Classification and automatic scoring of arousal intensity during sleep stages using machine learning.

Arousal during sleep can result in sleep fragmentation and various physiological effects, impairing cognitive function and raising blood pressure and heart rate. However, the current definition of arousal has limitations in assessing both amplitude and duration, making it challenging to measure sleep fragmentation accurately. Moreover, there is inconsistency among inter-raters in arousal scoring, which renders it susceptible to subjective variability. Therefore, this study aims to identify a highly accurate classifier for each sleep stage by employing optimized feature selection and machine learning models. According to electroencephalography (EEG) signals during the arousal phase, the intensity level was categorized into four levels. For control, the non-arousal cases were used as level 0 and referred as sham arousal, resulting in five arousal intensity levels. Wavelet transform was applied to analyze sleep arousal to extract features from EEG. Based on these features, we classified arousal intensity levels through machine learning algorithms. Due to the different characteristics of EEG in each sleep stage, the classification model was optimized for the four sleep stages. Excluding sham arousals, a total of 13,532 arousal events were used. The lowest intensity in the entire data, level 1, was computed to be 3107, level 2 was 3384, level 3 was 3472, and the highest intensity of level 4 was 3,569. The optimized classification model for each sleep stage achieved an average sensitivity of 82.68%, specificity of 95.68%, and AUROC of 96.30%. The sensitivity of the control, arousal intensity level 0, was 83.07%, a 1.25% increase over the unoptimized model and a 14.22% increase over previous research. This study used machine learning techniques to develop classifiers for each sleep stage, improving the accuracy of arousal intensity classification. The classifiers showed high sensitivity and specificity and revealed the unique characteristics of arousal intensity during different sleep stages. These findings represent a novel approach to arousal research and have implications for developing more accurate predictive models in sleep research.

Development of service standards and manpower calculation criteria for hospital clinical pharmacies in South Korea: a survey-based study.

BACKGROUND: After the revision of the Korean Pharmaceutical Affairs Act, the certification of specialized pharmacists is scheduled to be legally recognized in 2023. Considering that the specialized pharmacist certification was developed based on the working model of hospital clinical pharmacists, it is necessary to establish standards for clinical pharmacists in hospitals and to calculate appropriate manpower. Through this study, we aim to establish practical standards for clinical pharmacists and propose a method for calculating staffing levels based on an investigation of actual workloads. METHODS: This survey-based study consisted of two phases. In the first phase, a literature review was conducted to establish standards for clinical pharmacy services, and tasks in relevant literature were classified to identify clinical pharmacy service tasks that are applicable to the practice of Korean hospitals. Additionally, a preliminary survey was conducted to investigate the essential tasks. In the second phase of the investigation, a multicenter survey was conducted targeting pharmacists in facilities with more than 1,000 beds to explore their perceptions and actual workloads related to tasks. RESULTS: According to the standards for clinical pharmacists in Korea, clinical pharmacy services consist of a total of 23 tasks, of which 16 have been identified as essential tasks. Essential tasks accounted for 93% of the total tasks in clinical pharmacy services. The average full-time equivalent (FTE) through workload calculation was 2.5 ± 1.9 for each field, while the FTE allocated to actual practice was 2.1 ± 1.6. The distribution of each type of clinical pharmacy service was as follows: 77% for medication therapy management, 13% for medication education, 8% for multidisciplinary team activities, and 3% for medication use evaluation. CONCLUSION: This study identified essential tasks common to clinical pharmacy services across different healthcare institutions. However, the FTE of clinical pharmacists in actual practice was insufficient compared to the required amount. In order to establish and expand clinical pharmacy services in a hospital, it is necessary to ensure an adequate workforce for essential tasks.

A novel sleep aid device to reduce sleep latency using air-CO2 mixed gas.

