There are 30 million visits a year to America’s emergency departments by children. Most of these are to community hospitals rather than specialized children’s hospitals. A couple of months ago I wrote about what has been a steady trend in pediatric care — community hospitals are transferring an increasingly number of children who come to their emergency departments to other facilities for definitive care of their problems. At first glance one might ask if this could actually be a good thing. After all, isn’t a higher level of capability better for sick children? But it looks to me more as if they just don’t want to provide definitive pediatric care, even for fairly routine things. It seems highly unlikely that sicker and sicker kids are appearing in emergency departments. Now further data is available from a new study from the same research group that confirms this trend and extends its implications.
The first study looked at the past decade across America and identified those facilities that had moved up in their ability to provide definitive inpatient care to children and compared that to the number that had, for whatever reason, downgraded their capacity to care for hospitalized children. This is well shown in this graph:
The newer study has the advantage of studying three very common conditions that bring children to the emergency department: asthma, croup, and gastroenteritis. Critically ill children were excluded; the authors looked at what are bread-and-butter issues for any emergency department that sees children. They found significant increases in referrals out for all three conditions. Bear and mind these weren’t necessarily children who even needed to be admitted to the hospital. It’s not uncommon at all for them to be sent to another emergency department, evaluated and treated, and then sent home. That’s quite concerning.
For critically ill patients and some specialized conditions, there are clear benefits to transfer, which have informed national recommendations from the American Academy of Pediatrics for implementing regionalized emergency medical services. However, it is uncertain which children with lower-acuity conditions benefit from transfer to pediatric tertiary care hospitals. The majority of pediatric patients transferred between EDs are transferred for common conditions, and as many as one-third of children transferred are discharged from the hospital without requiring further intervention or subspecialty consultation.
The facilities are ranked from those who see less children (lowest) up to pediatric hospitals. Not surprisingly the rate of referral was highest in the facilities that cared for smaller numbers of children, but there were increases in all of them except for specialized children’s hospitals. What’s going on here? Should we be concerned about it? I’m concerned because, although serious and severe cases should not be cared for in facilities with limited pediatric experience, these three common conditions should be within the expertise of emergency department physicians, pediatricians, and family practitioners, particularly since these days it’s easy for physicians in smaller facilities to call larger ones for advice — I get such calls all the time, as do my pediatric hospitalist colleagues.
Increasing referral rates over time suggest decreasing provision of definitive care and regionalization of inpatient care for 3 common, generally straightforward conditions. . . . These findings provide further evidence of pediatric care regionalization occurring even for common conditions that do not routinely require specialty care.
This can be very hard on families. Being bounced from one facility to another for a common condition, especially if you are then sent home from the second facility, is traumatic and wasteful. I work in a regional facility and my colleagues in the emergency department not uncommonly see a child sent from a facility on hour or more drive away, only to be sent home. I suspect a major reason for this, as usual, is money. Children typically don’t require all the profit-generating tests and procedures adults do. But they do require that emergency department providers be competent in providing care of children. Clearly more and more facilities have decided it’s not worth the cost to provide this. I also think more than a few hospitals are trying to offload this cost under the guise of securing better care for children.
These articles may be behind paywalls except for the abstracts. If anybody is interested, contact me on the contact form on my homepage and I’ll send you the full articles.
I’m being sarcastic, of course, but that’s often how it feels some days. And, since I posted on this subject a few years ago, not much has changed. If anything it’s worse, with more and more clicks needed on the screen to find what I want to find. Those are days when I’ve been busy at patients’ bedsides all day and then struggle to get my documentation done later, typically many hours later. I jot notes to myself as I go along, but it can be hard to recall at 5 PM just what I did and why at 8 AM. This is a particular problem in critical care because some of our billing codes are time-based and we’re supposed to remember exactly, to the minute, how much time we spend on each patient.
