SciELO - Scientific Electronic Library Online

vol.15 issue5Metabolic syndrome prevalence in the town of Arjona, Colombia author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand



Related links

  • On index processCited by Google
  • Have no similar articlesSimilars in SciELO
  • On index processSimilars in Google


Revista Colombiana de Cardiología

Print version ISSN 0120-5633


GUERRERO, Fernando A; CABRERA, Diana P; BOHOQUEZ, Ricardo  and  RUIZ, Álvaro. Usefulness of clinical scores to improve prediction of significant coronary heart disease after conventional treadmill exercise testing. Rev. Col. Cardiol. [online]. 2008, vol.15, n.5, pp.207-214. ISSN 0120-5633.

Background: in the last AHA/ACC expert consensus document, clinical scores to improve sensitivity (68%) and specificity (77% of the exercise testing, diagnostic method considered a first line diagnostic method for coronary heart disease treatment (one of the main causes of mortality in Colombia and worldwide), are recommended. Nevertheless, few institutions in our country use them and they are difficult to apply in populations different to the ones for which they were developed. For this reason, a study to assess its performance in our environment, is needed. Materials and Methods: Morise and Duke treadmill scores were chosen to assess the reason for its validation in several populations, and were mentioned in the AHA/ACC consensus. The Morise and Duke scores classified patients in at low, middle and high risk for coronary heart disease. Primary objectives: validate the prediction scales for coronary heart disease and determine the best cutoff value for each score in a one year follow-up. Secondary objectives: determine the composite endpoint for acute myocardial infarction, cardiac death, angina requiring hospitalization, coronary obstruction >50% and/or angioplasty and stent implantation. Determine the best cutoff point through the ROC curves. Inclusion Criteria: patients >18 years old with suspected coronary heart disease. Exclusion Criteria: pregnant women with documented coronary heart disease, uninterpretable EKG, incapacity or contraindication for performing exercise stress test for any reason, ST depression < 1 mm in basal EKG, follow-up impossibility and incomplete data that might hinder the score calculation. Statistical analysis: the sample was calculated using alpha error < 0.05, beta error < 0.20 (power 80%), correct classification probability 0.4, kappa level for null hypothesis 0.85 and kappa level for alternate hypothesis 0,7. The Sample Size Program version 1.0 was used. 101 patients were required. Results: 127 patients were enrolled and 9 were excluded: 2 because of submaximal exercise testings and 7 because the follow-up was impossible. 118 patients were analyzed. Mean follow-up was 343 days (1 - 564). Mean age was 59 years (29 - 83). Women: 53% (63) and men: 47% (55). Mean age 59 and 57 years respectively. Other characteristics: cigarette smoking: 47% (55), dyslipidemia: 68% (80), body mass index > 27,5: 18% (45) and diabetes mellitus: 16% (19). Morise score classified 36% (43) patients at low risk, 52% (61) at intermediate risk and 12% (14) at high risk. According to Duke the results were 53% (63), 41% (48) and 6% (7) respectively. When interpreting an isolated exercise testing, cardiologists classified patients: 81% (95) negative, 8% (10) suggestive and 11% (14) positive. The composite endpoint appeared in 11% (14 patients). When comparing patients with and without outcomes, the first ones classified more frequently as having higher probability than those that had not, with statistically significant differences (Morise: p = 0,0002 and Duke: 0,0005). In the Morise score with cutoff value 48, 92% sensitivity and 68% specificity was achieved. In Duke, with cutoff value 38, it was 100% and 31% respectively. Discussion: concordance for Morise is better than for the other evaluated methods. The addition of clinical scores to the exercise testing interpretation improves its operative characteristics without any cost increment, achieving savings of 10% to 18%. Conclusions: clinical scores increase sensitivity and specificity, and for this reason they should be used as routine in the conventional exercise testing report. Nevertheless, it is necessary to look for solutions to improve its performance even further.

Keywords : coronary heart disease; treadmill stress test; clinical scores; validation.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )


Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License