JOURNAL OF THE
ROYAL MEDICAL SERVICES

Official Publication for the Jordanian Royal Medical Services


Pre- and Post-Test Probability of Coronary CT Angiography in Obstructive Coronary Artery Diagnosis


Sakher Abdullah Alsharo MD*,Mahmoud Hiffith Alhindawi,MD**, Rabaa Mustafa Sulaiman Rababah MD***,Haitham Suleiman Ahmad Al-Qaderi, MD***,Laith Ishaq Yousef Alazrai ,MD***,Anees Mahmoud Almasaafah,MD****


 

ABSTRACT

Aim: This study aimed to assess the change in the theoretical probability of obstructive coronary artery diseases in patients undergoing Coronary CT Angiography. among patients who presented with chest pain to Queen Alia Heart Institute’s Cardiology clinics, Royal Medical Services, Amman, Jordan between  January 2020 to  January 2021.

Methods: The pre-test probability of obstructive coronary artery diseases was determined for each patient using the predictive model of Genders, which is an updated and extended version of the previous model of Diamond and Forrester. The post-test probability was calculated for each patient using Bayes’ theorem, according to the equation: P(A|B)=[P(B|A)×P(A)]/P(B), in which P (A|B) is the post-test probability conditioned by the pre-test probability [P(A)] and P(B) is the probability determined by the test used. The receiver operating characteristic curve was thereafter used to evaluate the area under the receiver operating characteristic curve and performance indices of pre-test and post-test probabilities of CAD.

Results: The area under the curves were significantly higher for the pre-test Gender et al predictive model than post-test Gender et al model with Area±SEM of 0.902±0.006 (95% CI; 0.889-0.915) vs 0.798±0.010 (95% CI; 0.777-0.818), respectively. The Binary Logistic Regression of the two investigated prediction models against the stenosis probability of being ≥50% were formulated as [(Odd ≥50%=-5.362+0.087× Pre-test Gender et al probability) and (Odd of ≥50%=-21.61+0.256×Post-test Gender et al probability)].

Conclusion: The tested Jordanian patients' probabilities for having obstructive coronary artery diseases based on the extended and updated version of the Gender predictive model were undervalued after conditioning these patients' probabilities with the corresponding CT angiography-guided coronary obstructiveness percentages.

Keywords: Pre-Test probability; Post-Test probability; Obstructive coronary artery disease; Coronary CT angiography; Prediction models.

JRMS DECE 2024; 31 (3): 10.12816/0062037.

 

INTRODUCTION

In clinical practice, patients with suspected coronary artery disease (CAD) can be often evaluated by using invasive and non-invasive diagnostic tests. Traditionally, an exercise ECG, supported by other functional tests such as stress echocardiography or myocardial perfusion scintigraphy, is usually considered the first line in this context of diagnostic plan. Since each diagnostic procedural test being either invasive or non-invasive, each has  its pros and cons, so choosing the most beneficial diagnostic test for each patient is pivotal. [1-3]

In the recent years, coronary computed tomographic angiography (cCTA) has been increasingly used as a valid, reasonable, valuable, and affordable alternative, especially in patients with low-intermediate pre-test CAD risk probability. cCTA is a relatively safe non-invasive diagnostic procedure with a very high negative predictive value, making it especially valuable for ruling out obstructive CAD in the same session with a smaller procedural complication compared to invasive coronary angiography (ICA). [4-6]

The value of any diagnostic test depends on how the selected procedural findings will change the assessed patient’s pre-test probability, ideally either increasing it to a level that justifies ICA or reducing it to a level where the diagnosis can be excluded. Thus, the proper assessment of the true CAD is extremely beneficial to appropriately allocate patients for the optimal diagnostic procedure with the highest expected clinical benefit. The calculation of pre-test probability (PTP) of disease can facilitate prioritizing higher-risk CAD patients for invasive diagnostic procedures. [7-8]

The Diamond-Forrester PTP calculation model was introduced in 1979 and is based on three clinical parameters that have to be assessed in each patient (patient's gender, age, and type of angina presentation). This original Diamond-Forrester PTP calculation model was statistically updated by Genders in 2011 to improve the prediction of obstructive CAD, especially in women and patients with atypical angina presentation. [9-10] . Our study aimed to assess the change in the theoretical probability of the Genders  predictive model in patients undergoing cCTA as the first-line for patients who presented with clinical concerns for CAD to QAHI’s Cardiology service between  January 2020 to  January 2021.