It used to be very different, and that wasn’t always a good thing either. Years ago I spent months going through patient charts from the era of 1920-1950. They were all paper, of course, and the hospital charts were remarkably thin, even for complicated patients. I recall one chart in particular. It was for a young child who was clearly deathly ill. The physician progress notes for her already prolonged stay in the hospital consisted of maybe 2 sheets of paper. Most of the daily notes were a single line. I could tell from the graphs of the child’s vital signs — temperature, pulse, breathing rates, and blood pressure — that one night in particular was nearly fatal. The note the next morning was written by a very famous and distinguished physician. I knew him in his retirement and he was a very loquacious man in person. His note after the child’s bad night was this: “mustard plaster did not work.” If I were caring for a patient like that today there would be just for that day and night multiple entries probably totaling several pages on the computer screen.
Patient charts are burdened with several purposes that don’t always work together. The modern medical record as we know it was invented by Dr. Henry Plummer of the Mayo Clinic in the first decade of the twentieth century. Up until that time each physician kept his (only rarely her) case notes really as notes to themselves. When the multi-specialty group appeared, and Mayo was among the first, the notion of each physician having separate records for the same patient made no sense; it was far more logical to have a single record that traveled from physician to physician with the patient. That concept meant the medical record was a means for one physician to communicate with another. So progress notes were sort of letters to your colleagues. You needed to explain what you were thinking and why. Even today’s electronic medical records are intended to do this, although they do it less and less well.
Now, however, the record is also the principal way physicians document what they did so they can get paid for it. Patient care is not at all part of that consideration. The record is also the main source for defending what you did, say in court, if you are challenged or sued. The result is that documentation, doctors entering things in the record, has eaten more and more of our time. Patients and families know this well and the chorus of complaints over it is rising. Doctors may only rarely make eye contact these days as they stare at a computer screen and type or click boxes. But we don’t have much choice if we are to get the crucial documentation done. That’s how we (and our hospitals) are paid and payers are demanding more and more complex and arcane documentation. I don’t know what the answer is, but I do think we are approaching a breaking point. We are supposed to see as many patients as we can. But the rate-limiting step is documentation.
To some extent we brought this on ourselves. In our fee-for-service system physicians once more or less said to payers: “We did this — trust us, we did it — now pay us for it.” I can’t think of a formula more guaranteed to cause over-utilization or even rare cases of outright fraud. But there is only so much time in the day. In my world an ever smaller proportion of it is spent actually with the patient.
The demand for ever more detailed documentation is not the fault of the electronic medical record. One thing that helps in pediatric critical care is the notion of bundled care. This applies for all children up to age 5 — there are categories for 0-29 days of age, 29 days to 2 years, and 2 years to 5 years. The patient is charged a flat rate for the initial day of care, then a flat rate for each subsequent day of critical care. For anyone older than 5 you are supposed to keep track of the specific minutes you spend on them. My colleagues in hospital medicine bill according to the complexity of care, called evaluation and management or E&M codes, which on the face of things seems reasonable. What can become unreasonable is that there are all manner of odd and picky criteria, constantly changing, requiring the use of special magic language, to fulfill the requirements of the several different categories. For pediatric critical care (for those under 5) one only really needs to document how the child is critically ill and what you did about it. After that the daily global charge kicks in.
For myself, I only ask for a system in which the time we spend with patients is more than the time we spend with their medical record. Surely such a thing is possible?
This month’s edition of Pediatrics has some disturbing research: “Trends in Capability of Hospitals to Provide Definitive Acute Care for Children: 2008 to 2016.” What the paper really does is document what many of us who work in referral hospitals have noted for some time: more and more community hospitals are transferring children who appear in their emergency departments to other, larger facilities instead of admitting them to their own hospital for definitive care. At first glance one might ask if this could actually be a good thing. After all, isn’t a higher level of capability better for sick children? But it looks to me more as if they just don’t want to provide definitive pediatric care, even for fairly routine things.
We analyzed emergency department (ED) visits by children between 2008 and 2016 using the Nationwide Emergency Department Sample, a weighted sample of 20% of EDs nationally. For each hospital annually, we determined the Hospital Capability Index (HCI) to determine the frequency of definitive acute care, defined as hospitalization instead of ED transfer. Hospitals were classified annually according to 2008 HCI quartiles to understand shifts in pediatric capability.
The results showed a progressive and steady decrease in children admitted to the original hospital but rather transferred to another facility. This is illustrated in the graphs below, which show EDs stratified by their capability to render definitive care for children.
