Pandemic economics 4

Einleitung / Introduction

On July 14, 2021 heavy rainfalls hit Middle and Western Europe and caused a flood catastrophe in several areas. This study investigates, whether the severe floods following the rainfalls contributed to the spread of COVID-19 cases in the affected German areas.

Methode / Method

The analysis is based on a synthetic control approach in two steps. First, we contrast the average cumulative increase of COVID-19 cases for all German districts hit by the flood with a synthetic control constructed from all German districts not affected by the flood. Second, we apply an individual analysis by conducting a synthetic control design for different affected districts. In this way we can establish a closer relationship between the severity of being treated and the spread of COVID-19 in the affected regions. In both approaches, we attempt to control for differences in demographic, economic, health, and child care characteristics, as well as other potential confounders. We extend our analysis and check the robustness of our results by using different placebo specifications (e.g. placebo-int-time, placebo-in-space).

Ergebnisse / Results

Our preliminary results suggest a significant difference in the development of COVID-19 cases between the affected districts and their synthetic counterparts. Overall, between July 14 and August 1, we find that about 12 per 100,000 COVID-19 cases are not explained by district-level demographic, economic, health and child care characteristics. Additionally, the results seem to be related with the severity of the treatment. E.g. the severely damaged district of Ahrweiler reports a positive and significant difference of about 41 cases per 100,000 inhabitants on August 1.

Zusammenfassung / Conclusion

This paper quantifies the potential effect of severe floods in July 2021 on the spread of COVID-19 cases in the affected German areas. Given the preliminary results from our SCM approach, we cannot completely rule out that the flood disaster in Western Germany contributed to the spread of the coronavirus in the affected areas. Furthermore, the degree of severity seems to have had an influence. The interaction between those natural disasters and the spread of infectious diseases could be another potential channel for lasting negative effects as well as healths risks for the already affected population.

AutorInnen

Gerrit Stahn, Martin-Luther-Universität Halle-Wittenberg

Felix Krüger, Martin-Luther-Universität Halle-Wittenberg

Einleitung / Introduction

Since the declaration of COVID-19 as a pandemic of global concern by the World Health Organization on the 11th of March 2020, the pandemic experienced devastating impact on the world economies. The Middle East and North Africa (MENA) region is not an exception as most countries in this Region suffered from various negative effects on their economies. Actually the pandemic has affected every aspect of the economy in this Region including the agricultural, manufacturing and trade sectors, the financial and labour markets as well as the health and education sectors with serious implications on the welfare of the population especially the most vulnerable, such as informal sector, women and children. This might affect the progress of the MENA countries in achieving the SDGs unless rigorous measures are adopted in order to reverse the impact of the Pandemic on the Region. This paper aims at studying the socio economic impact of COVID-19 on the economy on both the macro and micro levels in selected MENA countries: Jordan, Egypt, Morocco and Saudi Arabia. The selected countries were chosen for a number of reasons. Jordan and Morocco are middle income countries and they represent the Mashrek (Eastern Mediterranean) and Maghreb countries (Western Mediterranean), while Egypt is considered to be a focal country and Saudi Arabia has been chosen as a representative of the Gulf rich countries. Furthermore, those countries are key trade partners for the EU. The socio-economic impact of the pandemic on the selected countries will be tackled through analyzing five main sectors, which are: Industrial, Trade, Tourism, Education, Health and Labour market.

Methode / Method

The paper will apply an analytical comparative approach to examine the case studies, where the socio-economic impact of the pandemic will be evaluated through analyzing key indicators of the previously mentioned sectors before and after the occurrence of the crisis exploring the reasons and implications of their observed trends and the policies adopted. To implement the previous analysis macro level data sources will be examined such as World Bank, Human Resource Development and OECD reports in addition to the National Economic Reports of the countries. For the micro level analysis the authors will conduct data analysis for the COVID-19 MENA Monitor Household survey of the Economic Research Forum. Furthermore, the OXFORD COVID-19 government response tracker will be used in order to explore the policies adopted by the governments of the selected countries to face the pandemic and to handle the after pandemic expected inflation.

AutorInnen

Heba Nassar, Faculty of Economics and Political Science, Cairo University

Pakinam Fikry, Faculty of Economics and Political Science, Cairo University

Einleitung / Introduction

As of November 2021, the Covid-19 pandemic caused by SARS-CoV-2 is still ongoing in most parts of the world. Community transmission triggered by asymptomatic and undetected SARS-CoV-2 carriers in general population presents a major challenge for the containment of this highly infectious disease. Active surveillance which tests everyone in general population and thus detect both symptomatic and asymptomatic SARS-COV-2 infected individuals present a promising strategy to monitor the community transmission of Covid-19. Given the high costs associated with mass testing, it remained unknown whether it is worthwhile to implement active surveillance for Covid-19. Using the novel test (isothermal RNA amplification-based method with saliva sample) and the innovative logistic solution for sample transportation based on postal services, the Virusfinder trial offered the unique opportunity to assess experimentally the economic costs and cost-effectiveness of the four different active surveillance strategies for Covid-19 in Germany. The four study strategies include (A1) direct testing of randomly drawn heterogenous individuals, (A2) direct testing of randomly drawn households of homogeneous individuals, (B1) testing conditioned on upstream COVID-19 symptom pre-screening of randomly drawn heterogenous individuals, and (B2) testing conditioned on upstream COVID-19 symptom pre-screening of randomly drawn households of homogenous individuals.

