According to the publication by Silverio et al, the challenges of implementing big data in cardiovascular care were mentioned, including missing data, selection bias, and data analysis and training. I proposed my suggestions to cope with those challenges.
Missing data
As the publication points out, there are various statistical techniques for handling missing data. The goal is to avoid creating biased estimations resulting in inaccurate outcomes. From my understanding, two main methods for handling missing data are data removal and imputation. The imputation technique replaces missing data with reasonable predictions. When the percentage of missing data is low, it is most helpful. If the portion of missing data is too high, the results lack natural variation that could result in an effective model. The other option is to remove data. This technique is a good option when dealing with data that is missing at random.
Selection bias
Selection bias occurs when the sample data doesn’t accurately reflect the larger population causing inaccuracies in conclusions and decision-making. Scrutinizing the sample’s representativeness can uncover the selection bias by comparing the characteristics of the sample with those of the larger population it’s intended to represent. Another method is to examine the data collection process. To cope with this challenge, randomized selection methods, stratified sampling, or oversampling of underrepresented groups can be utilized.
Data analysis and training
Data analysts have made it possible for businesses to make informed decisions by turning raw data into practical insights. I proposed that the organization should have both data analysts and data scientists who work together in the data-driven decision-making process.