Dying from COVID-19 or with COVID-19: a definitive answer through a retrospective analysis of mortality in Italy

Mortality Trends ◽

Male Mortality ◽

Definitive Answer ◽

Background: COVID-19 mortality was associated with several reasons, including conspiracy theories and infodemic phenomena. However, little is known about the potential endogenous reasons for the increase in COVID-19 associated mortality in Italy. Objective: This study aimed to search the potential endogenous reasons for the increase in COVID-19 mortality recorded in Italy during the year 2020 and evaluate the statistical significance of the latter. Methods: We analyzed all the trends in the timelapse 2011-2019 related to deaths by age, sex, region, and cause of death in Italy and compared them with those of 2020. Ordinary least squares (OLS) linear regressions and ARIMA (p, d, q) models were applied to investigate the predictions of death in the year 2020 as compared to death reported in 2020. Grubbs and Iglewicz-Hoaglin tests were used to identify the statistical differences between the predictors and observed death during the year 2020. The relationship between mortality and predictive variables was assessed using OLS multiple regression models. Results: Both ARIMA and OLS linear regression models predicted the number of deaths in Italy during the year 2020 is between 640,000 and 660,000 (95% confidence intervals range: 620,000 - 695,000) and these values were far from the observed deaths reported (n = 750,000). Significant difference in deaths at national level (P = 0.003), and higher male mortality than women (+18% versus +14%, P < 0.001 versus P = 0.01) was observed. Finally, higher mortality was strongly and positively correlated with latitude (R = 0.82, P < 0.001) Conclusions: Our findings suggest that the absence of historical endogenous reasons capable of justifying the increase in deaths and mortality observed in Italy in 2020. Together with the current knowledge on the novel coronavirus 2019, these findings provide decisive evidence on the devastating impact of COVID-19 in Italy. We suggest that this research be leveraged by government, health, and information authorities to furnish proof against conspiracy theorists. Moreover, given the marked concordance between the predictions of the ARIMA and OLS regression models, we suggest that these models be exploited to predict mortality trends.

Dying From COVID-19 or With COVID-19: A Definitive Answer Through a Retrospective Analysis of Mortality in Italy (Preprint)

10.2196/preprints.36022 ◽

2021 ◽

Author(s):

Alessandro Rovetta ◽

Akshaya Srikanth Bhagavathula

Keyword(s):

Regression Models ◽

Current Knowledge ◽

Statistical Significance ◽

National Level ◽

Ordinary Least Squares ◽

Mortality Trends ◽

Male Mortality ◽

Definitive Answer ◽

BACKGROUND COVID-19 mortality was associated with several reasons, including conspiracy theories and infodemic phenomena. However, little is known about the potential endogenous reasons for the increase in COVID-19 associated mortality in Italy. OBJECTIVE This study aimed to search the potential endogenous reasons for the increase in COVID-19 mortality recorded in Italy during the year 2020 and evaluate the statistical significance of the latter. METHODS We analyzed all the trends in the timelapse 2011-2019 related to deaths by age, sex, region, and cause of death in Italy and compared them with those of 2020. Ordinary least squares (OLS) linear regressions and ARIMA (p, d, q) models were applied to investigate the predictions of death in 2020 as compared to death reported in the same year. Grubbs and Iglewicz-Hoaglin tests were used to identify the statistical differences between the predicted and observed deaths. The relationship between mortality and predictive variables was assessed using OLS multiple regression models. RESULTS Both ARIMA and OLS linear regression models predicted the number of deaths in Italy during 2020 to be between 640,000 and 660,000 (95% confidence intervals range: 620,000 – 695,000) and these values were far from the observed deaths reported (above 750,000). Significant difference in deaths at national level (P = 0.003), and higher male mortality than women (+18% versus +14%, P < 0.001 versus P = 0.01) was observed. Finally, higher mortality was strongly and positively correlated with latitude (R = 0.82, P < 0.001). CONCLUSIONS Our findings support the absence of historical endogenous reasons capable of justifying the increase in deaths and mortality observed in Italy in 2020. Together with the current knowledge on the novel coronavirus 2019, these findings provide decisive evidence on the devastating impact of COVID-19 in Italy. We suggest that this research be leveraged by government, health, and information authorities to furnish proof against conspiracy hypotheses. Moreover, given the marked concordance between the predictions of the ARIMA and OLS regression models, we suggest that these models be exploited to predict mortality trends.

