The Administration closes the National Guideline Clearinghouse: Bad for our health!

The Department of Health and Human Services (HHS) has closed the National Guideline Clearinghouse (NGC), which was housed in its Agency for Health Research and Quality (AHRQ), by pulling its funding. If one goes to the NGC website, it actually says that there is no longer funding for it. The NGC provided a major resource for consensus information, and closing it is a tremendous loss, not only for the physicians and other medical providers who depended upon its recommendations, but for the health of the American people. It is also, sadly, a loss for the patients of physicians and other medical providers who did not heed, and even opposed the existence of these guidelines. This is because  they will now have the cover that comes from the absence of the NGC when they do things that are not supported by the best evidence available.

Wait, wouldn’t my doctor want to use the best evidence available? Can’t I depend on her/him doing this? Do I have to become a medical expert?

You – we would all – hope that our doctors would use the best evidence available. This is not necessarily the newest stuff, for the obvious reason that it is new, and thus less tested. Frequently, when a test or intervention is new, it looks good, but later, when more people have been exposed to it and more data are available, negative information may emerge. (Sometimes this is because serious side effects may appear to a new drug because so many more people are using it; sometimes it is the result of straight-up fraud, as in the case of Theranos – see The Political is Personal: Corporate power, social isolation, and the health of the nation -- Part 2, April 22, 2018.) Conversely, the oldest tests and treatments are not necessarily the best either – a doctor who does not keep up after his or her training is also going to be way behind.

And guidelines are, of course, guidelines, not laws – they are meant to guide the practitioner to help make the best decisions by summarizing and presenting the evidence, for the whole population and for significant “sub”-populations. For example, older people may have a greater benefit from – or be at greater risk of harm from – a certain treatment than younger people. This could be from a different physiology (kidneys, for example, don’t function as well at cleaning out poisons in older people) or from a different risk/benefit profile (perhaps a bad outcome usually takes longer to appear than the life expectancy of a person; a 20-year lag time for something bad has a different meaning for a 50 year old than an 80 year old).

Not all guidelines have the same strength of evidence behind them. Sometimes the evidence is very strong, coming from multiple randomized controlled trials (RCTs) that are consistent with each other; at the other extreme, they can come from the opinions of a group of experts in the field who are gathered for the purpose of creating a guideline. However, in any case, the guidelines presented should be those for which the strongest evidence exists. The website for NGC still has available the inclusion criteria for the guidelines that used to exist there. The physician is always able to do something different if the recommendation is for some (good) reason not appropriate for a particular patient. But s/he should know, or be able to find – on the NGC, say – what the guidelines are for a particular course of action (test, treatment, etc.). Until just recently, they could. Not now. Which raises the question: why is it gone?

I, of course, do not know the answer, so am forced to speculate. There is some cost to maintaining the guidelines and doing the research, and many people do not like cost, but so much more is being spent on arguably less important things (if one considers the health of the people important; I do). Some people, versions of libertarians, do not like anything that even suggests government mandates, but I know few libertarians who advocate suppressing the truth. Some people are not surprised at any actions taken by the current administration and its executive departments, particularly those that seem to take on science, as has happened with climate change and the use of fossil fuels, but those seem, ultimately, to be based on money and the opportunity for profit. Thus, destroying pristine areas for fossil fuel exploration (the Arctic, the tar sands) and transport (the Keystone pipeline), as well as of the world’s climate from burning them, makes people money. Rich and powerful people.

So maybe that is where we should look for the closure of NGC. The fact is that if there are guidelines there is a rebuttable presumption that, barring differences in the individual patient that are relevant, they should be followed. If the evidence shows that a particular diagnostic test or a treatment (drug, device, etc.) is usually better, it is going to hurt the pocketbooks of the manufacturers of alternative drugs or devices or tests. And it can also hurt the pocketbooks of actual doctors if they make their money doing something that is no more effective than, or even less effective than, doing something cheaper or easier – especially if that is done by someone else. Sometimes the issue is standardization; it saves money for a hospital (and can improve quality) if only one or two types of, say, artificial joints are used. But this hurts (obviously) the manufacturers of the other brands, and perhaps is a negative for those surgeons who have learned how to use the non-preferred brands (see Atul Gawande, “Big Med”, New Yorker, August 13, 2012). It is even a bigger issue when the evidence demonstrates that the costly surgical option doesn’t work, or doesn’t work as well as, a much cheaper non-surgical option that exists. Well, then, you are threatening someone’s income – just like renewable energy threatens the income of companies that produce and sell fossil fuels.

