Please note that the text below is not a full transcript and has not been copyedited. For more insight and commentary on these stories, subscribe to the This Week in Cardiology podcast, download the Medscape app or subscribe on Apple Podcasts, Spotify, or your preferred podcast provider. This podcast is intended for healthcare professionals only.
In This Week’s Podcast
For the week ending July 26, 2024, John Mandrola, MD, comments on the following news and features stories: Health and income, high blood pressure (BP) in the hospital, more on subclinical atrial fibrillation (AF) and when to use anticoagulants (ACs).
Health and Income
It has become fashionable these days to use the phrase “social determinants of health.” That is, people with challenging social situations have worse health outcomes than those with better social situations.
For instance, when I visit Oslo or Copenhagen, it is clear to me that the gaps between the poor and rich are smaller than in US cities. I have spent time volunteering in a city clinic here in Louisville and it shocks me how little these patients gain from an electrophysiologist (EP). An EP cannot help them when their problem is having enough money or shelter or family support.
There is certainty in knowing that poverty and income are important factors in determining health. How best to help people in this situation is very much uncertain.
There are two big approaches to uncertainty. The first approach is to let smart people decree a policy, say school closures during the pandemic, or randomly assign patients to different policies and measure outcomes.
This week, two major randomized controlled trials (RCTs) were published regarding approaches to poverty and health.
JAMA published the first one, “Effect of Cash Benefits on Health Care Utilization and Health: A Randomized Study,” and Yale Nephrologist and Medscape contributor F, Perry Wilson covered it.
First author Sumit Agarwal, along with his colleagues at Harvard, led a clever RCT in which 2800 people participated. This was in a town in Massachusetts. People in a low-income community were randomly assigned via a lottery to receive a cash benefit of $400 per month for 9 months. One group of about 1700 people got the benefit and 1100 pts did not.
The primary endpoint was emergency department (ED) visits.
The mean age was 45 years. More than 70% were first-language Spanish. The mean family income was just $16,000 per year. Most were women.
The main outcome was positive. ED visits in the cash benefit group were 217 per 1000 person years vs 317 per 1000 person years in the no cash benefit group. This met statistical significance, though barely. This also led to fewer hospital admissions.
There were no differences in out-patient visits, visits to primary care, or out-patient behavioral health. Visits to specialty care were slightly higher in the active arm.
Of importance, there were no differences in health outcome measures, such as COVID-19 vaccine, BP, body mass index, HbA1c, or cholesterol level.
The authors concluded:
Study results suggest that policies that seek to alleviate poverty by providing income support may have important benefits for health and access to care.
Dr Wilson, my colleague at Medscape, also lauded the results:
The truth is, you probably have causality in both directions. Making people healthier will make them less poor. And, as this study suggests, making them less poor may make them healthier.
Before I say anything, let me tell you about the second study — which is much more robust, albeit it’s not published in a medical journal. It’s published by an economics realm, the National Bureau of Economic Research (NBER).
The title is, “Does Income Affect Health? Evidence from a Randomized Controlled Trial of a Guaranteed Income,” first author Sarah Miller (University of Michigan Business School). The study was called the Open Research Unconditional Income Study (ORUS).
This is an NBER working paper. The first thing to say about economic reports of RCTs is that they don’t have word or reference limits, and the thing is massive. It’s the size of an entire issue of the New England Journal of Medicine (NEJM).
The basic goal was to make a causal estimate between giving $1000 per month vs $50 per month to low-income adults in two US cities.
The randomization was 1:2 so it was 1000 vs 2000 adults. The study lasted 3 years. Right there, note the differences from the Boston experiment, which was just $400 and just 9 months.
The main findings were:
The cash transfer resulted in large but short-lived improvements in stress and food security;
Greater use of hospital and ED care; and
Increased medical spending by about $20 per month in the treatment relative to the control group.
These findings suggested that the use of other office-based care — particularly dental care — may have increased as a result of the transfer of money.
But.
They found no effect of the transfer across several measures of physical health as captured by multiple well-validated survey measures and biomarkers derived from blood draws.
We can rule out even very small improvements in physical health and the effect that would be implied by the cross-sectional correlation between income and health lies well outside our confidence intervals.
We also find that the transfer did not improve mental health after the first year and by year 2 we can again reject very small improvements.
We also find precise null effects on self-reported access to health care, physical activity, sleep, and several other measures related to preventive care and health behaviors.
They concluded that more targeted interventions may be more effective at reducing health inequality between high- and low- income people.
Comments. I have written on this topic before. In 2019, I reviewed an interesting book co-authored by the George Mason economist Robin Hansen, The Elephant in the Brain .
Hansen has long held that empirical data don’t really support the notion that more healthcare or more healthcare access leads to better health. It’s a provocative concept. I am drawn to it, because it requires us (the medical profession) to be humble about what we contribute to society.
