A fiducial-based confidence interval for the linear combination of multinomial probabilities.

Across a broad set of applications, system outcomes may be summarized as probabilities in confusion or contingency tables. In settings with more than two outcomes (e.g., stages of cancer), these outcomes represent multinomial experiments. Measures to summarize system performance have been presented as linear combinations of the resulting multinomial probabilities. Statistical inference on the linear combination of multinomial probabilities has been focused on large-sample and parametric settings and not small-sample settings. Such inference is valuable, however, especially in settings such as those resulting from pilot or low-cost studies. To address this gap, we leverage the fiducial approach to derive confidence intervals around the linear combination of multinomial parameters with desirable frequentist properties. One of the original arguments against the fiducial approach was its inability to extend to multiparameter settings. Therefore, the great novelty of this work is both the derived interval and the logical framework for applying the fiducial approach in multiparameter settings. Through simulation, we demonstrate that the proposed method maintains a minimum coverage of 1 - α $1 - \alpha$ , unlike the bootstrap and large-sample methods, at comparable interval lengths. Finally, we illustrate its use in a medical problem of selecting classifiers for diagnosing chronic allograph nephropathy in postkidney transplant patients.

The Relationship between Pet Ownership, Social Support, and Internalizing Symptoms in Students from the First to Fourth Year of College.

Internalizing symptoms are prevalent in students as they enter and complete college. Considering research suggesting mental health benefits of pet ownership, this study explores the relationship between pet ownership, social support (SS), and internalizing symptoms (IS) in a cohort of students across their 4-year college experience. With no differences at college entry, students growing up with pets had greater IS through the fourth year, and greater SS through the third year, than those without pets. Currently living with a pet, gender, SS and personality predicted IS in the fourth year. Females experiencing higher IS in their first year are more likely to live with pets in their fourth year, and fourth year females living with pets or greatly missing absent pets have higher IS than females without pets or missing pets less. Findings suggest a unique relationship between IS in female students and their pet relationships not seen in males.

Choosing between the BP and BN sequential strategies.

Cost and burden of diagnostic testing may be reduced if fewer tests can be applied. Sequential testing involves selecting a sequence of tests, but only administering subsequent tests dependent on results of previous tests. This research provides guidance to choosing between single tests or the believe the positive (BP) and believe the negative (BN) sequential testing strategies, using accuracy (as measured by the Youden Index) as the primary determinant. Approximately 75% of the parameter combinations examined resulted in either BP or BN being recommended based on a higher accuracy at the optimal point. In about half of the scenarios BP was preferred, and the other half, BN, with the choice often a function of the value of the ratio of standard deviations of those without and with disease (b). Large values of b for the first test of the sequence tended to be associated with preference for BN as opposed to BP, while small values of b appear to favor BP. When there was no preference between sequences and/or single tests based on the Youden Index, cost of the sequence was considered. In this case, disease prevalence plays a large role in the selection of strategies, with lower values favoring BN and sometimes higher values favoring BP. The cost threshold for the sequential strategy to be preferred over a single, more accurate test, was often quite high. It appears that while sequential strategies most often increase diagnostic accuracy over a single test, sequential strategies are not always preferred.

OR Time and Sacral Pressure Injuries in Critically Ill Surgical Patients.

Critically ill patients are at risk for developing pressure injuries during operative and other invasive procedures. The purpose of this secondary analysis was to explore the relationship of OR time to sacral pressure injuries in critically ill patients using high frequency ultrasound as a method of assessment. The 41 participants examined in this study had both time in the OR and up to eight days of pressure injury data. The multivariable model containing OR bed time, body mass index, and Braden Scale score produced the best prediction of pressure injury (area under the curve = 0.859). A higher body mass index (P = .09), shorter OR bed time (P = .01), and lower Braden Scale score (P = .05) were associated with a greater chance of pressure injury. These results suggest that use of high frequency ultrasound may identify tissue changes before observable skin changes, leading to earlier pressure injury prevention strategies.

Lack of Association of High Backrest With Sacral Tissue Changes in Adults Receiving Mechanical Ventilation.

