lived for four years (area under the receiver-operating-characteristic curve = 0.849). The treadmill score was a better discriminator than the clinical data and was even more useful for outpatients than it had been for inpatients. Approximately two thirds of the outpatients had treadmill scores indicating low risk (greater than or equal to +5), reflecting longer exercise times and little or no ST-segment deviation, and their four-year survival rate was 99 percent (average annual mortality rate, 0.25 percent). Four percent of the outpatients had scores indicating high risk (less than -10), reflecting shorter exercise times and more severe ST-segment deviation; their four-year survival rate was 79 percent (average annual mortality rate, 5 percent). CONCLUSIONS: The treadmill score is a useful and valid tool that can help clinicians determine prognosis and decide whether to refer outpatients with suspected coronary disease for cardiac catheterization. In this study, it was a better predictor of outcome than the clinical assessment.

Adult↗
PubMed1991

Using ordinal logistic regression to estimate the likelihood of colorectal neoplasia.

The utility of ordinal logistic regression in the prediction of colorectal neoplasia was demonstrated in a group of 461 consecutive patients undergoing colonoscopy in a community practice. One hundred twenty-nine patients had adenomatous polyps and 34 had colorectal adenocarcinoma. An ordinal logistic regression model developed in a random subset (292 patients) identified five predictors of colorectal neoplasia. Colorectal neoplasia risk could be predicted using the patient's age, sex, hematocrit, fecal occult blood test result and indication for colonoscopy. The risk of colorectal neoplasia in the remaining subset of patients (169) could be reliably estimated from the model. Ordinal logistic regression analysis in this select group of patients can accurately estimate the likelihood of colorectal neoplasia. Because the generalizability of our findings are unknown, the model should not be applied to other patients. However, application of this technique to an unselected group of patients not already referred for colonoscopy could provide unbiased estimates of colorectal neoplasia risk in individual patients.

Adenocarcinoma↗
PubMed1991

Determinants of early versus late cardiac death in patients undergoing coronary artery bypass graft surgery.

Most analyses of risk factors affecting survival after coronary artery bypass graft surgery have not differentiated among factors that influence early and late survival. For this reason, a multiphase model was applied to survival data from 2,967 patients undergoing a first coronary artery bypass graft at the Duke University Medical Center between 1969 and 1984. There were 709 deaths during follow-up to 19.6 years. The data were analyzed using a multivariable survival model that separates the underlying hazard function into as much as three different phases, each incorporating separate risk factors. Two distinct phases were detected. One phase dominated early survival (0-1 year), and the second phase dominated late survival (greater than 1 year). Surgery performed earlier in our experience was associated with elevated risk of dying in both phases but with different magnitudes, whereas lower ejection fraction, greater extent of coronary disease, older age, conduction abnormality, and history of hypertension were associated with elevated risk of dying similarly in both phases (p less than 0.05). Severity of angina symptoms and lower weight were associated with an elevated risk of dying only in the early phase (p less than 0.05; because few of the patients were obese, estimates of the relative risk of morbid obesity could not be estimated), whereas vascular disease, diabetes, and extent of myocardial damage were associated with an elevated risk of dying only in the late phase (p less than 0.05). These data illustrate both the differential influence of risk factors over time and the importance of multiphase models.

Cohort Studies↗
PubMed1991

Estimating the likelihood of severe coronary artery disease.

PURPOSE: To determine which clinical characteristics obtained by a physician during an initial clinical examination are important for estimating the likelihood of severe coronary artery disease, and to determine whether estimates based on these characteristics remain valid when applied prospectively and in different patient groups. PATIENTS AND METHODS: We examined clinical characteristics predictive of severe disease in 6,435 consecutive symptomatic patients referred for suspected coronary artery disease between 1969 and 1983. RESULTS: Eleven of 23 characteristics were important for estimating the likelihood of severe coronary artery disease. A model using these characteristics accurately estimated the likelihood of severe disease in an independent sample of 2,342 patients referred since 1983. The model also accurately estimated the prevalence of severe disease in large series of patients reported in the literature. CONCLUSIONS: These findings suggest that the clinician's initial evaluation can identify patients at high or low risk of anatomically severe coronary artery disease. Cost-conscious quality care is encouraged by identifying patients at higher risk for severe coronary artery disease who are most likely to benefit from further evaluation.

Adult↗
PubMed1990

Effect of Type A behavior on exercise test outcome in coronary artery disease.

The outcome of the diagnostic exercise test depends on such patient-related factors as age, maximum exercise heart rate, exercise time and severity of the underlying coronary artery disease (CAD). This study examined the hypothesis that type A behavior would affect the amount of effort expended, as indicated by the exercise time and the maximum heart rate achieved, thereby resulting in differences in exercise test outcome. A total of 1,260 patients with suspected CAD, all of whom had coronary angiography, a structured interview to assess type A behavior and a treadmill exercise test, participated. Of these patients, 818 (65%) had significant CAD, and 852 (68%) were type A. There were no differences between type A and B patients in either maximum heart rate or total exercise time. Among both type A and B subjects, 36% of treadmill tests were positive. Exercise test sensitivity was similar for both groups (69% for type A vs 72% for type B, p = 0.39). Similarly, specificity was similar for both groups (87% for type A vs 80% for type B, p = 0.09). Results did not change after using logistic regression to control for potential confounding factors. Thus, type A behavior does not need to be taken into account when interpreting exercise test outcome.

Adult↗
PubMed1990

Clinical judgment and therapeutic decision making.

Clinical decision making is under increased scrutiny due to concerns about the cost and quality of medical care. Variability in physician decision making is common, in part because of deficiencies in the knowledge base, but also due to the difference in physicians' approaches to clinical problem solving. Evaluation of patient prognosis is a critical factor in the selection of therapy, and careful attention to methodology is essential to provide reliable information. Randomized controlled clinical trials provide the most solid basis for the establishment of broad therapeutic principles. Because randomized studies cannot be performed to address every question, observational studies will continue to play a complementary role in the evaluation of therapy. Randomized studies in progress, meta analyses of existing data, and increased use of administrative and collaborative clinical data bases will improve the knowledge base for decision making in the future.

Cardiology↗