5.5 Smoking by-law, 2004 Ottawa: retrospective

The City of Ottawa has historically been a leader in municipal policy to reduce tobacco smoking at the population level. A key milestone was the introduction of a 2004 by-law that limited smoking in bars/restaurants.

We want to know how many smoking attributable deaths were prevented due to this by-law.

Assumption: We assume that the by-law lead to decrease in current smokers either through individuals quitting smoking or individuals not starting to smoke.

For this example we will:

A. Transform 2003 CCHS PUMF variables to 2013 CCHS PUMF variables

B. Determine the prevalence of current smokers in 2003

C. Determine the prevalence of current smokers in 2013

D. Estimate how many smoking attributable deaths were prevented due to enacting the by-law

For this example we will use the sample data set data.sample2003.csv and data.sample2013.csv, which can be downloaded at https://github.com/Big-Life-Lab/PBL-Planning-Tool-Case-Examples. This link also contains all of the R code for this case example. Although we have included the R code any other statistics software (SAS, STATA, etc) can be used to complete this case example.

5.5.1 Part A: Transform 2003 CCHS PUMF variables to 2013 CCHS PUMF variables

The Project Big Life Planning Tool only works on 2013-2014 CCHS variable names only. Therefore if you are using a CCHS data set from previous years you will have to transform the variables to the 2013-2014 format.

  1. Use data.sample2003.csv

Note: Although the data.sample2003.csv is based on the 2003/2004 Canadian Community Health Survey Public Use Microdata File, data.sample2003.csv is a completely fabricated data set and can only be used for exemplary purposes.

With the data.sample2003 data set you will need to:

  • Rename 2003 CCHS PUMF variables to their corresponding 2013 CCHS PUMF variables,

  • Create variable CCC_091 from variables CCCC_91A and CCCC_91B, and

  • Recode DHHGAGE variable

Note: The Project Big Life team is developing a R package that will make this step easier for you! The package transforms CCHS variables from 2001 to 2014.

We’ll update the guidance as soon as the package is ready!

5.5.2 Part B: Determine the prevalence of current smokers in 2003

Current smokers are defined by variable SMKDSTY: 1 – Daily, 2 – Occasional, 3 – Always Occasional.

In 2003, current smokers accounted for 26% of the Ottawa population.

###Part C: Determine the prevalence of current smokers in 2013.

The prevalence of smokers in 2013 will be the scenario target in Part D.

  1. Use data.sample.csv

Note: Although the data.sample.csv is based on the 2013/2014 Canadian Community Health Survey Public Use Microdata File, data.sample.csv is a completely fabricated data set and can only be used for exemplary purposes.

In 2013, current smokers accounted for 21% of the Ottawa population.

5.5.3 Part D: Estimate how many smoking attributable deaths were prevented due to enacting the by-law

Use the Project Big Life Planning Tool for the following steps:

  1. Load your data file: data.sample2003.csv to the Project Big Life Planning Tool.

  2. Select initial calculation: Summary Measure – Deaths (5 years)

  3. Add Filter: GEODPMF - 35951

  4. Click: Scenarios, and select Intervention

  5. Click Smoking then Select “Prevalence of current smokers”

  6. Select Target

  7. Type in 21 (prevalence of smokers from 2013) into the text box.

  8. Name your calculation: Smoking attributable deaths 2003

  9. Click Calculate