 

MATERIALS AND METHODS

This study was performed as a pilot retrospective study for patients with stable chest pain and intermediate risk of coronary artery disease and was conducted in patients clinically referred for CT which was performed at the department of Queen Alia Heart Institute. Patients with high-risk CAD or with documented CAD, particularly those with a history of myocardial infarction or revascularization, or coronary stenosis ≥50% on previous invasive coronary angiography were excluded from the study. We also excluded patients being  not in sinus rhythm, or pregnant ladies, or on hemodialysisfrom the study cohort. Finally, all eligible patients with typical, atypical, and non-specific mild-moderate chest pain and who underwent coronary CT angiography were included in this retrospective study. Because of the retrospective nature of our research, a signed consent form was waived.

Based on the CTA result, studied patients were categorized into two primary cohorts; attended patients with typical, atypical, and non-specific mild-moderate chest pain and whose CTA stenosis <50% cohort (Cohort I) and attended patients with typical, atypical, and non-specific mild-moderate chest pain and whose CTA stenosis ≥50% cohort (Cohort II). We performed the statistical analysis by using SPSS version 25. A p-value of ≤ 0. 05 was defined to indicate statistical significance. Tested categorical variables were compared across the two cohorts via the Chi-Square test. Analyzed outcomes were presented as numbers (percentages), Pearson Chi-Square (χ 2) and Likelihood Ratio (G 2), odd ratio (OD), and the p-value.

The pre-test probability of obstructive CAD was determined for each patient using the predictive model of Genders et al., which is the updated and extended version of the previous model of Diamond and Forrester. The probability function, estimated by the logistic regression model used, is expressed as f(z) = e−z / (1 + e−z), in which z .

represents the contribution of each of the variables involved and is equal to −4.37 + 0.04×age (in years) + 1.34 (in men) + 1.9 in the case of typical angina or 0.64 in the case of atypical angina. The typicality of angina presentation was based on the presence of the following three criteria: retrosternal localization of pain, precipitation by exertion, and prompt relief by rest or after nitroglycerin.

The post-test probability was calculated for each patient using Bayes’ theorem, according to the equation: P(A|B)=[P(B|A)×P(A)]/P(B), in which P (A|B) is the post-test probability conditioned by the pre-test probability [P(A)] and P(B) is the probability determined by the test used. To compare the diagnostic accuracy of the pre-test predictive model of Genders et al versus CTA based post-test predictive model of Genders et al, we performed a logistic regression analysis with the outcome (CAD or no CAD) as a dependent variable. The receiver operating characteristic (ROC) curve was thereafter used to evaluate the AUROC and performance indices of pre-test and post-test probabilities of CAD. P-value < 0.05 was considered statistically significant. Confidence intervals of 95% were calculated. The analyses were done using SPSS (SPSS Statistics Version 25 IBM).

 

 

RESULTS

The ROC test was utilized to explore the area under the ROC curves (AUROCs) for the changes in the theoretical probability of the Genders predictive model in patients, who presented with stable chest pain, undergoing cCTA as the first-line at QAHI; the reference pre-test predictive Genders model (red color) versus the post-test probability of the Gender predictive model for the coronary artery obstructiveness percentage ≥ 50% (blue color). The AUROCs were significantly higher for the pre-test Gender predictive model than post-test Gender model with Area±SEM of 0.902±0.006 (95% CI; 0.889-0.915) vs 0.798±0.010 (95% CI; 0.777-0.818), respectively. The Receiver Operating Characteristic (ROC) test illustration was presented in Figure 2.

 

 

 

Table 1. Comparatively studied variables across stenosis grade cohorts who presented with stable chest pain to QAHI’s Cardiology clinics, RMS, Amman, Jordan between 1 Jan 2020 to 1 Jan 2021 

 

Stenosis grade<50%

Cohort I

(1320, 67.97%)

Stenosis grade≥50%

Cohort II

(622, 32.03%)

Total

(N=1942)

χ 2

G 2

Odd ratio

p-Value

Gender

 

Female

590 (44.7%)

66 (10.6%)

656 (33.8%)

219.595

248.181

6.81,95%CI;

(5.16-8.98)

0.000

Male

730 (55.3%)

556 (89.4%)

1286 (66.2%)

Female: Male

1.24: 1

8.42: 1

1.96: 1

Age categorization

 

18-29

123 (9.3%)

2 (0.3%)

125 (6.4%)

310.903

331.807

 

NA

0.000

30-39

337 (25.5%)

70 (11.3%)

407 (21.0%)

40-49

451 (34.2%)

109 (17.5%)

560 (28.8%)

50-59

295 (22.3%)

262 (42.1%)

557 (28.7%)

60-69

76 (5.8%)

139 (22.3%)

215 (11.1%)

>=70

38 (2.9%)

