Analysis of pre-stroke modified Rankin Scale

Distributions of mRS (modified Rankin Scale) are shown for:

• All patients

• Ischaemic and haemorrhagic stroke

• Ischaemic patients with NIHSS of 0-10 (surrogate for non large vessel occlusions)

• Ischaemic patients with NIHSS of 11+ (surrogate for large vessel occlusions)

# import libraries
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

# Change default colour scheme:
plt.style.use('seaborn-colorblind')

# import data
data = pd.read_csv(
    './../data/2019-11-04-HQIP303-Exeter_MA.csv', low_memory=False)

Get distribution of mRS before stroke for all patients

count, division = np.histogram(data['S2RankinBeforeStroke'], bins=range(0,7))

results = pd.DataFrame()
results.index.name='mRS'
results['all'] = count / count.sum()
results

	all
mRS
0	0.533043
1	0.156302
2	0.107781
3	0.121422
4	0.063581
5	0.017870

Get distribution of mRS before stroke by stroke type

# Infarction
mask = data['S2StrokeType'] == 'Infarction'
count, division = np.histogram(
    data[mask]['S2RankinBeforeStroke'], bins=range(0,7))
results['infarction'] = count / count.sum()

# Haemorrhage
mask = data['S2StrokeType'] == 'Primary Intracerebral Haemorrhage'
count, division = np.histogram(
    data[mask]['S2RankinBeforeStroke'], bins=range(0,7))
results['haemorrhage'] = count / count.sum()

# Show results
results

	all	infarction	haemorrhage
mRS
0	0.533043	0.534923	0.521374
1	0.156302	0.157958	0.144905
2	0.107781	0.108075	0.106007
3	0.121422	0.119199	0.137272
4	0.063581	0.062649	0.068735
5	0.017870	0.017196	0.021706

mRS for ischaemic stroke, by stroke severity

# Filter data
mask = data['S2StrokeType'] == 'Infarction'
data = data[mask]

# NIHSS 0-10 is a surrogate for non large vessel occlusions
mask = data['S2NihssArrival'] < 11 
count, division = np.histogram(
    data[mask]['S2RankinBeforeStroke'], bins=range(0,7))
results['nlvo'] = count / count.sum()

# NIHSS 11+ is a surrogate for non large vessel occlusions
mask = data['S2NihssArrival'] > 10
count, division = np.histogram(
    data[mask]['S2RankinBeforeStroke'], bins=range(0,7))
results['lvo'] = count / count.sum()

# Show results
results

	all	infarction	haemorrhage	nlvo	lvo
mRS
0	0.533043	0.534923	0.521374	0.582771	0.407796
1	0.156302	0.157958	0.144905	0.163513	0.143538
2	0.107781	0.108075	0.106007	0.103988	0.120133
3	0.121422	0.119199	0.137272	0.101109	0.166050
4	0.063581	0.062649	0.068735	0.041796	0.118023
5	0.017870	0.017196	0.021706	0.006822	0.044460

fig = plt.figure(figsize=(5,5))
ax = fig.add_subplot()
width = 0.17
x = results.index
ax.bar(x - width * 2, results['all'], width, label='all')
ax.bar(x, results['haemorrhage'], width, label='haemorrhage')
ax.bar(x + width, results['nlvo'], width, label='nLVO (NIHSS 0-10')
ax.bar(x + width * 2, results['lvo'], width, label='LVO (NIHSS 11+)')
ax.set_ylabel('Fraction of patients')
ax.set_xlabel('mRS before stroke')
ax.legend()
plt.savefig('./output/mrs_before_by_stroke_type.jpg', dpi=300)
plt.show()

fig = plt.figure(figsize=(5,5))
ax = fig.add_subplot()
width = 0.17
x = results.index
#ax.plot(results['all'], label='all')
ax.plot(results['haemorrhage'], marker='o', label='haemorrhage')
ax.plot(results['nlvo'], marker = 's', label='nLVO (NIHSS 0-10')
ax.plot(results['lvo'], marker = '^', label='LVO (NIHSS 11+)')
ax.grid()
ax.set_ylabel('Fraction of patients')
ax.set_xlabel('mRS before stroke')
ax.legend()
plt.savefig('./output/mrs_before_by_stroke_type.jpg', dpi=300)
plt.show()