To examine the costs and benefits of using stents following percutaneous transluminal coronary angioplasty (PTCA) compared to using PTCA alone and compared to medical treatment and coronary artery bypass grafting (CABG) in the invasive treatment of native coronary artery disease.

The following databases were searched (search strategies provided): MEDLINE (1993-Aug 1998), BIDS ISI (to Aug 1998), EMBASE (1991-Mar 1997), the Cochrane Library 1998 Issue 3. In addition conference proceedings were searched (no further details) and the bibliographies of retrieved articles were checked. Relevant reviews were also obtained through personal contact with other clinical review specialists. A separate search of MEDLINE, the Cochrane Library, bibliographies and conference proceedings was carried out to identify economic evaluations. Only articles that were published in peer-reviewed journals and not solely in the form of conference abstracts were included in the review. No language restrictions were reported.

Study designs of evaluations included in the review

Randomised controlled trials (RCTs) of stents after PTCA versus PTCA alone. Studies reported in the review had a follow-up period of one year or less.

Specific interventions included in the review

Stent with PTCA, PTCA alone, medical treatment (e.g. anti-anginal drugs such as nitrates, beta blockers and calcium antagonists; anticoagulants such as aspirin, ticlopidine, heparin and warfarin; streptokinase and tissue plasminogen activator; exercise advice; anti-hypertensive treatments) and CABG. Eight trials used the Palmaz-Schatz stent, two the Gianturco-Roubin stent, two the Wiktor stent and one the AVE stent. One trial used three different stents.

Participants included in the review

Individuals with native coronary artery disease. Participants reported in the review included individuals with stable and unstable angina, a previous myocardial infarction (MI), and those with a chronic artery occlusion.

Outcomes assessed in the review

Outcomes were not specified a priori. Outcomes reported in the review included: death, CVA, acute myocardial infarction, event free survival (no adverse outcome), presence of angina on follow up and the need for repeat intervention (CABG or repeat PTCA). Bleeding and vascular complications, duration of hospital stay and angiographic outcomes were also reported.

How were decisions on the relevance of primary studies made?

The authors do not state how the papers were selected for the review, or how many of the reviewers performed the selection.

The following factors were considered: similarity of participants in terms of baseline characteristics, method of randomisation, timing of outcome, drop-out and failure rates, loss to follow-up, intention to treat analysis, validity of conclusions based on presented results.The authors do not state how the papers were assessed for validity, or how many of the reviewers performed the validity assessment.

Two individuals extracted data and differences were resolved through discussion.

How were the studies combined?

Trials were grouped in terms of single new lesion trials, chronic coronary occlusion trials and following acute myocardial infarction trials. Event rates for death, MI (Q wave and Non Q wave combined where specified) and CABG were then pooled used a fixed-effect model. Relative risk (RR) and 95% confidence intervals (95% CI) were quoted. The pooled RRs (95% CI) of no adverse outcome, angina free status, repeat PTCA and target vessel revascularisation were similarly calculated. The authors also commented on whether the use of a random-effects model would affect the findings, although no data were presented.

How were differences between studies investigated?

Heterogeneity was assessed using a chi-squared analysis.

Eleven RCTs (2718 participants) compared PTCA and stents to PTCA alone. No studies were identified comparing PTCA and stents to medical treatment or CABG.

No comparisons of stents versus medical interventions or CABG were identified. The following comparisons of stent with PTCA versus PTCA alone were reported: Risk of death (9 studies): Pooled RR=0.71 (95% CI: 0.38, 1.31) Chi-square=2.64, df=8, Z=1.11 Risk of CABG (9 studies): Pooled RR=0.95 (95% CI: 0.65, 1.40) Chi-square=6.87, df=8, Z=0.25 Risk of MI (8 studies): Pooled RR=0.76 (95% CI: 0.52, 1.11) Chi-square=5.18, df=7, Z=1.42 Risk of repeat PTCA (9 studies): Pooled RR=0.57 (95% CI: 0.47, 0.70) in favour of PTCA and stents Chi-square=8.48, df=8, Z=5.42 Risk of target vessel revascularisation (8 studies): Pooled RR=0.48 (95% CI: 0.37, 0.63) in favour of PTCA and stents Chi-square=11.14, df=7, Z=5.37 Risk of no adverse outcome (8 studies): Pooled RR=1.11 (95% CI: 1.07, 1.16) in favour of PTCA and stents Chi-square=6.50, df=7, Z=5.01 Risk of angina free status (6 studies): Pooled RR=1.09 (95% CI: 1.04, 1.14) in favour of PTCA and stents Chi-square=21.64, df=5, Z=3.37 Angiographic results (11 studies): All of the studies showed significant differences in the angiographic results between the stent with PTCA group and the PTCA group immediately post intervention. Quality of the studies: The studies looked at different subgroups of coronary artery disease, patient characteristics and stents. There were no significant baseline differences between the control and intervention groups, apart from in two of the 11 studies. In six studies randomisation was only carried out after a successful PTCA, where as in the other five trials randomisation was carried out before the intervention. In five of the trials the few patients who dropped out after randomisation were not included in the results and analyses; one trial included the patients in the analyses and five trials had no drop-outs. There was no long-term follow-up of patients for longer than one year.

