Janelle Klar

Adequacy of sample size in health studies

Part 1 Statistical methods for sample size determination: the one sample problem the two sample problem sample size for case-control studies sample size determination for cohort studies lot quality assurance sampling the incidence density sample size for continuous response variables sample size for sample surveys. Part 2 Foundations of sampling and statistical theory: the population the sample sampling distribution characteristics of estimates of population parameters hypothesis testing two sample confidence intervals and hypothesis tests epidemiologic study design basis sampling concepts.

The Logistic Organ Dysfunction (LOD) System

The assessment of the severity of organ dysfunction in the ICU is a critical tool for conducting clinical trials, especially sepsis trials. The evaluation of new therapies cannot be successfully achieved without controlling for the degree of organ dysfunction. It is not adequate to assess severity, or to describe a patient’s condition, by simply counting the number of dysfunctioning organ systems.

Mortality probability models for patients in the intensive care unit for 48 or 72 hours: A prospective, multicenter study

ObjectiveTo develop models in the Mortality Probability Model (MPM II) system to estimate the probability of hospital mortality at 48 and 72 hrs in the intensive care unit (ICU), and to test whether the 24-hr Mortality Probability Model (MPM24), developed for use at 24 hrs in the ICU, can be used on a daily basis beyond 24 hrs. DesignA prospective, multicenter study to develop and validate models, using a cohort of consecutive admissions. SettingSix adult medical and surgical ICUs in Massachusetts and New York adjusted to reflect 137 ICUs in 12 countries. PatientsConsecutive admissions (n = 6,290) to the Massachusetts/New York ICUs were studied. Of these patients, 3,023 and 2,233 patients remained in the ICU and had complete data at 48 and 72 hrs, respectively. Patients <18 yrs of age, burn patients, coronary care patients, and cardiac surgical patients were excluded. Outcome MeasureVital status at the time of hospital discharge. ResultsThe models consist of five variables measured at the time of ICU admission and eight variables ascertained at 24-hr intervals. The 24-hr model demonstrated poor calibration and discrimination at 48 and 72 hrs. The newly developed 48− and 72-hr models—MPM48 and MPM72—contain the same 13 variables and coefficients as the MPM24. The models differ only in their constant terms, which increase in a manner that reflects the increasing probability of mortality with increasing length of stay in the ICU. These constant terms were adjusted by a factor determined from the relationship between the data from the six Massachusetts and New York ICUs and a more extensive data set, from which the ICU admission Mortality Probability Model (MPM0) and MPM24 were developed. This latter data set was assembled from ICUs in 12 countries. The MPM48 and MPM72 calibrated and discriminated well, based on goodness-of-fit tests and area under the receiver operating characteristic curve. ConclusionsModels developed for use among ICU patients at one time period are not transferable without modification to other time periods. The MPM48 and MPM72 calibrated well to their respective time periods, and they are intended for use at specific points in time. The increasing constant terms and associated increase in the probability of hospital mortality exemplify a common clinical adage that if a patients clinical profile stays the same, he or she is actually getting worse. (Crit Care Med 1994; 22:1351–1358)

Outcome prediction for individual intensive care patients: useful, misused, or abused?

Probabilities of hospital mortality provide meaningful information in many contexts, such as in discussions of patient prognosis by intensive care physicians, in patient stratification for analysis of clinical trial data by researchers, and in hospital reimbursement analysis by insurers. Use of probabilities as binary predictors based on a cut point can be misleading for making treatment decisions for individual patients, however, even when model performance is good overall. Alternative models for estimating severity of illness in intensive care unit (ICU) patients, while demonstrating good agreement for describing patients in the aggregate, are shown to differ considerably for individual patients. This suggests that identifying patients unlikely to benefit from ICU care by using models must be approached with considerable caution.

Prospective study of clinical bleeding in intensive care unit patients

We investigated prospectively clinical bleeding in 1,328 consecutive patients admitted to a medical/surgical ICU over 1 yr. One hundred thirty-eight (10.4%) patients bled after ICU admission, and an additional 388 (29.2%) bled coincident with admission. The upper GI tract was the site of bleeding in 34.8% of patients whose bleeds commenced in the ICU, and accounted for 22% of total sites. Patients with clinical bleeding after ICU admission had a significantly (p less than .001) higher likelihood of death than those who did not bleed, and those with multiple bleeding sites had a higher mortality (54.9%) than those with single sites (31%) (p less than .006). Multiple logistic regression analyses revealed that risk ratios (RR) for bleeding after ICU admission were mechanical ventilation (RR = 1.82), nutritional failure (RR = 3.45), acute renal failure (RR = 3.36), antiulcer medication (RR = 3.36), and anticoagulants (RR = 4.19). No antibiotics could be specifically incriminated. This study defines the scope, characteristics, and importance of bleeding in ICU patients and establishes risk factors.

