Creating symptom-based criteria for diagnostic testing: a case study based on a multivariate analysis of data collected during the first wave of the COVID-19 pandemic in New Zealand.

BACKGROUND: Diagnostic testing using PCR is a fundamental component of COVID-19 pandemic control. Criteria for determining who should be tested by PCR vary between countries, and ultimately depend on resource constraints and public health objectives. Decisions are often based on sets of symptoms in individuals presenting to health services, as well as demographic variables, such as age, and travel history. The objective of this study was to determine the sensitivity and specificity of sets of symptoms used for triaging individuals for confirmatory testing, with the aim of optimising public health decision making under different scenarios. METHODS: Data from the first wave of COVID-19 in New Zealand were analysed; comprising 1153 PCR-confirmed and 4750 symptomatic PCR negative individuals. Data were analysed using Multiple Correspondence Analysis (MCA), automated search algorithms, Bayesian Latent Class Analysis, Decision Tree Analysis and Random Forest (RF) machine learning. RESULTS: Clinical criteria used to guide who should be tested by PCR were based on a set of mostly respiratory symptoms: a new or worsening cough, sore throat, shortness of breath, coryza, anosmia, with or without fever. This set has relatively high sensitivity (> 90%) but low specificity (< 10%), using PCR as a quasi-gold standard. In contrast, a group of mostly non-respiratory symptoms, including weakness, muscle pain, joint pain, headache, anosmia and ageusia, explained more variance in the MCA and were associated with higher specificity, at the cost of reduced sensitivity. Using RF models, the incorporation of 15 common symptoms, age, sex and prioritised ethnicity provided algorithms that were both sensitive and specific (> 85% for both) for predicting PCR outcomes. CONCLUSIONS:  If predominantly respiratory symptoms are used for test-triaging,  a large proportion of the individuals being tested may not have COVID-19. This could overwhelm testing capacity and hinder attempts to trace and eliminate infection. Specificity can be increased using alternative rules based on sets of symptoms informed by multivariate analysis and automated search algorithms, albeit at the cost of sensitivity. Both sensitivity and specificity can be improved through machine learning algorithms, incorporating symptom and demographic data, and hence may provide an alternative approach to test-triaging that can be optimised according to prevailing conditions.

A longitudinal linkage study of occupation and ischaemic heart disease in the general and Maori populations of New Zealand.

OBJECTIVES: Occupation is a poorly characterised risk factor for cardiovascular disease (CVD) with females and indigenous populations under-represented in most research. This study assessed associations between occupation and ischaemic heart disease (IHD) in males and females of the general and Maori (indigenous people of NZ) populations of New Zealand (NZ). METHODS: Two surveys of the NZ adult population (NZ Workforce Survey (NZWS); 2004-2006; n = 3003) and of the Maori population (NZWS Maori; 2009-2010; n = 2107) with detailed occupational histories were linked with routinely collected health data and followed-up until December 2018. Cox regression was used to calculate hazard ratios (HR) for IHD and "ever-worked" in any of the nine major occupational groups or 17 industries. Analyses were controlled for age, deprivation and smoking, and stratified by sex and survey. RESULTS: 'Plant/machine operators and assemblers' and 'elementary occupations' were positively associated with IHD in female Maori (HR 2.2, 95%CI 1.2-4.1 and HR 2.0, 1.1-3.8, respectively) and among NZWS males who had been employed as 'plant/machine operators and assemblers' for 10+ years (HR 1.7, 1.2-2.8). Working in the 'manufacturing' industry was also associated with IHD in NZWS females (HR 1.9, 1.1-3.7), whilst inverse associations were observed for 'technicians and associate professionals' (HR 0.5, 0.3-0.8) in NZWS males. For 'clerks', a positive association was found for NZWS males (HR 1.8, 1.2-2.7), whilst an inverse association was observed for Maori females (HR 0.4, 0.2-0.8). CONCLUSION: Associations with IHD differed significantly across occupational groups and were not consistent across males and females or for Maori and the general population, even within the same occupational groups, suggesting that current knowledge regarding the association between occupation and IHD may not be generalisable across different population groups.

Ischaemic Heart Disease and Occupational Exposures: A Longitudinal Linkage Study in the General and Maori Populations of New Zealand.

