Analysis of health economics assessment reports for pharmaceuticals in France – understanding the underlying philosophy of CEESP assessment

Figures & data

Figure 1. Overview of the bodies involved in the pricing and reimbursement process in France.

Table 1. Description of the two-level classification of methodological concerns from the HAS guidelines.

Level 1	Level 2
Structural choices	Economic analysis and choice of outcome criterion
	Time horizon
	Comparison of strategies
	Perspective
Measurement and valuation of health states and costs	Costs Measurement and calculation Cost type Valuation
	Health states Measurement (methods and data) Valuation
Modelling	Transition estimates Data extrapolation Estimation method Reporting Efficacy data source
	Modelled population
	Model structure
	Programming errors
Study objective	–
Results presentation and sensitivity analyses	Sensitivity analyses Deterministic/probabilistic Scenario
	Presentation
	Validity (internal and external) and study limitations

Table 2. Methodological concerns expressed in CEESP reports.

Brand name	INN	CEESP report publication date	Total methodological concerns	Minor methodological concerns	Important methodological concerns	Major methodological concerns	Major overall methodological concerns
Adempas®	Riociguat	14/10/2014	14	6	4	4
Botox®	Botulinum toxin	25/11/2014	10	8	2	0
Daklinza®	Daclatasvir	03/02/2015	12	4	8	0
Defitelio®	Defibrotide	18/02/2014	14	2	10	2
Entyvio®	Vedolizumab	25/11/2014	11	7	4	0
Harvoni®	Ledipasvir /sofosbuvir	26/05/2015	11	3	7	1
Kadcyla®	Trastuzumab emtansine	11/03/2014	2	2	0	0
Nexplanon®	Etonogestrel	15/09/2015	9	9	0	0
Nplate®	Romiplostim	03/02/2015	6	1	5	0
Olysio®	Simeprevir	14/10/2014	17	12	4	1
Rotarix®	Rotavirus	22/07/2014	16	11	5	0
Rotateq®	Rotavirus	16/09/2014	15	10	5	0
Sovaldi®	sofosbuvir	15/04/2014	17	8	8	1
Tecfidera®	dimethyl fumarate	24/06/2014	23	20	3	0
Tivicay®	Dolutegravir	27/05/2014	11	10	1	0
Viekirax® /Exviera®	Ombitasvir/ paritaprevir/ ritonavir Dasabuvir	09/06/2015	11	5	5	1
Xolair®	Omalizumab	25/11/2014	16	9	7	0
Xtandi®	enzalutamide	09/06/2015	12	9	2	1
Zostavax®	Zoster vaccine	15/04/2014	16	15	1	0

Table 3. The number of methodological concerns by level.

	Type of methodological concern					Type of methodological concern
Level 1	Major	Important	Minor	Total level 1	Level 2	Major	Important	Minor	Total level 2
Structural choices	1 (4%)(9%)*	12 (46%)(15%)*	13 (50%)(9%)*	26 (11%)	Economic analysis and choice of outcome criterion	1 (50%)	0 (0%)	1 (50%)	2 (8%)
					Time horizon	0 (0%)	1 (50%)	1 (50%)	2 (8%)
					Comparison of strategies	0 (0%)	11 (52%)	10 (48%)	21 (81%)
					Perspective	0 (0%)	0 (0%)	1 (100%)	1 (4%)
Measurement and valuation of health states and costs	1 (1%)(9%)*	16 (21%)	58 (77%)	75 (31%)	Costs	0 (0%)	6 (14%)	36 (86%)	42 (56%)
		(20%)*	(38%)*		· Measure and calculation	0 (0%)	3 (14%)	19 (86%)	22 (52%)
					· Cost type	0 (0%)	1 (20%)	4 (80%)	5 (12%)
					· Valuation	0 (0%)	2 (13%)	13 (87%)	15 (36%)
					Health states	1 (3%)	10 (30%)	22 (67%)	33 (44%)
					· Measure (methods and data)	1 (4%)	10 (37%)	16 (59%)	27 (82%)
					· Valuation	0 (0%)	0 (0%)	6 (100%)	6 (18%)
Modelling	3 (4%)(27%)*	27 (34%)	49 (62%)	79 (33%)	Transition estimates	1 (3%)	9 (26%)	24 (71%)	34 (43%)
		(34%)*	(32%)*		· Data extrapolation	0 (0%)	4 (67%)	2 (33%)	6 (18%)
					· Estimation method	1 (8%)	0 (0%)	12 (92%)	13 (38%)
					· Reporting	0 (0%)	0 (0%)	5 (100%)	5 (15%)
					· Efficacy data source	0 (0%)	5 (50%)	5 (50%)	10 (29%)
					Modelled population	2 (20%)	2 (20%)	6 (60%)	10 (13%)
					Model structure	0 (0%)	13 (48%)	14 (52%)	31 (39%)
					Programming errors	0 (0%)	3 (75%)	1 (25%)	4 (5%)
Study objective	0 (0%)(0%)*	2 (100%)(3%)*	0 (0%)(0%)*	2 (1%)
Results presentation and sensitivity analyses	6 (10%)(55%)*	23 (38%)	32 (52%)	61 (25%)	Sensitivity analyses	3 (8%)	17 (47%)	16 (44%)	36 (59%)
		(29%)*	(21%)*		· Deterministic /Probabilistic	3 (10%)	14 (45%)	14 (45%)	31 (86%)
					· Scenario	0 (0%)	3 (60%)	2 (40%)	5 (14%)
					Presentation	3 (25%)	3 (25%)	6 (50%)	12 (20%)
					Validity (internal and external) and study limitations	0 (0%)	3 (23%)	10 (77%)	13 (21%)
Total	11 (4.5%)	80 (32.9%)	152 (62.6%)	243 (100%)		11 (4.5%)	80 (32.9%)	152 (62.6%)	243 (100%)

*The indicated percentage is calculated by reservation type (minor, important and major). All other percentages are calculated by category level.

Table 4. Summary of the items most frequently criticised by CEESP.

Category	Most frequently reported criticisms
Health states and cost	Robustness of clinical data Extrapolation of survival curve from immature data while more mature data is available Management of adverse events in the model questionable No discussion on similarity of adverse events and their costs between the modelled strategies Consumed resources are derived from expert opinion, without the experts being documented Inconsistent utility data Validation poorly described
Modelling	Hypotheses and extrapolation choices for probability estimates are insufficiently justified Poor or unsatisfactory documentation of structural hypotheses used in the model Transition probabilities extrapolation not conservative Adverse events are not included in the model or not taken into account in utilities or cost calculations Aggregation of heterogeneous health states
Results presentation and sensitivity analyses	Incomplete scenario presented Transferability of results in clinical practice is highly uncertain Lack of sensitivity analyses on some parameters Ranges for parameters used in deterministic sensitivity analysis are not adequately justified Distributions of parameters used in probabilistic sensitivity analysis are not presented or not justified in a comprehensive manner
Structural choice	Choice of comparators used in the analysis does not correspond to comparators used in the indication Insufficient justification for the failure to take into account some of the comparators Time horizon is not appropriate with regard to the evolution of the pathology Perspective selected in the reference case analysis does not match the one recommended by HAS.