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.

Global economic burden of unmet surgical need for appendicitis.

BACKGROUND: There is a substantial gap in provision of adequate surgical care in many low- and middle-income countries. This study aimed to identify the economic burden of unmet surgical need for the common condition of appendicitis. METHODS: Data on the incidence of appendicitis from 170 countries and two different approaches were used to estimate numbers of patients who do not receive surgery: as a fixed proportion of the total unmet surgical need per country (approach 1); and based on country income status (approach 2). Indirect costs with current levels of access and local quality, and those if quality were at the standards of high-income countries, were estimated. A human capital approach was applied, focusing on the economic burden resulting from premature death and absenteeism. RESULTS: Excess mortality was 4185 per 100 000 cases of appendicitis using approach 1 and 3448 per 100 000 using approach 2. The economic burden of continuing current levels of access and local quality was US $92 492 million using approach 1 and $73 141 million using approach 2. The economic burden of not providing surgical care to the standards of high-income countries was $95 004 million using approach 1 and $75 666 million using approach 2. The largest share of these costs resulted from premature death (97.7 per cent) and lack of access (97.0 per cent) in contrast to lack of quality. CONCLUSION: For a comparatively non-complex emergency condition such as appendicitis, increasing access to care should be prioritized. Although improving quality of care should not be neglected, increasing provision of care at current standards could reduce societal costs substantially.

The effect of eligibility for antiretroviral therapy on body mass index and blood pressure in KwaZulu-Natal, South Africa.

We use a regression discontinuity design to estimate the causal effect of antiretroviral therapy (ART) eligibility according to national treatment guidelines of South Africa on two risk factors for cardiovascular disease, body mass index (BMI) and blood pressure. We combine survey data collected in 2010 in KwaZulu-Natal, South Africa, with clinical data on ART. We find that early ART eligibility significantly reduces systolic and diastolic blood pressure. We do not find any significant effects on BMI. The effect on blood pressure can be detected up to three years after becoming eligible for ART.

Combined RT-qPCR and pyrosequencing of a Spike glycoprotein polybasic cleavage motif can uncover pediatric SARS-CoV-2 infections associated with heterogeneous presentation.

BACKGROUND: Reverse transcription of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (+)RNA genome and subgenomic RNAs (sgRNAs) and subsequent quantitative polymerase chain reaction (RT-qPCR) is the reliable diagnostic gold standard for COVID-19 diagnosis and the identification of potential spreaders. Apart from clinical relevance and containment, for specific questions, it might be of interest to (re)investigate cases with low SARS-CoV-2 load, where RT-qPCR alone can deliver conflicting results, even though these cases might neither be clinically relevant nor significant for containment measures, because they might probably not be infectious. In order to expand the diagnostic bandwidth for non-routine questions, particularly for the reliable discrimination between negative and false-negative specimens associated with high CT values, we combined the RT-qPCR workflow with subsequent pyrosequencing of a S-gene amplicon. This expansion can help to confirm SARS-CoV-2 infections without the demand of confirmative antibody testing, which requires to summon patients again for blood sampling few to several weeks after symptom onset. RESULTS: We successfully established a combined RT-qPCR and S-gene pyrosequencing method which can be optionally exploited after routine diagnostics. This allows a reliable interpretation of RT-qPCR results in specimens with relatively low viral loads and close to the detection limits of qPCR. After laboratory implementation, we tested the combined method in a large pediatric cohort from two German medical centers (n=769). Pyrosequencing after RT-qPCR enabled us to uncover 5 previously unrecognized cases of pediatric SARS-CoV-2-associated diseases, mainly exhibiting mild and heterogeneous presentation-apart from a single case of multisystem inflammatory syndrome in children (MIS-C) associated with SARS-CoV-2, who was hospitalized in the course of the study. CONCLUSIONS: The proposed protocol allows a specific and sensitive confirmation of SARS-CoV-2 infections close to the detection limits of RT-qPCR. The tested biotinylated primers do not negatively affect the RT-qPCR pipeline and thus can be optionally applied to enable deeper inspection of RT-qPCR results by subsequent pyrosequencing. Moreover, due to the incremental transmission of SARS-CoV-2 variants of concern, we note that the used strategy can uncover (Spike) P681H allowing the pre-selection of SARS-CoV-2 B.1.1.7 candidate specimens for deep sequencing.

Mental distress and its association with sociodemographic and economic characteristics: community-based household survey in Aceh, Indonesia.

BACKGROUND: The role of sociodemographic and economic characteristics in mental distress has been rarely investigated in Indonesia. AIMS: To investigate the prevalence of common mental disorders (CMD) and identify any associations between mental distress and sociodemographic and economic characteristics among communities living in urban and rural (peri-urban) areas. METHOD: A community-based household survey was conducted in the province of Aceh, Indonesia, in 2018. The 20-item Self Reporting Questionnaire (SRQ-20) screening tool was used to measure symptoms of CMD. Information on sociodemographic characteristics, family functioning, labour market outcomes and healthcare costs was collected. Multivariate regressions were conducted to analyse the relationships between the measures of mental distress and sociodemographic and economic characteristics. RESULTS: We found that 14% of the respondents had CMD symptoms. SRQ-20 scores were higher for female, older and lower-educated individuals. CMD prevalence was higher among non-married participants and clustered within families. Participants with CMD perceive their families as performing significantly better in the dimensions of affective involvement and behaviour control compared with their counterparts. Their work was more often affected by negative feelings; they were also twice as likely to report a recent physical or mental health complaint and faced twice the treatment costs compared with their non-affected counterparts. CONCLUSIONS: The prevalence of mental disorders is especially high in disadvantaged population groups. Moreover, mental distress is associated with a lower perceived productivity and a higher physical health burden.