Category: Statistics

N Z Med J. 2021 Feb 19;134(1530):12-20.

ABSTRACT

AIM: Burnout has a damaging effect on both the wellbeing of medical professionals and patients alike. Empathy is an important part of the therapeutic relationship and could be damaged by burnout. We aimed to describe the prevalence of burnout, assess levels of empathy and explore the relationship between burnout and empathy among senior medical officers (SMOs). We hypothesised that there would be a negative correlation between empathy and burnout.

METHOD: This was a cross-sectional observational study involving SMOs from a variety of specialities. The focus is on SMOs with relatively prolonged contact times with patients. Email invitations were sent out requesting participation in an electronic survey on the QuestionPro platform. The survey comprised 42 questions enquiring about demographics, empathy (Jefferson Scale of Physician Empathy) and burnout (Copenhagen Burnout Inventory). Correlational analyses were performed.

RESULTS: Three hundred and fourteen invitations were sent out and 178 responses were received (56.7% response rate). Forty-five percent of SMOs surveyed were experiencing high levels of personal burnout. There was a statistically significant negative correlation between empathy and patient-related burnout (p=0.018).

CONCLUSIONS: The results show high levels of personal burnout among SMOs and suggest that empathy reduces as patient-related burnout increases. The nature of this relationship is a complex one, and other contributing variables should be considered.

PMID:33651773

PLoS One. 2021 Mar 2;16(3):e0247046. doi: 10.1371/journal.pone.0247046. eCollection 2021.

ABSTRACT

Interacting Particle Systems (IPSs) are used to model spatio-temporal stochastic systems in many disparate areas of science. We design an algorithmic framework that reduces IPS simulation to simulation of well-mixed Chemical Reaction Networks (CRNs). This framework minimizes the number of associated reaction channels and decouples the computational cost of the simulations from the size of the lattice. Decoupling allows our software to make use of a wide class of techniques typically reserved for well-mixed CRNs. We implement the direct stochastic simulation algorithm in the open source programming language Julia. We also apply our algorithms to several complex spatial stochastic phenomena. including a rock-paper-scissors game, cancer growth in response to immunotherapy, and lipid oxidation dynamics. Our approach aids in standardizing mathematical models and in generating hypotheses based on concrete mechanistic behavior across a wide range of observed spatial phenomena.

PMID:33651796 | DOI:10.1371/journal.pone.0247046

Clin Orthop Relat Res. 2021 Feb 26. doi: 10.1097/CORR.0000000000001685. Online ahead of print.

ABSTRACT

BACKGROUND: Vertebral fractures are the most common osteoporotic fractures in older individuals. Recent studies suggest that the performance of artificial intelligence is equal to humans in detecting osteoporotic fractures, such as fractures of the hip, distal radius, and proximal humerus. However, whether artificial intelligence performs as well in the detection of vertebral fractures on plain lateral spine radiographs has not yet been reported.

QUESTIONS/PURPOSES: (1) What is the accuracy, sensitivity, specificity, and interobserver reliability (kappa value) of an artificial intelligence model in detecting vertebral fractures, based on Genant fracture grades, using plain lateral spine radiographs compared with values obtained by human observers? (2) Do patients’ clinical data, including the anatomic location of the fracture (thoracic or lumbar spine), T-score on dual-energy x-ray absorptiometry, or fracture grade severity, affect the performance of an artificial intelligence model? (3) How does the artificial intelligence model perform on external validation?