Introduction: Sleep is an indispensable component of human life. However, in modern times, the number of people suffering from sleep disorders, such as insomnia and sleep deprivation, has increased significantly. Therefore, to alleviate the discomfort to the patient due to lack of sleep, sleeping pills and various sleep aids are being introduced and used. However, sleeping drugs are prescribed only to a limited extent due to the side effects posed by them and resistance to such drugs developed by patients in the long term, and the majority of sleep aids are scientifically groundless products. This study aimed to develop a device that induced sleep by spraying a mixed gas of carbon dioxide and air to create an environment that could induce sleep, similar to the inside of a sealed vehicle, to control oxygen saturation in the body. Methods: Based on the stipulated safety standards and the human tidal volume, the target concentration of carbon dioxide was determined to be of three types: 15,000, 20,000, and 25,000 ppm. After analyzing diverse structures for safely mixing gases, the most appropriate shape, the reserve tank, was selected as the best suited structure. Various variables, such as spraying angle and distance, flow rate, atmospheric temperature, and nozzle length, were comprehensively measured and tested. Furthermore based on this aspect, diffusion simulation of carbon dioxide concentration and actual experiments were conducted. To secure the stability and reliability of the developed product, an accredited test was performed to investigate the error rate of carbon dioxide concentration. Furthermore, clinical trials comprising polysomnography and questionnaires confirmed the effectiveness of the developed product not only in reducing sleep latency but also in enhancing the overall sleep quality. Results: When the developed device was put to use in reality, sleep latency was decreased by 29.01%, on average, for those with a sleep latency of 5 min or more, compared to when the device was not in use. Moreover, the total sleep time was increased by 29.19 min, WASO was decreased by 13.17%, and sleep efficiency was increased by 5.48%. We also affirmed that the ODI and 90% ODI did not decrease when the device was used. Although different questions may be raised about the safety of using a gas such as carbon dioxide (CO2), the result that tODI was not reduced shows that sleep aids using CO2 mixtures do not adversely affect human health. Discussion: The results of this study suggest a new method that can be used to treat sleep disorders including insomnia.

Application of various machine learning techniques to predict obstructive sleep apnea syndrome severity.

As the incidence of obstructive sleep apnea syndrome (OSAS) increases worldwide, the need for a new screening method that can compensate for the shortcomings of the traditional diagnostic method, polysomnography (PSG), is emerging. In this study, data from 4014 patients were used, and both supervised and unsupervised learning methods were used. Clustering was conducted with hierarchical agglomerative clustering, K-means, bisecting K-means algorithm, Gaussian mixture model, and feature engineering was carried out using both medically researched methods and machine learning techniques. For classification, we used gradient boost-based models such as XGBoost, LightGBM, CatBoost, and Random Forest to predict the severity of OSAS. The developed model showed high performance with 88%, 88%, and 91% of classification accuracy for three thresholds for the severity of OSAS: Apnea-Hypopnea Index (AHI) [Formula: see text] 5, AHI [Formula: see text] 15, and AHI [Formula: see text] 30, respectively. The results of this study demonstrate significant evidence of sufficient potential to utilize machine learning in predicting OSAS severity.

Association between Statin Use and Sensorineural Hearing Loss in Type 2 Diabetic Patients: A Hospital-Based Study.

Statins have emerged as protective agents against sensorineural hearing loss (SNHL) associated with dyslipidemia, but the effects of statins on SNHL are not consistent. The purpose of this study was to investigate the association between statin use and the risk of SNHL using a hospital cohort. This nested case-control study included type 2 diabetic patients over the age of 18 years without a history of hearing loss. Of these, 1379 patients newly diagnosed with SNHL or tinnitus were classified as cases, and 5512 patients matched to the cases based on age, sex, and index year were classified as controls. Chi-squared tests were used to compare categorical variables between the two groups. Odds ratios (ORs) and adjusted odds ratios (AOR) were calculated from univariate and multivariable unconditional logistic regression analyses, respectively. There was a significant difference in the prevalence of statin use between the cases and controls (53.7% vs. 61.2%, respectively; p < 0.001). The use of statins in type 2 diabetic patients significantly reduced the risk of SNHL or tinnitus by 24.8% (95% CI 14.2-34.1%, p < 0.001) after controlling for confounders. Similar results were found for the association between statin use and SNHL (AOR = 0.706; 95% CI 0.616-0.811, p < 0.001). The protective effects of statins against SNHL were consistent regardless of age and sex. The use of statins for type 2 diabetic patients was significantly associated with a reduced risk of SNHL, regardless of age and sex. Further studies are needed, especially large cohort studies, to evaluate the long-term protective effects of statins.

Impact of a clinical decision support system for high-alert medications on the prevention of prescription errors.

OBJECTIVE: To evaluate the impact of a high-alert medication clinical decision support system called HARMLESS on point-of-order entry errors in a tertiary hospital. METHOD: HARMLESS was designed to provide three kinds of interventions for five high-alert medications: clinical knowledge support, pop-ups for erroneous orders that block the order or provide a warning, and order recommendations. The impact of this program on prescription order was evaluated by comparing the orders in 6 month periods before and after implementing the program, by analyzing the intervention log data, and by checking for order pattern changes. RESULT: During the entire evaluation period, there were 357,417 orders and 5233 logs. After HARMLESS deployment, orders that omitted dilution fluids and exceeded the maximum dose dropped from 12,878 and 214 cases to 0 and 9 cases, respectively. The latter nine cases were unexpected, but after the responsible programming error was corrected, there were no further such cases. If all blocking interventions were seen as errors that were prevented, this meant that 4137 errors (3584 of which were 'dilution fluid omitted' errors) were prevented over the 6-month post-deployment period. There were some unexpected order pattern changes after deployment and several unexpected errors emerged, including intramuscular or intravenous push orders for potassium chloride (although a case review revealed that the drug was not actually administered via these methods) and an increase in pro re nata (PRN; administer when required) orders for most drugs. CONCLUSION: HARMLESS effectively implemented blocking interventions but was associated with the emergence of unexpected errors. After a program is deployed, it must be monitored and subjected to data analysis to fix bugs and prevent the emergence of new error types.