In the March 17, 2023, podcast I discussed many of the empirical studies suggesting that providing more healthcare to people does not lead to better health. There were the RAND, Oregon, and India health insurance studies, which were all (essentially) null for providing more health insurance. There have also been two trials looking at eliminating co-payments or even free meds for cardiovascular (CV) health. MI-FREEE, published in NEJM, famously found that free statins, beta-blockers, angiotensin-converting–enzyme inhibitors, or angiotensin-receptor blockers did not change outcomes.
Rishi Wadhera’s group at Harvard Medical School, even found that Medicaid expansion led to no significant differences in the treatment of CV risk factors.
My take-home therefore is that cracking the problem of inequalities in health in our population is super-complex. Direct cash transfers did not seem to do much. Even substantial boosts of income in the case of the ORUS study had no significant effects.
The previous RAND, Oregon, India studies, and the slew of cardiac studies like MI-FREEE paint a humbling picture. That is, those of us treating patients have the chance to improve the health of the person we are seeing.
But it is a much different matter when it comes to improving the health of populations. This is a far more complex issue. And the empirical data suggest (surprisingly) that it surely requires more than just providing access to care.
This is one of the reasons I have come to believe that the best policy solutions would be to advocate not for one or other policy but a culture of randomization.
Elevated BP in the Hospital
The vast majority of topics covered on #TWIC podcast stem from studies. This week, I want to speak on the problem of elevated BP in the hospital.
JAMA-Internal Medicine has published an elegant and short clinical insights paper from Drs. Zachary Jacobs and Timothy Anderson.
I am approaching three decades working in hospitals, and little has changed around the folly of treating in-hospital asymptomatic elevated BP. I still see patients who have been harmed by overzealous therapy of asymptomatic high BP readings.
The process goes something like this: A nurse notifies an on-call person about an elevated BP. The on-call person hardly knows the patient and is likely covering a hundred or more patients that night. The BP is high, like 180/110 mmHg. So, a drug is given. Many times nothing bad happens. But sometimes that person gets BP, which may cause a fall, or acute kidney injury. Or even a stroke.
Drs Jacob and Anderson review the evidence, which is mostly observational but mostly suggests caution rather than aggression in treating elevated readings.
They offer a very simple six-step plan for approaching the patient with an elevated BP in the hospital
First: Assess for end-organ damage. If you see it, then treat it as a hypertensive-emergency in the ICU. If you don’t move to step 2.
Second: Ensure appropriate BP technique, like the right cuff size. Then remeasure after a period of rest.
Third: Identify and treat factors that may exacerbate BP, such as pain, volume excess, alcohol withdrawal, or drug intoxication. Pain is the most common in my experience.
Fourth: Review home and hospital medicines that could worsen BP. Consider stopping these; things such as nonsteroidal anti-inflammatory drugs, decongestants, stimulants, and steroids.
Fifth: If, after these, the BP is still elevated, consider patient-specific factors before deciding to intensify home BP therapy.
Factors that argue against intensifying therapy are good baseline control of BP at home, frailty, limited life expectancy, or cognitive impairment. Most patients do not want med intensification in the hospital.
Consider medication increase when there is patient readiness to make a change. And, in selected high-risk patients, you can initiate a guideline-concordant BP meds.
Step 6 is important. It is to develop a transitional care plan. Whether the choice is watch and not add meds, or add meds, it is crucial that this patient gets proper follow-up and instruction on home blood pressure.
Comments. I love this article and the extremely medically conservative approach. This practice would save many people from potential iatrogenic harm. I would propose that it not only be taught in residencies and fellowships, but also in nursing schools, because some of the tension in treating numbers not symptoms stems from worry on the part of nurses.
Kudos to JAMA-IM and the authors. Please help me spread the word on the give-peace-a-chance to inpatient asymptomatic high BP readings.
Short-duration Subclinical AF
Once again, we are back to the matter of subclinical AF. It’s almost an everyday decision for me; 3 hours of AF, 6 hours of AF, 30 minutes of AF post-stroke.
ARTESiA investigators have published this week an important post-hoc analysis of their trial of apixaban vs aspirin (ASA) for device detected AF (DDAF) based on CHADSVASC score.
Before I tell you about this study, which the authors make strong conclusions from, we need to set out the results of ARTESiA main trial, as well as the fact that there is another similar trial in this space, called NOAH, which randomly assigned with DDAF to edoxaban or placebo.
First the main results of ARTESIA – 4000 patients with DDAF — which was stopped early for slow enrollment and lower than expected stroke rates.
Apixaban reduced the rate of the primary outcome, stroke or systemic embolism, by 37%; hazard ratio (HR): 0.63; 95% confidence interval (CI): 0.45-0.88, at a cost of a 36% increased incidence of major bleeding (1.53% vs 1.12% per year; HR: 1.36; 95% CI: 1.01-1.82).
