Application of a system dynamics model to inform investment in smoking cessation services in New Zealand

Abstract

Objectives: We estimated the long-term effects of smoking cessation interventions to inform government decision-making regarding investment in tobacco control.

Methods: We extracted data from the 2006 New Zealand Tobacco Use Survey and other sources and developed a system dynamics model with the iThink computer simulation package. The model derived estimates of population cessation rates from smoking behaviors and applied these over a 50-year period, from 2001 to 2051, under business-as-usual and enhanced cessation intervention scenarios.

Results: The model predicted larger effects by 2051 with the enhanced cessation than with the business-as-usual scenario, including: an 11% greater decline in adult current smoking prevalence (9 versus 10 per 100 people), 16% greater decline in per capita tobacco consumption (370 versus 440 cigarette equivalents per year), and 11% greater reduction in tobacco-attributable mortality (3000 versus 3300 deaths per year).

Conclusions: The model generated reliable estimates of the effects on health and on tobacco use of interventions designed to enhance smoking cessation. These results informed a decision announced in May 2007 to increase funding for smoking cessation by NZ $42 million over 4 years.

FIGURE 1

Causal structure of the New Zealand tobacco system dynamics model. Note. AS = adult smoker; R1 = peer feedback; R2 = parental feedback; XS = ex-smoker; YS = youth smoker. Boxes represent population stocks; arrows with valve symbols represent physical flows; cloud symbols represent births and deaths; single-line arrows represent causal links; circles represent other factors and influences; hexagons indicate tobacco control policies. The actual model is structured into 10-year age groups rather than in youths and adults. Tobacco control policies in general can influence smoking initiation rates and excess mortality risks, as well as quitting behaviors and smoking intensity (the parameters affected by smoking cessation policies).

FIGURE 2

Smoking cessation tree diagram. Note. C = proportion of smokers who made 1 or more quit attempts per year; A = proportion of quit attempts that were assisted; R = proportion of quitters who relapsed; S = short-term unassisted quitting success rate; E = rate ratio S_assisted/S_unassisted.

FIGURE 3

New Zealand tobacco system dynamics model base case projections for business-as-usual (BAU) and enhanced cessation (EC) simulations for (a) smoking prevalence, (b) per-capita tobacco consumption, and (c) tobacco attributable mortality count: 2001–2051. Note. The base case represents smoking prevalence in 2001 to 2004.