The panel on the left (A) shows the trend in EDs able to provide definitive care. You can see there was a quite dramatic rise in the number of EDs unable to provide this (orange line — lowest quintile). The blue line shows the highest capability EDs, and there was a drop in those. So overall there has been a shift in ED capability from higher to lower capability. The panel on the right (B) plots ED visits by children over time. The only real change was that number of visits to the lowest capability EDs actually increased over time. So more kids are appearing at facilities unable (or unwilling) to care for them. Note that the number of highly capable EDs in panel A actually has gone down, adding more stress on the system.
This graph presents the same trend in a little different but useful way. It simply plots the raw number of EDs that went up in capability or went down over time. Far more went down than up.
So what’s happening? Is this a manifestation of a good regionalization of pediatric care, of providing children better care? I doubt it myself. It would require one to postulate pediatric care a decade ago was not very good and now it’s much better. One could argue the standard of pediatric care has risen and that smaller hospitals cannot meet this higher standard, so they appropriately transfer children. But that explanation seems doubtful to have developed over such a short time — over my 40 years of practice, probably yes. But not over just 10 years. Not surprisingly, this trend hits rural hospitals particularly hard. Transfer distances are long, can take hours, and are expensive and not risk-free. They also can be hard on families. An accompanying editorial, “Emergency and Definitive Care for Children in the United States: The Perfect Storm,” is succinct:
These results portend a “perfect storm” of events for care of children in the United States health care system. Increasing pediatric ED visits, poor access to EDs ready to care for children, reduced inpatient capability of hospitals, and increased transfers create increased risk for poor outcomes. Pediatric inpatient capacity across most general community hospitals is decreasing, shifting the burden of pediatric inpatient care to regional pediatric centers, often freestanding children’s hospitals with a high Medicaid-insured population of patients. These hospitals, which compose 5% of hospitals yet are responsible for more than one-third of pediatric discharges and are relied on to care for children with complex medical conditions, suffer significant financial losses from pediatric inpatient care.
I have my own anecdotal take on this. I spent much of my career at a large tertiary facility 40 miles away from the town I grew up in. My father was a pediatrician in that small town for 50 years and I myself worked in that hospital during my college years. Occasionally in my capacity as transport director at the tertiary facility I would get a transfer request from my home town. Several of the older nurses, who knew me and my father well, would often say something like: “Your father never would have transferred this child. How times have changed.”
What we don’t know, of course, is if this trend leads to better outcomes. I do hope the authors will continue their work to answer that key question. But in their discussion they doubt it. So do I. Me, I think it’s mostly about money, as many things are. It’s hard to make money off children admitted as inpatients, especially if they are on Medicaid. Administrative costs are high because admissions tend to be short and children typically don’t require all the profit-generating tests and procedures adults do. I think more than a few hospitals are trying to offload this cost under the guise of securing better care for children. This research demonstrates it’s not about patient volume. It’s possible physician unwillingness to admit these children also plays a role. It adds to their already heavy workload. But whatever the cause, or more likely causes, it can’t go on.
Regular readers of the blog know I have written before about the potential effects of screen time on brain development. I think it’s an important issue and represents a kind of ongoing experiment in our children for which we don’t know the results. But what we do know is that excessive screen time is bad for development. The problem is we don’t know what “excessive” means in this situation. This new study brings further information to the question. The somewhat ominous title is: “Associations between screen-based media use and brain white matter integrity in preschool-aged children.”
The authors used MRI scanning, a way of imaging the brain in detail, to assess the fine details of brain structure in children who were chronically exposed to more screen time than that recommended by the American Academy of Pediatrics. (You can find the AAP recommendations here.) They studied 47 children ages 3 to 5 years. That’s not a huge number, but it’s still quite a few, considering the cost and difficulties in doing MRIs on young children. You don’t need to be a developmental neurologist to understand the implications of the findings:
In this cross-sectional study of 47 healthy prekindergarten children, screen use greater than that recommended by the American Academy of Pediatrics guidelines was associated with (1) lower measures of microstructural organization and myelination of brain white matter tracts that support language and emergent literacy skills and (2) corresponding cognitive assessments.
Although this is a small study, the take-home message to me is to take seriously the recommendations of the AAP regarding screen time for small children, certainly until we know more about what this vast, uncontrolled experiment in our children — the proliferation of screens in our daily lives — represents.