But would doctors actually do such a thing? Resist a cheaper and more effective alternative because it would cost them money? You betcha. Not always, of course, and not all doctors, but it has happened. “In the late 1990s,” the New York Times observes in its excellent editorial on the subject,  “when it [AHRQ] endorsed nonsurgical interventions for back pain, the back surgeon lobby waged an attack that resulted in huge funding cuts and placed a permanent target on the Agency for Healthcare Research and Quality (A.H.R.Q.), the agency that houses the database.”[1] They attacked the existence of the Agency because they didn’t like (REALLY didn’t like, since it would hit them in the pocketbook!) the evidence.

Yup. It happened. I remember it. It was shocking to me (I must have been more naïve!). I lived in San Antonio at the time, and a San Antonio congressman, Henry Bonilla, was leading the charge against AHRQ, to the embarrassment of the non-back-surgeon San Antonio medical community. But they were, and are, a rich and powerful lobby. And they are still doing thousands of surgical interventions for back pain of the type that AHRQ recommended against in the ‘90s. And making a lot of money on it. Maybe you had such a procedure? Did it help? I hope so. If you are trying to decide, for this or any other complicated treatment (surgery, cancer treatment, etc.) and wanted to know what was recommended for your problem, you could have gone to the NGC website and looked it up. Not now.

One issue with the evidence, of course, is that it depends upon what research was done, and on what populations. And diagnostic and treatment plans that make money for doctors and hospitals are only of use to them if they get paid, and paid well, so they are less likely to be done to poor and uninsured or underinsured people. Maybe this is one time when not having good insurance is a health benefit!

But it shouldn’t be. We should all be covered for necessary diagnosis and treatment. And whether providers or manufacturers can make money should not be a criterion for recommending it. And now we have less access to finding out what the recommendations are.

---

[1]By the way, how did Fox News cover this story? I could only find one link, to Fox Vegas, which used the CNN coverage, which was quite critical of the elimination of NGC. National Fox didn’t seem to think it worth covering.

Maternal mortality in the US and UK: Why do we tolerate paying so much more for so much worse outcomes?

Last year, ProPublica, in association with NPR, published Why Giving Birth Is Safer in Britain Than in the U.S. In typical journalistic style, it starts by grabbing your attention with a case report of a woman in England who almost died of post-partum hemorrhage, but did not. The lesson is presumably that she did not because the physicians, midwives, and others attending her followed a rigorous set of established protocols for addressing post-partum hemorrhage that are implemented nationally in the UK.

Of course, there is a possibility that this individual woman could have died, or had to undergo more invasive surgical procedures further down the protocol’s algorithm, but the real point is that, overall, the system is working. The evidence is in that the maternal mortality rate (deaths/100,000 women delivering) is 8.9 in Britain, while in the US the rate in 2015 was 25.1, three times that of the UK! What makes this more dramatic is that the disparity has developed only since 1990; until then the maternal mortality rates in the US and UK had been declining in parallel since the 1950s.

The article cites several reasons for this difference. One, a very important one, is that the UK collects data on maternal mortality nationally and develops guidelines based upon this data which are implemented nationally. In contrast, the US collects data at best by state, or even by hospital, and

There is no federal-level scrutiny of maternal deaths, and only 26 states have an established committee (of varying methodology and rigor) to review them. Nor do all U.S. hospitals routinely examine whether a death could have been avoided. Procedures for treating complications such as preeclampsia, and for responding to emergencies such as hemorrhage, vary from one doctor, hospital and state to the next.

