Abstract:Adequate forecasts of future population developments that are based on cohortcomponent methods demand an age-and sex-specific analysis; otherwise, the structure of the future population cannot be specified correctly. Age-specific demographic measures are both highly correlated and highly dimensional. Thus, a methodology that not only considers the correlations between the random variables but also reduces the effective dimensionality of the forecasting problem is needed: principal component analysis serves bot… Show more
“…Figure 1 shows the loadings of the first PC (PC1). The loadings can be interpreted as correlations between the PCs and original variables (Vanella, 2018 ) and in this case, the logit-WASCSMRs. Fig.…”
Section: Methodsmentioning
confidence: 99%
“…They simply serve as correction factors to systematic under/over-estimation of the Fourier series, which systematically underestimates mortality in winter and overestimates mortality in summer. The residuals from the quantified model are fitted by using a seasonal autoregressive integrated moving average (SARIMA) model, which is chosen by using a series of tests following Vanella ( 2018 ). Figure 3 illustrates the fit of model 3 (continuous line) to the data (dots).…”
The COVID-19 outbreak has called for renewed attention to the need for sound statistical analyses to monitor mortality patterns and trends over time. Excess mortality has been suggested as the most appropriate indicator to measure the overall burden of the pandemic in terms of mortality. As such, excess mortality has received considerable interest since the outbreak of COVID-19 began.Previous approaches to estimate excess mortality are somewhat limited, as they do not include sufficiently long-term trends, correlations among different demographic and geographic groups, or autocorrelations in the mortality time series. This might lead to biased estimates of excess mortality, as random mortality fluctuations may be misinterpreted as excess mortality.We propose a novel approach that overcomes the named limitations and draws a more realistic picture of excess mortality. Our approach is based on an established forecasting model that is used in demography, namely, the Lee-Carter model. We illustrate our approach by using the weekly age- and sex-specific mortality data for 19 countries and the current COVID-19 pandemic as a case study. Our findings show evidence of considerable excess mortality during 2020 in Europe, which affects different countries, age, and sex groups heterogeneously. Our proposed model can be applied to future pandemics as well as to monitor excess mortality from specific causes of death.
“…Figure 1 shows the loadings of the first PC (PC1). The loadings can be interpreted as correlations between the PCs and original variables (Vanella, 2018 ) and in this case, the logit-WASCSMRs. Fig.…”
Section: Methodsmentioning
confidence: 99%
“…They simply serve as correction factors to systematic under/over-estimation of the Fourier series, which systematically underestimates mortality in winter and overestimates mortality in summer. The residuals from the quantified model are fitted by using a seasonal autoregressive integrated moving average (SARIMA) model, which is chosen by using a series of tests following Vanella ( 2018 ). Figure 3 illustrates the fit of model 3 (continuous line) to the data (dots).…”
The COVID-19 outbreak has called for renewed attention to the need for sound statistical analyses to monitor mortality patterns and trends over time. Excess mortality has been suggested as the most appropriate indicator to measure the overall burden of the pandemic in terms of mortality. As such, excess mortality has received considerable interest since the outbreak of COVID-19 began.Previous approaches to estimate excess mortality are somewhat limited, as they do not include sufficiently long-term trends, correlations among different demographic and geographic groups, or autocorrelations in the mortality time series. This might lead to biased estimates of excess mortality, as random mortality fluctuations may be misinterpreted as excess mortality.We propose a novel approach that overcomes the named limitations and draws a more realistic picture of excess mortality. Our approach is based on an established forecasting model that is used in demography, namely, the Lee-Carter model. We illustrate our approach by using the weekly age- and sex-specific mortality data for 19 countries and the current COVID-19 pandemic as a case study. Our findings show evidence of considerable excess mortality during 2020 in Europe, which affects different countries, age, and sex groups heterogeneously. Our proposed model can be applied to future pandemics as well as to monitor excess mortality from specific causes of death.
“…Our first aim is to identify the role of the age structure of the cases in the overall CFRs, and to derive age-specific and age-standardised CFR estimates for all study countries. Based on the cumulative age-specific case numbers Nijk of the 11 European countries, we calculate age-specific CFR estimates (Vanella, 2018). Working with PCs allows us to cover indirectly simultaneous trends and sensitivities of the case numbers to the testing strategy.…”
Section: Step I: Adjustment To Demographicsmentioning
Across European countries, there have been large differences in COVID-19 case fatality risk (CFR) estimates, and considerable variation in these estimates over time. CFR estimates vary depending on both the method used for estimation and countryspecific characteristics. While crude methods simply use cumulative total numbers of cases and deaths, the CFR can be influenced by the demographic characteristics of the cases, the case detection rates, the time lags between the reporting of infections and deaths and infrastructure characteristics, such as healthcare capacities. We use publicly available weekly data for 11 European countries on the COVID- 19 case and death numbers by age group for the year 2020. Moreover, we use data on national weekly test rates to adjust the case numbers, and to investigate the effects of different time lags between the reporting of cases and deaths on the estimation of CFRs. Finally, we describe the association between case fatality rates and the demand for hospital and intensive care unit beds for COVID-19 cases, while taking into account national bed capacities. The crude CFR estimates differ considerably across the investigated countries. In the crude international CFR time series, the differences are smaller when adjusting for the demographics of the cases. Differences in testing policies significantly affect the CFR estimates as well. However, the question of precisely how these testing procedures should be adjusted requires further investigation. Lag adjustments of CFRs do not lead to improvements in estimates of COVID-19 CFRs, and no connection between hospital capacities and CFRs can be found for the countries included in our study.
“…The principal components are derived iteratively by singular value decomposition. E.g., [32] gives a mathematical introduction to PCA; more applied recourses, among others, can be found in [33] or [34], the latter with special emphasis on forecasting in demography. PCA is the method of choice for dealing with high dimensional data and cross-correlations among the time series.…”
Section: Conceptual Approachesmentioning
confidence: 99%
“…Modern approaches apply this method to migration forecasting as well [17,35]. Nevertheless, it is essential not to blindly trust the forecast results but to also check their qualitative plausibility [34].…”
Population projections serve various actors at subnational, national, and international levels as a quantitative basis for political and economic decision-making. Usually, the users are no experts in statistics or forecasting and therefore lack the methodological and demographic background to completely understand methods and limitations behind the projections they use to inform further analysis. Our contribution primarily targets that readership. Therefore, we give a brief overview of different approaches to population projection and discuss their respective advantages and disadvantages, alongside practical problems in population data and forecasting. Fundamental differences between deterministic and stochastic approaches are discussed, with special emphasis on the advantages of stochastic approaches. Next to selected projection data available to the public, we show central areas of application of population projections, with an emphasis on Germany.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