Linear regression models for biomass table construction, using cluster samples

Canadian Journal of Forest Research ◽

10.1139/x89-103 ◽

1989 ◽

Vol 19 (5) ◽

pp. 664-673 ◽

Cited By ~ 1

Author(s):

Andrew J. R. Gillespie ◽

Tiberius Cunia

Keyword(s):

Least Squares ◽

Regression Models ◽

Predictor Variable ◽

Ordinary Least Squares ◽

Cluster Sampling ◽

Parameter Estimates ◽

Least Squares Regression ◽

Estimation Procedures ◽

Biased Estimates

Biomass tables are often constructed from cluster samples by means of ordinary least squares regression estimation procedures. These procedures assume that sample observations are uncorrelated, which ignores the intracluster correlation of cluster samples and results in underestimates of the model error. We tested alternative estimation procedures by simulation under a variety of cluster sampling methods, to determine combinations of sampling and estimation procedures that yield accurate parameter estimates and reliable estimates of error. Modified, generalized, and jack-knife least squares procedures gave accurate parameter and error estimates when sample trees were selected with equal probability. Regression models that did not include height as a predictor variable yielded biased parameter estimates when sample trees were selected with probability proportional to tree size. Models that included height did not yield biased estimates. There was no discernible gain in precision associated with sampling with probability proportional to size. Random coefficient regressions generally gave biased point estimates with poor precision, regardless of sampling method.

Properties of the ordinary least squares and stein-rule predictions in linear regression models with proxy variables

Statistical Papers ◽

10.1007/bf02925525 ◽

1993 ◽

Vol 34 (1) ◽

pp. 27-41 ◽

Cited By ~ 1

Author(s):

V. K. Srivastava ◽

M. Dube

Keyword(s):

Linear Regression ◽

Least Squares ◽

Regression Models ◽

Ordinary Least Squares ◽

Proxy Variables

A Comparative Analysis on Some Estimators of Parameters of Linear Regression Models in Presence of Multicollinearity

Asian Journal of Probability and Statistics ◽

10.9734/ajpas/2018/v2i228773 ◽

2018 ◽

pp. 1-8

Author(s):

Warha, Abdulhamid Audu ◽

Yusuf Abbakar Muhammad ◽

Akeyede, Imam

Keyword(s):

Linear Regression ◽

Least Squares ◽

Regression Models ◽

Mean Squared Error ◽

Least Squares Method ◽

Ordinary Least Squares ◽

Least Square ◽

Independent Variables ◽

Different Levels

Linear regression is the measure of relationship between two or more variables known as dependent and independent variables. Classical least squares method for estimating regression models consist of minimising the sum of the squared residuals. Among the assumptions of Ordinary least squares method (OLS) is that there is no correlations (multicollinearity) between the independent variables. Violation of this assumptions arises most often in regression analysis and can lead to inefficiency of the least square method. This study, therefore, determined the efficient estimator between Least Absolute Deviation (LAD) and Weighted Least Square (WLS) in multiple linear regression models at different levels of multicollinearity in the explanatory variables. Simulation techniques were conducted using R Statistical software, to investigate the performance of the two estimators under violation of assumptions of lack of multicollinearity. Their performances were compared at different sample sizes. Finite properties of estimators’ criteria namely, mean absolute error, absolute bias and mean squared error were used for comparing the methods. The best estimator was selected based on minimum value of these criteria at a specified level of multicollinearity and sample size. The results showed that, LAD was the best at different levels of multicollinearity and was recommended as alternative to OLS under this condition. The performances of the two estimators decreased when the levels of multicollinearity was increased.

Association of body mass index with morbidity in patients hospitalised with COVID-19

BMJ Open Respiratory Research ◽

10.1136/bmjresp-2021-000970 ◽

2021 ◽

Vol 8 (1) ◽

pp. e000970

Author(s):

Maria Plataki ◽

Di Pan ◽

Parag Goyal ◽

Katherine Hoffman ◽

Jacky Man Kwan Choi ◽

...