This is true despite the fact that the methods used by the British to collect and analyze this data were developed in the US. While there has been a well-documented 30-year effort to improve quality and to reduce preventable deaths (a category into which most maternal mortality falls) in the US, led by such organizations as the Institute for Healthcare Improvement (IHI) and embraced by such other organizations as the American Hospital Association (AHA), National Center for Quality Assurance (NCQA), the Joint Commission for the Accreditation of Healthcare Organizations (TJC), the National Academy of Medicine, and on and on, no compulsory national approach to this problem has developed. This reflects a common, and often knee-jerk, opposition to centralized approaches to almost everything, even when they have been determined to have an important effect on reducing death. It is actually parallel to efforts within hospitals to standardize care, to require, for example, all surgeries to go through a series of prescribed steps (“timeouts”) before operating, or limiting the number of different devices implanted to those needed by different types of patients rather than by the preference of the individual surgeon (for an interesting discussion, see A. Gawande, “Big Med”, New Yorker, August 13, 2012). That is, it is effective where it has been done, but it is not mandated to be done everywhere and comprehensive national data is not even collected.

Another big part of the successful UK approach to the reduction of maternal mortality has been collecting detail on what happened and why; this goes beyond “there was a death”, or “there was a death from hemorrhage”, or “there was a death from sepsis”, to identifying why it happened, particularly if the reason was something that is relatively easily addressed. Marian Knight, head of MBRRACE-UK, the group that collects the data and makes the guidelines, says

It’s all very well to know a woman died of sepsis, but to know that she died of sepsis because nobody measured her temperature, as they had no thermometers on the postnatal ward, that’s where the instruction Put a thermometer on your postnatal ward might make a difference. It’s not just the what, it’s the why.

Can you believe that there might be post-partum units where there is no thermometer? Apparently it has happened. And having a rule that it must be present is a way of preventing it from happening again.

There are at least three other important dimensions. One is that, as the report states, “These U.S. deaths are not spread equally. Women who are poor, African American or live in a rural area are more likely to die during and after pregnancy.”  This is not a big surprise.  Poor women, minority women, rural women, and uninsured women do much worse, and are much more likely to die from complications of delivery. It is the familiar song in the US, as in so many areas, of health and of every aspect of society. Is it not true in Britain? The article goes on to say

In the U.K., while inequalities persist when it comes to serious complications, according to 2012-2014 data, there is no statistically significant difference in mortality rates between women in the highest and lowest socioeconomic groups. All British women have equal access to public medical services, including free care and prescriptions from pregnancy through the postpartum period.

This also has two components; the greater equity of the quality of healthcare delivered to all segments of the population, a result of having a national health care system, and less disparity in the “social determinants of health”, the actual quality of the lives of people (women, in this case) before they access care. Especially for rural women, some of the problem the lack of an adequate number of health professionals. Over 20 years ago, family medicine residencies in Texas were surveyed and only a small percent of residents were interested in providing rural obstetric care, but no OB/Gyn residents were! This has not improved; while a recent study published by Tong, et al., in Family Medicine (Characteristics of Graduating Family Medicine Residents Who Intend to Practice Maternity Care) found that 22% of FM residents planned to deliver babies, they cite his 2012 study that showed that only 9.1% of FM residency graduates were delivering babies 1-10 years out into practice despite an intention to do so of 24%, comparable to the current study.

Another dimension is that the difference in maternal mortality reflects a greater focus on the health – and life -- of the pregnant woman in Britain, while in the US the focus is more on the health of the fetus and the newborn. This goes far beyond the issue of abortion, although the focus on children rather than women is one that is also characteristic of the anti-choice movement in the US. It is so deeply ingrained in our culture that even many health professionals (including midwives, doctors, and nurses) who see themselves as “pro-choice” and would never want to see a maternal mortality, are still more focused on the fetus and baby. It results in a practice, if not a belief, that considers the woman but a vehicle for producing a child.

Finally, there is cost, ironically but again unsurprisingly much higher in the US. Much higher. The total cost for a normal vaginal delivery in the US is about $30,000, and about $50,000 for a Caesarean section. In Britain, the cost for a normal vaginal delivery or planned Caesarean is about $2500, or less than 1/10^th the cost in the US, perhaps rising to $3400 for complicated cases (such as the one that leads off the article). Thus, the US charges far more, but has much worse outcomes for maternal mortality (as for many other conditions). This is not a side note; it is not just an interesting contradiction that our care costs more but has worse outcomes. And it is not by any means limited to pregnancy care or maternal mortality, but is present in our healthcare system at almost all levels and for most conditions. What ties these two components, cost and quality, tightly and inversely together, is that our “healthcare system” is only secondarily about delivering quality healthcare, and primarily about being a profit-making business.

