Sunday, February 07, 2010

How Technology Is Straining the Doctor-Patient Relationship

Technology is an incredible thing.
Technology is expensive.

Technology saves lives.
Technology can bankrupt.

When there's no technology, are you a "bad" doctor for not following guidelines?
When technology's used, are you a "bad" doctor because the patient has multiple comorbidities and the benefit for the implanted technology is questionable?

It's become the yin and yang of medicine. An inconvenient truth.

Medicine's technology is incredibly expensive, but incredibly valuable.

But if the struggle isn't enough, along comes the press to skew the debate by "raising awareness" with our patients.

Doctor, you need to "Get with the Guidelines." The subtitle with such an industry-sponsored trial and press report should be, "Oh, and business is off."

The journal article at the heart of the Chicago Tribune piece (referenced below) suggests the underpenetrated market of defibrillators (ICDs) was partially caused by three factors:
Adjusted analyses revealed lack of adherence for ICD use most notably with advancing age (odds ratio: 0.87; 95% confidence interval: 0.82 to 0.93 per 10 years), black race (odds ratio: 0.75; 95% confidence interval: 0.60 to 0.94), and lack of insurance (odds ratio: 0.45; 95% confidence interval: 0.26 to 0.78).
But other factors exist, they claim, like geography and available expertise:
Practices in the Northeast U.S. were more likely to adhere to guidelines (P <.001), as were those with a dedicated HF clinic (P = .004) and electrophysiologists on staff (P <.001).
These data are indeed valuable, even for an industry-sponsored trial. But patients should be aware that six of these devices must be implanted to save one life in properly selected populations of patients. Not to say that the cost-effectiveness of this approach hasn't been extensively reviewed, it has. But referring doctors and patients have also been barraged with the problems with these technologies. No doubt the chart reviews in the study cited probably didn't account for the rash of recalls whose influence continues today.

We must also place a jaded eye at the manufacturer's earlier press release about this trial that 35,000 charts had been reviewed, rather than less than half of that (15,381). Small error? Not so much.

Medicine is a complicated, non-linear profession. But as patients continue to shoulder more of their health care bills, doctors are finding themselves in the increasingly difficult position of recommending very expensive life-saving technology that might bankrupt their patients. Unless industry acknowledges that very real price pressures are straining this doctor-patient relationship, there will remain a reluctance to completely "Get With the Guidelines" and implant the technology, even when doing so stands to benefit the doctor.


Ref: Evidence of clinical practice heterogeneity in the use of implantable cardioverter-defibrillators in heart failure and post–myocardial infarction left ventricular dysfunction: Findings from IMPROVE HF. HeartRhythm Dec 2009, 6(12), Pp 1727-1734.


Unknown said...

Great post. The problem with guidelines are that they are just that - guidelines and need to be applied after appropriate consideration of patient factors. Most guidelines are based on trials done on patients with specific conditions and after excluding patients with other diseases. When the evidence from those trials is used in real life patients with multiple problems the situation calls for customization of care.
Another problem is that guidelines thought to be based on rock solid evidence often end up having to change as new evidence comes along, the recent case of aspirin in DM comes to mind.

As EHRs make it easy to automate the process of "Chart reviews" and "Data" becomes easily accessible, we are heading towards very interesting times.

Anonymous said...

Wow. This is just ugly. ICDs are one of the shadiest areas of manufacturer-medical interaction due to the huge sums of money and ever-expanding "indications". How about the guidelines in other countries? I'm guessing they're not quite so ridiculously liberal with "ICD every single person who may have some benefit in survival". But then, in other countries they're a bit more realistic about the role of money in all this.

Jay said...

ICDs sure do get beat up a lot on this blog.

Honestly, I have a hard time understanding this.

--The technology is sound.

--The data showing benefit is high quality and has withstood peer review, as well as the test of time.

--The cost effectiveness is comparable to well accepted standards.

--There has been no evidence to refute the guidelines, nor is there even a plausible theory that devices "don't work."

I agree that ICDs are expensive. Compare this however to "cheap" technologies that get applied to larger populations. Consider mammography in 40-50 year old women. Consider drug eluting stents and Plavix. Consider branded statins. All of these are a greater drain on health care resources with less cost effectiveness.

How about other expensive technologies that don't have legitimate cost effectiveness? Where was the outcry over health care reform advocate Ted Kennedy's controversial surgery for malignant glioblatoma? How about $55,000 for an Avastin course for breast cancer that may or may not offer a couple of months of life in the best of circumstances?

When held up to the dispassionate scrutiny that comparative effectiveness research affords, I think ICD therapy could do quite well.

No doubt, there are "aggressive" or even unethical doctors abusing the technology. I bet, however, that for every one inappropriately implanted patient, there are a bunch never offered the option. Despite all the "evil" direct to consumer marketing and "greedy" doctors, this still remains an underserved population. I don't think this fact is in dispute.

If you had CHF with an EF less than 35% after appropriate treatment, wouldn't you want to have a reasoned conversation with an expert about your options? If ICD therapy's not for you, then you'd be free to decline.

Those who bash ICDs should be prepared to draw a rationing line around other expensive but effective treatments. Are you ready for this?


DrJohnM said...


Very important topic.