It’s widely known socioeconomic status correlates with measures of health; rich people have better overall health and even longer lifespans than do poorer people. Of course there are several reasons this might be the case, including better access to healthcare for chronic problems, better diet opportunities (Google “food deserts” to learn more about that), and better living environments. Using Medicaid as a surrogate for socioeconomic status, it’s been shown children on Medicaid are much more likely to end up in a PICU than are more affluent children. I’ve written about that before. The Medicaid data are sort of a 50,000 foot high view of the issue. Now a recent study from the Cincinnati PICU entitled “Neighborhood Poverty and Pediatric Intensive Care Use” focuses on a specific local region — a view from the ground. It provides a useful case study of the issue.
The authors looked at 4,071 admissions to the PICU that led to a total of 12,297 patient days. They only evaluated children from Hamilton County, the county around Cincinnati. They then matched the children to the poverty rates in the neighborhoods they lived in using census tracts. It’s a pretty crude, yet straightforward measure to answer the question they’re asking. The results are a bit noisy and can best be appreciated with simple scatter plots:
Child poverty was significantly associated with PICU admission (p < 0.001). When the PICU admission days were grouped into quintiles, the most affluent quintile had 23 days per 1,000 children and the poorest quintile had 82 days per 1,000 children. That’s a pretty striking difference — 350% higher in the poorest children.
I think the strength of this simple study is that it puts a local face on a phenomenon that has previously been studied by state or nationally. The implication is clear: if we want healthier children and lower healthcare costs, we should focus on childhood poverty. The returns on investing in this would be huge.
There is a troubling entity in pediatrics. Sometimes infants appear to have suffered some catastrophic problem, only to recover within minutes. Forty years ago these were termed “near miss sudden infant death syndrome events.” The notion was that they represented part of the spectrum of SIDS — sudden infant death syndrome — that had, for some reason, been averted. In the mid-1980s researchers realized these spells weren’t really related to SIDS because infants who experienced them were not at higher risk for having a real SIDS event later. A new term was coined: “Apparent life threatening event (ALTE).” The working definition was this:
An episode that is frightening to the observer and that is characterized by some combination of apnea [pause or cessation of breathing], color change (usually cyanotic [blue] or pallid but occasionally erythematous [red, flushed] or plethoric), marked change in muscle tone (usually marked limpness), choking, or gagging. In some cases the observer fears that the infant has died.
ALTE events are not only terrifying to parents; physicians are also worried and uncertain about what to do. I know — I’ve cared for many infants over the years who fit this description. It’s easy to know what to do if the infant looks abnormal when you examine him because whatever abnormality you find guides your course of action. But typically these babies look fine by the time the physician sees them. If the parents called 911, the paramedics also find a normal-appearing baby. Usually we admit such babies to the hospital and place them on heart and breathing monitors to see if they do whatever it was again. In my experience, they typically don’t. So now what? In at most half the cases we identify a probable cause, the most common of which is reflux of feedings from the stomach to the mouth, after which a small amount gets in the airway. When that happens a common infant reflex is to stop breathing — stimulating them gets them to start again. Respiratory infections (particularly RSV) and seizures, convulsions, are also potential causes. But at least half the time we have no idea what happened.
What is the risk of an ALTE recurring? It turns out we have no idea about that, either. A review of nearly 37 research articles spanning 40 years concluded such a prediction can’t be made, largely because the definitions used for what is or is not an ALTE have been quite variable in spite of the above consensus statement. A huge issue is that infants who are clearly abnormal, but who arrive for evaluation with an ALTE, are often lumped in with children who appear normal after the event. Those are very different groups.
To try to analyze this troubling condition the American Academy of Pediatrics has issued a clinical practice guideline about what to call these events and what to do about them. They did add to the alphabet soup by coining a new term: “Brief, resolved, unexplained event (BRUE).” I’m not sure how useful that term will be, but the intent is to separate out those children in which a detailed conversation with the parents and a thorough evaluation do not identify any potential cause. For those infants the committee (and it was a committee, with all the problems that can bring) recommended not doing so many tests. One thing is clear; a home apnea monitor does not at all reduce the risk of future harm, and so using one is not recommended. A BRUE (or ALTE) is not associated with risk for SIDS — that’s a key observation that has stood the test of time. We do know one thing that is clearly associated with SIDS: putting babies to sleep on their stomachs. Ever since the introduction of the “back to sleep program,” of instructing parents to put their babies to sleep lying on their backs, the incidence of SIDS has dropped remarkably, although it still happens.