BACKGROUND: Although higher backrest elevation may be a theoretical risk for integrity of sacral tissues, few data support use of high backrest elevation. OBJECTIVE: To describe the effect of backrest elevation on the integrity of sacral tissue in critically ill adults receiving mechanical ventilation. METHODS: Patients from 3 critical care units (surgical trauma, medical respiratory, and neuroscience) who were expected to have mechanical ventilation for at least 24 hours were intubated and mechanical ventilation was started. Participants were enrolled in the study within 24 hours of intubation. Backrest elevation was continuously measured by using mechanical system- based accelerometers. Integrity of sacral tissue was evaluated by using high-frequency sonography. RESULTS: Data for 84 patients who had measurements of both backrest elevation and skin integrity were available for analysis. General linear models indicated no significant difference among the proportions of time spent at less than 20° (P values: .57 the first 24 hours, .17 the first 48 hours, .81 the first 72 hours), 20° to 30° (P values: .25 the first 24 hours, .08 the first 48 hours, .25 the first 72 hours), or greater than 30° (P values: .62 the first 24 hours, .28 the first 48 hours, .68 the first 72 hours) among participants with no injury, no change in injury, improvement in injury, or injury that worsened. CONCLUSIONS: Level of backrest elevation is not associated with changes in tissue integrity. Body positioning in critically ill patients receiving mechanical ventilation may not be as important or as effective as once thought.

High-Frequency Ultrasound: Obtaining Optimal Images and the Effect of Image Artifacts on Image Quality.

Objective: High-frequency ultrasound (HFUS) images are being researched for use in the prevention, detection, and monitoring of pressure injuries in patients at risk. This seminal longitudinal study in mechanically ventilated adults describes image quality, the incidence of image artifacts, and their effect on image quality in critically ill subjects. Approach: Mechanically ventilated subjects from three adult intensive care units were enrolled, and multiple sacral images from each subject were obtained daily. Using a subset of best image per patient per day, artifacts were grouped, and their effect on image quality was statistically evaluated. Results: Of a total of 1761 images collected from 137 subjects, 8% were rated as poor. In the subset, 70% had good quality ratings. Four groups of artifacts were identified as follows: "bubbles," "texture problems," "layer nondifferentiation," and "reduced area for evaluation." Artifacts from at least one group were found in 83% of images. Bubbles were most frequently seen, but artifacts with adverse effect on image quality were "layer nondifferentiation," "texture problems," and "reduced area for evaluation." Innovation: HFUS image evaluation is still in the development phase with respect to tissue injury use. Artifacts are generally omnipresent. Quickly recognizing artifacts that most significantly affect image quality during scanning will result in higher quality images for research and clinical applications. Conclusion: Good quality images were achievable in study units; although frequent artifacts were present in images, in general, they did not interfere with evaluation. Artifacts related to "layer nondifferentiation" was the greatest predictor of poor image quality, prompting operators to immediately rescan the area.

High-Frequency Ultrasound: Description of Sacral Tissue Characteristics in Healthy Adults.

OBJECTIVES: The purpose of this study was to describe selected sacral tissue characteristics in a convenience sample of healthy volunteer subjects. DESIGN: Descriptive. SAMPLE AND SETTING: Fifty healthy volunteers in a clinical learning center in a school of nursing. METHODS: Sacral scans were obtained using a 20-MHz ultrasound scanning system in 3 positions: prone and 60° and 90° side-lying from the back. The images were analyzed by software in the ultrasound program using quantitative parameters of dermal thickness and density (dermal median intensity and derived number of low-echogenic pixels to total pixels [LEP:TP] ratio). RESULTS: In general, average values were as follows: dermal thickness between 2.32 and 2.65 mm; median pixel intensity between 102 and 112; and the LEP:TP ratio between 0.39 and 0.56. There were significant differences in sacral tissue characteristics between measures of thickness and dermal density (median intensity and LEP:TP ratio) by subject side-lying position (60° and 90°) versus prone position, with all P values less than .0001. CONCLUSIONS: Overall, the ranges were consistent across measures of thickness and dermal density except for systematic differences between side-lying and prone positions. When comparing thickness, median intensity, or LEP:TP ratio, it is important to report subject position. To best recognize tissue inflammation indicative of pressure injuries before surface changes are seen, it is useful to understand healthy high-frequency ultrasound sacral tissue characteristics. It is anticipated that quantitative assessment of dermal thickness, density, and LEP:TP ratio could help identify individuals with incipient pressure injury.

