A systematic literature review of methods of incorporating mortality in cost-effectiveness analyses of lipid-lowering therapies

Figures & data

Table 1. Study inclusion criteria applied when screening articles and abstracting data.

Component	Inclusion	Exclusion
Population	Adults with hypercholesterolemia and mixed dyslipidemia	Children <18 years of age
Publication date	2000–2017	Prior to 2000
Language restrictions	English	Non-English
Intervention and comparators	Statins Ezetimibe	Surgical procedures Lifestyle or dietary modifications
Outcome	Cardiovascular event reduction LYs gained QALYs gained Cost-effectiveness and/or cost-utility results	Costs alone (no efficacy)

LY, life year; QALY, quality-adjusted life year.

Figure 1. Flow diagram depicting the literature review process and studies identified. HTA, health technology assessment.

Table 2. Results: frequency of statistically significant mortality data utilized within cost effectiveness analyses of lipid-lowering therapies.

	n (%)
	Assumed CVD mortality benefit in the model	No RCT evidence for statistically significant benefit in CVD mortality^a
All (n = 100)	99/100 (99%)	87/99 (88%)
Active controlled studies^b (n = 41)	40/41 (98%)	40/40 (100%)
Placebo controlled studies (n = 59)	59/59 (100%)	47/59 (80%)

^a Defined as individual RCTs that did not demonstrate statistically significant benefit in CVD mortality, even though, in certain cases, statistical significance can be achieved through ad-hoc sub-group analysis, non-RCTs, or meta-analysis of RCTs.

^b Includes dose intensification studies.

Figure 2. Results: source of mortality data and use of composite and individual end-points within CEA that incorporated a direct effect.

Table 3. Results: quantitative assessment of mortality data sources and format of data used among studies modeling a direct mortality effect.

Parameter	n (%)
	Among studies based on meta-analysis	Among studies based on individual trials
All studies that used a direct effect  on mortality	17/60 (28%)	43/60 (72%)
End-point applied:
Individual end-point (e.g. CHD death)	12/17 (71%)	23/43 (53%)
Composite end-point (e.g. fatal and   non-fatal CHD)	5/17 (29%)	20/43 (47%)

CHD, coronary heart disease.

Table 4. Sensitivity analysis of studies that incorporated a direct CVD mortality benefit.

Parameter	n (%)
	Used direct evidence of CVD mortality benefit in the model among all studies	No RCT evidence for statistically significant benefit in CVD mortality^a among studies using direct evidence
All (n = 100)	60/100 (60%)	48/60 (80%)
Active controlled studies (n = 41)	19/41 (46%)	19/19 (100%)
Placebo controlled studies (n = 59)	41/59 (69%)	29/41 (71%)

^a Defined as individual RCTs that did not demonstrate statistically significant benefit in CVD mortality, even though, in certain cases, statistical significance can be achieved through ad-hoc sub-group analysis, non-RCTs, or meta-analysis of RCTs.