Is Age-targeted full-field digital mammography screening cost-effective in emerging countries? A micro simulation model

Souza, Fabiano Hahn; Polanczyk, Carísi Anne

doi:10.1186/2193-1801-2-366

SpringerPlus

Table 1 Main parameters used in the base case and sensitivity analyses

From: Is Age-targeted full-field digital mammography screening cost-effective in emerging countries? A micro simulation model

Variables	Screening test performance			Distribution/comments	Reference
	Mean	Minimum	Maximum
Mammography coverage	18%	10%	30%	Uniform	(Ministério_Saúde_Brasil, DATASUS 2011)
Mammography coverage ^ǁ	70%	55%	85%	Uniform	(Lilliu et al. 2002)
Sensitivity of SFM (40–49 years)	76%	60%	85%	Effectiveness data from large population	(Kerlikowske et al. 2011)
Sensitivity of FFDM (40–49 years)	82%	65%	90%	Effectiveness data from large population	(Kerlikowske et al. 2011)
Sensitivity of SFM (50–59 years)	85%	65%	90%	Effectiveness data from large population	(Kerlikowske et al. 2011)
Sensitivity of FFDM (50–59 years)	80%	65%	90%	Effectiveness data from large population	(Kerlikowske et al. 2011)
Sensitivity of SFM (60–69 years)	83%	65%	90%	Effectiveness data from large population	(Kerlikowske et al. 2011)
Sensitivity of FFDM (60–69 years)	90%	65%	95%	Effectiveness data from large population	(Kerlikowske et al. 2011)
Treatment complication (yearly) - Chemotherapy	16%	10%	20%	Resource utilization database	(Hassett et al. 2006)
Treatment complication (yearly) – Endocrine therapy	5%	1%	10%	Resource utilization database	(Hassett et al. 2006)
Overdiagnosis	5%	0	30%	Systematic review estimate	(Smith & Duffy 2011)
	Cancer stage distribution
	Mean	CI ^* 95%
DCIS (clinical diagnostic)	6.1%	4.9–7.3%		Beta (α = 97; β = 1494)	(INCA 2009b; Martins et al. 2009)
State 1 (clinical diagnostic)	14%	13.1–16.6%		Beta (α = 232; β = 1329)	(INCA 2009b; Martins et al. 2009)
State 2 (clinical diagnostic)	38.6%	36.5–40.5%		Beta (α = 915; β = 1455)	(INCA 2009b; Martins et al. 2009)
State 3 (clinical diagnostic)	34.7%	32.4–37.1%		Beta (α = 546; β = 1028)	(INCA 2009b; Martins et al. 2009)
State 4 (clinical diagnostic)	10.8%	NA		Complementary	(INCA 2009b; Martins et al. 2009)
CDIS (screening diagnostic)	6.1%	NA		Dynamic range ^∫	(Kerlikowske et al. 2011)
State 1 (screening diagnostic)	58%^Ξ	NA		Effectiveness data from large population	(Kerlikowske et al. 2011)
State 2 (screening diagnostic)	32.4%^Ξ	NA		Effectiveness data from large population	(Kerlikowske et al. 2011)
State 3 (screening diagnostic)	8.3%^Ξ	NA		Effectiveness data from large population	(Kerlikowske et al. 2011)
State 4 (screening diagnostic)	1.3%^Ξ	NA		Effectiveness data from large population	(Kerlikowske et al. 2011)
	Transition probabilities
BC Recurrence	Mean	Range	Local	Regional/systemic
CDIS	0.008/y	0.002–0.014/y	50–98%	2–50%	(Baxter et al. 2004; Meijnen et al. 2008)
Stage 1	0.030/y	NA	16–47%	53–84%	(Hirsch et al. 2011a; Hirsch et al. 2011b)
Stage 2	0.087/y	NA	19–56%	44–81%	(Wapnir et al. 2006)
Stage 3	0.283/y	0,11–0,28/y	19–56%	19–56%	(Wapnir et al. 2006)
BC Death	Mean	Range
CDIS	0.002/y	0.002–0.003/y			(Ernster et al. 2000)
Stage 1	0.009/y	NA			(de Oliveira et al. 2009)
Stage 2	0.031/y	NA			(de Oliveira et al. 2009)
Stage 3	0.090/y	NA			(de Oliveira et al. 2009)
Stage 4	0.270/y	0.20–0.34			(de Oliveira et al. 2009)
	Relative risk		Distribution/comments
	Mean
Adjuvant Taxane chemotherapy^§	0.86		Log-Normal (μ = −0.15;σ=0.07)		(Peto et al. 2012)
Adjuvant Aromatase inhibitor^§¶	0.82		Log-Normal (μ = −0.20;σ=0.12)		(Dowsett et al. 2010)