High frequency ultrasound sacral images in the critically ill: Tissue characteristics versus visual evaluation.

PURPOSE: High frequency ultrasound (HFUS) systems may identify tissue injury. We compared HFUS tissue characteristics (dermal thickness and dermal density) with visual image examination. METHODS: Longitudinal study in critically ill mechanically ventilated adults, from three ICUs (Surgical Trauma, Medical Respiratory, Neuroscience) enrolled within 24hours of airway intubation. Sacral HFUS images were obtained daily for up to seven days. Expert evaluation of the best image per day was completed and compared to HFUS generated tissue characteristics (dermal thickness and dermal density). RESULTS: Of the113 subjects with 1614 comparisons analysed, 73.2% to 84% were normal, and 6.3% to 11.8% of the comparisons had injury present but no change was noted in the injury observed. There were no significant differences in one-day comparisons among type of injury and mean dermal thickness (p=0.6645) or dermal median intensity (adjusted p=0.06-0.17). All other day-to-day comparisons were similarly non-significant. CONCLUSIONS: We found no association among dermal density, dermal thickness and visual examination of changes in sacral HFUS images for any day-to-day comparison. The use of sacral HFUS as a screening tool for the development of tissue injury is in its infancy. Additional comparative studies should be conducted to identify its future clinical usefulness.

Tissue interface pressure and skin integrity in critically ill, mechanically ventilated patients.

OBJECTIVE: To describe tissue interface pressure, time spent above critical pressure levels and the effect on skin integrity at seven anatomical locations. DESIGN, SETTING, PATIENTS: Descriptive, longitudinal study in critically ill mechanically ventilated adults, from Surgical Trauma ICU-STICU; Medical Respiratory ICU-MRICU; Neuroscience ICU-NSICU in a Mid-Atlantic urban university medical centre. Subjects were enroled in the study within 24hours of intubation. MEASUREMENTS: Tissue interface pressure was measured continuously using the XSENSOR pressure mapping system (XSENSOR Technology Corporation, Calgary, Canada). Skin integrity was observed at all sites, twice daily, using the National Pressure Ulcer Advisory Panel staging system, for the first seven ICU days and at day 10 and 14. RESULTS: Of the 132 subjects, 90.9% had no observed changes in skin integrity. Maximum interface pressure was above 32mmHg virtually 100% of the time for the sacrum, left and right trochanter. At the 45mmHg level, the left and right trochanter had the greatest amount of time above this level (greater than 95% of the time), followed by the sacrum, left and right scapula, and the left and right heels. Similarly, at levels above 60mmHg, the same site order applied. For those six subjects with sacral skin integrity changes, maximum pressures were greater than 32mmHg 100% of the time. Four of the six sacral changes were associated with greater amounts of time above both 45mmHg and 60mmHg than the entire sample. CONCLUSIONS: Maximum tissue interface pressure was above critical levels for the majority of the documented periods, especially in the sacrum, although few changes in skin integrity were documented. Time spent above critical levels for mean pressures were considerably less compared to maximum pressures. Maximum pressures may have reflected pressure spikes, but the large amount of time above the critical pressure levels remains substantial.

Backrest Elevation and Tissue Interface Pressure by Anatomical Location During Mechanical Ventilation.