40 (6.4%)

78 (4.0%)

Co-Morbidities

 

<3

1203 (91.1%)

187 (30.1%)

1390 (71.6%)

775.077

767.490

23.92, 95% CI;

(18.52-30.89)

0.000

 >=3

117 (8.9%)

435 (69.9%)

552 (28.4%)

Heart score MACE

 

Low risk

1267 (96.0%)

238 (38.3%)

1505 (77.5%)

807.730

798.606

38.57, 95%CI;

(28.04-53.06)

0.000

Moderate risk

53 (4.0%)

384 (61.7%)

437 (22.5%)

Angina class

 

Typical Chest Pain

409 (31.0%)

607 (97.6%)

1016 (52.3%)

752.826

932.276

NA

0.000

Atypical Chest Pain

468 (35.5%)

15 (2.4%)

483 (24.9%)

Non-Specific Chest Pain

443 (33.6%)

0 (0.0%)

443 (22.8%)

Pre-test stenosis %_Genders et al

 

<50%

996 (75.5%)

61 (9.8%)

1057 (54.4%)

734.567

806.590

28.27, 95% CI; (21.10-37.88)

0.000

 >=50%

324 (24.5%)

561 (90.2%)

885 (45.6%)

Pre-test stenosis grade Genders et al

 

Minimal 1%-24%

664 (50.3%)

0 (0.0%)

664 (34.2%)

856.698

1024.187

 

NA

0.000

Mild 25%-49%

332 (25.2%)

61 (9.8%)

393 (20.2%)

Moderate 50%-69%

222 (16.8%)

200 (32.2%)

422 (21.7%)

Severe >=70

102 (7.7%)

361 (58.0%)

463 (23.8%)

Stenosis grade_CTA

 

Minimal 1%-24%

433 (32.8%)

0 (0.0%)

433 (22.3%)

1942.000

2435.604

NA

0.000

Mild 25%-49%

887 (67.2%)

0 (0.0%)

887 (45.7%)

Moderate 50%-69%

0 (0.0%)

622 (100%)

622 (32.0%)

Genders et al Model vs CTA

 

Under-Estimation

987 (74.8%)

101 (16.2%)

1088 (56.0%)

587.970

620.968

15.29, 95%CI;

(11.95-19.57)

0.000

Over-Estimation

333 (25.2%)

521 (83.8%)

854 (44.0%)

Data results of the comparative variables between the 2 tested cohorts were statistically analyzed by Chi-Square Test (at p-value< 0.05) and expressed as Numbers (Percentage). The strength of associations was also described as odd ratios (OR). The Pearson chi-square statistic (χ 2) involves the squared difference between the observed and the expected frequencies. The Goodness of Fit (G-Test of independence) uses the log of the ratio of two likelihoods and tests the goodness of fit of observed frequencies to their expected.

 

 

 

 

 

 

 

 

 

Fig 1. Bar charts illustrations for the tested variables across the two 2 tested cohorts; Attended patients with stable chest pain and whose coronary CT angiography stenosis was <50% (Cohort I) versus Attended patients with stable chest pain and whose coronary CT angiography stenosis were ≥50% (Cohort II), for patients who presented to Queen Alia Heart Institute’s Cardiology clinics, Royal Medical Services, Amman, Jordan between 1 Jan 2020 to 1 Jan 2021.

 

 

 

Fig 2. The Receiver Operating Characteristic (ROC) test was conducted to explore the area under the ROC curves (AUROCs) for the changes in the theoretical probability of the Genders et al predictive model in patients, who presented with stable chest pain, undergoing cCTA as the first-line at Queen Alia Heart Institute's Cardiology clinics, Royal Medical Services, Amman, Jordan between 1 Jan 2020 to 1 Jan 2021; the reference pre-test predictive Genders et al model (red color) versus the post-test probability of the Gender et al predictive model for the coronary artery obstructiveness percentage ≥ 50% (blue color). The AUROCs were significantly higher for the pre-test Gender et al predictive model than post-test Gender et al model with Area±SEM of 0.902±0.006 (95% CI; 0.889-0.915) vs 0.798±0.010 (95% CI; 0.777-0.818), respectively.

 

 

Table 2. The Binary Logistic Regression analysis for the changes in the theoretical probability (Pre-test vs Post-test) of the Genders et al predictive model in patients, who presented with stable chest pain, undergoing cCTA as the first-line at Queen Alia Heart Institute’s Cardiology clinics, Royal Medical Services, Amman, Jordan between 1 Jan 2020 to 1 Jan 2021.

Tested predictors

B±SEM

Wald

Sig.

