Outcome switching in research

Vox points to another issue in the evidence-base.

For years, the drug company GlaxoSmithKline illegally marketed paroxetine, sold under the brand name Paxil, as an antidepressant for children and teenagers. It did so by citing what’s known as Study 329 — research that was funded by the drug company and published in 2001, claiming to show that Paxil is “well tolerated and effective” for kids.

That marketing effort worked. In 2002 alone, doctors wrote 2 million Paxil prescriptions for children and adolescents.

Years later, after researchers reanalyzed the raw data behind Study 329, it became clear that the study’s original conclusions were wildly wrong. Not only is Paxil ineffective, working no better than placebo, but it can actually have serious side effects, including self-injury and suicide.

So how did the researchers behind the trial manage to dupe doctors and the public for so long? In part, the study was a notorious example of what’s called “outcome switching” in medical research.

Before researchers start clinical trials, they’re supposed to pre-specify which health outcomes they’re most interested in. For an antidepressant, these might include people’s self-reports on their mood, how the drug affects sleep, sexual desire, and even suicidal thoughts.

The idea is that researchers won’t just publish positive or more favorable outcomes that turn up during the study, while ignoring or hiding important results that don’t quite turn out as they were hoping.

How widespread is the issue?

Well, Dr. Ben Goldacre has a group of medical students reviewing recently published studies in major medical journals.

It’s pretty bad.

COMPare  Tracking switched outcomes in clinical trials.png

Of course, there’s nothing wrong with reporting on unexpected outcomes. I imagine that a lot of hugely important innovations  have come from sharing unexpected findings.

However, I cannot see a reason to not report on the specified outcomes. Withholding “specified” outcomes while sharing new outcomes, for whatever reason, seems like a half-truth and anti-science.

Why would researchers do this, and why would a journal publish it?

1 Comment

Filed under Uncategorized

Most popular posts of 2015 – #10 – A great loss for the field

10786_10152985539074029_8054043966267382162_nFrom Jim Balmer, Dawn Farm’s President:

We lost the great Ernest Kurtz last night – and many of us have lost a wonderful friend. Ernie was a brilliant and inquisitive man who helped countless people understand both AA and spirituality in new ways. What a privilege to have known him.

I did not realize that Ernie recently started blogging: http://ernestkurtz.com/blog/

As posted last week, Bill White recently started an Ernest Kurtz archive: http://www.williamwhitepapers.com/ernie_kurtz/

From his website:

Ernest Kurtz received his Ph.D. in the History of American Civilization from Harvard University in 1978. His doctoral dissertation was published as Not-God: A History of Alcoholics Anonymous. Since then, he has written Shame and Guilt and, with Katherine Ketcham, The Spirituality of Imperfection. He has also published articles and has lectured nationally and internationally on subjects related to the academic study of spirituality. Some of his articles appear in the book, The Collected Ernie Kurtz.

Dr. Kurtz taught American History and the History of Religion in America at the University of Georgia and Loyola University of Chicago. From 1978 to 1997, he served on the faculty of the Rutgers University Summer School of Alcohol Studies and from 1987 to 1997 at the University of Chicago School of Social Service Administration. After a stint as Director of Research and Education at Guest House, then an alcoholism treatment facility for Catholic clergy, Ernie retired to Ann Arbor, Michigan, where he lives with his wife,Linda Farris Kurtz, who is also an educator and author. He continued to travel and lecture widely until 1997, when a botched medical procedure led to back surgeries that only partially restored his ability to stand and walk.

Confined since early 2012 to wheelchair and walker, Ernie devotes time to reading, research, and writing. Since 2002 he has held an appointment as Adjunct Assistant Research Scientist in the Department of Psychiatry at the University of Michigan School of Medicine. His latest book, again with Katherine Ketcham, is Experiencing Spirituality, published in May 2014.

Comments Off on Most popular posts of 2015 – #10 – A great loss for the field

Filed under Uncategorized

Most popular posts of 2015 – #11 – Not good enough

This is good:

The opiate-blocker naloxone is one of the year’s most celebrated drugs, breaking into the mainstream as a magic-bullet antidote that yanks overdose victims from the brink of death with a shot of nasal spray or an intravenous injection. Police take it on patrols. Emergency medical technicians keep it in their ambulances. Ordinary Americans are stocking their medicine cabinets with it. Because of it, hundreds of people who might have died this year from taking too much heroin, Oxycontin or similar painkillers remain alive.

This is not good:

But the lifesaving medication is not a cure. After it has done its job, overdose survivors are left with their cravings intact. Without follow-up care, they are likely to keep feeding their habits, putting them at risk of another overdose, one that could kill them. Treatment, however, can be very difficult to find.

Lying in the emergency room after being revived, many addicts say they experience a fleeting moment of clarity that makes them receptive to help. But that potential is often lost in a patchwork healthcare system that gives survivors little incentive to change. Many walk out of the hospital with just a list of treatment options on their discharge papers, researchers and health care workers say.

