Skip to main content

Table 1 Summarized characteristics of included studies

From: Are big data analytics helpful in caring for multimorbid patients in general practice? - A scoping review

Authors

Year

Country

Aims

No. of used observations

Method of data analysis

Outcome

Andriopoulou F, et al. [19]

2013

Greece

Managing patients suffering from chronic conditions.

30

Random Forest

Framework that identifies the necessity to deliver personalized health services by specialists when they are most appropriate.

Schäfer I, et al. [20]

2014

Germany

Depicting which diseases are associated with each other on person-level in multimorbid patients and which ones are responsible for the overlapping of multimorbidity clusters.

98.619 (72.548 for replication analyses)

Analysis based on clustering techniques

Model for the association of diseases to each other. Identification of diseases that form a multimorbidity cluster as well as the identification of diseases responsible for overlapping multimorbidity clusters.

Marx P, et al. [21]

2015

Hungary

Investigating a systems-based approach for the use of separated large-scale multimorbidity data to explore common latent factors of related diseases.

117.803 (subset of the UK Biobank)

MCMC on a Bayesian network

Bayesian, multivariate, system-based approach to identify shared latent factors that could cause multi-morbid diseases without interpreting these factors.

Boshuizen HC, et al. [22]

2017

Netherlands

Determining the magnitude of the difference in the burden of a risk factor with different calculation methods.

Not defined. Study based on the Global Burden of disease database.

Temporal counterfactual reasoning

Dynamic modelling with the DYNAMO-HIA Method estimates the gain in Disability Adjusted Life Years (DALYs) obtained by eliminating exposure to a risk factor more accurately than other established methods.

Kalgotra P, et al. [23]

2017

USA

Addressing the co-occurrences of diseases using network analysis while putting a special focus in disparities by gender.

22.1 million

Network analysis

Identification of different multimorbidity clusters for male and female patients with a prevalence of higher comorbidities in females than males.

Nicholson K, et al. [24]

2017

Canada

Development of the Multimorbidity Cluster Analysis Tool and Toolkit to identify distinct clusters within patients living with multimorbidity.

75.000

Analysis based on clustering techniques

Downloadable Toolkit for analysis and description of combination and permutation of diseases in large datasets of multimorbid patients.