So, that is the bottom line. The US has three times the maternal mortality rate of the UK despite charging ten times as much for delivery. The reasons are the absence of a national strategy to identify and remedy the causes of maternal mortality and the lack of a national healthcare system to provide the access necessary for women at risk. It causes the suffering and death to disproportionately affect those women who are already at greatest risk: the poor, minority, rural, and uninsured. The root cause is that our healthcare system is composed of poorly interconnected components, most of which are run as businesses to make money.

It is inequitable, and it is irrational from a health perspective if not from a business one. It causes unnecessary death and excessive cost, and seems to not be improving. It is not acceptable. We need to change it.

Conservative, Liberal, or Relevant Blood Pressure? We need studies that are really relevant...

The following is a guest post from Robert Bowman, MD

One set of guidelines says to loosen up blood pressure control to prevent consequences such as falls. Almost in reaction there appears a new study that indicates a need to, perhaps, tighten up control to reduce the potential for stroke or heart disease. When British GPs were paid more to address BP there did not seem to be gains, other than better pay for participating GPs. Time after time we seem to have cutting edge studies that are pro-intervention. They remain until longer and bigger studies come out years later indicating problems or limitations involving the once “cutting edge” pro-intervention studies.

The lives of millions of people can be impacted by hypertension guidelines. How can we best deal with blood pressure issues?

The answer is not likely to be found in guidelines, quality measures, reports, or the latest studies.

The answer is more about process and less about print.

The problem with guidelines that are too liberal or too conservative is the same – they are too distant from the real world. The real problem remains the failure to include many of those most important to the process.

The real world in health care is what is happening at home. What is the real world for elderly patients? What happens when you get up at night to go to the bathroom? Do you get up at night more because of your medications or because of how you take them? Do your physicians coordinate the addition and subtraction of medications – including drugs that impact blood pressure or fluid volumes? With more and more medications there are more potential interactions. What happens when you get up from a chair or sofa? When you lose 10 or 15 pounds, do your medications drop your blood pressure to dangerous levels? Do you even have contact with a health care professional to help look out for the problem of weight loss and medications that are suddenly too effective?

We know about the benefits of blood pressure control, but do we know about the consequences of too much control?

Sadly we know less about the risk of blood pressure that is kept too low, the risk that a fall that may belife or lifestyle ending will occur.For example, should our blood pressure goals be more aggressive in men or in those with proven bone stability while being less aggressive with those likely to suffer greater consequences becaue they have osteopenia, osteoporosis, or other conditions? We often assume linear increase in risk, but this may not be the case.

What Is Missing from “Guidelines?”

With control of BP or coagulation or other treatments that defeat how the body adjusts to change, the patient and family must have the best understanding regarding what the drugs do and how the drugs impact their particular body and situations.

Large scale studies can be helpful, but the studies need to be relevant. Studies should reflect the real world. After three hundred recent home visits, observations indicated a few with perhaps lax control that may, just may have problems in 1, 5, or 10 years. There are also a different few of the 300 who are having falls or symptoms that suggest the potential for falls - with the potential of immediate consequences. Many large studies examine only the outcome of interest to the researchers – such as the rate of strokes in people with uncontrolled blood pressure – rather than the overall rate of harm or death to the people involved. In addition, many studies exclude from participation people who have the risk factors most likely to cause them harm. Studies that do examine all outcomes (“all-cause morbidity and mortality”) are of more value, and we need to be sure that the people we are treating in the “real world” are similar to those studied.

And then there is the problem of getting reliable BP measurements at the office or at home. How many people do we overtreat based on office-only measurements when lower home BP measurements are more relevant? How do we best use ambulatory measurements? Are these accessed and priced in ways that can make a difference?

A final reflection may be more relevant. In the 5 or 6 minutes of face-to-face time that is often all that exists, given current payment design, how do we get to know our patients well enough in our office environment to optimize BP control while minimizing the consequences to them as they live in their world? How do we teach them enough to loosen up medication when necessary to help prevent falls?

Addressing Problems and Solutions

One thing is certain. As long as research is distant and irrelevant, guidelines will contribute to too much negative consequence and not enough gain.

Practice-based research could contribute. But the real potential involves home-based research. The irrelevance of the academic setting and even the office was suggested by the founders of family medicine. What matters most is far away from university hospitals and NIH researchers. Lest we forget, the home – and the community -- is where it all happens.

