Variola

• Orthopox virus
• Smallpox, la petite vérole, isorokko, smittkoppor
• One of the deadliest diseases in history
  • 30 % mortality
  • 300 million dead in 20th century
  • Killed 20–30 % of Finnish children in the worst periods
• Wiped out by vaccination campaign in 1980

Vaccination

• Jennerian vaccination in the late 1700s
• Mandatory vaccination for children in Finland in 1883
• Smallpox no longer endemic in Finland after 1890s

However…

• Widespred vaccine hesitancy
• Kuhnean antivaccine movement popular in the 1910s
• Smallpox still endemic in Russia proper

WWI

• Before the revolutions of 1917 WWI affected Finland very little
• Several diseases spread near frontlines
  • German army suffered 1 dead from disease for every 10 KIA
• Influenza pandemic of 1918–19 killed 500 million people world wide

Finnish Civil War

• Peacetime medical infrastructure breaks down
• Reds had much worse medical corps than Whites
  • Did not trust doctors
• Large movements of people
  • Armies
  • Refugees
  • Economic migrants due to famine conditions
  • Prisoner transfers
• People crammed in barracks and in prison camps

Smallpox in Civil War

• Local smallpox cases in Karelian Isthmus in 1916
• Several outbreaks during and after the civil war
• Smallpox spread among Red troops and then in prison camps
  • 3000 cases and 900 deaths

Civilian cases

• Recruits and prisoners returning spread cases
• In many areas civilians did not recognise the disease as smallpox
• Origin sometimes unknown
  • Doctors suspected Variola had spread from Russia
• One large wave at the end of the Civil War
• Second, more dispersed wave in early 1919

Prevention of smallpox

• Quarantining was the most important way of prevention
• Increased vaccination, targeted vaccination efforts
  • Similar to later WHO eradication campaign!
  • White troops and red prisoners vaccinated
• Desinfections of barracks

Eastern Border

• Refugees from Russia often had smallpox infections
• Mandatory checks at the border, quarantine hospital

Healthcare districts

• Finland was divided into 53 medical districts piirilääkäripiiri
• The area of districts evolved over time
• Each had a district medical officer (piirilääkäri) charged with reporting on local conditions
  • Medical statistics do not correspond to other statistical units
• Cities had their own medical boards

Medical data

• Annual reports of the Central Medical Board (Lääkintöhallitus) report diseases by medical districts and cities
• These do not include the many cases in garrisons and prison camps
• No good data on vaccine coverage

Spatial data

• Finnish municipalities in 1918 digitized, then combined into medical districts
• Cities are included in the district that surrounds them, in reality they were independent units

Distance to Russia and Variola

• x_i distance to Russian border (100 km) measured for each medical district i
• y_i incidence of Variola

Bayesian spatial Poisson regression

• Incidence is modelled with a type of Generalised Linear Model (GLM), where the response variable is Poisson-distributed
• Bayesian regression:
  • Distribution of model parameters are estimated based on data
  • Data is a given, model parameters vary
  • Prior distributions incorporate prior knowledge

Model (1/2)

\[ y_i \sim Poisson(\lambda_i) \] \[ \log{\lambda_i} = \beta_0 + \beta_1 x_i + \theta_i + \phi_i \] \[ \beta_0 \sim Normal(\mu=4, \sigma=0.05) \] \[ \beta_1 \sim Normal(\mu=0, \sigma=0.05) \]

Model (2/2)

\[ \theta_i \sim Normal(\mu=0, \tau=\tau_{independent}) \] \[ \phi | \phi_{j~i} \sim Normal(\mu=\alpha \sum^{n_i}_{j=1} \phi_j, \tau=\tau_{spatial}) \] \[ \tau_{independent} \sim Gamma(\alpha=1, \beta=1) \] \[ \tau_{spatial} \sim Gamma(\alpha=1, \beta=1) \]

Results

• 95 % credible interval (CI) for β_1 indicates significant effect of distance to Russia (x_i)
  • 1918 CI: from -0.0064 to -0.00078
  • 1919 CI: from -0.014 to -0.0028
  • very small numbers are explained by the fact x_i is measured in hundreds of km

1918 vs 1919

• Two distinct peaks in reported civilian smallpox cases
• End of the Civil War, May 1918
• Winter and Spring 1919

Local patterns

1. Two peaks
2. Peak in 1918, no peak in 1919
3. No peak in 1918, peak in 1919
4. Very few or no cases

1. Two peaks

• Karelian Isthmus and Ladoga coast
• Areas most dependent on Petrograd before 1917
• Tampere region
  • The city itself had high vaccination rate
• Kuopio: first wave from prisoners, second wave from North Karelia

2. Peak in 1918, no peak in 1919

• Historic Tavastia and Birkaland, except Tampere

3. No peak in 1918, peak in 1919

• Areas outside of the most intensive fighting in the Civil War
• Mostly areas influenced by Viipuri and Petrograd (before 1917)
• Pori district in the west

4. Very few or no cases

• Northernmost Finland was safe, likely due to the very low population density
• Western islands protected by Baltic sea
• Smallpox needed higher population density to survive than the Spanish flu

Russian influence

• Distance to Russia explains east-west differences within Finland
• Finnish civil war explains outbreaks in the south
• Quarantine hospital at border prevented many outbreaks

Smallpox

• The problems faced by medical professionals were almost exactly the same as the ones during COVID-19
• Data quality and lack of digitisation are huge problems
• Variola virus is the only pathogen ever to be extinguished by humans