What do I think? We don’t know what causes these spells. However, premature infants are at higher risk for them and we know such infants often have disordered breathing and heart reflexes. I have no data at all about it, but my guesswork opinion is these spells represent immaturity of infants’ brains such that they respond to a variety of stimuli by pausing or stopping breathing or slowing their heart rate. The tendency goes away with growing older because we only see this problem in infants. I hope this new categorization leads to useful research about what is really happening during these mysterious events. Of course simply renaming ALTE a BRUE doesn’t really help our understanding of all this.
The debate over the safety of giving birth at home, both for the mother and for the infant, has continued for years. I’ve written about the issue myself. From time immemorial until about 75 years ago or so most babies were born at home. Now it’s around 1% in the USA, although it’s much higher than that in many Western European countries. The shift to hospital births paralleled the growth of hospitals, pediatrics, and obstetrics. With that shift there has been a perceived decrease in women’s autonomy over their healthcare decisions. There has also been an unsurprising jump in the proportion of caesarian section deliveries, an operative procedure, and various other medical interventions in labor and delivery, even though current data suggests the recent jump in caesarian delivery (now around 30%) has not added any benefits. The debate over whether the dominance of hospital births is a good thing or a bad thing (or neither) is much more than a medical debate; it’s also a social and political one. It’s also to some extent an issue of medical power, a struggle between physician obstetricians who deliver babies in the hospital and nurse midwives who often deliver babies at home. I’m very interested in the social and political aspects, but as a pediatrician I’m particularly concerned with the safety question: Is it more dangerous for your baby to be born at home?
One problem in answering this question is that most of the studies about the safety of home birth came from abroad. But a few years ago we got some data from the USA, published in the New England Journal of Medicine, entitled “Planned out of hospital births and birth outcomes.”
One big problem with evaluating previous data has been that vital statistics from birth certificates counted home births and hospital births, but did not identify as a separate category those women who planned to deliver at home, but then were admitted to a hospital to deliver there because of some issue with the pregnancy. Such women were just counted as hospital births. Also, the recent growth of birthing centers has introduced a location kind of intermediate between home and hospital. The large study linked above was from Oregon, using the years 2012 and 2013. It gives some useful information.
The bottom line is that children born to women who intended to give birth at home had an infant mortality rate of 3.9 deaths per 1,000 deliveries. This was significantly higher than the death rate of infants born in a hospital, which was 1.8 deaths per 1,000 deliveries. Not surprisingly, women who delivered in the hospital had a far high rate of some kind of intervention, such as caesarian section.
What should we make of this? Thinking about risk can be difficult, and it’s important to understand the difference between relative and absolute risk. (I’ve written about that, too.) Media reports often obscure this key point. For example, in this study the risk of infant mortality increased 100% with home birth. 100%!! But twice a very small number is still a very small number. The absolute risk of a baby dying in a home delivery is very small. Still, it is higher.
What this means is that a woman deciding to deliver at home should understand all the facts. Some will not want to accept this increased risk, however small it is in absolute terms. Some will accept it. The same issue of the Journal had a good editorial discussing how to think about the issue. It’s a very good summary of the fundamental question. It’s all about the issue of acceptable risk, and how that varies with the person. The conclusion:
Ultimately, women’s choices for place of delivery will be determined by the extent of their tolerance for risk and which risks they most want to avoid.
All of us are aware of what has been termed our “obesity epidemic.” The current prevalence of obesity among adults in the US is around 40%, a dramatic increase over the past 50 years; it was about 15% in 1970. Rates are also increasing across first world countries, so we are not alone in this. Obesity is defined as a body mass index (BMI) of greater than 30. Values of 25 – 30 are termed overweight. BMI is weight in kilograms divided by height in meters squared.
