Medical decision-making competence

This week’s Lancet contains a study by Raymont et al investigating the prevalence of mental incapacity in medical inpatients and associated risk factors (see also this item in Shrinkette’s blog ).

In order to judge medical decision-making competence, the authors used two tools: the MacArthur competence assessment tool for treatment (MacCAT-T), and clinical vignettes (based on something called the “Thinking Rationally About Treatment” research method).

After excluding 143 patients for severe cognitive impairment, altered level of consciousness and refusal to participate, 159 patients were interviewed. Of these, 50 (31%) were deemed lacking in capacity to make medical decisions.

Lack of capacity was correlated with increasing age and with a poor score on the Mini Mental Status Exam. The median MMSE score was 29 (out of a maximum of 30) in patients with capacity and 22 in those without capacity.

Significantly, when the clinical team caring for the patients and close relatives were asked to judge the patients’ capacity, those with capacity were correctly assessed almost 100% of the time, but those deemed lacking in capacity by the authors were judged incompetent only 25% of the time by the physicians and similarly by the relatives.

The conclusion to the abstract: “Mental incapacity is common in acutely ill medical inpatients, and clinicians tend not to recognise it. Screening methods for cognitive impairment could be useful in detecting those with doubtful capacity to consent”.
 
 
Are we really underestimating patients’ decision-making capacity as badly as this article suggests? Possibly, but I have a few reservations about the study.

The MacCAT-T tool used to assess medical decision-making capacity is a relatively new one. How valid is it? If three-quarters of physicians caring for the patients and three quarters of their relatives judged the patients who “failed” the MacCAT-T assessment to be competent, it makes me wonder as much about the validity of the tool as about the physicians’ and relatives’ assessments.

According to the authors:

MacCAT-T is a semi-structured interview that measures: (1) understanding of the disorder and its treatment, including associated benefits and risks; (2) appreciation of the disorder and its treatment–ie, how the patient understands they could be specifically affected, which usually entails some level of insight; (3) reasoning, which assesses the processes behind the decision and ability to compare alternatives in view of their consequences; and (4) the ability to express a choice.

I am no expert in such assessment tools, but it seems to me that this sort of assessment might bias, for example, against people who have crackpot medical ideas. Are they necessarily incompetent?

Since the assessment of medical competence correlated well with the MMSE assessment in this study, wouldn’t we be better off relying on the MMSE itself, which is not limited to medical decision making capacity? The determination of a patient’s specific competence to make medical decisions seems somehow more paternalistic and prone to bias than a more general assessment of mental status.

Endovascular repair of abdominal aortic aneurysms

An article in the October 14, 2004 New England Journal of Medicine reports on the results of the Dutch Randomized Endovascular Aneurysm Management (DREAM) trial, which compares endovascular abdominal aortic aneurysm repair with open surgical repair. 345 patients with AAA of at least 5 cm in diameter and who were eligible for both endovascular and open approaches were randomized. In the short term, the endovascular approach yielded lower peri-operative mortality (1.2% vs. 4.6%) and lower peri-operative mortality plus severe complications (4.7% vs 9.8%). The numbers did not quite reach statistical significance, but are convincing and in line with a second, similar, recently published study from England.

An important note of caution is sounded, however, in an editorial by F. Lederle. He does not dispute the short term results, but suggests that the long term results remain in question, and require longer follow-up of these trials. He notes that two European registries have reported a higher annual failure rate (3% vs. 0.3%) for endovascular repair compared to surgery. Lederle also mentions an article about the problem of late graft failure that was initially accepted for publication by the Journal of Vascular Surgery and then suppressed after industry pressure.

An editorial describing this matter in the August, 2004 Journal of Vascular Surgery is very enlightening. It describes how the article in question, based on data on the Medtronic AVE Aneurx stent graft system and written by four authors from the FDA, was submitted to the Journal of Vascular Surgery in December, 2003 and accepted for publication in March, 2004. This article was based on data that the FDA had made public on December 17, 2003, in a Public Health Notification, concerning aneurysm-related mortality rates.

According to the editorial, in May, 2004, Medtronic threatened legal action if the article was published because it supposedly revealed confidential data; the FDA then requested that the article be withdrawn because “the conclusions drawn in the article went beyond the information provided in the Public Health Notification and, therefore, did not reflect the FDA’s current position regarding AneuRx related mortality.”

The authors of the Journal of Vascular Surgery editorial clearly do not believe that the confidentiality issues invoked by Medtronic were valid, and believe that the article’s conclusions are within the bounds of reasonable scientific discussion. They encourage readers to make themselves aware of the data that is available at the FDA website. There are some more comments on this issue in the Letters to the Editor of the following month’s (September) Journal of Vascular Surgery.

This controversy was also publicized in a long front-page article in the July 9, Wall Street Journal, which is not freely available online.

I would encourage readers interested in endovascular approaches to abdominal aortic aneurysms to take a look at the above references.