Wednesday, January 28, 2015

Guest Post: Apples, oranges, and treating mild hypertension

- Stephen Martin, MD

A recent systematic review published in the Annals of Internal Medicine (and discussed by Dr. Lin in a previous post) revisited the potential efficacy of drug treatment for mild hypertension. The authors concluded that this treatment “in the primary preventive setting is likely to reduce the risk for several important adverse health outcomes.” This is an interesting finding, given that a 2012 Cochrane review found no evidence of mortality or morbidity benefits from drug treatment of mild hypertension. What changed?

The Annals authors explain their different conclusions as follows (italics mine): “Besides the inclusion of persons with diabetes or prior antihypertensive treatment, the disparity between the conclusions of this review and the one immediately preceding it is primarily attributable to statistical power. The present review nearly doubled the number of patients, quadrupled the number of cardiovascular events, and provides data on end points not available in the prior meta-analysis.”

Having recently examined this literature for our BMJ piece, “Mild hypertension in people at low risk,” my interpretation is different. In order to gain statistical power, the new review combined high-risk and low-risk patients with mild hypertension. This is equivalent of mixing apples—e.g., higher-risk people with diabetes and prior treatment—with oranges— people without diabetes or prior treatment of hypertension.

To summarize this systematic review a bit differently, I derived the table below from Annals Appendix Table 3 and assigned the sources of patients as being oranges (Cochrane) or apples (BPLTTC Trials):

Cochrane (Oranges)
BPLTTC Trials (Apples)
Difference (Apples/Oranges)
Active Participants
Diabetes (%)
Previous antihypertensive treatment (%)
Total Deaths
(#, events per participant)
77 (0.018)
230 (0.068)
3.78 (0.068/0.018)
Cardiovascular Deaths
96 (0.029)
10 (0.002)
89 (0.026)
13 (0.026/0.002)
Coronary Events
71 (0.016)
114 (0.034)    
2.13 (0.034/0.016)
Heart Failure

* NA = “Not Available,” which occurred for individual trials in the Cochrane and BPLTTC, more so the former (see Appendix Table 3 link above)

Unfortunately, these subtleties were absent from the journal’s communication to the mainstream media, which simply stated: “Blood pressure drugs likely to prevent stroke and death in patients with mild hypertension.”

There is hope yet. The paper buried a worthy point when it concluded that “… estimation of cardiovascular risk may aid prioritization in this patient group.” Rather than combining apples and oranges, we need to treat them as different fruit. Apples may benefit from treatment of mild hypertension while oranges do not. Blending them serves neither fruit well.

For now, I’ll continue to generally avoid drug treatment for my low-risk patients with mild hypertension.


Dr. Martin is a family physician and an Assistant Professor at University of Massachusetts Medical School.

Monday, January 26, 2015

Think twice before taking Tamiflu for seasonal influenza

As the Centers for Disease Control and Prevention (CDC) reported that people who received this season's influenza vaccine were only 23 percent less likely to be diagnosed with influenza than unvaccinated persons, CDC director Tom Frieden was publicly urging high-risk patients and their physicians to use antiviral medications to prevent complications and disease transmission:

People who are sick with flu, if they're very sick in the hospital or if they have underlying, chronic medical conditions, like asthma, diabetes, heart disease, women who are pregnant, children under two and people over the age of 65 - all of these people, if they get flu, should get treated with antiviral drugs. The evidence indicates that it will shorten how long you're sick, might keep you out of the hospital and could even save your life. ... There is some evidence that suggests that taking antiviral medications may reduce the risk that you'll spread the disease to others in your family so it may be helpful for others as well as for yourself.

A previous AFP Community Blog post discussed a 2012 Cochrane review which cast doubt on the ability of the neuraminidase inhibitors oseltamivir (Tamiflu) and zanamivir (Relenza) to prevent influenza transmission, noting that after including data from unpublished, industry-sponsored, randomized controlled trials, the difference was not statistically significant.

Also in that year, a meta-analysis by Dr. Mark Ebell and colleagues concluded: "There is no evidence that oseltamivir reduces the likelihood of hospitalization, pneumonia or the combined outcome of pneumonia, otitis media and sinusitis." The authors noted that two large unpublished trials of oseltamivir in older persons and persons with chronic disease did not even show that the drug reduced the duration of symptoms, suggesting that oseltamivir might be less effective at treating influenza in high-risk groups.

