Related Experiment Videos

A predictive method for estimating the late angiographic results of coronary intervention despite incomplete

R E Kuntz1, K M Keaney, C Senerchia

  • 1Charles A. Dana Research Institute, Boston, MA.

Circulation
|March 1, 1993
PubMed

Insights

Coronary restenosis trials with incomplete angiographic follow-up are subject to selection bias. A new predictive method using clinical data can provide a more accurate estimate of restenosis rates for the entire patient population.

Area of Science:

  • Cardiovascular Medicine
  • Interventional Cardiology
  • Biostatistics

Background:

  • Coronary restenosis studies often rely on late angiographic endpoints (4-6 months).
  • Incomplete angiographic follow-up (50-80% of patients) can lead to biased estimates of restenosis rates.
  • Existing methods do not adequately correct for differences between restudied and non-restudied patients.

Purpose of the Study:

  • To develop and validate a method to correct for selection bias in coronary restenosis trials caused by incomplete angiographic follow-up.
  • To assess the impact of clinical indications for restudy on restenosis rates.
  • To compare a novel predictive method with traditional estimation techniques.

Main Methods:

  • Analyzed clinical indications for angiographic restudy in 301 lesions (Palmaz-Schatz stenting or directional coronary atherectomy).
  • Classified follow-up as 'elective' (no clinical indication) or 'nonelective' (recurrent symptoms, positive exercise tests).
  • Developed a 'predictive' model using clinical data to estimate restenosis in non-angiographically studied lesions.

Main Results:

  • Nonelective lesions had significantly higher restenosis rates (53% vs. 13%) and percent stenosis (50% vs. 27%) compared to elective lesions.
  • Even with 83% angiographic follow-up, selection bias was evident, with traditional methods estimating 29.1% restenosis versus 26.3% by the predictive method.
  • Nonelective patients were more likely to undergo repeat angiography (97% vs. 75% for elective, p < 0.001).

Conclusions:

  • Coronary restenosis trials with less than 90% angiographic follow-up are susceptible to selection bias.
  • Traditional analysis methods fail to account for the confounding influence of clinical status in non-restudied patients.
  • A predictive method utilizing readily available clinical information offers a more accurate estimation of population-wide restenosis.
Abstract

Related Concept Videos