An Exposure Prevention Rating Method for Intervention Needs Assessment and Effectiveness Evaluation

This article describes a new method for (1) systematically prioritizing needs for intervention on hazardous substance exposures in manufacturing work sites, and (2) evaluating intervention effectiveness. We developed a checklist containing six unique sets of yes/no variables organized in a 2 x 3 matrix of exposure potential versus protection (two columns) at the levels of materials, processes, and human interface (three rows). The three levels correspond to a simplified hierarchy of controls. Each of the six sets of indicator variables was reduced to a high/moderate/low rating. Ratings from the matrix were then combined to generate a single overall exposure prevention rating for each area. Reflecting the hierarchy of controls, material factors were weighted highest, followed by process, and then human interface. The checklist was filled out by an industrial hygienist while conducting a walk-through inspection (N = 131 manufacturing processes/areas in 17 large work sites). One area or process per manufacturing department was assessed and rated. Based on the resulting Exposure Prevention ratings, we concluded that exposures were well controlled in the majority of areas assessed (64% with rating of 1 or 2 on a 6-point scale), that there is some room for improvement in 26 percent of areas (rating of 3 or 4), and that roughly 10 percent of the areas assessed are urgently in need of intervention (rated as 5 or 6). A second hygienist independently assessed a subset of areas to evaluate inter-rater reliability. The reliability of the overall exposure prevention ratings was excellent (weighted kappa = 0.84). The rating scheme has good discriminatory power and reliability and shows promise as a broadly applicable and inexpensive tool for intervention needs assessment and effectiveness evaluation. Validation studies are needed as a next step. This assessment method complements quantitative exposure assessment with an upstream prevention focus.

How to Assess Organ Dysfunction in the Intensive Care Unit? The Logistic Organ Dysfunction (LOD) System

The assessment of the severity of organ dysfunction in the ICU is a critical tool for conducting clinical trials, especially sepsis trials. The evaluation of new therapies cannot be successfully achieved without controlling for the degree of organ dysfunction. It is not adequate to assess severity, or to describe a patient’s condition, by simply counting the number of dysfunctioning organ systems [1]. In many scoring systems, each organ dysfunction is graded from 1 to 4 points, or from 1 to 6 points, and a score is produced by adding the points. These systems cannot adequately re_ect patient severity. Not only are the ranges de~ning the levels different from those found using statistical methods, but weighting each organ system in the same way does not take into account the differential pronostic signi~cance of the involved organs. Comparing the discriminating power of the OSF system of Knaus [2] and the APACHE III equation, Wagner et al found ROC curves of 0.66 and 0.88, respectively [1].

Organ system failure: How to create an objective system using logistic regression

SummaryTill now all the proposed OSF systems are subjective. Besides they often include therapeutic measures that may vary from an ICU to another. We propose here an objective organ system failure using logistic regression.

A Characterization of Occupational Static Magnetic Field Exposures at a Diaphragm-Cell and a Mercury-Cell Chlor-Alkali Facility

Abstract Static magnetic field levels from direct current were studied in two chlor-alkali facilities, one a diaphragm-cell and one a mercury-cell technology facility. At the diaphragm-cell facility, static magnetic fields ranged from 1.0 to 173.2 Gauss (G), with an average of 82.3 G. At the mercury-cell facility, levels ranged from 4.1 to 182.9 G, with an average of 46.8 G. At the diaphragm-cell facility, field strengths were highest at the mid-height (3.5 feet) level. There was no difference in field strength by height at the mercury-cell facility. In both facilities, field strengths were higher at sampling locations closer to the electrical buss bar. Static magnetic field strengths at the property lines were comparable to expected background levels (below detection to 2.5 G). Measurements near the cell building (25–45 feet away) were slightly higher and ranged from 0.4 to 15.0 G. All measurements inside and outside the cell buildings were below the American Conference of Governmental Industrial Hygieni...

Use of a Probability Model for Predicting ICU Outcome

During the past decade, a number of different intensive care unit (ICU) mortality prediction systems have been developed [1–3]. These techniques provide information related to the severity of illness of ICU patients as well as to the likelihood of success of treatment in the ICU. As a result, they have been suggested for use in the assessment of quality of care within institutions, for the comparison of patient outcomes among ICU in different institutions, and as a means of controlling for severity of illness in studies of ICU patients. These systems may also be useful for issues of cost reimbursement since they provide a quantitative measure of severity of illness. Of considerable debate is whether it is appropriate to use such techniques to assist in patient management decisions including whether to admit patients to the ICU, to triage patients after admission to the ICU, and to continue support of patients who have a poor prognosis.