Follow up to come:

Why the Home is the Best Unit of Analysis for Research

Guidelines, bias, and your health

One of the challenges in educating health-care professionals is the massive – and growing -- amount of information about diseases, diagnosis, and especially treatments that are available. This is not only an issue for students and residents, but for continuing education; if a doctor or nurse or pharmacist knows only what they learned in school, it will be outdated in a very few years. Thus, the health professions have committed to the concept of “life-long learning”, but how this happens, and how effective and accurate it is, is the challenge.

With the development of the Internet, information is extraordinarily widely shared very quickly. It makes information available not only to health professionals, but to everyone; in the “old days”, perhaps someone with symptoms, or a new diagnosis, might go to the library, or find an outdated copy of Morris Fishbein’s “Handy Home Medical Encyclopedia” or a similar book. Now they go to the internet and are overwhelmed with information. The problem is that this information may be correct, partially correct, or wrong; it may be complete, incomplete, or valueless. There is a difference between information and truth, and certainly between information and wisdom. Wisdom, and truth, require information, but much  of the information is chaff or worse.

If keeping up with information and sifting through it to discover what is true, relevant, and important is hard for a person who wants to find out about their one condition, it is truly daunting for health professionals who need to keep up on many. There are so many drugs! And new ones all the time! And they are all better, improved, more effective (and certainly more expensive!) than the old ones. Until, of course, we find out that they are not. Or, worse, dangerous. Sometimes so dangerous that they have to be pulled from the market.

In his Op-Ed piece in the New York Times on June 12, 2013, “Healing the overwhelmed physician”, Dr. Jerry Avorn of Harvard Medical School calls this the “Stendhal syndrome”, after the French author who, on an 1817 visit to Florence, “…was seized by palpitations, dizziness and a feeling of being overwhelmed by the abundance of great art surrounding him”; the over-abundance of information, too much to grasp, integrate, and categorize, not to mention utilize effectively.  How can doctors and other health professionals get through this to “truth”? To be able to most effectively help, and, primum non nocere,  not hurt you?

One traditional method, adopted by our friends in the drug industry, has been “detailing”. Pharmaceutical representatives (called, regardless of gender, “detailmen”) fanned out, visiting physicians, giving out free samples of their wares as well as food and pens and sticky-note pads and clocks and calendars and little scale models of whatever organ their drug worked on, along with information about the drug. The information might have been accurate and complete, but often was skewed, and almost always was research funded by the manufacturer. The detailpersons were easily identifiable, even when not pushing a cart full of pizza for “grand rounds” because they were all good-looking and so much more well-dressed than run of the mill health care workers. What a great system! You get information about new drugs, direct from the person selling them, and little scale models of the urinary system and donuts too! Isn’t that how we all hoped our doctors were keeping up with the field of caring for our health?

Well, ethics rules have limited the ubiquity of detailpersons, and free gifts, and even samples; the “educational cruises” to the Caribbean just for being a high-utilizer of a company’s drugs are all but gone. A resident asking a drug rep to fly their whole residency class to Las Vegas for graduation is so last millennium! But an incredibly wealthy industry like the pharmaceutical industry will not give up there. This is the real thrust of Avorn’s op-ed piece. He discusses the influence that these companies have on the development of “clinical guidelines” by groups of experts in a particular medical field.

Clinical guidelines can be really good and helpful. The information glut is real, sifting through all the data is almost impossible for all the conditions that exist, especially ones that are seen rarely by a particular practitioner. It makes sense that a group of experts in the field review all the evidence in that narrow area and make recommendations that we can count on as accurate and can utilize effectively. But we may be reaching a point where there are too many guidelines, often with conflicting recommendations. Avorn focuses on recent recommendations from the American Association of Clinical Endocrinologists (AACE) on caring for patients with diabetes:

“The A.A.C.E.’s latest guidelines elevate many second- or third-line drugs to more prominent positions in the prescribing hierarchy, rivaling once uncontested go-to medications like metformin, an inexpensive generic. They also emphasize the riskiness of established treatments like insulin and glipizide, which now carry yellow warning labels in the A.A.C.E. summary. Several of the now promoted drugs are expensive newcomers that lack the track records of clinical effectiveness and safety held by the older, potentially displaced treatments. The changes were made, ostensibly, to give physicians more treatment choices for their patients. But there is also concern that they could have been influenced by another factor: the manufacturers of some of these new drugs financially supported the development of the guidelines, and many of the authors are paid consultants to some of those companies.”