Exp

(B)

95% C.I.for EXP(B)

χ2

(df)

 

Variation

range

%Cases

Lower

Upper

oCAD≥50%

e (-5.362+0.087×Pre-test Gender et al)/[1+ e (-5.362+0.087×Pre-test Gender et al)]

Pre-test

0.087±0.004

469.534

0.000

1.091

1.082

1.099

(8)

218.339

41.1%-57.9%

78.4%

 

Constant

-5.362±0.253

449.655

0.000

0.005

 

 

oCAD≥50%

e (-21.61+0.256×Post-test Gender et al)/[1+ e (-21.61+0.256×Post-test Gender et al)]

Post-test

0.256±0.013

359.156

0.000

1.292

1.258

1.326

(8)

621.504

30.1%-42.1%

73.5%

 

Constant

-21.61±1.127

367.530

0.000

0.000

 

 

  • The Binary Logistic Regression analysis was conducted for the 2 tested predictive obstructive coronary artery disease (oCAD) models; The reference tested predictive Genders et al model and our constructed Multiple Logistic Regression (MLgR) predictive models, against oCAD≥50% as documented by the coronary CT angiography.
  • The Binary Logistic Regression Test was conducted to explore the degree of correlations, how much of the total variations in the dependent variable can be explained by the independent variables, and the quality of the prediction of the dependent variable.
 

Table 2. The Binary Logistic Regression analysis for the changes in the theoretical probability (Pre-test vs Post-test) of the Genders et al predictive model in patients, who presented with stable chest pain, undergoing cCTA as the first-line at Queen Alia Heart Institute’s Cardiology clinics, Royal Medical Services, Amman, Jordan between 1 Jan 2020 to 1 Jan 2021.

Tested predictors

B±SEM

Wald

Sig.

Exp

(B)

95% C.I.for EXP(B)

χ2

(df)

 

Variation

range

%Cases

Lower

Upper

oCAD≥50%

e (-5.362+0.087×Pre-test Gender et al)/[1+ e (-5.362+0.087×Pre-test Gender et al)]

Pre-test

0.087±0.004

469.534

0.000

1.091

1.082

1.099

(8)

218.339

41.1%-57.9%

78.4%

 

Constant

-5.362±0.253

449.655

0.000

0.005

 

 

oCAD≥50%

e (-21.61+0.256×Post-test Gender et al)/[1+ e (-21.61+0.256×Post-test Gender et al)]

Post-test

0.256±0.013

359.156

0.000

1.292

1.258

1.326

(8)

621.504

30.1%-42.1%

73.5%

 

Constant

-21.61±1.127

367.530

0.000

0.000

 

 

  • The Binary Logistic Regression analysis was conducted for the 2 tested predictive obstructive coronary artery disease (oCAD) models; The reference tested predictive Genders et al model and our constructed Multiple Logistic Regression (MLgR) predictive models, against oCAD≥50% as documented by the coronary CT angiography.
  • The Binary Logistic Regression Test was conducted to explore the degree of correlations, how much of the total variations in the dependent variable can be explained by the independent variables, and the quality of the prediction of the dependent variable.
 

 

Table 3. The optimal cut-off points, sensitivities, specificities, positive and negative predictive values, likelihoods ratios, and Youden and accuracy indices for the changes in the theoretical probability (Pre-test vs Post-test) of Genders et al predictive model in patients, who presented with stable chest pain, undergoing cCTA as the first-line at Queen Alia Heart Institute's Cardiology clinics, Royal Medical Services, Amman, Jordan between 1 Jan 2020 to 1 Jan 2021.

Prognostic Indicator

Cutoff

TPR

FPR

YI

TNR

PPV

NPV

NLR

PLR

AI

Pre-test

40.855%

99.2%

26.6%

72.61%

73.41%

63.74%

99.49%

1.10%

373.05%

81.67%

Post-test

78.87%

99.2%

32.9%

66.32%

67.12%

58.71%

99.44%

1.20%

301.70%

77.39%

• The area under the receiver operating characteristic (ROC) analysis was constructed for the changes in the theoretical probability of the Genders et al predictive model in patients, who presented with stable chest pain, undergoing cCTA as the first-line at Queen Alia Heart Institute's Cardiology clinics, Royal Medical Services, Amman, Jordan between 1 Jan 2020 to 1 Jan 2021; the reference pre-test predictive Genders et al model versus the post-test probability of the Gender et al predictive model for the coronary artery obstructiveness percentage ≥ 50%; oCAD<50% (0) vs oCAD≥50% (1). The sensitivity analysis was thereafter processed on a total of 1942 processed cases; 622 cases were processed as positive actual state, 1320 cases were processed as negative actual state, and 0 cases were treated as missing data. Higher values of the test result variable(s) indicate stronger evidence for a positive actual state. The positive actual state in our study was set for the oCAD probability was ≥50%.