OBJECTIVES: This study assessed associations between occupational exposures and ischaemic heart disease (IHD) for males and females in the general and Maori populations (indigenous people of New Zealand). METHODS: Two surveys of the general adult [New Zealand Workforce Survey (NZWS); 2004-2006; n = 3003] and Maori population (Maori NZWS; 2009-2010; n = 2107), with information on occupational exposures, were linked with administrative health data and followed-up until December 2018. Cox proportional hazards regression (adjusted for age, deprivation, and smoking) was used to assess associations between organizational factors, stress, and dust, chemical and physical exposures, and IHD. RESULTS: Dust [hazard ratio (HR) 1.6, 95%CI 1.1-2.4], smoke or fumes (HR 1.5, 1.0-2.3), and oils and solvents (HR 1.5, 1.0-2.3) were associated with IHD in NZWS males. A high frequency of awkward or tiring hand positions was associated with IHD in both males and females of the NZWS (HRs 1.8, 1.1-2.8 and 2.4, 1.1-5.0, respectively). Repetitive tasks and working at very high speed were associated with IHD among NZWS females (HRs 3.4, 1.1-10.4 and 2.6, 1.2-5.5, respectively). Maori NZWS females working with vibrating tools and those exposed to a high frequency of loud noise were more likely to experience IHD (HRs 2.3, 1.1-4.8 and 2.1, 1.0-4.4, respectively). Exposure to multiple dust and chemical factors was associated with IHD in the NZWS males, as was exposure to multiple physical factors in males and females of the NZWS. CONCLUSIONS: Exposures associated with an elevated IHD risk included dust, smoke or fumes, oils and solvents, awkward grip or hand movements, carrying out repetitive tasks, working at very high speed, loud noise, and working with tools that vibrate. Results were not consistently observed for males and females and between the general and Maori populations.

A survey of current practice in reporting third trimester fetal biometry and Doppler in Australia and New Zealand.

INTRODUCTION: Inconsistent reporting practices in third trimester ultrasound, the choice of reference charts in particular, have the potential to misdiagnose abnormal fetal growth. But this may lead to unnecessary anxiety and confusion amongst patients and clinicians and ultimately influence clinical management. Therefore, we sought to determine the extent of variability in choice of fetal biometry and Doppler reference charts and reporting practices in Australia and New Zealand. METHODS: Clinicians performing and/or reporting obstetric ultrasound were invited to answer questions about fetal biometry and Doppler charts in a web-based survey. RESULTS: At least four population-based charts are in current use. The majority of respondents (78%) report the percentile for known gestational age (GA) alongside measurements and 63% using a cut-off of estimated fetal weight (EFW) < 10th percentile when reporting small for gestational age (SGA) and/or fetal growth restriction (FGR). The thresholds for the use of fetal and maternal Doppler in third trimester ultrasound varied in terms of the GA, EFW cut-off, and how measures were reported. The majority of respondents were not sure of which Doppler charts were used in their practice. CONCLUSION: This survey revealed inconsistencies in choice of reference chart and reporting practices. The potential for misdiagnosis of abnormal fetal growth remains a significant issue.

The Prevalence of Cardiovascular Risk Factors in Different Occupational Groups in New Zealand.

OBJECTIVES: Although cardiovascular disease (CVD) risk has been shown to differ between occupations, few studies have specifically evaluated the distribution of known CVD risk factors across occupational groups. This study assessed CVD risk factors in a range of occupational groups in New Zealand, stratified by sex and ethnicity. METHODS: Two probability-based sample surveys of the general New Zealand adult population (2004-2006; n = 3003) and of the indigenous people of New Zealand (Maori; 2009-2010; n = 2107), for which occupational histories and lifestyle factors were collected, were linked with routinely collected health data. Smoking, body mass index, deprivation, diabetes, high blood pressure, and high cholesterol were dichotomized and compared between occupational groups using age-adjusted logistic regression. RESULTS: The prevalence of all known CVD risk factors was greater in the Maori survey than the general population survey, and in males compared with females. In general for men and women in both surveys 'Plant and machine operators and assemblers' and 'Elementary workers' were more likely to experience traditional CVD risk factors, while 'Professionals' were less likely to experience these risk factors. 'Clerks' were more likely to have high blood pressure and male 'Agricultural and fishery workers' in the general survey were less likely to have high cholesterol, but this was not observed in the Maori survey. Male Maori 'Trades workers' were less likely to have high cholesterol and were less obese, while for the general population survey, this was not observed. CONCLUSIONS: This study showed differences in the distribution of known CVD risk factors across occupational groups, as well as between ethnic groups and males and females.