BACKGROUND: Backrest elevations less than 30° are recommended to reduce pressure ulcers, but positions greater than 30° are recommended during mechanical ventilation to reduce risk for ventilator-associated pneumonia. Interface pressure may vary with level of backrest elevation and anatomical location (eg, sacrum, heels). OBJECTIVE: To describe backrest elevation and anatomical location and intensity of skin pressure across the body in patients receiving mechanical ventilation. METHODS: In a longitudinal study, patients from 3 adult intensive care units in a single institution receiving mechanical ventilation were enrolled within 24 hours of intubation from February 2010 through May 2012. Backrest elevation (by inclinometer) and pressure (by a pressure-mapping system) were measured continuously for 72 hours. Mean tissue interface pressure was determined for 7 anatomical areas: left and right scapula, left and right trochanter, sacrum, and left and right heel. RESULTS: Data on 133 patients were analyzed. For each 1° increase in backrest elevation, mean interface pressure decreased 0.09 to 0.42 mm Hg. For each unit increase in body mass index, mean trochanter pressure increased 0.22 to 0.24 mm Hg. Knee angle (lower extremity bent at the knee) and mobility were time-varying covariates in models of the relationship between backrest elevation and tissue interface pressure. CONCLUSIONS: Individual factors such as patient movement and body mass index may be important elements related to risk for pressure ulcers and ventilator-associated pneumonia, and a more nuanced approach in which positioning decisions are tailored to optimize outcomes for individual patients appears warranted.

A nonparametric fiducial interval for the Youden index in multi-state diagnostic settings.

The Youden index is a commonly employed metric to characterize the performance of a diagnostic test at its optimal point. For tests with three or more outcome classes, the Youden index has been extended; however, there are limited methods to compute a confidence interval (CI) about its value. Often, outcome classes are assumed to be normally distributed, which facilitates computational formulas for the CI bounds; however, many scenarios exist for which these assumptions cannot be made. In addition, many of these existing CI methods do not work well for small sample sizes. We propose a method to compute a nonparametric interval about the Youden index utilizing the fiducial argument. This fiducial interval ensures that CI coverage is met regardless of sample size, underlying distributional assumptions, or use of a complex classifier for diagnosis. Two alternate fiducial intervals are also considered. A simulation was conducted, which demonstrates the coverage and interval length for the proposed methods. Comparisons were made using no distributional assumptions on the outcome classes and for when outcomes were assumed to be normally distributed. In general, coverage probability was consistently met, and interval length was reasonable. The proposed fiducial method was also demonstrated in data examining biomarkers in subjects to predict diagnostic stages ranging from normal kidney function to chronic allograph nephropathy. Published 2015. This article is a U.S. Government work and is in the public domain in the USA.

Effects of position and operator on high-frequency ultrasound scan quality.

OBJECTIVES: High-frequency ultrasound may evaluate those at risk for pressure ulcers. Images may be obtained by clinicians with limited training. The prone position is recommended for obtaining sacral scans but may not be feasible in the critically ill. This study investigated image quality using multiple operators and a variety of patient positions. RESEARCH METHODOLOGY: Sacral scans were performed in three randomised positions in 50 volunteers by three different investigators using a 20 MHz ultrasound system. General linear models and ANOVA random effects models were used to examine the effects of operator and position on image quality rating, and measures of dermal thickness and dermal density. RESULTS: The best scan for each position and operator was used for analysis (n=447 images). Image rating varied by operator (p=0.0004), although mean ratings were 3.5 or above for all operators. Dermal thickness was less for the prone position than in 90° or 60° side-lying positions (p=0.0137, p=0.0003). Dermal density was lower for the prone position than for the 90° or 60° positions (p<0.0001 for both). CONCLUSIONS: These data show that overall scan quality was acceptable in all positions with all operators. However, differences were found between side-lying positions and the prone for dermal thickness and dermal density measures.

Use of high frequency ultrasound to detect changes in skin integrity: An image evaluation validation procedure.