The graph shows the trends over the past decades and has some interesting features. Note that the percent of the population that is obese or extremely obese (BMI > 40) has increased but the percent classified as overweight has not. This suggests to me, although I haven’t seen anything written about it, that overweight and obese patients are two separate groups; the overweight are not destined to become obese. There are even some recent data that suggest being mildly overweight may actually be a good thing as you age.
Many explanations have been offered for the progressive increase in adult obesity, including increased intake of calories, often in the form of soft drinks, and sedentary lifestyle. The simple calculation of excess calories consumed versus calories burned offers a partial explanation, and certainly that’s what I was taught in medical school in the 1970s; obesity was simple arithmetic. It turns out things are more complicated than that. Genetics, for example, plays a large role, as do various hormonal systems.
I don’t follow the enormous medical literature on obesity closely, but this recent study really intrigued me. It was in a journal I haven’t seen before, Economics and Human Biology. This seems appropriate, since the economic effects of the obesity epidemic are massive and getting larger all the time. The authors studied annual sugar consumption in the US population and compared it with obesity rates later. Now, that approach is pretty reductionist in that it ignores many other kinds of calories that aren’t sugar, but the results are interesting. Their findings suggest that, among today’s adults, obesity correlates with global sugar intake during their childhood years in the 1970s and 1980s. If this is the case, one would predict a decrease in obesity among adolescents and young adults now because sugar intake in the US has decreased by 25% in the last decade. In fact, adolescent obesity prevalence, after a steady and seemingly inexorable rise, may actually have plateaued over the past 5 years or so.
The usual caveat of correlation not indicating causation need to be kept in mind, of course. Yet it makes biological sense to me. I think our metabolic state could have a certain kind of “memory” about the milieu it experienced during early growth and development and have responded to that it ways that could persist for many years.
I have written previously on Kevin Pho’s useful KevinMD site about the alarming statistic that gunshot injuries are now the second most common cause of death among children. Between 2011 and 2015 there were over 21,000 children killed by guns. This recent study in the Pediatrics, the journal of the American Academy of Pediatrics, further analyzes the question; it compares pediatric firearm fatality rates among the various states and then tests for correlation between children killed and the degree of strictness or not of the state’s gun laws. There are extensive data on whole populations showing that stricter laws correlate with lower rates of firearms injuries, but pediatric fatalities have not been specifically investigated. Of course as a PICU physician, somebody who takes care of children shot by guns, the latter question is of great interest to me.
Central to the work is developing some sort of grading scale for strictness of gun laws. The authors didn’t use a scale they developed themselves — they used the 2011–2015 Gun Law Scorecard system from the Brady Campaign to Prevent Gun Violence. The higher the state gun law score, the stricter the firearms legislation. The authors also used what they termed secondary exposure variables. These were the presence or not of individual laws previously associated with lower mortality rates in the total population of adults and children: universal background checks for firearm purchase, universal background checks for ammunition purchase, and identification requirement for firearms (microstamping, ballistic fingerprinting). Their findings are best summarized in this graph from the paper. It plots gun law score against the rate of children killed by guns.
The trend line for that graph is pretty striking. It’s clear increasing strictness of gun laws correlates with less children killed by guns. This also fits with the experience in other Western countries. See the example of Australia tightening their gun laws, which greatly reduced gun deaths.
The issue, of course, is that America has a gun culture unique in the Western world. Our courts have interpreted the rights of citizens to own guns quite broadly owing to the Second Amendment. It’s known the great majority of guns are owned by a minority of citizens. So the question is: How many dead children represent an acceptable price to pay for loose gun laws? Because it’s clear that, all slogans aside (e.g. “guns don’t kill people, people kill people”), looser laws lead to more deaths. The research is there. Now we have to decide if we want to do anything about it.
Every day we get bombarded in the news with health statistics. Coffee causes cancer! Coffee cures cancer! And so on. Many of these are meant to grab headlines (and, these days, web page clicks) and the articles they accompany are often very poor at telling the reader what they mean. They often have statistics, and health statistics can be complicated. Sad to say, even many physicians are pretty poor and sorting out the hype from the helpful. This article is a very helpful guide to finding your way when you are reading the health news. It’s called “Helping Doctors and Patients Make Sense of Health Statistics,” and it does just that. I’d bookmark it or even print it out for future reference, if you’re more old school. Don’t be put off by the somber looking first page — it’s actually quite readable. I should point out here that, although I took statistics courses long ago and have used simple statistical tests in my own research career, I am by no means expert. I always consulted a real statistician before submitting any research for publication.