A comparison of sensitivity-specificity imputation, direct imputation and fully Bayesian analysis to adjust for exposure misclassification when validation data are unavailable.

Purpose: Measurement error is an important source of bias in epidemiological studies. We illustrate three approaches to sensitivity analysis for the effect of measurement error: imputation of the 'true' exposure based on specifying the sensitivity and specificity of the measured exposure (SS); direct imputation (DI) using a regression model for the predictive values; and adjustment based on a fully Bayesian analysis. Methods: We deliberately misclassify smoking status in data from a case-control study of lung cancer. We then implement the SS and DI methods using fixed-parameter (FBA) and probabilistic (PBA) bias analyses, and Bayesian analysis using the Markov-Chain Monte-Carlo program WinBUGS to show how well each recovers the original association. Results: The 'true' smoking-lung cancer odds ratio (OR), adjusted for sex in the original dataset, was OR = 8.18 [95% confidence limits (CL): 5.86, 11.43]; after misclassification, it decreased to OR = 3.08 (nominal 95% CL: 2.40, 3.96). The adjusted point estimates from all three approaches were always closer to the 'true' OR than the OR estimated from the unadjusted misclassified smoking data, and the adjusted interval estimates were always wider than the unadjusted interval estimate. When imputed misclassification parameters departed much from the actual misclassification, the 'true' OR was often omitted in the FBA intervals whereas it was always included in the PBA and Bayesian intervals. Conclusions: These results illustrate how PBA and Bayesian analyses can be used to better account for uncertainty and bias due to measurement error.

Compositional data in neuroscience: If you've got it, log it!

BACKGROUND: Compositional data sum to a constant value, for example, 100%. In neuroscience, such data are common, for example, when estimating the percentage of time spent for a behavioural response in a limited choice situation or a neurochemical within brain tissue. Compositional data have a distinct structure which complicates analysis and makes inappropriate standard statistical analyses such as general linear model analyses and principal components or factor analysis (whether Q-mode or R-mode), as a result of the correlation of the components, the dependence of the pairwise covariance on which other components are included in the analysis, and the bounded nature of the data. NEW METHOD: This problem has been recognised in disciplines such as geology and zoology for decades, where log ratio methods have been successfully applied. The isometric log ratio (ilr) method has some particular advantages. COMPARISON WITH EXISTING METHOD: Classical statistical methods such as t-tests, ANOVAs, and multivariate analyses are invalid when applied to compositional data. CONCLUSIONS: The compositional data analysis methods developed by statisticians and used by geologists and zoologists should be considered for compositional data analysis in neuroscience.

A multivariate statistical and data mining analysis of spatial memory-related behaviour following bilateral vestibular loss in the rat.

Vestibular dysfunction in animals and humans is associated with a variety of cognitive and anxiety disorders, and it has been difficult to determine how the different symptoms may be related to one another. The aim of this study was to determine the extent to which the spatial memory deficits that occur following bilateral vestibular deafferentation (BVD) in rats can be attributed to other behavioural symptoms. Spatial memory was measured using a spatial T maze alternation task (STM), while locomotor activity and anxiety were measured using open field, elevated plus and T mazes, respectively. Using multiple linear and random forest regression, we determined that the best predictors of performance in the STM were whether the animals had received a BVD or sham lesion, and the duration of rearing. Using linear discriminant analysis, random forest classification, support vector machines and cluster analysis, we found that BVD animals could be clearly distinguished from sham controls by their behavioural syndrome, in particular their decreased duration of rearing in the open field maze (suggesting reduced exploration), decreased time spent in the outer zone of the open field maze ('reduced thigmotaxis', suggesting increased risk taking), and spatial memory deficits in the STM. These results suggest that the poor performance of rats with BVD in spatial memory tasks is largely due to spatial memory deficits themselves rather than a result of other changes in locomotor activity or anxiety.