OBJECTIVES: High frequency ultrasound (HFUS) scanning may be used for prevention, detection and monitoring of pressure ulcers in patients at risk and is amenable for portable, bedside use by a variety of clinicians. Limited data are available about the criteria to determine an ideal image or measures of tissue changes representative of tissue injury. We developed and evaluated criteria for overall image quality and measures of tissue integrity. METHODS: In 40 mechanically ventilated adults in 3 ICUs, 241 HFUS sacral images were evaluated for agreement using criteria for overall image quality and tissue changes (dermal, hypodermal layer thickness and layer density). RESULTS: HFUS sacral images (N=241) were evaluated in three analyses and showed poor agreement in all three analyses using the specific criteria for global quality, however when criteria were collapsed agreement was good to substantial. Evaluator agreement for layer thickness and layer density was also good. CONCLUSIONS: A global rating is adequate for identifying good images. Agreement for measurements of layer thickness and density were also good and may be useful to identify early changes in tissue integrity leading to tissue injury. Additional data are needed concerning the association of changes in layer thickness and layer density to eventual tissue injury.

Predictors of agitation in critically ill adults.

BACKGROUND: Agitation in critically ill adults is a frequent complication of hospitalization and results in multiple adverse outcomes. Potential causes of agitation are numerous; however, data on factors predictive of agitation are limited. OBJECTIVES: To identify predictors of agitation by examining demographic and clinical characteristics of critically ill patients. METHODS: A medical record review was performed. Documentation of agitation was indicated by scores on the Richmond Agitation-Sedation Scale or the use of an agitation keyword. Records of 200 patients from 1 medical and 1 surgical intensive care unit were used for the study. Risk factors were determined for 2 points in time: admission to the intensive care unit and within 24 hours before the first episode of agitation. Data on baseline demographics, preadmission risk factors, and clinical data were collected and were evaluated by using logistic multivariable regression to determine predictors of agitation. RESULTS: Predictors of agitation on admission to intensive care were history of use of illicit substances, height, respiratory and central nervous system subscores on the Sequential Organ Failure Assessment, and use of restraints. Predictors of agitation within 24 hours before the onset of agitation were history of psychiatric diagnosis, height, score on the Sequential Organ Failure Assessment, ratio of Pao2 to fraction of inspired oxygen less than 200, serum pH, percentage of hours with restraints, percentage of hours of mechanical ventilation, pain, and presence of genitourinary catheters. CONCLUSIONS: Predictors of agitation on admission and within 24 hours before the onset of agitation were primarily clinical variables.

Agitation onset, frequency, and associated temporal factors in critically ill adults.

BACKGROUND: Agitation is a frequent complication in critically ill adults, can result in life-threatening events for patients or care providers, and extends the hospital length of stay, thereby increasing hospital costs. OBJECTIVES: To describe the incidence, onset, and temporal factors related to agitation in critically ill adults. METHODS: Data were collected for the first 5 days of stay of all adult patients consecutively admitted to a medical respiratory intensive care unit and a surgical trauma intensive care unit during a 2-month period. Agitation was documented by using scores on the Richmond Agitation-Sedation Scale or notation of agitation in the medical record. The hour was used as the documentation epoch, and data were summarized by hour, 4-hour block, and day for each patient. RESULTS: Data were collected on 200 patients, 100 from each unit. Among the sample, 118 (59%) were agitated at some time during the 5 days. The overall agitation rate was 7.8% of the total hourly time. Mean onset of agitation was 11.6 hours from time of admission to the unit. Of the 118 patients who were agitated at some time, 102 (86%) had agitation on day 1. Compared with patients in the surgical trauma unit, patients in the medical respiratory unit had significantly more hours of agitation the first day and first hour of admission and significantly earlier onset of agitation. CONCLUSIONS: Agitation was present in more than one-half of the patients in the sample, typically developed on the first day, and involved consecutive days.

Confidence intervals around Bayes Cost in multi-state diagnostic settings to estimate optimal performance.