Avorn goes on to cite previous examples of expert-generated and industry-sponsored guidelines that have led to overuse of drugs with very bad outcomes for patients. He also cites reliable sources of clinical guidelines, which are based solely on review of the evidence, not the opinions (possibly influenced by money) of experts, “…including the Institute of Medicine, the American College of Physicians and the Cochrane Collaboration, an international network of experts that evaluates clinical research.” He also refers to the practice of “academic detailing”, which he helped to develop[1] [2], in which academic physicians and pharmacists try to counter the impact of drug-industry detailpersons by going out to practicing physicians and presenting evidence-based recommendations.

There are at least two other sources of bias in addition to any direct influence exerted by drug company funding of research, guidelines development, and individual experts. One is the fact that many treatments, particularly when they involve procedures, can increase the income of the individual physicians doing them. Unfortunately, we have seen (and continue to see) the persistence of many procedures for which there is weak evidence overall or for as broad a population as it is being used on, presumably for this reason. The other, less overtly money-driven and thus more insidious, is that many experts are, almost by definition, specialists (well described in this context by the more pejorative “partialists”) who are focused on the impact of a treatment on “their” disease and may miss the potential for negative impact on the patient. An example may be very tight control of blood sugar in people with diabetes; it may improve the outcomes from diabetes, but if it also leads to potentially dangerous episodes of low blood sugar, it may not be a good idea.

Guidelines issued by different organizations also may be in conflict. An excellent example is mammographic screening for breast cancer, discussed in two outstanding “Viewpoint” articles in a recent JAMA. Michael Marmot addresses potential benefits and harms from different guidelines, focusing on UK recommendations[3], and Kachalia and Mello look at the differing US recommendations from the US Preventive Services Task Force (USPSTF), American Cancer Society (ACS), and American College of Radiology (ACR).[4] If a radiologist reads a mammogram as normal, but, following ACR guidelines recommends another in one year, while the clinician is following USPSTF guidelines for less frequent screening, does that place the clinician at increased legal risk? They argue at least for consistent guidelines to be utilized within a single institution.

Finally, there is the potential for unintended outcomes. One of the most reliable evidence-based sources of recommendations is the USPSTF. Its recommendations are based solely on review of the science, which has generated controversy before when those recommendations were unpopular with powerful groups (see mammography guidelines). Their reliability led them to be written into the health reform law, ACA, which mandates insurers pay for preventive services with an “A” or “B” recommendation from USPSTF. However, as articulately described by USPSTF members Steven Woolf and Doug Campos-Outcalt in another recent JAMA[5], this creates the unintended consequence of turning USPSTF into a group that decides whether companies and doctors will make money; you can imagine the lobbying by a company to try to prove that its product deserves a “B” rating!

You sure can imagine it. It is not what we need. Essentially, as long as potential for big profit exists in health care, for providers, drug companies, device manufacturers, insurers, and others, there will be plenty of opportunities -- and actual occurrences – of this, not scientific evidence, guiding medical practice.

And I doubt that this is what you want for your health.

---

[1] Soumerai SB, Avorn J,  “Principles of educational outreach ('academic detailing') to improve clinical decision making.”, JAMA. 1990 Jan 26;263(4):549-56. PMID: 2104640

[2] Fischer MA, Avorn J, “Academic detailing can play a key role in assessing and implementing comparative effectiveness research findings,” Health Aff (Millwood). 2012 Oct;31(10):2206-12. doi: 10.1377/hlthaff.2012.0817.

[3] Marmot M, “Sorting through the arguments on breast screening” JAMA. 2013;309(24):2553-2554. doi:10.1001/jama.2013.6822

[4] Kachalia A, Mello MM, “Breast cancer screening: conflicting guidelines and medio-legal risk”, JAMA. 2013;309(24):2555-2556. doi:10.1001/jama.2013.7100

[5] Woolf SH, Campos-Outcalt D, “Severing the link between coverage policy and the US Preventive Services Task Force”, JAMA. 2013 May 8;309(18):1899-900. doi: 10.1001/jama.2013.3448.