Adjuvant Trastuzumab therapy ^§‡	0.61		Log-Normal (μ = −0.49;σ=0.06)		(Perez et al. 2011)
Screening vs. non-screening cancer cases^Φ	0,62		Log-Normal (μ= − 0.48;σ=0.12)		(Mook et al. 2011)
Advanced disease - Luminal A vs. Luminal B^¥	1.42		Log-Normal (μ = 0.34;σ=0.12)		(Kennecke et al. 2010)
Advanced disease - Luminal A vs. HER2 + ^¥	1.90		Log-Normal (μ=0.64;σ=0.11)		(Kennecke et al. 2010)
Advanced disease - Luminal A vs. Triple negative^¥	1.62		Log-Normal (μ = 0.48;σ=0.11)		(Kennecke et al. 2010)
	Relative odds ratio		Distribution/comments
	Mean
Diagnostic cancer downstage (FFDM under 50 years)	0.54		Log-Normal (μ= − 0.654;σ=0.307)		(Souza et al. 2013)
	Mean	Minimum	Maximum
Discount rate	5%	0%	10%	Brazilian Health Economic Guidelines	(Ministério_Saúde_Brasil 2009)
Costs (Brazilian Real)
	Mean	Minimum	Maximum
Medical visit	10	5	25	DATASUS	(Ministério_Saúde_Brasil, DATASUS 2011)
FFDM	68	45	90	Estimated^∀	(Souza 2012)
SFM	45	30	60	DATASUS	(Ministério_Saúde_Brasil, DATASUS 2011)
Biopsy	429	150	700	Gamma (α = 14.93; λ = 0.03)	(Souza 2012)
Recall SFM	152	50	250	Aggregate costs	(Souza 2012)
Recall FFDM	197	100	300	Aggregate costs	(Souza 2012)
Staging early BC^Ψ	509	250	750	Gamma (α = 3.09 λ=0.01)	(Souza 2012)
Staging locally and advanced cancer^Δ	592	200	800	Gamma (α = 2.52 λ=0.04)	(Souza 2012)
Invasive cancer stage 1 (first year)	6,502	2,500	11,500	Aggregate costs	(Souza 2012)
Invasive cancer stage 2 (first year)	15,610	6,500	24,500	Aggregate costs	(Souza 2012)
Invasive cancer stage 3 (first year)	18,638	9,500	27,500	Aggregate costs	(Souza 2012)
Invasive cancer stage 4 (first year)	12,452	6,500	20,500	Aggregate costs	(Souza 2012)
Invasive cancer stage 1 (≥ 2 year)	602	200	1,000	Aggregate costs	(Souza 2012)
Invasive cancer stage 2 (≥ 2 year)	677	200	1,200	Aggregate costs	(Souza 2012)
Invasive cancer stage 3 (≥ 2 year)	742	200	1,600	Aggregate costs	(Souza 2012)
Invasive cancer stage 4 (≥ 2 year)	12,439	4000	20,000	Aggregate costs	(Souza 2012)
	Utilities
	Mean	CI ^* 95%
Healthy woman	0.800	NA		South of Brazil population^⊥	(Cruz 2010)
Healthy woman – false positive mammography	0.795	NA		Estimated^∴	(Cruz 2010)
Non metastatic BC^χ – follow-up	0.772	0.63–0.90		Normal distribution	(Souza 2012; Cruz 2010)
Early BC^χ – Adjuvant Endocrine Therapy	0.762	0.62–0.91		Normal distribution	(Souza 2012; Cruz 2010)
Early BC^χ – Adjuvant Chemotherapy	0.739	0.61–0.87		Normal distribution	(Cruz 2012; Cruz 2010)
Clinical Stage 3 – Adjuvant Endocrine Therapy	0.760	0.59–0.95		Normal distribution	(Souza 2012; Cruz 2010)
Clinical Stage 3 – Adjuvant Chemotherapy	0.700	0.63–0.78		Normal distribution	(Souza 2012; Cruz 2010)
Clinical Stage 4 – Advanced disease	0.680	0.57–0.80		Normal distribution	(Souza 2012; Cruz 2010)

^ǁ Screening strategies; NA: not applicable; ^∫ time and screening coverage-dependent (increase in the DCIS rate with the introduction of the screening program); ^Ξ relative to invasive cancer (excluding DCIS); ^§ Relative risk of BC death in clinical stage 2 and 3 patients; ^¶ hormone-positive patients; ^‡ HER2-positive patients; ^Φ Relative risk of BC death; ^¥ Relative risk of BC death in advanced disease (stage 4) according to prognostic subtype; ^∀ Plausible estimate 50% above SFM reimbursement value; ^Ψ clinical stages 1 and 2; ^Δ clinical stages 3 and 4; ^* confidence interval; ^⊥ Porto Alegre city; ^∴Considering the mean of non-metastatic BC utility (0.77) and a false positive as a 2-month period of disutility (0.80–0.77= [(0.03)*(0.16 year)=0.005] →0.80–0.005=0.795; ^χin situ, stage 1, stage 2, and stage 3 patients.

Back to article page