This data visualization (dataviz) infographics presents six plots based on COVID-19 data from hospitalized patients during the pandemic Spanish first wave. Each graphic explores different statistical concepts.
This is a part of the GRBIO Divulga initiative.
This Balloon plot is a contingency table for the frequency of the different diagnoses and the status, individually and jointly. Each balloon reflects the relative magnitude of the corresponding component. The larger and, therefore, the purplish is the balloon the higher is the frequency. For instance, the most common status was discharge when is combined with a breathing difficulty diagnosis. The smaller and, therefore, the yellowish is the balloon the lower is the frequency. For example, the less common frequency is the combination of coughing and a transfer status.
Among the total number of COVID-19 hospitalizations per day during March-Abril 2020 what was the proportion of women?
The Coxcomb or Rose diagram was designed by Florence Nightingale. This plot is a variation of a pie chart highlighting the evolution of COVID19 hospitalizations during March and April of 2020. The area of a section represents the value of the corresponding category. This current plot allows us to visualize: i) the total number of hospitalizations per day and ii) the proportion of hospitalized women.
A barplot shows the relation between a numeric and a categoric variable. Each category of the categoric variable is represented as a bar and the size of the bar represents the numeric value that it takes. In this case, the numeric variable is the median of the length of ICU stay and the categoric variable is the age of the patient. On this barplot we can see that there is no relation between the median of the length of ICU stay and the age of the patient.
What is the relationship between patient characteristics, such as oxygen saturation, temperature and days in the ICU?
The Correlation plot represents associations between factors through ellipses. An ellipse close to a circumference indicates lack of relationship, while ellipses more elongated indicate a direct relationship if they grow to the right or an inverse relationship otherwise. In this plot two measurements of O2 Sats at different times were highly correlated (direct relationship). In turn, these saturations were moderately inversely related to the days in ICU (thehigher the saturation, the fewer days in the ICU)
Calendar plot represents an indicator over time using a color scale associating more intense colors with larger magnitudes. On March 14th, 2020, the lockdown began in Spain. The plot represents the number of hospitalizations in HM hospitals by fortnight, weekday and time slot. There was a rebound in hospitalizations in the first fortnight post-confinement that was reversed in the subsequent fortnight. The time slots with the highest number of hospitalizations were located between noon and 10pm. and there were slightly fewer hospitalizations on weekends. In the worst scenario, more than 25 hospitalizations were recorded every 2 hours, that is, one every 5 minutes.
Sankey Diagram visualizes the number of individuals that move from one state to another in a fixed period of time. The width of the flows is proportional to the number of individuals that pass through it. In this diagram, and over a 2-month period during the first wave of the COVID-19 pandemic, there was a higher percentage of patients who were discharged among those who did not enter the ICU (78%, 1538 out of 1964) than among those who did enter (38%, 78 out of 203).