ICD use/misuse is a pet peeve of mine.

Had too many words for the comment section so I made it a topic for my longest post to date.

Thanks for getting me started.


Anonymous said...

I agree -- ugly. P4P runs the risk of making us all idiots. Is it appropriate that someone dying of cancer have their blood pressure controlled or cholesterol controlled?

I helped take care of a patient with metastatic cancer who was found to have an EF of 15%. The patients performance status was incredibly good but at the time of the diagnosis was already on 4th line chemotherapy with progressive disease. The patient went to a referral center, and was placed on treatment that gave stable disease for over 8 months. I saw her 6 weeks after she missed two treatments. She had been told by her cardiologist that she had to have an ICD and the tertiary oncologist had said that she had to be off treatment for 3 months before she could get one. Her tumor marker tripled in that time. We placed her back on treatment but two weeks later the patient died of complications of the cancer.

Despite the fact that most would have given her a 6 month survival (based on her heavy pre-treatment), her cardiologist felt compelled to put in an ICD because "the patient was living too long and I can't defend myself if he dies of an arrhythmia." Really? I had discussions with the patient of how that can be a "good" way to die. In the end, hind-sight is 20/20 and one can argue that the rapidity of this cancer's relapse would portend impeding failure of the treatment. Somehow, the whole thing didn't sit well with me.

I'll bet those studies didn't include patients with metastatic cancer....

Jay said...


Per guidelines, ICDs are contraindicated in patient's with expected life expectancy less than one year.

The patient you cite has a cardiologist that seems for all intents and purposes to be a greedy coward.

I and any other ethical electrophysiologist would have refused ICD therapy for this and any similar patient. If fact, if your patient already had an ICD placed prior to the cancer diagnosis, I'd be talking about turning therapies off. We should and do extend similar discretion to many other more "borderline" cases as well. Every ICD I implant is done so only after a measured and thoughtful discussion about the pros and cons of the therapy.

Your example is clearly tragic. Is the compensated post MI patient who dies suddenly when never offered the option of indicated ICD less tragic? The data would suggest that this latter scenario is more common.

Any treatment has the potential for abuse. Guidelines give a framework to counter abuses rather than promote them.

The problem you cite is with the doctor, not the ICD.



Anonymous said...

ICD's are not perfect nor are the physicians who implant or refer for them, or the companies who make them. However that does not explain the utilization disparities shown in this study based on race and gender. If the impact of recalls and physicians not buying into the mortality benefit were the only factor at work you would not expect to see such significant variation based on these demographics.

Lastly, ICD's are not "shady" just because it is a large industry. They are one of the few medical device technologies with proven, incontrovertable evidence for a mortality benefit - compared to every interventional cardiology technology, orthopedics, spinal technology, etc.... Remind me again how many patients need to receive stents to save a life?...

DrWes said...

Jay and Anony 11:41PM-

The point of this post was not that ICD's aren't effective or cost-efficient. The point here is that we are running head-long into the very real consideration of price of that technology. ICD companies are under constant pressure from hospitals to lower prices - I get that. Hospitals still see this technology as a major source of revenue and often double and triple those prices to offset THEIR expenses (and those of the insurers) in other less profitable areas of the hospital. As a result, it is not uncommon for an simple ICD implant to cost over six-digits for an overnight stay in the hospital. Should the treating doctor ignore this fact in discussions with the patient who might be saddled with 20% of that cost?

I recently had a patient comment that he felt so much better after his biV ICD implanted - at least until he got his bill. "That," he said, "nearly killed me."

Jay said...


I get your point, but your blog post did take pains to point out issues like the NNT, device recalls, and the integrity of the data in IMPROVE-HF You have also been a consistent opponent of direct to consumer ICD marketing (which you may know, I have had a hand in helping develop). All of this seems to be an attack on the "value" of ICD therapy in general.

I agree that patients without insurance or co-insurance are going to have to think long and hard about ICD therapy. Most won't be able to afford it and that is an unfortunate reality. This is perhaps analogous to an uninsured or underinsured patient developing cancer or being involved in a bad motor vehicle accident. In general, the sicker you are, the more expensive it is to get medical care.

Doctors need to continue to exert pressure on their hospitals and vendors to lower costs. They should not, however, let this get in the way of delivering high quality care if it is affordable and acceptable to the patient. (i.e Covert Rationing a la DrRich).

I share your concerns about costs and have felt this way my entire career. If you want to see a bigger cost offender, look at the lack of "Courage" in our interventional cardiology colleagues as discussed in today's WSJ:

Thanks for letting me ventilate. Wanna pick this up over a beer in Denver this May??


Anonymous said...

Dr Wes, At some point couldn't Microvolt T-Wave Alternans Be another tool that would help a patient and/or their family make a very expensive choice?

Anonymous said...

MTWA in risk stratification of CDM patients: can it help to better select candidates for ICD implantation? First experience of a single Belgian Center (Belgium Congress of Cardiology 2010)

Conclusion: MTWA ia an accurate NON-INVASIVE test to select ICD patients for primary prevention of arrhythmic events and sudden cardiac death. Even in our small cohort, a normal TWA test clearly identifies patients at low risk who have a good prognosis and are unlikely to benefit from primary prevention ICD implantationin in a long term follow-up...