Dealing with discreteness: making `exact’ confidence intervals for                 proportions, differences of proportions, and odds ratios more exact

Agresti, Alan

doi:10.1191/0962280203sm311ra

Cited by 83 publications

(59 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The result of ICA was used as the reference standard to calculate the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of CTA. The influence of heart rate and calcium score was evaluated using the binominal distribution test to calculate the 95% confidence interval [26]. The McNemar test was performed for both CTA and ICA results.…”

Section: Discussionmentioning

confidence: 99%

Evaluation of CT coronary artery angiography with 320-row detector CT in a high-risk population

Sun

Liu²,

Li³

et al. 2012

BJR

View full text Add to dashboard Cite

Objectives: The aim of this article was to prospectively evaluate the accuracy and radiation dose of 320-detector row dynamic volume CT (DVCT) for the detection of coronary artery disease (CAD) in a high-risk population. Methods: 60 patients with a high risk of CAD underwent DVCT without preceding heart rate control and also underwent invasive coronary angiography (ICA), which served as the standard reference. Results: On a per segment analysis, overall sensitivity was 95.3%, specificity was 97.6%, positive predictive value was 90.6%, negative predictive value was 98.8% and Youden index was 0.93. In both heart rate subgroups, diagnostic accuracy for the assessment of coronary artery stenosis was similar. The accuracy of the subgroup with an Agatston score $100 was lower than that for patients with an Agatston score ,100. However, the difference between DVCT and ICA results was not significant (p50.08). The mean estimated effective dose of CT was 12.5¡9.4 mSv. In those patients with heart rates less than 70 beats per minute (bpm), the mean radiation exposure of DVCT was 5.2¡0.9 mSv. The effective radiation dose was significantly lower than that of ICA (14.1¡5.9 mSv) (p,0.001). When the heart rate was .70 bpm, a significantly higher dose was delivered to patients with DVCT (22.6¡5.2 mSv, p,0.001) than with ICA (15.0¡5.3 mSv, p,0.001). Conclusion: DVCT reliably provides high diagnostic accuracy without heart rate/ rhythm control. However, from a dosimetric point of view, it is recommended that heart rate should be controlled to ,70 bpm to decrease radiation dose. The small diameter of the coronary segments, their complex three-dimensional geometry and their rapid movement throughout the cardiac cycle represent the major challenges for artefact-free coronary CT angiography (CTA). With each scanner generation, motion artefacts re-appear as a major cause of image quality degradation during coronary CTA [1][2][3][4][5][6][7][8][9][10]. Coronary CTA studies of each coronary artery with four-multidetector CT (MDCT) at a gantry rotation time of 500 ms had significantly decreased image quality with increasing mean heart rates [3]. Using 16-MDCT at a gantry rotation time of 420 ms, Hoffmann et al [2] found a significant negative correlation between overall image quality and mean heart rate. Even using 64-section CT, with its gantry rotation speed of 330 ms, elevated and irregular heart beats were found to cause relevant degradation of image quality [1,4,9,11]. Using dual-source CT (DSCT) with an increased temporal resolution of 83 ms, there was no significant correlation between mean heart rate and the overall image quality for any coronary segment or for any individual coronary artery. Nonetheless, irregular heart rates still slightly affect the image quality of non-invasive coronary angiography, even with DSCT [10,12].The 320-detector row dynamic volume CT (DVCT) is characterised by 320 slice detectors with a thickness of 0.5 mm and gantry rotation time of 350 ms. With a wide coverage of 16 cm in the z-axis, the whole hea...

show abstract

Section: Discussionmentioning

confidence: 99%

Evaluation of CT coronary artery angiography with 320-row detector CT in a high-risk population

Sun

Liu²,

Li³

et al. 2012

BJR

View full text Add to dashboard Cite

show abstract

“…Even though these two confidence interval methods are similar to large-sample formulas for means, both the W and WCC confidence intervals behave poorly in terms of zero width intervals and overshoot (Beal, 1987;Vollset, 1993;Newcombe, 1998;Pires, 2002;Rieczigel, 2003;Agresti, 2003). For instance, when r=0 or n, W and WCC have zero width or degenerate confidence intervals.…”

Section: Methodsmentioning

confidence: 99%

Better Binomial Confidence Intervals

Reed

2007

J. Mod. App. Stat. Meth.

View full text Add to dashboard Cite

The construction of a confidence interval for a binomial parameter is a basic analysis in statistical inference. Most introductory statistics textbook authors present the binomial confidence interval based on the asymptotic normality of the sample proportion and estimating the standard error -the Wald method. For the one sample binomial confidence interval the Clopper-Pearson exact method has been regarded as definitive as it eliminates both overshoot and zero width intervals. The Clopper-Pearson exact method is the most conservative and is unquestionably a better alternative to the Wald method. Other viable alternatives include Wilson's Score, the Agresti-Coull method, and the Borkowf SAIFS-z.

show abstract

“…When K=1 and  1 =1, the objective is to make inferences about one proportion (as in Agresti and Coull, 1998). When K=2, there may be several objectives: the difference between the two proportions if  1 =1 and  2 =+1 (as in Agresti and Caffo, 2000); the sum of two proportions if  1 =+1 and  2 =+1 (as in Pham-Gia and Turkkan, 1994); the ratio  of two proportions if  1 = and  2 =+1 (as in Agresti, 2003); or a linear combination of two proportions with  1 <0 (as in Phillips, 2003). Cases with K>2 are historically rather less frequent, although in recent years they have received more and more attention due to their great practical interest (Newcombe, 2001;Price and Bonett, 2004;Schaarschmidt et al 2008;Tebbs and Roths, 2008;Agresti et al, 2008;Zou et al, 2009 andMartín et al, 2010).…”

Section: Introductionmentioning

confidence: 99%

The optimal method to make inferences about a linear combination of proportions

Andrés

Tejedor

Hernández

2012

Journal of Statistical Computation and Simulation

View full text Add to dashboard Cite

Asymptotic inferences about a linear combination of K independent binomial proportions are very frequent in applied research. Nevertheless, until quite recently research had been focused almost exclusively on cases of K2 (particularly on cases of one proportion and the difference of two proportions). This article focuses on cases of K>2, which have recently begun to receive more attention due to their great practical interest.In order to make this inference, there are several procedures which have not been (which is a generalization of the Newcombe method). The article describes a new procedure (P0) based on the classic Peskun method, modifies the previous methods giving them continuity correction (methods S0c, W3c, N0c and P0c respectively) and, finally, a simulation is made to compare the eight aforementioned procedures (which are selected from a total of 32 possible methods). The conclusion reached is that S0c method is the best, although for very small samples (n i  10, i) the W3 method is better. The P0 method would be the optimal method if one needs a method which is almost never too liberal, but this entails using a method which is too conservative and which provides excessively wide confidence intervals. The W3 and P0 methods have the additional advantage of being very easy to apply.A free programme which allows the application of the S0 and S0c methods (which are the most complex) can be obtained at

show abstract

Dealing with discreteness: making `exact’ confidence intervals for proportions, differences of proportions, and odds ratios more exact

Cited by 83 publications

References 29 publications

Evaluation of CT coronary artery angiography with 320-row detector CT in a high-risk population

Evaluation of CT coronary artery angiography with 320-row detector CT in a high-risk population

Better Binomial Confidence Intervals

The optimal method to make inferences about a linear combination of proportions

Contact Info

Product

Resources

About