A critical feature of diagnostic testing is correctly classifying subjects based upon specified thresholds of some measure. The commonly employed Youden index determines a test's optimal thresholds by maximizing the correct classification rates for a diagnostic scenario. An alternative to the Youden index is the cost function, Bayes Cost (BC). BC determines a test's optimal setting by minimizing the sum of all misclassification rates from the test. Unlike the Youden index, BC can consider a priori costs of all the diagnostic outcomes including class specific misclassifications regardless of the number of classes. Delta method approximate confidence intervals around BC are derived under the assumption of normally distributed classes as a means for quantifying a test's performance and comparing classifiers at their optimal settings in a multi-state diagnostic framework. A simulation study is conducted to demonstrate the performance of the derived confidence intervals that are found to perform well, especially for sample sizes of 50 or larger in each diagnostic class. Finally, the proposed methods are applied to a four-class breast tissue classification problem, where four possible discriminatory features are compared under varying decision cost structures. Using the confidence intervals around BC, the best feature for classification is selected, and the optimal thresholds and their 95% confidence intervals are determined.

Reducing cost in sequential testing: a limit of indifference approach.

In noninferiority studies, a limit of indifference is used to express a tolerance in results such that the clinician would regard such results as being acceptable or 'not worse'. We applied this concept to a measure of accuracy, the Receiver Operating Characteristic (ROC) curve, for a sequence of tests. We expressed a limit of indifference for the range of acceptable sensitivity values and examined the associated cost of testing within this range. In doing so, we generated the minimum cost maximum ROC (MCMROC) curve, which reflects the reduced sensitivity and cost of testing. We compared the MCMROC and its associated cost curve between limits of indifference set to 0.999 [a 0.1% reduction in true positive rate (TPR)], 0.95 (a 5% reduction in TPR), and 1 (no reduction in TPR). The limit of indifference tended to have less of an effect on the MCMROC curves than on the associated cost curves that were greatly affected. Cost was reduced at high false positive rates (FPRs) at higher limit of indifference (0.999) and at small FPRs as the limit of indifference decreased (0.95). These patterns were also observed as applied to sequential strategies used to diagnose diabetes in the Pima Indians.

Pain and inflammation in women with early-stage breast cancer prior to induction of chemotherapy.

CONTEXT: Pain is a commonly experienced and distressing symptom in women with breast cancer (BCA), and recent evidence suggests that immune activation may be associated with pain and other co-occurring symptoms. However, no studies to date have explored the relationships among perceived pain and biomarkers of inflammation in women with early-stage BCA during the initial course of treatment. OBJECTIVES: The purpose of this research study was to examine the relationships among pro- and anti-inflammatory biomarkers and the presence of pain and other symptoms (anxiety, depression, fatigue, and sleep disorder) prior to induction of chemotherapy. METHOD: This was a secondary analysis of data that measured perceived symptoms, including the presence of pain and pain interference, and plasma levels of pro- and anti-inflammatory cytokines and C-reactive protein (CRP) in women with early-stage BCA (N = 32) at 1 month postsurgery but prior to induction of chemotherapy. RESULTS: Women experiencing pain had significantly higher levels of CRP (p < .01), interleukin (IL) 13 (p < .02), and IL-7 (p < .02) and more pain interference (p < .01), depression (p < .01), and sleep disturbance (p < .01) compared to women reporting no pain. CONCLUSION: The presence of pain during the initial course of treatment in women with early-stage BCA was associated with significantly higher levels of CRP, IL-7, and IL-13, suggesting a potential role of immune activation in perceived pain. Further research to examine the precise effects of these biological factors in modulating pain is needed. Perceived pain was also associated with multiple co-occurring symptoms, and this finding has important implications for symptom management.

A pilot study of a weekend retreat intervention for family survivors of homicide.

Homicide causes negative unintended consequences for family survivors. Family survivors face complicated grief and overwhelming loss with minimal support from others. The authors offered a retreat intervention as a way to ameliorate the effects of the homicidal death for family survivors of homicide. An exploratory longitudinal pilot study examined the feasibility and acceptability of the intervention and explored the impact of the TOZI© Healing intervention on participants' distress symptoms. Eight family members participated in the 2-day retreat and completed surveys at five time intervals over 30 months. Descriptive statistics and correlations were used to analyze the data. Although sample sizes were too small to achieve statistical significance, changes on selected holistic health outcomes, supported by overwhelmingly positive focus group responses to the intervention, affirm the need for further study.