METHODS: Between 2016 and 2018, 1019 patients older than 60 years were treated for vertebral fractures in our institution. Seventy-eight patients were excluded because of missing CT or MRI scans (24% [19]), poor image quality in plain lateral radiographs of spines (54% [42]), multiple myeloma (5% [4]), and prior spine instrumentation (17% [13]). The plain lateral radiographs of 941 patients (one radiograph per person), with a mean age of 76 ± 12 years, and 1101 vertebral fractures between T7 and L5 were retrospectively evaluated for training (n = 565), validating (n = 188), and testing (n = 188) of an artificial intelligence deep-learning model. The gold standard for diagnosis (ground truth) of a vertebral fracture is the interpretation of the CT or MRI reports by a spine surgeon and a radiologist independently. If there were any disagreements between human observers, the corresponding CT or MRI images would be rechecked by them together to reach a consensus. For the Genant classification, the injured vertebral body height was measured in the anterior, middle, and posterior third. Fractures were classified as Grade 1 (< 25%), Grade 2 (26% to 40%), or Grade 3 (> 40%). The framework of the artificial intelligence deep-learning model included object detection, data preprocessing of radiographs, and classification to detect vertebral fractures. Approximately 90 seconds was needed to complete the procedure and obtain the artificial intelligence model results when applied clinically. The accuracy, sensitivity, specificity, interobserver reliability (kappa value), receiver operating characteristic curve, and area under the curve (AUC) were analyzed. The bootstrapping method was applied to our testing dataset and external validation dataset. The accuracy, sensitivity, and specificity were used to investigate whether fracture anatomic location or T-score in dual-energy x-ray absorptiometry report affected the performance of the artificial intelligence model. The receiver operating characteristic curve and AUC were used to investigate the relationship between the performance of the artificial intelligence model and fracture grade. External validation with a similar age population and plain lateral radiographs from another medical institute was also performed to investigate the performance of the artificial intelligence model.

RESULTS: The artificial intelligence model with ensemble method demonstrated excellent accuracy (93% [773 of 830] of vertebrae), sensitivity (91% [129 of 141]), and specificity (93% [644 of 689]) for detecting vertebral fractures of the lumbar spine. The interobserver reliability (kappa value) of the artificial intelligence performance and human observers for thoracic and lumbar vertebrae were 0.72 (95% CI 0.65 to 0.80; p < 0.001) and 0.77 (95% CI 0.72 to 0.83; p < 0.001), respectively. The AUCs for Grades 1, 2, and 3 vertebral fractures were 0.919, 0.989, and 0.990, respectively. The artificial intelligence model with ensemble method demonstrated poorer performance for discriminating normal osteoporotic lumbar vertebrae, with a specificity of 91% (260 of 285) compared with nonosteoporotic lumbar vertebrae, with a specificity of 95% (222 of 234). There was a higher sensitivity 97% (60 of 62) for detecting osteoporotic (dual-energy x-ray absorptiometry T-score ≤ -2.5) lumbar vertebral fractures, implying easier detection, than for nonosteoporotic vertebral fractures (83% [39 of 47]). The artificial intelligence model also demonstrated better detection of lumbar vertebral fractures compared with detection of thoracic vertebral fractures based on the external dataset using various radiographic techniques. Based on the dataset for external validation, the overall accuracy, sensitivity, and specificity on bootstrapping method were 89%, 83%, and 95%, respectively.

CONCLUSION: The artificial intelligence model detected vertebral fractures on plain lateral radiographs with high accuracy, sensitivity, and specificity, especially for osteoporotic lumbar vertebral fractures (Genant Grades 2 and 3). The rapid reporting of results using this artificial intelligence model may improve the efficiency of diagnosing vertebral fractures. The testing model is available at http://140.113.114.104/vght_demo/corr/. One or multiple plain lateral radiographs of the spine in the Digital Imaging and Communications in Medicine format can be uploaded to see the performance of the artificial intelligence model.

LEVEL OF EVIDENCE: Level II, diagnostic study.

PMID:33651768 | DOI:10.1097/CORR.0000000000001685

Shock. 2021 Feb 25. doi: 10.1097/SHK.0000000000001766. Online ahead of print.

ABSTRACT

BACKGROUND: The weekend effect is the increased mortality in hospitalized patients admitted on the weekend. The aim of this study was to examine the effect of weekend admissions on septic shock patients.

METHODS: This is a retrospective observational study of the 2014 Nationwide Emergency Department Sample Database (NEDS). Septic shock patients were included in this study using ICD-9-CM codes. Descriptive analysis was done, in addition to bivariate analysis to compare variables based on admission day. Multivariate analysis was conducted to examine the association between admission day and mortality in septic shock patients after adjusting for potential confounding factors.