  • TPR: True positive rate (sensitivity).
  • FPR: False positive rate.
  • YI: Youden index.
  • TNR: True negative ratio (specificity).
  • NLR: Negative likelihood ratio.

 

  • PPV: Positive predictive value.   
  • NPV: Negative predictive value.
  • AI: Accuracy index.
  • PLR: Positive likelihood ratio.
 

 

 

Odd of oCAD≥50%=-5.362+0.087×Gender et al Probability

 

Odd of oCAD≥50%=-21.61+0.256×Post-test Gender et al

 

Fig 3. The Binary Logistic Regression illustrations for the changes in the theoretical probability of Genders et al predictive model in patients, who presented with stable chest pain, undergoing cCTA as the first-line at Queen Alia Heart Institute’s Cardiology clinics, Royal Medical Services, Amman, Jordan between 1 Jan 2020 to 1 Jan 2021. the reference pre-test predictive Genders et al model versus the post-test probability of the Gender et al predictive model for the coronary artery obstructiveness percentage ≥ 50%; oCAD<50% (0) vs oCAD≥50% (1)

 
 
 
 

he Binary Logistic Regression of the two investigated prediction models against the stenosis probability of being ≥50% were formulated as [(Odd of oCAD≥50%=-5.362+0.087× Pre-test Gender probability) and (Odd of oCAD≥50%=-21.61+0.256×Post-test Gender probability)]. The Binary Logistic Regression analysis results and illustrations were fully described in Table 2 and Figure 3.

The sensitivity analysis was thereafter formulated on a total of 1942 processed cases; 622 cases were processed as positive actual state, 1320 cases were processed as negative actual state, and 0 cases were treated as missing data. Higher values of the test result variable(s) indicate stronger evidence for a positive actual state. The positive actual state in our study was set for the oCAD probability was ≥50%. The sensitivity indices’ results for the changes in the theoretical probability (Pre-test vs Post-test) of the Genders predictive model, including sensitivity, youden’s index, specificity, positive and negative predictive values, negative and positive likelihood ratios, and accuracy index, were 99.2%, 72.61%, 73.41%, 63.74%, 99.49%, 1.10%, 373.05%, and 81.67% versus 99.2%, 66.32%, 67.12%, 58.71%, 99.44%, 1.20%, 301.70%, and 77.39%, respectively. The explored optimal operating cutoff points for the 2 investigated predictive models in our study were 40.85% and 78.87%, respectively.

The overall tested gender ratio (male to female ratio) in this study was assigned to 1.96: 1 with significant distributions across Cohort I-II [1.24: 1 and 8.42: 1, respectively]. Regarding the tested patients' age categorization, the age range (40-49) years and (50-59) years had significantly the highest proportional distribution among the other age range categories [560 (28.8%) and 557 (28.7%), respectively] followed by age ranges of (30-39) years and (60-69) years [407 (21.0%) and 215 (11.1%), respectively].

The overall heart score risk was higher for lower risk cohort compared to the moderate risk cohort [1505 (77.5%) vs 437 (22.5%)] with significant distribution across the 2 comparative groups [1267 (96.0%) vs 53 (4.0%) and 238 (38.3%) vs 384 (61.7%), respectively] Also, the overall co-morbidity burden was higher for lower burden (<3 co-morbidity number) compared to higher burden (≥3 co-morbidity number) [1390 (71.6%) vs 552 (28.4%), respectively] with significant distribution across the Cohort I-II [1203 (91.1%) vs 117 (8.9%) and 187 (30.1%) vs 435 (69.9%), respectively].

 

The presented chest pain classes for the attended tested patients were highest for the typical class followed by atypical and lastly by the non-specific chest pain class [1016 (52.3%) vs 483 (24.9%) vs 443 (22.8%), respectively] with significant distributions across the 2 tested comparative groups [409 (31.0%), 468 (35.5%), and 443 (33.6%) vs 607 (97.6%), 15 (2.4%), and 0 (0.0%), respectively]. 

The tested patients’ oCAD risks based on the Gender model had significant distributions across the 2 categorized cohorts [664 (50.3%), 332 (25.2%),222 (16.8%), and 102 (7.7%) vs 0 (0.0%), 61 (9.8%), 200 (32.2%), and 361 (58.0%), respectively] for the estimated obstructiveness of coronary arteries of 1%-24% (Minimal), 25%-49% (Mild), 50%-69% (Moderate), and >=70 (Severe), respectively.