Prevalence of psoriasis among adults in the U.S.: 2003-2006 and 2009-2010 National Health and Nutrition Examination Surveys.

BACKGROUND: A 2010 CDC-sponsored consultation of psoriasis, psoriatic arthritis, and public health experts developed a public health agenda for psoriasis and psoriatic arthritis indicating that additional population-based research is needed to better characterize psoriasis in the population. PURPOSE: To better characterize the burden of psoriasis in the U.S. using recent population-based, cross-sectional data in this 2012 analysis. METHODS: A subset of 10,676 adults aged 20-59 years from the 2003-2006 and 2009-2010 National Health and Nutrition Examination Surveys was used to examine psoriasis prevalence, severity, disparities, health-related quality of life, and selected comorbidities. RESULTS: The overall prevalence of psoriasis was 3.1% (95% CI=2.6, 3.6); extrapolating to older adults suggests that 6.7 million adults aged ≥20 years are affected. Psoriasis was significantly more prevalent among non-Hispanic whites than other race/ethnicity subgroups, as well as among those with arthritis. Approximately 82% reported no/little or mild disease; the impact of psoriasis on daily life increased with disease severity (p=0.0001 for trend). Those with psoriasis reported significantly more frequent mental distress or mild to severe depression than those without psoriasis. Psoriasis was also significantly associated with obesity and former smoking status. CONCLUSIONS: Psoriasis is a large public health problem. Further characterizing psoriasis from a public health perspective will require better survey questions and inclusion of these questions in national surveys.

Lessons learned from the development of health applications in a tertiary hospital.

BACKGROUND: Adoption of smart devices for hospital use has been increasing with the development of health applications (apps) for patient point-of-care and hospital management. To promote the use of health apps, we describe the lessons learned from developing 12 health apps in the largest tertiary hospital in Korea. MATERIALS AND METHODS: We reviewed and analyzed 12 routinely used apps in three categories-Smart Clinic, Smart Patient, and Smart Hospital-based on target users and functions. The log data for each app were collected from the date of release up until December 2012. RESULTS: Medical personnel accessed a mobile electronic medical record app classified as Smart Clinic an average of 452 times per day. Smart Hospital apps are actively used to communicate with each other. Patients logged on to a mobile personal health record app categorized as Smart Patient an average of 222 times per day. As the mobile trend, the choice of supporting operating system (OS) is more difficult. By developing these apps, a monitoring system is needed for evaluation. CONCLUSIONS: We described the lessons learned regarding OS support, device choice, and developmental strategy. The OS can be chosen according to market share or hospital strategic plan. Smartphones were favored compared with tablets. Alliance with an information technology company can be the best way to develop apps. Health apps designed for smart devices can be used to improve healthcare. However, to develop health apps, hospitals must define their future goals and carefully consider all the aspects.

Prevalence of smoking associated with pregnancy in three Southern Ontario Health Units.

OBJECTIVES: The objectives of this study were to determine the prevalence of pregnancy-associated smoking among women residing in three Southern Ontario Health Units and to examine potential risk factors for smoking during pregnancy, using an existing data collection mechanism. METHODS: During May 2001, questions about pregnancy-associated smoking were asked during the telephone follow-up of postpartum women living in the three health units in Southern Ontario; this follow-up is routinely conducted by public health nurses. Sociodemographic data were also obtained. Data from 1,134 women were analyzed concerning smoking before and after the occurrence of the pregnancy was known, during each trimester, and immediately postpartum. RESULTS: The rates of smoking before and after the pregnancy was known, in the first, second, and third trimesters, and immediately postpartum were 17.8%, 10.4%, 9.6%, 8.7%, 8.1%, and 7.9%, respectively. For all six estimates of smoking, Canadian-born women had rates 2.5 to 4 times higher than those of women born outside Canada. Age less than 25 years and lower educational attainment were also independent risk factors for smoking during pregnancy. CONCLUSIONS: The Ontario Tobacco Strategy goal of eliminating smoking in pregnancy has not yet been realized. Ongoing smoking cessation programs among pregnant women are needed as part of a comprehensive strategy to reduce the overall prevalence of smoking. In planning such programs, particular attention should be paid to the needs of women who are Canadian-born, have lower educational attainment, and are under the age of 25.