An automated patient monitoring using discrete‐time wireless sensor networks

In recent years, Internet of Things (IoT) devices are used for remote health monitoring. For remotely monitoring a patient, only the health information at different time points are not sufficient; predicted values of biomarkers (for some future time points) are also important. In this article, we propose a powerful statistical model for an efficient dynamic patient monitoring using wireless sensor nodes through Bayesian Learning (BL). We consider the setting where a set of correlated biomarkers are measured from a patient through wireless sensors, but the sensors only report the ordinal outcomes (say, good, fair, high, or very high) to the sink based on some prefixed thresholds. The challenge is to use the ordinal outcomes for monitoring and predicting the health status of the patient under consideration. We propose a linear mixed model where interbiomarker correlations and intrabiomarker dependence are modeled simultaneously. The estimated and the predicted values of the biomarkers are transferred over the internet so that health care providers and the family members of the patient can remotely monitor the patient. Extensive simulation studies are performed to assess practical usefulness of our proposed joint model, and the performance of the proposed joint model is compared to that of some other traditional models used in the literature.     

Is evidence-based practice diverse enough? Philosophy of science considerations.

In its policy rationale for evidence-based practice in psychology (EBPP), the APA Presidential Task Force on Evidence-Based Practice (see record 2006-05893-001) claims to have constituted itself with "scientists and practitioners from a wide range of perspectives and traditions, reflecting the diverse perspectives within the field" (p. 273). We applaud this attention to diversity but contend that an entire perspective of the debate was omitted in the Task Force's newly approved policy and its underlying report. The failure to consider a philosophy of science perspective led the Task Force to make a number of epistemological assumptions that are not based on evidence or rationale and that thus violate the very spirit of evidence-based decision making. In this comment, we reveal a few of these assumptions and discuss their detrimental consequences. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Predictors of outcome in LAAM, buprenorphine, and methadone treatment for opioid dependence.

This study examined (1) predictors of treatment outcome for opioid-dependent participants in a single-site controlled trial comparing methadone, buprenorphine, and LAAM treatments and (2) the extent to which various subpopulations of patients may have more successful outcomes with each medication. The relationships between patient demographics, drug use history, and psychological status and outcome measures of treatment retention, opiate use, and cocaine use were assessed. We believe this study to be the first to demonstrate that predictors of treatment success appear to be largely similar in LAAM, buprenorphine, and methadone treatment for opioid dependence. We did not find any factors that would strongly guide selection of one medication over others. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Differences in Dialysis Practice Are the Main Reasons for the High Mortality Rate in the United States Compared to Japan

The cumulative survival of Japanese hemodialysis patients is more than 2.5 times better than that of dialysis patients in the United States (U.S.). The difference is particularly pronounced in older patients, being 4 times better in patients over the age of 50 years. The mortality in U.S. patients has increased from 10 to 25% over the last three decades, but has remained stable at around 10% in Japan.
There is no obvious difference in patient selection. The Japanese accept almost as high a proportion of diabetic patients as does the United States, and the mean age of incident patients is higher in Japan.
Renal transplantation, virtually absent in Japan, should increase mortality in U.S. dialysis patients by removing patients with the highest probability of survival, but even if one adds surviving transplant patients and studies prevalent populations, the survival rate is much better in Japan. Genetic factors are unlikely to explain differences in mortality, as older Americans live much longer than older Japanese.
We speculate that the difference lies in the practice of dialysis. Patients in the United States are generally treated by much faster and shorter dialysis than in Japan. This puts a severe burden on the cardiovascular system of older patients, leading to the poorer survival rate. Japanese physicians also appear to be better trained in dialysis and to spend more time with their patients. The nursing shortage in the United States may also contribute to the increased mortality. Whatever the explanations, the U.S. dialysis community must work to equal and, hopefully, surpass the now superior survival of Japanese dialysis patients.     

重症监护病房紧急状况预警算法

重症监护病房中的病人身体状况通常很不稳定,常出现各种需要医护人员介入治疗的紧急状况。由于医疗资源有限,医护人员可能无法及时发现并处理这些紧急状况,给病人的存活率带来严重的负面影响。如果可以预测这些紧急状况的发生,并及时通知相关医护人员,将大大提高病人的存活率。常见重症监护病房紧急状况包括突然死亡、败血症、肺部感染、急性低血压、以及器官衰竭等。紧急状况预警建模主要采用病人的长时间生命体征监测数据,预测在一定时间之后发生某种紧急状况的可能性。预警模型所采用的监测数据分为静态数据、事件数据和时间序列数据等三类。静态数据具有容易采集、但预测准确性偏低的特点。事件数据或时间序列数据、以及多种类型数据的混合数据对于紧急状况预警模型的预测性能的提高有重要作用,将会获得更广泛的应用。     

考虑多样化需求的诊疗设备维护和患者调度联合优化研究

针对诊疗设备维护和住院患者候诊难以有效协调的问题,建立了联合优化模型。首先,假设设备具有离散的劣化状态,将设备的劣化过程建模为连续时间马尔可夫链;其次,考虑到患者对诊疗设备的不同功能频率需求,以及不同劣化状态对患者治疗时间、费用的影响,以患者就诊顺序、检查策略、修复策略为决策变量建立了设备维护和患者调度模型;最后,采用改进后的非支配排序遗传算法对多目标问题进行了求解。实验结果验证了设备维护与患者调度联合优化模型的有效性。改进后的算法提高了整体和局部的搜索能力,且具有鲁棒性。     

A review of the use of human factors classification frameworks that identify causal factors for adverse events in the hospital setting

Various human factors classification frameworks have been used to identified causal factors for clinical adverse events. A systematic review was conducted to identify human factors classification frameworks that identified the causal factors (including human error) of adverse events in a hospital setting. Six electronic databases were searched, identifying 1997 articles and 38 of these met inclusion criteria. Most studies included causal contributing factors as well as error and error type, but the nature of coding varied considerably between studies. The ability of human factors classification frameworks to provide information on specific causal factors for an adverse event enables the focus of preventive attention on areas where improvements are most needed. This review highlighted some areas needing considerable improvement in order to meet this need, including better definition of terms, more emphasis on assessing reliability of coding and greater sophistication in analysis of results of the classification.

Practitioner Summary: Human factors classification frameworks can be used to identify causal factors of clinical adverse events. However, this review suggests that existing frameworks are diverse, limited in their identification of the context of human error and have poor reliability when used by different individuals.