The tested patients had significant stenosis grade distributions across Cohort I-II [433 (32.8%), 887 (67.2%), and 0 (0.0%) vs 0 (0.0%), 0 (0.0%), and 622 (100%), respectively] for which the Gender et al predictive model had an under-estimation propensity of 1088 (56.0%) and an over-estimation propensity of 854 (44.0%) when compared with the referenced oCAD' s diagnostic test in this study (CTA), that were significantly distributed across Cohort I-II [987 (74.8%) and 333 (25.2%) vs 101 (16.2%) and 521 (83.8%), respectively].

Comparably, our proposed MLgR predictive model had a higher under-estimation probability when compared to the comparative referenced Gender model [1330 (68.5%) vs 1088 (56.0%), respectively]. On the other hand, our proposed MLR model had a lower over-estimation probability compared to the Gender model [533 (27.4%) vs 854 (44.0%), respectively]. Interestingly, our proposed model had a matched probability of 4.1% (79 cases) in regards to the CTA reference. The comparatively studied variables and the bar charts’ visualizations for the studied patients across Cohort I-II were summarized in Table 1 and Figure 1.

 

 

 

DISCUSSION

Owing up to maximally 41% of oCAD' suspected patients who underwent elective ICA will have a confirmed oCAD diagnosis, most cardiovascular guidelines, including the European and American guidelines, galvanize the practical mandatory for the initial pre-testing and risk discrimination of the attended patients with stable chest pain before advancing into higher levels of non-invasive and invasive diagnostic procedures. [11-12]

Because decisions on diagnostic testing are usually made before CTA is performed, the clinical usefulness of diagnostic test-based CTA, in particular, the coronary calcium score is limited. As CTA has a high sensitivity/ specificity with a reasonable overall diagnostic performance for significant oCAD, we used CTA results equally as ICA results in this study. If ICA was followed after CTA, the ICA result was selected for the final post-test analysis. If the CTA was performed without ICA following, we consider the CTA results for the post-test estimation of the investigated predictive model of the Gender model. In this study, only 17.2% of the all-eligible investigated patients (327 cases) followed by ICA confirmation procedure after the initial CTA procedure, while 82.8% (1579 cases) underwent only CTA diagnostic procedure. Of note, the CTA results were visually and quantitatively assessed by qualified and trained specialists for CTA and cardiologists. [13-14]

Although the extended and updated version of the D-F predictive model, the Gender model, was originally constructed in a population with a high prevalence of oCAD and who underwent ICA, it was validated in a population with a low prevalence of CAD and who underwent CTA. Depending on these explored predictive models, as recommended by NICE chest pain guidelines, the conclusions of pre-test and/ or post-test calculations may result in up to 2/3 of the triaged patients being excluded from further advanced cardiac investigations. While the Gender predictive model was considered as the extended and updated multiple logistic version of the D-F predictive model, the D-F model was primarily constructed more than 30 years ago on United State populationwhose ages ranged between 30-70 years, which raises the debate regarding its generalizability in the Jordanian cohort whose attended chief complaint is substernal chest pain. In this study, we investigated that the valid Gender predictive model had AUROCs significantly higher for the pre-test Gender predictive model than the post-test Gender model with Area±SEM of 0.902±0.006 (95% CI; 0.889-0.915) vs 0.798±0.010 (95% CI; 0.777-0.818), respectively. [15-16]

To the best of our knowledge, recent data for affected oCAD patients from a large-volume center over 5 years in Portugal revealed that approximately 43% of 1892 patients undergoing diagnostic CTA didn’t have oCAD. And wider use of Bayes theorem-based CTA constructed probabilistic models, when used appropriately, could help in minimizing this practical issue. [17-18] Additionally, the results of our study support the latest and updated clinical guidelines for the stable oCAD diagnosis, especially with the UK National Institute for Health and Clinical Excellence (NICE), which propose diagnostic CTA in patients with a pre-test probability of coronary stenosis of 10–29%, ischemia imaging for those with a pre-test probability of 30–60%, and direct referral for ICA in those with a pre-test probability of >60%. [19-22]

The fact that the CTA procedure has some access restrictions, relatively higher cost-related expenditures, and possible ionizing radiation and iodinated contrast risk to the employees, its clinical utilities, and cost-effectiveness had been a subject of debate. [23-24] It should be borne in mind that the predictive performance of the Gender model was undervalued in our study after the estimated Gender based pre-test probabilities of the tested Jordanian cohort were conditioned by the CTA-guided coronary obstructiveness percentages.