Keyword(s):

Body Mass Index ◽

New York ◽

Body Mass ◽

Regression Models ◽

Hospital Length ◽

Hospital Length Of Stay ◽

Discharge Disposition ◽

Home Discharge ◽

PurposeTo evaluate the association between body mass index (BMI) and clinical outcomes other than death in patients hospitalised and intubated with COVID-19.MethodsThis is a single-centre cohort study of adults with COVID-19 admitted to New York Presbyterian Hospital-Weill Cornell Medicine from 3 March 2020 through 15 May 2020. Baseline and outcome variables, as well as lab and ventilatory parameters, were generated for the admitted and intubated cohorts after stratifying by BMI category. Linear regression models were used for continuous, and logistic regression models were used for categorical outcomes.ResultsThe study included 1337 admitted patients with a subset of 407 intubated patients. Among admitted patients, hospital length of stay (LOS) and home discharge was not significantly different across BMI categories independent of demographic characteristics and comorbidities. In the intubated cohort, there was no difference in in-hospital events and treatments, including renal replacement therapy, neuromuscular blockade and prone positioning. Ventilatory ratio was higher with increasing BMI on days 1, 3 and 7. There was no significant difference in ventilator free days (VFD) at 28 or 60 days, need for tracheostomy, hospital LOS, and discharge disposition based on BMI in the intubated cohort after adjustment.ConclusionsIn our COVID-19 population, there was no association between obesity and morbidity outcomes, such as hospital LOS, home discharge or VFD. Further research is needed to clarify the mechanisms underlying the reported effects of BMI on outcomes, which may be population dependent.

Quantifying the effects of temporal autocorrelation on climatological regression models using geostatistical techniques

Canadian Journal of Forest Research ◽

10.1139/x26-094 ◽

1996 ◽

Vol 26 (5) ◽

pp. 864-871 ◽

Cited By ~ 1

Author(s):

Ian B. Strachan ◽

L. Edward Harvey

Keyword(s):

Time Series ◽

Stomatal Conductance ◽

Regression Models ◽

Temporal Structure ◽

Time Lag ◽

Ordinary Least Squares ◽

Data Sets ◽

Temporal Autocorrelation ◽

Geostatistical Techniques

When time-dependent data are used in regression models, temporal autocorrelation violates ordinary least squares assumptions and impedes their proper testing and interpretation. The problem of temporal autocorrelation is exacerbated by the uneven temporal spacing inherent in many data sets. Using simple linear regression models of stomatal conductance as examples, we compare the effectiveness of two methods for removing temporal autocorrelation from regression models (first-differencing and Cochrane–Orcutt) and we introduce the geostatistical technique of semivariograms as a method for quantifying temporal autocorrelation in uneven time series. The Cochrane–Orcutt method proved more effective than first-differencing at removing autocorrelation and produced regression models without changing the significance of the independent variables. Semivariograms were used to quantify the time dependence of the unevenly spaced stomatal conductance time series. This technique revealed the dominant autocorrelation at the minimum time lag (0.5 h) and the 24-h periodicity caused by the climatological variables used in the model. We conclude that geostatistical techniques provide a robust method for quantifying temporal structure and periodicity in unevenly spaced time series.

Comparison of initial and recertification QOPI scores: Maintaining compliance through direct surveyor and practice interaction.

Journal of Clinical Oncology ◽

10.1200/jco.2017.35.8_suppl.246 ◽

2017 ◽

Vol 35 (8_suppl) ◽

pp. 246-246

Author(s):

Anne C. Chiang ◽

Constance Barysauskas ◽

Tara Conti-Kalchik ◽

Terry Gilmore ◽

Carolyn Bennett Hendricks ◽

...