In 2014, after gaining access to the complete clinical study reports from the manufacturers, the Cochrane Collaboration updated its previous review and drew the following conclusions: "Oseltamivir and zanamivir have small, non-specific effects on reducing the time to alleviation of influenza symptoms in adults, but not in asthmatic children." Taking either of the drugs did not reduce hospitalizations or serious complications, but did increase risk of nausea, headaches, and psychiatric symptoms. The CDC responded that their recommendations for antiviral medications "remained unchanged," arguing that observational studies not considered in the Cochrane review supported prescribing neuraminidase inhibitors for high-risk patients with influenza. They provided no rationale for dismissing the results of the randomized trials.

The bottom line for influenza sufferers is that at best, antiviral medications have modest benefits that need to be balanced against potential adverse effects (nausea and vomiting) in deciding whether be treated. Also, existing evidence does not support the CDC's contention that antivirals are more likely to be beneficial in high-risk populations.


This post first appeared in slightly modified form on the AFP Community Blog.

Saturday, January 24, 2015

The future of medicine is low-tech and high-touch

A student told me about his experience at TEDMED, the future-oriented medical conference that bills itself as "a celebration of human achievement and the power of connecting the unconnected in creative ways to change our world in health and medicine." He recounted how one speaker showed off the Remote Presence Virtual + Independent Telemedicine Assistant, which news outlets quickly dubbed the "Robo-Doc." This high-priced gadget is designed to provide remote medical services to patients who wouldn't otherwise be able to see real-life doctors, but my student told me that the presentation didn't talk about that much. Instead, he felt, the speaker's message seemed to be: "Robots are cool, so let's make more of them."

Along similar lines, a blog post on the smartphone physical described how a team led by an enterprising Johns Hopkins University medical student created a virtual "checkup" from a combination of smartphone-powered devices. These devices measure standard physical examination parameters such as body mass index, blood pressure, and visual acuity; and less routine tests such as oxygen saturation, electrocardiography, lung function testing, and carotid artery visualization. In addition to collecting far more data than the traditional checkup, the smartphone physical touts the advantage of using devices that are "smaller and less invasive" - no more "fumbling" to take a patient's blood pressure, for example. One commenter gushed, "Getting a smart phone physical was so fun. I got an EKG and an ultrasound of my carotids in under three minutes." So what's not to like?

I'm hardly a Luddite when it comes to adopting the latest in medical technologies, including electronic health records and smartphone apps. But I think it's worth asking how likely it is that high-tech innovations such as robo-docs and smartphone physicals will actually improve patients' health outcomes. For the latter, the answer is not likely at all. A comprehensive review of the evidence on traditional checkups in adults found that they increase the number of diagnoses per patient, but have no effects on hospitalization, disability, worry, absence from work, morbidity, or mortality. Based on other systematic reviews, the U.S. Preventive Services Task Force recommends against doing EKGs, spirometry, or carotid artery ultrasound in healthy adults because the harms of these screening tests outweigh any benefits. False positive results on that 3-minute EKG and ultrasound may lead to an unstoppable cascade of costly cardiovascular stress testing and invasive coronary or carotid angiograms, which can cause serious adverse effects.

If you really want to see the future of medicine, skip TEDMED and head over to Camden, New Jersey, where a family physician named Jeffrey Brenner showed that providing intensive primary care to patients with the most complex illnesses dramatically improved disease outcomes, quality of life, and health care costs. Then hit the road for Lancaster, Pennsylvania, where a similar program empowers "super-utilizer" patients to take control of their health care by providing them with comprehensive, multi-disciplinary case management and social support. On your way, stop by Health Quality Partners in Doylestown, PA, which has improved outcomes and reduced hospitalizations and costs for the sickest Medicare patients through the revolutionary innovation of - wait for it - regular nurse home visits. Here's what Dr. Brenner told Washington Post Wonkblog columnist Ezra Klein when asked what he thought of the Doylestown program:

"There is a bias in medicine against talking to people and for cutting, scanning and chopping into them. If this was a pill or or a machine with these results it would be front-page news in the Wall Street Journal. If we could get these results for your grandmother, you’d say, ‘Of course I want that.’ But then you’d say, what are the risks? Does she need to have chemotherapy? Does she need to be put in a scanner? Is it a surgery? And you’d say, no, you just have to have a nurse come visit her every week."