Our study had certain limitations. First, it was essentially dependent on certain assumptions, particularly in terms of the sensitivity (98%) and specificity (85%) of the CTA which was retrieved from the Ollendorf metanalysis. [25] Second, only 17.2% of the all-eligible investigated patients (327 cases) in this study were followed by an ICA confirmation procedure after the initial CTA procedure, it was not possible to assess the true sensitivity and specificity of each of the tests in this population. Third, the sample may not have been representative of all Jordanian population with suspected oCAD, since generalizing conclusions from this study should be considered with caution. Last, it should also be borne in mind that the CTA procedure’s functional and anatomical findings do not necessarily correlate closely with the patients’ attended symptoms. Large-sample, multiinstitutional, and prospective studies are needed to identify the post-test probability after conditioning predictive model pre-test probability by the CTA obstructiveness percentages.

 

CONCLUSIONS

In our study, we demonstrated that the CTA procedure reduced the predictive probability performances of the validated and the guidelines approved pre-test predictive model of oCAD. So, depending on pre-test probability of coronary stenosis between 10%-29% may be the optimal eligibility for pursuing CTA diagnostic procedures.

 

REFERENCES

  1. Taylor AJ, Cerqueira M, Hodgson JM, et al ACCF/ SCCT/ ACR/ AHA/ ASE/ ASNC/ NASCI/ SCAI/ SCMR 2010 appropriate use criteria for cardiac computed tomography. J Cardiovasc Comput Tomogr. 2010;4:407.e1---133.
  2. Genders TS, Steyerberg EW, Alkadhi H, et al. A clinical prediction rule for the diagnosis of coronary artery disease: validation, updating, and extension. Eur Heart J. 2011;32:1316---30.
  3. Diamond GA, Forrester JS. Analysis of probability as an aid in the clinical diagnosis of coronary---artery disease. N Engl J Med. 1979;300:1350---8.
  4. Detrano R, Yiannikas J, Salcedo EE, et al. Bayesian probability analysis: a prospective demonstration of its clinical utility in diagnosing coronary disease. Circulation. 1984;69:541---7.
  5. Detrano R, Guppy KH, Abbassi N, et al. Reliability of Bayesian probability analysis for predicting coronary artery disease in a veterans hospital. J Clin Epidemiol. 1988;41:599---605.
  6. Patel MR, Peterson ED, Dai D, et al. Low diagnostic yield of elective coronary angiography. N Engl J Med. 2010;362:886---95. 10. Cheng VY, Berman DS, Rozanski A, et al. Performance of the traditional age, sex, and angina typicality-based approach for estimating pretest probability of angiographically significant coronary artery disease in patients undergoing coronary computed tomographic angiography: results from the multinational coronary CT angiography evaluation for clinical outcomes: an international multicenter registry (CONFIRM). Circulation. 2011;124:2423---32.
  7. Maffei E, Seitun S, Martini C, et al. CT coronary angiography and exercise ECG in a population with chest pain and low to-intermediate pre-test likelihood of coronary artery disease. Heart. 2010;96:1973---9.
  8. Borges Santos M, Ferreira AM, Monteiro A, et al. Coronariografia para avaliac¸ão de doenc¸a coronária obstrutiva: estamos a selecionar bem os doentes? Rev Port Cardiol. 2011;30 Suppl. I:21. 13. Chow BJ, Abraham A, Wells GA, et al. Diagnostic accuracy and impact of computed tomographic coronary angiography on utilization of invasive coronary angiography. Circ Cardiovasc Imaging. 2009;2:16---23.
  9. Myers J, Prakash M, Froelicher V, et al. Exercise capacity and mortality among men referred for exercise testing. N Engl J Med. 2002;346:793---801.
  10. Chow BJ, Small G, Yam Y, et al. Incremental prognostic value of cardiac computed tomography in coronary artery disease using CONFIRM: COroNary computed tomography angiography evaluation for clinical outcomes: an International Multicenter registry. Circ Cardiovasc Imaging. 2011;4:463---72.
  11. Bamberg F, Sommer WH, Hoffmann V, et al. Meta-analysis and systematic review of the long-term predictive value of assessment of coronary atherosclerosis by contrast-enhanced coronary computed tomography angiography. J Am Coll Cardiol. 2011;57:2426---36. 17. Dedic A, Genders TS, Ferket BS, et al. Stable angina pectoris: a head-to-head comparison of the prognostic value of cardiac CT and exercise testing. Radiology. 2011;261:428---36.
