BACKGROUND
Intravascular ultrasound (IVUS) is the generic name for any ultrasound technology used in vivo within the blood vessels. More specifically, intracoronary ultrasound enables imaging of the coronary arteries from within the lumen. This review concentrates on the role of intracoronary ultrasound as an adjunct to interventional cardiology.
OBJECTIVES
(1) To identify the literature on IVUS for guiding coronary interventions, and to synthesise evidence about outcomes compared with outcomes when IVUS guidance has not been used. (2) To use this evidence, together with other information about costs and outcomes, to model the cost effectiveness of IVUS guidance. (3) To synthesise the evidence on the reproducibility of measurements of cross-sectional area made using IVUS.
METHODS
DATA SOURCES
(1) Electronic searches of MEDLINE, EMBASE, Science Citation Index, Index to Scientific and Technical Proceedings, Engineering Compendex, Engineering Page One, Cochrane Library, Inside (British Library), 1990-98. (2) Contacting experts and centres of expertise, 1990-99. (3) Internet search, 1990-99.
METHODS
STUDY SELECTION
Studies of IVUS-guided coronary interventions performed on humans were included in the review. Non-English language studies were also included when they covered IVUS-guided stenting or angioplasty. Control evidence regarding outcomes without IVUS guidance was sought only from randomised controlled trials (RCTs). Studies investigating the reproducibility of measurements of cross-sectional area were included only if the results were expressed in terms of the mean and standard deviation of paired differences.
METHODS
DATA EXTRACTION
Checklists that covered study details, patient characteristics and results were completed independently by three reviewers. Consensus was reached on any disagreements. Local data were gathered on the costs of IVUS-guided stenting.
METHODS
DATA SYNTHESIS
Overall event rates were calculated by pooling patient results from the included studies. A decision-analytic model was used to combine information from the literature with cost estimates, in order to predict cost-effectiveness in terms of cost per restenosis event avoided by the use of IVUS guidance. The analysis was performed from the perspective of the healthcare provider. Sensitivity analysis was undertaken. A simple extrapolation was made to long-term outcome so that cost-utility (using quality-adjusted life years (QALYs)) could be estimated. The minimum detectable change in cross-sectional area was estimated from the reproducibility results.
RESULTS
Only one study on IVUS-guided angioplasty satisfied the inclusion criteria, and there were no studies on IVUS-guided atherectomy or other IVUS-guided interventions that satisfied the inclusion criteria. Of the 15 articles on IVUS-guided stenting that satisfied the inclusion criteria, seven presented data on outcomes at 6 months post-intervention. The angiographic restenosis rate was 16 +/- 1%. This compared with 24 +/- 2% derived from five articles on stenting without IVUS guidance. Data for follow-up periods longer than 6 months were presented in only two studies. Data from a total of five studies were included in the decision-analytic model. The cost per restenosis event avoided was 1545 pound sterling. After extrapolation to long-term outcome, the calculated cost per QALY was 6438 pound sterling. The baseline QALY gain was only 0.03 years. Sensitivity analysis resulted in large differences between the best- and worst-case scenarios, for example, from a saving of 5000 pound sterling to a cost of 24,000 pound sterling restenosis event avoided. The smallest changes in cross-sectional area that could be measured were 1.6 mm2 by a single observer and 1.9 mm2 by different observers.
CONCLUSIONS
Implications for healthcare: The evidence available is too weak for there to be any reliable implications for clinical practice. (ABSTRACT TRUNCATED)