Table 1.

Review of articles

Author & Date	Place of Study	Methods	Main Findings	Author’s Conclusions
Vermeer F, et al. 198815	The Netherlands	A 12 month follow up of 269 patients allocated to thrombolytic treatment and of 264 allocated to conventional treatment.	Quality adjusted mean survival for inferior infarction patients was 307 days (out of 365days) vs 300 days in the thrombolysis & control group; in patients with anterior infarction, it was 38 days longer in the thrombolysis group.	The higher costs of patients receiving thrombolytic were mainly the result of acute catheterization.
		The costs of medical treatment, including medication, hospital stay, cardiac catheterisation, coronary angioplasty, and bypass surgery were considered.	Medical cost: after MI averaged Dfl 21000 for interior infarction & Dfl 20000 for anterior infarction in the control group; Dfl28000 & Dfl29000 in the thrombolysis group.	Thrombolytic increased life expectancy particularly in patients with anterior infarction (2.4 years).
		Mean survival was calculated and survival was adjusted for impaired quality of life.	Additional costs per year of life gained were Dfl 3800 in patients with anterior infarction and Dfl 10000 in inferior infarction.
Levin LA, et al. 199216	Sweden	Randomized double-blinded placebo-controlled study comparing rt-PA with placebo in patients with suspected AMI.	The direct costs were significantly higher among rt-PA group than the placebo group due to the cost of thrombolytic drug.	Cost-effectiveness of intravenous thrombolysis treatment in suspected MI is higher than other medical treatment of coronary heart disease.
		ASSET data and the specific economic data are applied.	The indirect costs are 4190 SEK lower in the rt-PA group as a larger proportion of rt-PA patients returned to work during the follow-up period.
		Direct medical costs & indirect costs (productivity loss) were computed.	The rt-PA mean patient cost(total cost) is 5700 SEK higher than the placebo group. Rt-PA therapy inceses life expectancy by about 1.25 years with a 6% mortality rate and it increases as the mortality rate falls.
		The health benefits are analyzed by the treatment effects on mortality and the effects of treatment on the patient’s quality of life which are measured with the NHP questionnaire.	The cost-utility ratio of rt-PA varies from 3260 SEK to 6310 SEK per QALY gained & the cost-effectiveness ratio varies from 3090 to 5970 SEK.
		The life expectancy is estimated by the DEALE method.
Machecourt J, et al, 199317	France	Double-blind, doubledummy procedure. 180 patients were randomized in a 12-month period with either anistreplase or alteplase.	The cost ranged from 6570 ECU to 6050 ECU per patient, without any significant thrombolytic agent related difference: the total cost of the hospital phase was 6460 ECU for alteplase, 6570 ECU for anistreplase and 6050 ECU for streptokinase (NS). The cost/efficacy ratio was 548 ECU for alteplase, 570 ECU for anistreplase and 405 ECU for streptokinase.	No difference observed in efficacy between the three thrombolytic agents for the three left ventricular parameters and for the patency of the infarctrelated artery.
		Analysis of efficacy including secondary effects of treatments, cost assessment, and variance analysis were conducted	The cost for deceased patients was lower than for those who survived with the total costs of 4466 ECU vs 6512 ECU.	The ICER is similar for anistreplase, alteplase and streptokinase, with a slight advantage for streptokinase.
Rawles JM, 199718	Scotland	Double-blind, randomized, placebo-controlled small trial.	Age, treatment delay and time of presentation determine the outcome at 30 months according to logistic regression.	The pre-hospital therapy of anistreplase is more effective than in-hospital use of streptokinase.
		“Home injection” & “hospital injection” was given. Multivariate analysis with logistic regression.	In the 5-year-follow up period, patients in the home groups live more than 6 months longer on average.
		Comparative costeffectiveness analysis with Streplase from GUSTO trial.
Franzosi MG, et al. 199819	Italy	RCCT (2*2 factorial design)- patients were randomly assigned to receive oral lisinopril or open control &to receive nitroglycerin or open control	Costs for most of the concomitant treatments were $US67472 vs $US32677 (control group vs lisinopril group).	The comparative costeffectiveness ratio for the use of lisinopril was $US2080 per life saved.
			The comparative cost-effectiveness ratio for the use of lisinopril was $US2080 per life saved at 6 weeks post MI.
		Only direct costs regarding publicly financed healthcare were considered	The cost-effectiveness ratios varies from $US1121 to $US40910 per life saved, after conducting the initial sensitivity analysis.	The cost-effectiveness ratios were more favorable in subgroups with higher absolute survival benefit.
		CEA is based on the crude rate of survivors from the 6-week treatment of AMI with Lisinopril (no discounting). Sensitivity analysis was performed.	After the second sensitivity analysis, the comparative cost-effectiveness ratio for the use of lisinopril increased to $US4530 per life saved.
			The number of lives saved per 1000 patients treated with lisinoprilwas greater for older patients than for younger patients.
Lorenzon R, et al. 199820	Italy, UK Germany, USA	CEA of results from the 30-day GUSTO trial (no discounting of costs& benefits):	The cost for each extra life saved in Germany, Italy, U.S.A., is 31%, 45%, 97%, higher than that in the U.K.	Cost-effective analysis is an “affair of state”.
		The incremental costs for each life saved for the ageselective and site-selective protocols were considered.	Age-selective protocol: there would be 64 deaths & 9 patients saved per 1000 patients treated.	Age-selective use of rt-PA is inappropriate; siteselective use of tPA in anterior AMI halves the costs for each extra life saved.
		Sensitivity analysis on the results performed No discounting for costs and effects.	Cost for each of these extra lives saved would be $U.S 144,126, $U.S 159,883, $U.S 109,848, $U.S 216,142 in Germany, Italy, UK and the US respectively Site-selective protocol: 65 deaths/1000 patients treated&8 patients per 1000 patients treated saved, and the cost for each of these extra lives saved would be $U.S 71,858, $U.S 79,715, $U.S 54,769, $U.S 107,764 in Germany, Italy, UK and the US
Marcoff L, et al. 200921	France	ExTRACT-TIMI 25 – a large, randomized, multinational trial at 674 sites in 48 countries.	The net clinical benefit compared with UFH: relative risk reduction 17%, 95% CI:0.10-0.23, p< 0.001; absolute risk reduction 2.1%.	Enoxaparin is both effective and cost effective compared with UFH – also confirmed by probabilistic sensitivity analysis.
		QALY and mean cost compared between Enoxaparin & UFH group. discounted at 3% annually.	There is 90% probability of the enoxaparin being cost effectiveness at the $50,000 threshold.
		Cost and effectiveness analysis were conducted for subgroups measured by NMB &NHB Probabilistic Sensitivity analysis of costs and LYG.
Welsh RC et al. 200922	Canada	Randomized, double-blind, double-dummy, parallel group, of 20,506 patients in at 674 sites in 48 countries, and 118 patients in Canada. Enoxaparin or matching placebo.	When considering the marginal time horizon and allowing clinical benefit to be accounted for, enoxaparin was found to be cost-effective with an ICER of $4,930 and 99% probability of the costeffectiveness ratio being less than $20,000.	Long-term clinical data are required to confirm the assumption that the difference in survival between arms of the study does not widen or close after 30 days.
		LYG was used as the outcome measure; societal perspective (5% discounting rate). Costs of treatment, ICER, LYL were computed.	A reduction of treatment duration reduced the ICER to 1,176/LYG.
			A drop of 15% in marginal costs resulted in an ICER of $4,191;an increase of 15% resulted in an ICER of $5,670.
		Sensitivity analysis performed.