Methode / Method

We adopted health system perspective and conducted our costing and economic evaluation alongside the trial. Our costing study followed the activity-based micro-costing approach and estimated the economic costs for each strategy for two distinct phases (start-up and implementation) between 15 September 2020 until 30 December 2020. Data on resource consumption were collected both prospectively and retrospectively using digitalized individual database, daily time records, key informant interviews and direct observation. Our cost-effectiveness analysis evaluated the four primary outcomes estimated by the trial (contact reached, complete participant, sample tested, and case detected) and calculated the average cost-effective ratios (ACERs) of the four study strategies in relation to the status quo of having no active surveillance. We performed sensitivity analysis on two major drivers of the cost and cost-effectiveness estimates: response rate and prevalence.

Ergebnisse / Results

Both our base case analysis and the sensitivity analyses at different response rates (20%, 36,6%, 50%) and different prevalence (0,004, 0,001, 0,01, 0,04) consistently showed that strategy A2 had the lowest ACERs for two primary outcomes (cost per sample tested and cost per case detected at 56 EURO and 17.169 EURO respectively) which was closely followed by strategy A1 (cost per sample tested and cost per case detected at 67 EURO and 19.314 EURO). Strategies B1 and B2 have lower cost per contact reached (10 EURO and 17 EURO) but much higher cost per sample tested (256 EURO and 192 EURO) as well as per case detected (23.650 EURO and 28.665 EURO respectively).

Zusammenfassung / Conclusion

The strategy A2 appeared to the most cost-effective strategy among the four strategies tested in the Virusfinder trial. Future studies should assess the costs and cost-effectiveness of the strategy A2 in routine settings, in other geographical areas and adopt the broader societal perspective to better inform policy making.

AutorInnen

Hoa Thi Nguyen, Heidelberg Institute of Global Health and Division of Tropical Medicine, Heidelberg University Hospital

Manuela De Allegri, Heidelberg Institute of Global Health, Heidelberg University Hospital

Andreas Deckert, Heidelberg Institute of Global Health, Heidelberg University Hospital

Claudia Denkinger, Division of Tropical Medicine, Heidelberg University Hospital

Einleitung / Introduction

The COVID-19 pandemic has substantially disrupted health systems globally. In Europe, the first wave of the pandemic led to overwhelmed hospitals and clinics and placed a tremendous strain on health care professionals. These disruptions increase the risk that individuals miss essential care – either because they actively decide against seeking care or because their care is denied or cancelled by health professionals. Such missed visits can have strong negative health impacts and likely contribute to excess mortality that is not directly the result of COVID-19. Prediction-based approaches can help health administrators to identify individuals at risk of missing care and target retention efforts where they are most needed, which is especially important in the context of strained health systems.

Methode / Method

We use data from 52,830 participants spanning 27 countries from the Survey of Health, Aging and Retirement in Europe (SHARE). We use data on missed essential health care visits from the SHARE COVID-19 survey and sociodemographic and health data from SHARE waves 1-8. We compare four different machine learning algorithms to predict missed health care visits: Stepwise selection, group LASSO, Random Forest, and Neural Networks. We employ 5-fold cross-validation to test the prediction accuracy, sensitivity, and specificity of the selected models. For our models, we focus on individual characteristics that would be available to health facilities and health insurance providers and thus could be used in real-world settings to identify which individuals are at risk of missing essential care.

Ergebnisse / Results

Within our sample, 16% of the respondents reported any missed essential health care visit. Compared to the other algorithms, Random Forest outperformed all other models by far. The Random Forest model correctly identified 86% of all respondents who truly missed essential health care, while misclassifying less than 1% of those who did not miss essential health care. All other models identify only about one third of positive cases (stepwise selection and group LASSO: 30%, Neural Networks: 34%) while falsely targeting more individuals without missed essential health care (stepwise selection: 17%, group LASSO: 16%, Neural Networks: 21%).

Zusammenfassung / Conclusion

Pandemics such as COVID-19 require rapid and efficient responses to reduce disruptions in health care. Based on characteristics available to health insurance providers, machine learning algorithms can be used to efficiently target efforts to reduce missed essential care.

AutorInnen

Anna Reuter, Heidelberg Institute of Global Health, Universität Heidelberg

Nikkil Sudharsanan, TU München

Till Bärnighausen, Heidelberg Institute of Global Health, Universität Heidelberg