  12. Dewey M, Hamm B. Cost-effectiveness of coronary angiography and calcium scoring using CT and stress MRI for diagnosis of coronary artery disease. Eur Radiol. 2007;17:1301---9. 20. Ko BS, Cameron JD, Defrance T, et al. CT stress myocardial perfusion imaging using multidetector CT --- a review. J Cardiovasc Comput Tomogr. 2011;5:345---56. 21. Bettencourt N, Rocha J, Ferreira N, et al. Incremental value of an integrated adenosine stress-rest MDCT perfusion protocol for detection of obstructive coronary artery disease. J Cardiovasc Comput Tomogr. 2011;5:392---405.
  13. Koo BK, Erglis A, Doh JH, et al. Diagnosis of ischemia causing coronary stenoses by noninvasive fractional flow reserve computed from coronary computed tomographic angiograms. Results from the prospective multicenter DISCOVER-FLOW (Diagnosis of Ischemia-Causing Stenoses Obtained Via Non-invasive Fractional Flow Reserve) study. J Am Coll Cardiol. 2011;58:1989---97.
  14. Sousa PJ, Gonc¸alves PA, Marques H, et al. Radiation in cardiac CT: predictors of higher dose and its reduction over time. Rev Port Cardiol. 2010;29:1655---65.
  15. Earls JP, Berman EL, Urban BA, et al. Prospectively gated transverse coronary CT angiography versus retrospectively gated helical technique: improved image quality and reduced radiation dose. Radiology. 2008;246:742---53.
  16. Gopal A, Mao SS, Karlsberg D, et al. Radiation reduction with prospective ECG-triggering acquisition using 64-multidetector computed tomographic angiography. Int J Cardiovasc Imaging. 2009;25:405---16. 26. Kaufmann PA, Knuuti J. Ionizing radiation risks of cardiac imaging: estimates of the immeasurable. Eur Heart J. 2011;32:269---71.
  17. M. Borges Santos, A.M. Ferreira, A. Monteiro, et al. Coronariografia para avaliação de doença coronária obstrutiva: estamos a selecionar bem os doentes?. Rev Port Cardiol, 30 (2011), pp. 21.
  18. B.J. Chow, A. Abraham, G.A. Wells, et al. Diagnostic accuracy and impact of computed tomographic coronary angiography on utilization of invasive coronary angiography. Circ Cardiovasc Imaging, 2 (2009), pp. 16-23.
  19. A. Dedic, T.S. Genders, B.S. Ferket, et al. Stable angina pectoris: head-to-head comparison of prognostic value of cardiac CT and exercise testing. Radiology, 261 (2011), pp. 428-436.
  20. J.S. Skinner, L. Smeeth, J.M. Kendall, et al. NICE guidance. Chest pain of recent onset: assessment and diagnosis of recent onset chest pain or discomfort of suspected cardiac origin. Heart, 96 (2010), pp. 974-978.
  21. B.J. Chow, M.R. Freeman, J.M. Bowen, et al. Ontario multidetector computed tomographic coronary angiography study: field evaluation of diagnostic accuracy. Arch Intern Med, 171 (2011), pp. 1021-1029.
  22. N.R. Mollet, F. Cademartiri, C. van Mieghem, et al. Adjunctive value of CT coronary angiography in the diagnostic work-up of patients with typical angina pectoris. Eur Heart J, 28 (2007), pp. 1872-1878.
  23. Skinner JS, Smeeth L, Kendall JM, et al. NICE guidance. Chest pain of recent onset: assessment and diagnosis of recent onset chest pain or discomfort of suspected cardiac origin. Heart. 2010;96:974---8. 28. Chow BJ, Freeman MR, Bowen JM, et al. Ontario multidetector computed tomographic coronary angiography study: field evaluation of diagnostic accuracy. Arch Intern Med. 2011;171: 1021---9.
  24. Mollet NR, Cademartiri F, van Mieghem C, et al. Adjunctive value of CT coronary angiography in the diagnostic work-up of patients with typical angina pectoris. Eur Heart J. 2007;28: 1872---8. 30. Shaw LJ, Peterson ED, Shaw LK, et al. Use of a prognostic treadmill score in identifying diagnostic coronary disease subgroups. Circulation. 1998;98:1622---30
  25. .Ollendorf DA, Kuba M, Pearson SD. The diagnostic performance of multi-slice coronary computed tomographic angiography: a systematic review. J Gen Intern Med. 2011;26:307---16.

About
The Journal

The Journal of the Royal Medical Services (JRMS) is an open access journal and it is the official publication for the Royal Medical Services of the Jordanian Armed Forces... Read More

Subscribe to OUR
newsletter

To receive updates on new issues

JRMS Journal

Articles Archive

Archive

Previous Issues

Volume 25
April 2018

Volume 24
December 2017

Volume 24
August 2017

Volume 24
March 2017