The absolute stroke reduction was tiny, 0.78% vs 1.24% per year. Note that the stroke rate in the control arm was 1.2% which is way less than the 2.2% rate we like to associate with a CHADSVASC of 2.
The main results of NOAH were similar, with 2500 patients.
Edoxaban reduced the primary endpoint of stroke, systemic embolism, and CV death.
The trial was stopped early for increased bleeding, and the 19% reduction in the composite did not reach statistical significance.
The bleeding rates were 31% higher in the edoxaban arm.
Also true was that absolute stroke rates were low in the placebo arm at 1.1% per 100 patient years.
In sum, just with the main studies, it’s not clear whether we should treat DDAF of short duration. Yes, you get some stroke reduction, it’s small in absolute terms, but you also get increased bleeding, also a small increase in risk in absolute terms. Some of the ARTESiA investigators have said, rightly, that stroke is worse than bleeding and we should lean towards AC. But that is definitely just an opinion.
The question is now whether sub-studies can provide better answers.
The Journal of the American College of Cardiology published the ARTESiA CHADSVASC substudy. I’ve covered it as an abstract, but let’s re-review because it’s such an everyday decision.
They broke the subgroup into 3 groups: CHADSVASC < 4, = 4, and >4.
They had about 1000 to 1500 patients in each group.
The main finding:
The only group that sustained a significantly lower HR (0.44) for apixaban over ASA was the CHADSVASC > 4 arm; 18 vs 38 strokes. This was a nearly 4% absolute risk reduction.
They then looked at bleeding in the three CHADSVASC groups. They found no significant difference. Apixaban vs ASA had basically the same increase: 27% vs 48% higher.
Apixaban therefore seemed to have a heterogeneous treatment effect for stroke reduction but not bleeding.
The authors concluded:
One in four patients in ARTESiA with subclinical atrial fibrillation had a CHADSVASC score > 4 and a stroke/systemic embolism risk of 2.2% per year. For these patients, the benefits of treatment with apixaban in preventing stroke/systemic embolism are greater than the risks. The opposite is true for patients with CHADSVASC score < 4. A substantial intermediate group (CHADSVASC = 4) exists in which patient preferences will inform treatment decisions.
Comments. This makes it seem easy. Just anticoagulate those with CHADSVASC > 4 who have DDAF. But I don’t think it is that easy.
First is always that this is a subgroup analysis from a main trial that had essentially null results in that stroke reduction was similar to major bleeding increase. Second, the P-for interaction of the subgroup was not signficiant.
Third, there is other data to consider.
Two months ago, the European Heart Journal published a rapid communication from the NOAH authors looking at edoxaban vs placebo based on age, sex, CV comorbidities, and kidney function.
They made two groups CHADSVASC ≤ 4 and CHADSVASC > 4. Recall that NOAH had 2500 patients vs 4000 in ARTESiA.
In the CHADSVASC > 4 group, the HR for the primary endpoint of stroke, systemic embolism, CV death of edoxaban vs placebo was 0.88 (CI 0.55-1.41). So, not significant. And even in this high-risk group, rates of stroke were low at 1.2 per 100 patient years.
In the total population, efficacy and safety outcome rates increased with increasing CHADSVASC scores without a treatment interaction
Another factor influencing the AC decision besides CHADSVASC is the duration of the episode.
ARTESIA investigator Bill McIntyre presented an analysis at European Heart Rhythm Association meeting (I have not seen the publication yet) on stroke risk according to baseline duration and frequency of AF episodes.
It was a surprising analysis.
The absolute risk of stroke was about 1% regardless of whether the episode was less than hour, 1 to 6 hours or > 6 hours. Seriously. It did not matter.
Apixaban (vs ASA) did not show any added benefit for the longer episodes.
Apixaban treatment vs ASA was not affected by the frequency of episodes.
Comments. So, despite the ARTESiA CHADSVASC paper, which suggests higher risk patients with DDAF should be treated, we have against that declaration
It’s a subgroup of an essentially non-significant trial.
The P value for the interaction is non-significant.
It’s not confirmed in the NOAH trial.
Further analysis of ARTESiA shows that duration or frequency of the episode has no bearing on decision.
The bottom line is that despite more than 6500 patients randomized, we really don’t know what to do with DDAF.
I believe that is because of something that Soren Diedrichsen and I have talked about and written about: device-detected short duration often asymptomatic AF is a different entity than the AF we diagnosed before we had these devices. It’s more ubiquitous, it’s lower risk. And treatment is not going to be the same.
What is the best approach? We don’t yet know, but NOAH and ARTESiA and their substudies clearly show that we need new thinking.
© 2024 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: Jul 26, 2024 This Week in Cardiology Podcast - Medscape - Jul 26, 2024.
Comments