Four cost-effectiveness studies were identified based on three of the eleven RCTs and two cost-utility studies; one based on the BENESTENT trial. Only the BENESTENT-II trial included costs and cost-effectiveness as part of the trial design. The cost-utility studies estimated additional costs per quality-adjusted life-year (QALY) gained at US$23,600 (£14,750) and £250,000. The BENESTENT-II trial effectiveness results were used as the basis for a new cost-utility analysis. The costs used were derived from a local Birmingham NHS trust's published tariff for treatment. The benefit calculation showed the increase in QALYs from PTCA to stent to be very small (0.04 QALYs per person). The cost-utility analysis (based on registry and case series data) derived an additional cost per QALY if a single stent is used instead of PTCA of £22,975 and if two stents are used of £41,500. Sensitivity analysis around the assumptions made gave an incremental cost per QALY gained of single stent over PTCA of between £13,000 and £53,000.

Meta analyses showed that there was a reduced risk of the need for repeat PTCA and target vessel revascularisation in the stent group compared to the PTCA group but no evidence that stents reduce the risk of death, MI or the need for CABG. For the stent with PTCA group there was also a small increased chance of being angina free at the end of the trials and of experiencing event free survival (no adverse event occurring during the one-year follow-up period).

Evidence from the economic analysis of the BENESTENT-II trial show that the quality adjusted life years gained from stent insertion is very similar to that from PTCA, in spite of the difference in numbers of repeat interventions performed. This is also suggested by the cost utility study reviewed in the economic evidence section.

[AC:There are two other systematic reviews that supersede this review and should be consulted in preference to this review. Stenting is a rapidly changing area and several randomised controlled trials have been published since this review. See 1. Perieth M, Kochs G. Stenting versus Ballondilatation bel koronarer Herzkrankheit. Systematische Ubersichten zur medizinischen Effektivat und zur Kosten-Effektivitat. Cologne: Deutsches Institut fur Medizinische Dokumentation und Information: 1999. 2. Meads C, Cummins C, Jolly K, Stevens A, Buris A, Hyde C. Coronary artery stents in the treatment of ischaemic heart disease: a rapid and systematic review. Health Technology Asssessment 2000;4(23)]

This review is based on a wide search of the literature, although no attempts were made to locate unpublished work and so there is potential for publication bias. The inclusion criteria for interventions and study design were clearly defined, but little detail was specified for the inclusion of participants and outcomes. The quality of individual studies was assessed, however the number of individuals involved in the process was not stated.

The studies were pooled using a fixed-effect model. A chi-squared test for hetereogeneity was performed and did detect heterogeneity in three of the seven comparisons. Given the fact that the studies do vary in a number of ways including the type of stent used, the types of patients treated and the length of follow-up, it may have been more appropriate to use a random-effects model or not to pool at all. The authors do not state on some occasions that using a random-effects model would not have significantly altered the results but no data were presented. In view of the heterogeneity between studies, the findings of the review should be treated with caution.

Practice: The authors state that 'there are some small advantages in stent insertion compared to PTCA for patients groups with new lesions in native coronary arteries, in chronic coronary occlusion and following myocardial infarction'. The review classified the support for the routine use of coronary artery stents as category III (i.e. limited support on the evidence available, and more research on medium and long-term outcomes is required).

Research: The authors state that further trials need to be assessed in order to establish the costs and benefits for emergency stent insertion and stenting in unfavourable lesion subsets, in particular, for saphenous vein grafts. In addition, long term effects of using coronary artery stents need to be established'. Also 'follow-up in the trials included in this review is for one year or less which is too short a time to properly evaluate the procedures and their associated costs'.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.