The article starts with the common problem one sees in media reports: the difference between absolute and relative risk. The authors used the example of a scare over birth control pills that happened in 1995 when the U.K. Committee on Safety in Medicines issued a warning about a newer version of pills. The committee sent a warning to all physicians that the newer pills were associated with a 100% rise in risk for serious blood clots. One hundred percent! Yikes! The warning led to many women stopping their pills and there was a predictable rise in unwanted pregnancies, accompanied by an estimated 15,000 more abortions the following year. The effects lasted for years. What was the truth about this new risk?
The truth was that the newer pills were associated with a risk for serious blood clots of 2 per 7,000 women. For comparison, the earlier pills had been associated with a 1 per 7,000 women risk. And 2 is 100% more than 1, so that’s the increase in relative risk. But the increase in absolute risk was an additional 1 woman in 7,000. (It should be noted here that pregnancy itself is associated with an increased risk of blood clots.) I see this particular misunderstanding often in news reports regarding risks of medical procedures. When you read these stories you need to examine not just the relative risk, which often makes for good headlines; you need to look at the actual number, not just the percent change.
The pill scare hurt women, hurt the National Health Service, and even hurt the pharmaceutical industry. Among the few to profit were the journalists who got the story on the front page.
There is also this helpful article, from the always readable Scientific American. I like it a lot, too. It reminds us how statistical significance doesn’t always mean real life significance.
Imagine if there were a simple single statistical measure everybody could use with any set of data and it would reliably separate true from false. Oh, the things we would know! Unrealistic to expect such wizardry though, huh? Yet, statistical significance is commonly treated as though it is that magic wand. Take a null hypothesis or look for any association between factors in a data set and abracadabra! Get a “p value” over or under 0.05 and you can be 95% certain it’s either a fluke or it isn’t. You can eliminate the play of chance! You can separate the signal from the noise! Except that you can’t. That’s not really what testing for statistical significance does. And therein lies the rub.
The article points out what the vaunted, and too often venerated, p value means is that it estimates the probability of getting roughly that result if the study hypothesis is assumed to be true. It can’t on its own tell you whether this assumption was right, or whether the results would hold true in different circumstances. It provides a limited picture of probability, taking limited information about the data into account and giving only “yes” or “no” as options.
If you are really interested in this topic you should read a bit about what’s called Bayesian statistics, named after an 18th century mathematician. The basic notion here is that we need to consider our prior knowledge about something before applying statistical tests, and that we should factor in this knowledge when we make our statistical comparisons. In other words, all possibilities are not intrinsically equal going into the analysis. The debates between Bayesian and what are termed “frequentist” statisticians go back and forth. But what we should take home from these debates is that the science of statistics, like other sciences, is subject to revision and change over time.
A final key point is to look at medical headlines of new medical breakthroughs and try to decide if the findings really are “significant” in real life. Is there only a tiny effect, really, even though the p value is “significant” at the 0.05 level? Also beware of what we call “data dredging,” in which multiple comparisons are made using the same data set. When you do that the chances of coming up with a significant, yet spurious association go up.
All of this has made some people call for some rudimentary statistical training to be part of the standard mathematics curriculum at the high school level. I think this is a good idea. I didn’t get introduced to any statistical concepts in high school, and I took all the math available. That should change if we expect the mass of our citizenry to be competent to judge things for themselves. Medical journalists definitely need this knowledge because currently many do a terrible job interpreting medical reports. The two articles I linked are a great place to start.
How Your Child Heals An Inside Look at Common Childhood Ailments This book takes you on a tour of the inner workings of a child's body as it heals from injury, illness, and common diseases.
How To Talk To Your Child's Doctor A Handbook for Parents "A great read for any parent. It really helps you get inside the mind of your child's doctor and figure out how best to communicate."
Your Critically Ill Child Life and Death Choices Parents Much Face Personal stories of children and their families and how they and their doctors together learned the best way to understand and take care of each other.