Small but Quantifiable Patient Preference for MRA Versus Catheter Angiography
To the Editor:
In our previous article1 we used decision analysis to examine the cost-effectiveness of various test strategies for both screening and diagnosing carotid artery disease. While our conclusions for screening were contingent on several factors (ie, prevalence of significant stenosis, benefits and risk of surgery, and quality of life with stroke), for presurgical patients the best test strategy was clear. 3D time-of-flight (TOF) magnetic resonance angiography (MRA) is more cost-effective than catheter angiography (CA), as long as the MRA sensitivity and specificity are at least 86% and 90%, respectively. Missing from this analysis, however, was data on patient preference for CA or MRA. Swan and Langlotz2 point out that a patient’s attitude about an intervention is an essential component in cost-effectiveness models. Therefore, we undertook a patient preference study of CA versus MRA in presurgical patients and report our results here.
Sixty-nine patients were recruited prospectively and randomized to first undergo either an MRA or CA examination. The second study was performed on the same day and followed by a patient questionnaire. The questionnaire was modeled after one developed by Mushlin et al3 in a study of multiple sclerosis patients. We first asked the patients which modality they would prefer if one had to be repeated. We then asked them to give the number of days with a severe headache that they would be willing to endure to avoid the less-desirable modality. Finally, the patient used a visual analog scale (VAS) to locate a “severe headache” on a scale from 0 (representing perfect health) to 10 (representing death). From these data we estimated the number of quality-adjusted life-years (QALYs) lost when a patient undergoes the less-desirable modality. QALYs were estimated as the product of the number of days (converted into years) of a severe headache and the severity factor from the VAS divided by 10.
For the 69 patients who gave consent and were randomized, there were technical problems on 3 MRA examinations, and 4 patients refused the MRA. These 4 patients did not appear to be claustrophobic, but rather refused MRA because it was purely for research purposes. Sixty-two patients successfully underwent both examinations and completed the questionnaire. Ten patients (16%) indicated no preference for MRA or CA. Of the remaining 52 patients, there was a statistically significant preference for MRA (P<0.0001): 46 (88%) preferred MRA over CA, and 6 (12%) preferred CA over MRA. There was no difference in average age, gender, or symptomology between those who preferred MRA versus those who preferred CA.
Five patients had complications on CA (2 had small hematomas, 1 a moderate hematoma with numbness on the right upper extremity, 1 a pseudoaneurysm, and 1 supraventricular tachycardia/chest pain that required overnight observation). Four of these patients preferred MRA; 1 patient with a small hematoma preferred CA.
Although most patients preferred MRA over CA, the disutility of CA was small. The average QALYs that a patient was willing to give up to avoid a CA was 0.0024 (95% CI 0.0014–0.0034), equivalent to <1 quality-adjusted day. The minimum was −0.0041 (1.5 days willing to be given up to avoid an MRA) and maximum 0.0201 (1 week to avoid a CA). When we factor into our decision analysis model an average reduction in QALYs of 0.0024 when a CA is performed instead of MRA, there is no effect on our overall conclusions.
3D-TOF-MRA is a cost-effective alternative to CA for presurgical evaluation because of its good accuracy, noninvasiveness, and modest price. It is the modality preferred by the majority of patients, though the magnitude of the preference is small.
This work was supported in part by National Institutes of Health grant 1RO2HL 43812-01A1.
- Copyright © 1999 by American Heart Association