Non-Communicable Disease Analytics
Monitor the health and wellness of your workforce
Non-communicable diseases (NCDs) such as cardiovascular diseases, cancer, diabetes and chronic respiratory diseases are responsible for 70% per cent of all deaths globally, according to the World Health Organisation (WHO). Reducing NCDs requires a precise understanding of the prevalence and incidence of risk factors and disease outcomes. Next, interventions must be monitored and evaluated to determine their effectiveness in reducing NCDs and their return on investment (ROI).
NCDs have a complex aetiology, with modifiable and non-modifiable risk factors. This has important implications for designing surveys and interventions with quantifiable impact. We offer expert advise on a wide range of study designs, from cross-sectional surveys, to quasi-experimental and randomised controlled studies. Systematic literature reviews, combined with statistical analysis and simulation studies can help you understand which of the modifiable risk factors have the largest impact and which interventions and target groups you should prioritise.
Assessment of risks and impact of NCDs for companies
improvement of productivity through wellness interventions
predicting the benEfits of lifestyle modification
CAUSAL inference of interventions that reduce NCD risk
Health-economic analysis and return on investment
A case study
CAN biomarkers predict the risk of failed liver transplantation?
Wimmy used its interdisciplinary expertise to support a research grant application by the Ghent University Hospital (Belgium) to the Research Foundation – Flanders. The hospital research team sought to validate the prognostic value of a biomarker test that predicts graft failure after liver transplantation.
The ideal biomarker accurately differentiates between patients with a high and low risk of experiencing graft loss at three months after liver transplantation. Wimmy’s team of data scientists and medical specialists were able to combine their expertise to produce bespoke statistical and machine learning models to meet the client’s specific requirements. In the course of the project, various state of the art techniques were implemented, including conditional tabular generative adversarial networks (CTGAN), latent class analysis, segmented regression analysis and supervised principal component analysis. The latter method was used to predict the probability of graft failure, using repeated measurements of biomarkers in the blood of transplant patients in the first 2 weeks after liver transplantation. In this way, we were able to classify patients into high, medium or low risk groups for graft failure. Finally, we calculated the model’s accuracy and the sample sizes that would be required to validate this new prognostic tool, under different power assumptions.