RESULTS: A total of 364,604 septic shock patients were included in this study. The average age was 67.19 years, and 51.1% were males. 73.0% of patients presented on weekdays. 32.3% of septic shock patients died during their hospital stay. After adjusting for confounders, there was no significant difference in the ED or in-hospital mortality of septic shock patients admitted on the weekend compared to those admitted during weekdays, (OR = 1.00 [95%CI: 0.97 – 1.03], p-value = 0.985).

CONCLUSION: There was no statistically significant difference in overall mortality between septic shock patients admitted on the weekend or weekday. Our results are contradictory to previous studies showing an increased mortality with the weekend effect. The previous observations which have been made may not stand up with current treatment protocols.

PMID:33651724 | DOI:10.1097/SHK.0000000000001766

J Strength Cond Res. 2021 Mar 2. doi: 10.1519/JSC.0000000000004001. Online ahead of print.

ABSTRACT

Gannon, EA, Higham, DG, Gardner, BW, Nan, N, Zhao, J, and Bisson, LJ. Changes in neuromuscular status across a season of professional men’s ice hockey. J Strength Cond Res XX(X): 000-000, 2021-To quantify changes in neuromuscular function over a full professional men’s ice hockey season, 27 players (n = 18 forwards and 9 defensemen) performed 3 countermovement jumps (CMJ) each week over 30 sessions separated into 4 phases: preseason, early-season, midseason, and late-season. Outcome variables represented jump performance (jump height), kinematics (mean velocity and peak velocity), and movement strategy (countermovement depth). Mixed models characterized relationships between positional group, season phase, and CMJ outcomes. Statistical significance was set at p ≤ 0.05. Concentric peak velocity (p = 0.02), jump height (p = 0.001), and countermovement depth (p < 0.001) displayed a significant reduction across the season. Peak velocity was lower during the early-season than the preseason (-0.10 ± 0.06 m·s-1, mean change ± 95% confidence limit, p = 0.05). Countermovement depth was reduced during the early-season (-0.06 ± 0.03 m, p = 0.02), midseason (-0.10 ± 0.04 m, p = 0.002), and late-season (-0.15 ± 0.04 m, p < 0.001) relative to the preseason. Reductions in CMJ variables from preseason to in-season ranged from trivial to large. Changes in countermovement depth differed for forwards and defensemen by the season phase (p = 0.04). A professional ice hockey season decreases CMJ performance, with the effects of fatigue most prominent during the late-season phase. Countermovement depth was most sensitive to fatigue and differentiated positional-group responses. Frequent CMJ testing is useful for identifying the neuromuscular status of team-sport athletes relative to season-specific phases. Fatigue monitoring should incorporate movement-strategy variables alongside traditional measures of performance and kinematics.

PMID:33651739 | DOI:10.1519/JSC.0000000000004001

Menopause. 2021 Mar 1. doi: 10.1097/GME.0000000000001754. Online ahead of print.

ABSTRACT

OBJECTIVES: First, to establish the respective ability of body mass index (BMI), waist circumference (WC), and relative fat mass index (RFM), to estimate body fat (BF%) measured by DXA (DXA-BF%) and correctly identify postmenopausal women living with obesity (BF% > 35). Second, to identify the best indicator of successful weight-loss intervention in postmenopausal women living with obesity.

METHODS: A total of 277 women (age: 59.8 ± 5.3 y; BF%: 43.4 ± 5.3) from five weight-loss studies with complete data for anthropometric measurements [BMI = weight/height (kg/m2); WC (cm)] and BF% were pooled together. Statistical performance indicators were determined to assess ability of RFM [64-(20 × height/waist circumference) + (12 × sex)], BMI and WC to estimate BF% before and after weight-loss intervention and to correctly identify postmenopausal women living with obesity.

RESULTS: Compared with RFM (r = 0.51; r2 = 0.27; RMSE = 4.4%; Lin’s CCC = 0.46) and WC (r = 0.49; r2 = 0.25; RMSE = 4.8%; Lin’s CCC = 0.41), BMI (r = 0.73; r2 = 0.52; RMSE = 3.7%; Lin’s CCC = 0.71) was the best anthropometric index to estimate DXA-BF% and correctly identify postmenopausal women living with obesity (sensitivity + specificity: BMI = 193; RFM = 152; WC = 158), with lower misclassification error, before weight-loss intervention. After weight-loss, the change in BMI was strongly correlated with change in DXA-BF%, indicating that the BMI is the best indicator of success weight-loss intervention.