Keyword(s):

Direct Observation ◽

Regression Models ◽

Statistical Significance ◽

Performance Status ◽

Sustainable Improvement ◽

Chemotherapy Administration ◽

Wilcoxon Rank Sum Test ◽

Site Evaluation ◽

And Performance

246 Background: Since 2006, the ASCO QOPI certification program has certified 322 practices, of which 197 practices have recertified. This retrospective study compares the number and type of standards passed at the time of initial and subsequent re-certification, and examines if on-site audits at the original certification influenced re-certification scores. Methods: 87 unique US practices that obtained QOPI certification with on-site audits at the original and re-certification between 2006 and 2014 were included. 17 QOPI certification standards were included in the analysis. Standards are metric based, except 3 standards that are observable. We defined total score per certification round as the total number of standards passed and used a Wilcoxon Rank Sum Test to test the concordance of standards passed between rounds. Linear regression models were used to identify factors related to higher recertification scores. A two-sided p<0.05 defined statistical significance. Results: 31 practices (36%) showed concordance of the 3 observable standards in the initial and re-certification rounds. For standards that assess policies, procedure and credentials of the practice, and do not require direct observation, 52 practices demonstrated improvement whereas 14 did not (p<0.0001). In contrast, for the standards that require direct observation, only 22 practices showed improvement versus 34 that did not (p=0.07). Three standards were most commonly missed in both rounds: initial chart documentation and at each clinical visit, and double verification of chemotherapy administration. There were no significant predictors of higher recertification scores. Conclusions: Many diverse oncology practices are voluntarily achieving and maintaining QOPI certification. Sustainable improvement is easier to identify in policy-based measures compared to directly observed ones. Three standards not usually passed at recertification highlight the need for assessment of psychosocial and performance status, comprehension of treatment goals, and chemotherapy double-verification. On-site evaluation of practices is key for targeting and sustaining quality efforts.

Mapping individuals' earthquake preparedness in China

Natural Hazards and Earth System Science ◽

10.5194/nhess-18-1315-2018 ◽

2018 ◽

Vol 18 (5) ◽

pp. 1315-1325 ◽

Cited By ~ 13

Author(s):

Guochun Wu ◽

Ziqiang Han ◽

Weijin Xu ◽

Yue Gong

Keyword(s):

Disaster Preparedness ◽

Regression Models ◽

Current Knowledge ◽

Mainland China ◽

National Sample ◽

Public Affairs ◽

Western China ◽

Earthquake Risk ◽

Earthquake Preparedness

Abstract. Disaster preparedness is critical for reducing potential impact. This paper contributes to current knowledge of disaster preparedness using representative national sample data from China, which faces high earthquake risks in many areas of the country. The adoption of earthquake preparedness activities by the general public, including five indicators of material preparedness and five indicators of awareness preparedness, were surveyed and 3245 respondents from all 31 provinces of Mainland China participated in the survey. Linear regression models and logit regression models were used to analyze the effects of potential influencing factors. Overall, the preparedness levels are not satisfied, with a material preparation score of 3.02 (1–5), and awareness preparation score of 2.79 (1–5), nationally. Meanwhile, residents from western China, which has higher earthquake risk, have higher degrees of preparedness. The concern for disaster risk reduction (DRR) and the concern for building safety and participation in public affairs are consistent positive predictors of both material and awareness preparedness. The demographic and socioeconomic variables' effects, such as gender, age, education, income, urban/rural division, and building size, vary according to different preparedness activities. Finally, the paper concludes with a discussion of the theoretical contribution and potential implementation.

In depth analysis of Cyprus-specific mutations of SARS-CoV-2 strains using computational approaches

BMC Genomic Data ◽

10.1186/s12863-021-01007-9 ◽

2021 ◽

Vol 22 (1) ◽

Author(s):

Anastasis Oulas ◽

Jan Richter ◽

Maria Zanti ◽

Marios Tomazou ◽

Kyriaki Michailidou ◽

...

