Table 2.

Summary of sensitivity analyses: prevalence rates (/1000) for COPD and severe + very severe COPD (GOLD grade 3–4) as predicted by the dynamic model for year 2025

Variable	Alternative hypothesis entered as baseline values for the sensitivity analysis	COPD		Severe+very severe airflow limitation
		Prevalence /1000	Relative % change	Prevalence /1000	Relative % change
Prevalence	10% relative increase in 2005 prevalence (~upper limit of the confidence interval)	100.09	+ 4.5%	20.62	+ 5.0%
	10% relative decrease in 2005 prevalence (~lower limit of the confidence interval)	91.42	−4.5%	18.65	−5.0%
	Variation in the distribution of prevalent cases by severity of airflow limitation [13]	90.82	−5.2%	18.98	−3.4%
	Variation in the distribution of prevalent cases by smoking status	94.89	−0.91%	19.40	−1.2%
Incidence	10% increase in incidence	100.36	+ 4.8%	20.50	+ 4.4%
	10% decrease in incidence	91.11	−4.9%	18.77	− 4.4%
	Increase of incidence in women, reaching incidence in men	120.81	+ 26.2%	24.22	+ 23.3%
	Increase from 4.5 to 14% of the proportion of incident cases with severe or very severe airflow limitation at diagnosis	97.01	+ 1.3%	22.04	+ 12.2%
Probabilities of transition between GOLD grades	Reduction of transition rate to a more severe category, from 10 to 20 to 5%	98.41	+ 2.8%	8.92	−54.6%
Mortality [17]	Lower limit of the 95% confidence intervals of hazards ratios of COPD-related mortality by severity of airflow limitation	104.77	+ 9.4%	22.10	+ 12.5%
	Upper limit of the 95% confidence intervals of hazards ratios of COPD-related mortality by severity of airflow limitation	85.31	−10.9%	16.98	−13.5%
Probabilities of transition between smoking status	No change in individual smoking status	93.18	−2.7%	18.95	−3.5%

Percentages indicate the % of difference between each sensitivity analysis and the reference analysis. Details on alternative hypotheses can be found in the Additional files 1, 2, 3, 4, 5 and 6