Later in the column, Klein reflects:

We’ve been conditioned by “Grey’s Anatomy” and hospital rooms to believe that saving lives is a complicated, heroic business. And it is — after people get very sick. But keeping them from getting very sick doesn’t necessarily require the discovery of new molecules. It requires someone who has a relationship with them to stop by once a week to see how they’re doing. The problem is, it’s hard to make money off it.

Pop culture and perverse financial incentives inherent in fee-for-service payment reinforce a bias for health care services that are high-tech and low-touch. Yes, robots and smartphones can and will play vital roles in the future of medicine. But if we really want sick patients to have the best chance to get better - and healthy patients to avoid getting sick in the first place - then we should do everything in our power to support low-tech and high-touch interventions too.


This post first appeared on Common Sense Family Doctor on May 8, 2013.

Monday, January 19, 2015

Remembering Dr. King

I have a dream that my four little children will one day live in a nation where they will not be judged by the color of their skin, but by the content of their character.

Of all the forms of inequality, injustice in health care is the most shocking and inhuman.

- Martin Luther King, Jr.

Saturday, January 17, 2015

Screening for hepatitis C and lung cancer: proceed with caution

Most of my patients don't know (or if they know, don't particularly care) that I blog, but after practicing for nearly three years at the same Washington, DC practice, I am starting to see a few who specifically chose me to be their doctor because of something I wrote in Common Sense Family Doctor or my health care philosophy of "conservative medicine." It isn't that I am averse to doing screening tests; for example, I still order PSA tests for men and mammograms for women in their 40s, provided that they have been informed of the likely harms as well as the possible benefits. What I can't abide are physicians who bully patients into having these tests, or in the case of prostate cancer screening, order the test without even bothering to tell them.

Speaking of screening tests, my first video commentary for Medscape Family Medicine addressed the key issues that primary care physicians will need to review with patients eligible for CT lung cancer screening. Recognizing that the decision to screen or not screen will be a close call for many, the Centers for Medicare & Medicaid Services plans to pay physicians to have a discussion about pros and cons. Although I am concerned that these discussions may not happen, I hope that my commentary makes a difference. Non-clinicians who don't have access to Medscape can read about the issues in one of my previous blog posts.

This week also saw the publication of the analysis, "Is widespread screening for hepatitis C justified?" that I co-authored in BMJ with Ron Koretz, John Ioannidis, and Jeanne Lenzer. In this paper, we argued that the rush to expand screening to all persons born between 1945 and 1965 (the "Baby Boomer" generation) regardless of health status or risk factors could do more harm than good because 80 percent of people with chronic hepatitis C do not develop liver complications and, therefore, cannot benefit from treatment; and the surrogate marker ("sustained viral response" after 12 weeks) that persuaded the U.S. Food and Drug Administration to approve the fantastically expensive new drugs may correlate imperfectly with the health outcomes that patients really care about:

Given the uncertainty about the validity of the surrogate markers, the lack of evidence regarding clinical outcomes of treatment or of screening strategies, and the adverse events caused by the newer regimens, screening may be premature. ... Given the converging recommendations from major organisations for widespread screening, the pressure on practitioners to adopt this policy is mounting. We have a limited window of opportunity to collect appropriate evidence on whether this is a good idea. Until then, physicians should not be pressured to enforce birth cohort screening strategies out of enthusiasm for new treatments that have not yet been shown to cause long term clinical improvement.

For additional context, or if you prefer listening to reading, Jeanne and I recorded a 15-minute podcast to accompany the article which nicely summarized these arguments.

Saturday, January 10, 2015

Can treating mild hypertension be too much medicine?