CONCLUSION: In the absence of more objective measures of adiposity, BMI is a suitable proxy measure for BF% in postmenopausal women, for whom a lifestyle intervention is relevant. Furthermore, BMI can be used as an indicator to assess success of weight-loss intervention in this subpopulation.

PMID:33651744 | DOI:10.1097/GME.0000000000001754

Am J Respir Crit Care Med. 2021 Mar 2. doi: 10.1164/rccm.202009-3481OC. Online ahead of print.

ABSTRACT

RATIONALE: Systemic sclerosis-pulmonary arterial hypertension (SSc-PAH) is one of the most prevalent and deadly forms of PAH. B cells may contribute to SSc pathogenesis.

OBJECTIVE: We investigated the safety and efficacy of B-cell depletion for SSc-PAH.

METHODS AND MEASUREMENTS: In an NIH-sponsored, multi-center, double-blinded, randomized, placebo-controlled, proof-of-concept trial, 57 SSc-PAH patients on stable-dose standard medical therapy received two infusions of 1000 mg of rituximab or placebo administered two weeks apart. The primary outcome measure was the change in six-minute walk distance (6MWD) at 24 weeks. Secondary endpoints included safety and invasive hemodynamics. We applied a machine learning approach to predict drug-responsiveness.

MAIN RESULTS: We randomized 57 subjects from 2010-2018. In the primary analysis, using data through week 24, the adjusted mean change in 6MWD at 24 weeks favored the treatment arm but did not reach statistical significance (23.6±11.1m vs. 0.5±9.7m, p=0.12). While a negative study, when data through week 48 were also considered, the estimated change in 6MWD at week 24 was 25.5±8.8m for rituximab and 0.4±7.4m for placebo (p=0.03). Rituximab treatment appeared to be safe and well tolerated. Low levels of rheumatoid factor (RF), IL-12, and IL-17 were sensitive and specific as favorable predictors of a rituximab response as measured by an improved 6MWD (ROC AUC 0.88-0.95).

CONCLUSIONS: B cell depletion therapy is a potentially effective and safe adjuvant treatment for SSc-PAH. Future studies in these patients can confirm whether the identified biomarkers predict rituximab-responsiveness. Clinical trial registration available at www.clinicaltrials.gov, ID: NCT01086540.

PMID:33651671 | DOI:10.1164/rccm.202009-3481OC

IEEE Trans Image Process. 2021 Mar 2;PP. doi: 10.1109/TIP.2021.3060909. Online ahead of print.

ABSTRACT

Existing part-aware person re-identification methods typically employ two separate steps: namely, body part detection and part-level feature extraction. However, part detection introduces an additional computational cost and is inherently challenging for low-quality images. Accordingly, in this work, we propose a simple framework named Batch Coherence-Driven Network (BCD-Net) that bypasses body part detection during both the training and testing phases while still learning semantically aligned part features. Our key observation is that the statistics in a batch of images are stable, and therefore that batch-level constraints are robust. First, we introduce a batch coherence-guided channel attention (BCCA) module that highlights the relevant channels for each respective part from the output of a deep backbone model. We investigate channel-part correspondence using a batch of training images, then impose a novel batch-level supervision signal that helps BCCA to identify part-relevant channels. Second, the mean position of a body part is robust and consequently coherent between batches throughout the training process. Accordingly, we introduce a pair of regularization terms based on the semantic consistency between batches. The first term regularizes the high responses of BCD-Net for each part on one batch in order to constrain it within a predefined area, while the second encourages the aggregate of BCD-Net’s responses for all parts covering the entire human body. The above constraints guide BCD-Net to learn diverse, complementary, and semantically aligned part-level features. Extensive experimental results demonstrate that BCD-Net consistently achieves state-of-the-art performance on four large-scale ReID benchmarks.

PMID:33651691 | DOI:10.1109/TIP.2021.3060909

J Med Internet Res. 2021 Feb 17. doi: 10.2196/24883. Online ahead of print.