Keyword(s):

Structure Prediction ◽

Regression Models ◽

Response Times ◽

National Level ◽

Nucleotide Position ◽

Single Mutation ◽

Computational Approaches ◽

Viral Virulence ◽

Depth Analysis

Abstract Background This study aims to characterize SARS-CoV-2 mutations which are primarily prevalent in the Cypriot population. Moreover, using computational approaches, we assess whether these mutations are associated with changes in viral virulence. Methods We utilize genetic data from 144 sequences of SARS-CoV-2 strains from the Cypriot population obtained between March 2020 and January 2021, as well as all data available from GISAID. We combine this with countries’ regional information, such as deaths and cases per million, as well as COVID-19-related public health austerity measure response times. Initial indications of selective advantage of Cyprus-specific mutations are obtained by mutation tracking analysis. This entails calculating specific mutation frequencies within the Cypriot population and comparing these with their prevalence world-wide throughout the course of the pandemic. We further make use of linear regression models to extrapolate additional information that may be missed through standard statistical analysis. Results We report a single mutation found in the ORF1ab gene (nucleotide position 18,440) that appears to be significantly enriched within the Cypriot population. The amino acid change is denoted as S6059F, which maps to the SARS-CoV-2 NSP14 protein. We further analyse this mutation using regression models to investigate possible associations with increased deaths and cases per million. Moreover, protein structure prediction tools show that the mutation infers a conformational change to the protein that significantly alters its structure when compared to the reference protein. Conclusions Investigating Cyprus-specific mutations for SARS-CoV-2 can lead to a better understanding of viral pathogenicity. Researching these mutations can generate potential links between viral-specific mutations and the unique genomics of the Cypriot population. This can not only lead to important findings from which to battle the pandemic on a national level, but also provide insights into viral virulence worldwide.

Association Between CLOCK 3111 T/C Polymorphism with Ghrelin, GLP-1, Food Timing, Sleep and Chronotype in Overweight and Obese Iranian Adults

10.21203/rs.3.rs-1092653/v1 ◽

2021 ◽

Author(s):

Sara Rahati ◽

Mostafa Qorbani ◽

Anoosh Naghavi ◽

Milad Heidari Nia ◽

Hamideh Pishva

Keyword(s):

Linear Regression ◽

Dietary Intake ◽

Sleep Duration ◽

Regression Models ◽

Essential Element ◽

Allele Carrier ◽

Fat Intake ◽

Wide Range ◽

Abstract Background Circadian Locomotor Output Cycles Kaput (CLOCK), an essential element of the positive regulatory arm in the human biological clock, is involved in metabolic regulation. The aim was to investigate the behavioral (sleep duration, food timing, dietary intake, appetite and chronobiologic characteristics) and hormonal (plasma ghrelin and Glucagon-like peptide-1 concentrations) factors that could explain the previously reported association between the CLOCK 3111T/C SNP and obesity. Methods This cross-sectional study included 403 subjects, overweight and/or obesity, aged 20- 50 years from Iran. The CLOCK rs1801260 data were measured by the PCR-RFLP method. Dietary intake, food timing, sleep duration, appetite and Chrono-type were assessed using validated questionnaires. Ghrelin and GLP-1 were measured by radioimmunoassay in plasma samples. Participants were also divided into three groups based on rs1801260 genotype and BMI. Logistic regression models and general linear regression models were used to assess the association between CLOCK genotype and study parameters. Univariate linear regression models were used to assess the interaction between CLOCK and VAS, Food timing, chronotype and sleep on food intakes. Results After controlling for confounding factors, there was a significant difference between genotypes for physical activity (P=0.001), waist circumference (P˂0.05), BMI (˂0.01), weight (P=0.001), GLP-1 (P= 0.02), ghrelin (P= 0.04), appetite (P˂0.001), chronotype (P˂0.001), sleep (P˂0.001), food timing (P˂0.001), energy (P˂0.05), carbohydrate (P˂0.05) and fat intake (P˂0.001). Our findings also show that people with the minor allele C who ate lunch after 3 PM and breakfast after 9 AM are more prone to obesity (P˂0.05). furthermore, there was significant interactions between C allele carrier group and high appetite on fat intake (Pinteraction=0.041), eat lunch after 3 PM on energy intake (Pinteraction=0.039) and morning type on fat intake (Pinteraction=0.021). Conclusion Sleep reduction, changes in ghrelin and GLP-1 levels, changes in eating behaviors and evening preference that characterized CLOCK 3111C can all contribute to obesity. Furthermore, the data demonstrate a clear relationship between the timing of food intake and obesity. Our results support the hypothesis that the influence of the CLOCK gene may extend to a wide range of variables related to human behaviors.