As part of a plan to improve our practice's quality of care for patients with high blood pressure, my office's nurse announces at every morning huddle which patients on that day's schedule had a blood pressure measurement of greater than 140/90 at their last visit. Most of these patients have measurements consistent with mild hypertension (systolic blood pressure less than 160 mm Hg and/or diastolic blood pressure less than 100 mm Hg). However, a previous AFP Community Blog post pointed out that the few randomized trials of anti-hypertensive medications in this population did not show improvements in cardiovascular outcomes or mortality.

In fact, some have gone as far to suggest that the identification and treatment of mild hypertension in persons at low cardiovascular risk may be "too much medicine," given the poor accuracy of office blood pressure measurements and that these patients typically receive drugs (and their accompanying side effects) rather than lifestyle change counseling. Last year's JNC 8 Guideline for the Management of High Blood Pressure in Adults added fuel to this controversy by acknowledging that no randomized trial evidence supports any systolic blood pressure treatment threshold in adults younger than 60 years, and recommending that adults 60 years and older receive medication only if their blood pressure is higher than 150/90.

In the first of a series of AFP editorials about overscreening, overdiagnosis, and overtreatment, Drs. Mark Ebell and Jessica Herzstein highlighted several examples of screening that does not improve patient-oriented outcomes: too-short intervals between screening colonoscopies; Pap smears in women after a hysterectomy for benign disease; and mammography in older women with dementia. Does looking for hypertension in adults without cardiovascular risk factors fall into this category as well?

new meta-analysis published in the Annals of Internal Medicine could tip the scales in favor of medicating adults with mild hypertension. Combining individual-patient data from a collaboration of blood pressure treatment trials with that from a previous Cochrane review, Dr. Johan Sundstrom and colleagues concluded that pharmacotherapy for mild hypertension reduced the relative risk of strokes, cardiovascular deaths, and total deaths by 22 to 28 percent after 5 years. (They estimated, however, that absolute risk reductions in contemporary primary care populations would be a modest 0.6 to 1.2 percent, or a number needed to treat of 83 to 167.)

Key take-home points are that the absolute benefits of treating otherwise healthy persons with mild hypertension are relatively small; lifestyle modification should generally precede medication; and blood pressure measurement should be performed and repeated carefully to ensure accurate identification of hypertensive patients. In a draft recommendation statement, the U.S. Preventive Services Task Force recently advised routine home blood pressure monitoring to confirm new hypertension diagnoses, which should hopefully limit overdiagnosis and overtreatment.


This post first appeared on the AFP Community Blog.

Tuesday, January 6, 2015

Why every screening test is a gamble

I'm in Las Vegas for the first time in 15 years to attend the International Consumer Electronics Show, and earlier today participated in a panel of "early-adopter" family physicians discussing our perspectives on consumer (patient) health technology such as apps and wearable health data collection devices. I'm staying at one of Vegas's many combination hotel and casinos, with a layout designed to funnel guests and other visitors through the gaming floor to get practically anywhere. While walking past a row of pulsating slot machines in the lobby this morning, I remembered the title of a terrific New York Times editorial I read a few years ago, "You Have to Gamble On Your Health."

Courtesy of

In this editorial, Dr. H. Gilbert Welch (whom I've lauded before for his work on the subject of overdiagnosis) explained why even though most people who receive screening tests for cancer think that they are playing it safe, every test has tradeoffs. Just as a gambler rarely hits the jackpot in Vegas, a patient who undergoes cancer screening is rarely the lucky one whose life is extended from the test, and much more likely to figuratively lose his or her shirt. Common harms of screening include false positive results, risks associated with subsequent diagnostic procedures, and possible unnecessary treatment (and associated side effects) for "cancer" that looks dangerous under the microscope but is actually destined to never cause health problems.

Courtesy of

The good news is that for a few very well-studied screening tests such as mammography, an informed patient can assess the odds of all of these outcomes and decide whether screening is a better choice for her than no screening. A mammography screening decision aid by Dr. Jill Jin that appeared in last month's JAMA did a great job of illustrating these tradeoffs, and may help to explain why a prominent health journalist recently announced that she had decided to forego mammography because she believed that "the numbers are in my favor."

Yes, every screening test is a gamble, but I give credit to my fellow physicians for providing increasingly sophisticated support for these tough decisions that you'd never get in Vegas.