ABSTRACT

BACKGROUND: Effective communication during a health crisis can ease public concerns and promote the adoption of important risk-mitigating behaviors. Public health agencies and leaders have served as the primary communicators of information related to COVID-19 and a key part of their public outreach has taken place on social media platforms.

OBJECTIVE: This study examined the content and engagement of COVID-19 Tweets authored by Canadian public health agencies and decision-makers. We propose ways that public health accounts can adjust their Tweeting practices during public health crises to improve risk communication and maximize engagement.

METHODS: We retrieved data from Tweets by Canadian public health agencies and decision-makers from January 1st, 2020 to June 30th, 2020. The Twitter accounts were categorized as either a public health agency, regional/local health department, provincial health authority, medical health officer, or minister of health. We analyzed trends in COVID-19 Tweet engagement and conducted a content analysis on a stratified random sample of 485 Tweets to examine the message functions and risk communication strategies used by each account type.

RESULTS: We analyzed 32,737 Tweets authored by 111 Canadian public health Twitter accounts, of which 6,982 Tweets were about COVID-19. Medical health officers authored the largest percentage of COVID-related Tweets (35%) relative to total Tweets, and averaged the highest number of retweets per COVID-19 Tweet (60 retweets per Tweet). Public health agencies had the highest frequency of daily Tweets about COVID-19 throughout the study period. Compared to Tweets containing media and user-mentions, hashtags and URLs were used in Tweets more frequently by all account types, appearing in 69% and 68% of COVID-related Tweets, respectively. Tweets containing hashtags also received the highest average retweets (47 retweets per Tweet). Our content analysis revealed that of the three Tweet message functions analyzed (information, action, community), Tweets providing information constituted the largest share (42%); however, Tweets promoting actions from users received higher average retweets (55 retweets per Tweet). When examining Tweets that received one or more retweet (n=359), the difference between mean retweets across the message functions was statistically significant (P<.001). Risk communication strategies that we examined were not widely used by any account type, only appearing in 262 out of 485 Tweets. However, when these strategies were used, these Tweets received more retweets compared to Tweets that did not use any risk communication strategies (P<.001) (61 retweets versus 13 retweets on average).

CONCLUSIONS: Public health agencies and decision-makers ought to examine what messaging best meets the needs of their Twitter audiences to maximize shares of their communications. Public health accounts that do not currently employ risk communication strategies in their Tweets may be missing an important opportunity to engage with information users about the mitigation of health risks related to COVID-19.

PMID:33651705 | DOI:10.2196/24883

J Epidemiol Community Health. 2021 Mar 1:jech-2020-214714. doi: 10.1136/jech-2020-214714. Online ahead of print.

ABSTRACT

BACKGROUND: Physical and cognitive functioning in older age follows a socioeconomic gradient but it is unclear whether the strength of the association differs between populations. Using harmonised data from an international collaboration of cohort studies, we assessed socioeconomic inequalities in physical and cognitive functioning and explored if the extent of inequalities varied across countries based on their economic strength or wealth distribution.

METHODS: Harmonised data from 37 population-based cohorts in 28 countries were used, with an overall sample size of 126 765. Socioeconomic position of participants was indicated by education and household income. Physical functioning was assessed by self-reported mobility and activities of daily living; and cognitive functioning by memory and verbal fluency tests. Relative (RII) and slope (SII) index of inequality were calculated in each cohort, and their association with the source country’s Gross Domestic Product (GDP) and Gini-index was assessed with correlation and cross-level interaction in multilevel models.

RESULTS: RII and SII values indicated consistently higher risk of low physical and cognitive functioning in participants with lower education or income across cohorts. Regarding RII, there were weak but statistically significant correlations and interactions with GDP and Gini-index, suggesting larger inequalities in countries with lower Gini-index and higher GDP. For SII, no such correlations were observed.

CONCLUSION: This study confirms that socioeconomic inequalities in physical and cognitive functioning exist across different social contexts but the magnitude of these inequalities varies. Relative inequalities appear to be larger in higher-income countries but it remains to be seen whether such observation can be replicated.

PMID:33649052 | DOI:10.1136/jech-2020-214714