Predicting missed health care visits during the COVID-19 pandemic using machine learning methods: evidence from 55,500 individuals from 28 European countries.

BACKGROUND: Pandemics such as the COVID-19 pandemic and other severe health care disruptions endanger individuals to miss essential care. Machine learning models that predict which patients are at greatest risk of missing care visits can help health administrators prioritize retentions efforts towards patients with the most need. Such approaches may be especially useful for efficiently targeting interventions for health systems overburdened during states of emergency. METHODS: We use data on missed health care visits from over 55,500 respondents of the Survey of Health, Ageing and Retirement in Europe (SHARE) COVID-19 surveys (June - August 2020 and June - August 2021) with longitudinal data from waves 1-8 (April 2004 - March 2020). We compare the performance of four machine learning algorithms (stepwise selection, lasso, random forest, and neural networks) to predict missed health care visits during the first COVID-19 survey based on common patient characteristics available to most health care providers. We test the prediction accuracy, sensitivity, and specificity of the selected models for the first COVID-19 survey by employing 5-fold cross-validation, and test the out-of-sample performance of the models by applying them to the data from the second COVID-19 survey. RESULTS: Within our sample, 15.5% of the respondents reported any missed essential health care visit due to the COVID-19 pandemic. All four machine learning methods perform similarly in their predictive power. All models have an area under the curve (AUC) of around 0.61, outperforming random prediction. This performance is sustained for data from the second COVID-19 wave one year later, with an AUC of 0.59 for men and 0.61 for women. When classifying all men (women) with a predicted risk of 0.135 (0.170) or higher as being at risk of missing care, the neural network model correctly identifies 59% (58%) of the individuals with missed care visits, and 57% (58%) of the individuals without missed care visits. As the sensitivity and specificity of the models are strongly related to the risk threshold used to classify individuals, the models can be calibrated depending on users' resource constraints and targeting approach. CONCLUSIONS: Pandemics such as COVID-19 require rapid and efficient responses to reduce disruptions in health care. Based on characteristics available to health administrators or insurance providers, simple machine learning algorithms can be used to efficiently target efforts to reduce missed essential care.

Remote Healthcare During the COVID-19 Pandemic: Findings for Older Adults in 27 European Countries and Israel.

The COVID-19 pandemic exacerbated issues regarding access to healthcare for older people, by far the most vulnerable population group. In particular, older adults avoided seeking medical treatment for fear of infection or had their medical treatments postponed or denied by health facilities or health professionals. In response, remote medical services were recognized as an essential adjustment mechanism to maintain the continuity of healthcare provision. Using the SHARE Corona Survey data, we estimate logistic and multilevel regression models for the remote care of 44,152 persons aged 50 and over in 27 European countries and Israel. Our findings suggest that those aged 80+ were the least likely to use remote healthcare. However, women, better educated individuals, older adults who lived in urban areas, those with no financial strain, and active Internet users used remote medical consultations more often. Those who reported poor or fair health status, two or more chronic diseases, or hospitalization in the last 12 months were significantly more likely to use remote healthcare. Furthermore, remote medical consultations were more frequent for those who had their healthcare postponed or went without it due to fear of coronavirus infection. Finally, older adults used remote care more frequently in countries with less healthcare coverage and lower health expenditures. Health systems should prioritize vulnerable groups in maintaining continuity in access to healthcare, despite the availability of remote care. Policymakers should improve telemedicine regulation and offer incentives for providers of remote healthcare services by adapting reimbursement policies. Remote medical care could play an important role in maintaining healthcare access for older adults and increasing health systems' preparedness in future health emergencies.

Access to healthcare for people aged 50+ in Europe during the COVID-19 outbreak.

This paper combines SHARE Corona Survey and SHARE Wave 7 data for 25 European countries and Israel (N = 40,919) with institutional and epidemic-related country characteristics to investigate healthcare access for Europeans aged 50+ during the outbreak of COVID-19. We use a micro-macro approach to examine whether and to what extent barriers to accessing healthcare measured by reported unmet healthcare needs vary within and between countries. We consider various aspects of barriers and distinguish among: (1) respondents who forewent medical treatment because they were afraid of becoming infected with the Coronavirus; (2) respondents who had pre-scheduled medical appointments postponed by health providers due to the outbreak; and (3) respondents who tried to arrange a medical appointment but were denied one. Limited access to healthcare during the initial outbreak was more common for the occupationally active, women, the more educated and those living in urban areas. A bad economic situation, poor overall health and higher healthcare utilisation were robust predictors of unmet healthcare. People aged 50+ in countries of 'Old' Europe, countries with higher universal health coverage and stricter containment and closure policies were more likely to have medical services postponed. Policymakers should address the healthcare needs of older people with chronic health conditions and a poor socio-economic status who were made more vulnerable by this pandemic. In the aftermath of the health crisis, public health systems might experience a great revival in healthcare demand, a challenge that should be mitigated by careful planning and provision of healthcare services.

Living Alone in the United States and Europe: The Impact of Public Support on the Independence of Older Adults.

While we know that living alone is often associated with greater risk of financial hardship, we have limited knowledge on the possible link between the availability of public support and independent living. We use data from the 2014 Health and Retirement Study and the 2011-2015 Survey of Health, Ageing and Retirement in Europe to compare income and wealth profiles of the population aged 60 and above who are living alone in the United States and 19 European countries. We find that the likelihood of living alone is higher in generous welfare states, with social support and spending both positively associated with living alone. The relationship between personal resources and living alone has a smaller positive gradient in countries with robust welfare systems. The lack of adequate public support in less generous welfare states may constrain the ability of many low-income older adults without a partner to continue living independently.

Employment status and other predictors of mental health and cognitive functions in older Croatian workers.

The aim of this study was to examine mental health and cognitive functions in older Croatian workers (50-65 years) taking into account their employment status, self-assessed health, and a set of demographic characteristics. We analysed the data collected on 650 older workers (71 % employed) in the Wave 6 of the Survey of Health, Ageing and Retirement in Europe (SHARE). Unemployed workers reported symptoms of loneliness more often than the employed, while in rural areas unemployment was additionally associated with more pronounced symptoms of depression. Feeling of loneliness was also higher in those living without a partner in the household and in those with poorer health. In urban residents symptoms of depression were more severe in women, respondents with higher education, those living without a partner, and those who rated their health as poorer. As for cognitive functions, unemployment significantly predicted poorer subtraction in the rural subsample. Women in general showed less efficient numerical abilities. In the urban subsample poorer numerical abilities were also associated with lower education and living without a partner in the household. Better verbal recall was predicted by higher education and better self-rated memory. Higher scores in verbal fluency were predicted by urban residency and better self-rated health. Our results indicate that the protective factors for good mental health and cognitive functioning in older Croatian workers are being employed, having more education, living with a partner in the household, and being healthier. These findings stress the importance of implementing broader social policy strategies covering employment, education, and health.