Tables Contingency tables

Contingency tables & analyses

2x2 tables are commonly used to assess risk in epidemiology. The rows represent a risk factor, like exposure to a disease, or sex. The columns represent an outcome, like infection status, or whether the disease was severe or mild.

Create a 2x2 table

Both rows and columns must be binary. Note that you must tell epipy how you wish the table to be organized by providing a list of values.

import epipy
import pandas as pd

mers_df = epipy.get_data('mers_line_list')
table = epipy.create_2x2(mers_df, row='Sex', 'Health status',
                ['M', 'F'], ['Dead', 'Alive'])

table returns:

     Dead  Alive  All
M      46     54  101
F      16     44   60
All    70    114  185

Analyze a 2x2 table

2x2 tables are used to calculate odds ratios, relative risk, and chi square tests.

epipy.analyze_2x2(table)

returns:

Odds ratio: 0.57 (95% CI: (0.22, 1.46))
Relative risk: 0.69 (95% CI: (0.38, 1.26))

Attributable risk: -0.126 (95% CI: (-0.34, 0.08))
Attributable risk percent: -44.86% (95% CI: (-43.19, -46.517))
Population attributable risk: -0.089
Population attributable risk percent: -27.84%

Chi square: 2.91278015543
p value: 0.57252592908

Alternatively, you can call each function separately:

epipy.odds_ratio(table)
epipy.relative_risk(table)
epipy.chi2(table)

Back to documentation