See this video on after narcan.

Beyond Narcan Why Heroin Addicts Need More Than an Overdose Antidote NBC News


Filed under Uncategorized

Most popular posts of 2015 – #12 – Hope is created in community

hopeAcross the disciplines, we see a movement away from individually focused understandings of hope to more communally and relationally dependent models. Many focus on connectedness as a central aspect of hope. This takes the form of friendship, solidarity, and bearing witness as central relational aspects of hope. Within the recovery model and other models of care, the relationship with caregivers is central for engendering hope. Caregivers are often required to carry hope on behalf of those for whom they care. Hope exists as an interpersonal possibility reflecting the extent to which humans are made for relationship, for love. When we are living in relationships of love, hope is present. Isolation, lack of belonging, and lack of connectedness reflect that which distances from hope. The experience of connection with another, even in the midst of pain, opens up hope’s possibility. . . . Hope is created in community.

Pamela McCarroll

1 Comment

Filed under Uncategorized

Optimism? Or, is it low expectations?

Lowering_The_Bar_Cover_2010.09.22The feds recently published an article touting the long-term success of buprenorphine:

In the first long-term follow-up of patients treated with buprenorphine/naloxone (Bp/Nx) for addiction to opioid pain relievers, half reported that they were abstinent from the drugs 18 months after starting the therapy. After 3.5 years, the portion who reported being abstinent had risen further, to 61 percent, and fewer than 10 percent met diagnostic criteria for dependence on the drugs.

These studies are important. Long-term outcomes have been a big gap in the research.

This is great news, right? 50% abstinent at 18 months! 61% abstinent at 3.5 years! Fewer than 10% dependent at 5.5 years!


Not so fast

There are a couple of problems here.

  • They were only able to do follow-up with 38% of subjects at 18 months and 52% at 3.5 years.
    • So, that 50% abstinent at 18 months is really more like 19%.
    • The 61% abstinent at 3.5 years is more like 32%

Still, 19% abstinent at 18 months and 32% abstinent at 3.5 years is pretty good, right?

Pump the brakes

There are a couple of problems here too.

  • They are only reporting on abstinence from illicit opioid use, not other drugs.
  • Buried in the article, they mention that they are reporting on being abstinent for the last 30 days. This doesn’t tell us much about how they’ve been doing over the previous 18 months or 3.5 years, does it?
  • Same thing for the reporting on diagnostic criteria for dependence. That was also based only on the previous 30 days.

Taking their conclusions at face value

Further, their conclusions open the door to some interesting questions:

In the first study examining long-term treatment outcomes of patients with prescription opioid dependence, our results were more encouraging than short-term outcomes from POATS suggested. As reported in our 18-month follow-up study (Potter et al., 2014), and consistent with other literature (Moore et al., 2007, Nielsen et al., 2013 and Potter et al., 2013), patients with prescription opioid dependence may have a more promising long-term course, compared with expectations based on long-term follow-up studies of heroin users (Darke et al., 2007, Flynn et al., 2003, Grella and Lovinger, 2011, Hser et al., 2001 and Vaillant, 1973). Indeed, a history of occasional heroin use at POATS entry was the only prognostic indicator 42 months later, associated with a higher likelihood of meeting symptomatic criteria for current opioid dependence. Our results are consistent with research on heroin dependence in supporting the value of opioid agonist therapy for prescription opioid dependence; however, half of the follow-up participants reported good outcomes without agonist therapy.

This begs a couple of important questions.

  • First, many medication assisted treatment advocates have argued that opioid addiction is unique in that it creates long-term or permanent brain dysfunction that requires opioid replacement. Do these findings undermine this theory?
  • Second, half of their follow-up subjects doing well without opioid replacement. Can we assume that opioid replacement is responsible for their good outcomes?

This is the basis for the federal and media push for MAT?

It would appear so.

This not quite what you imagined when they reported 61% abstinent, is it? Why would they present it in a manner that many of us would consider misleading?

It’s also hard to understand their certainty, isn’t it?

I mean, when they talk about this being “treatment that works”, “evidence-based treatment” or “science-based treatment”, don’t most member of the public assume that expressions like “works”, “evidence-based” and “science-based” mean that there’s a body of research indicating that these treatments provide a good chance of getting well?

Instead, these studies suggest that these treatments help make people less sick.

If that’s what patients and their families want, there’s nothing wrong with that. But, they ought to know what they’re getting. (The same goes for communicating the limitations or gaps in evidence for other treatments.)

1 Comment

Filed under Uncategorized

One step taken. More to go. 911 Good Samaritan Legislation in Michigan


Good news

From the Detroit Free Press:

Lawmakers gave final, and unanimous, passage to a bill Wednesday that they hope will help lower the number of drug overdose deaths from prescription drugs.

The Good Samaritan bill, which passed the state Senate on a 38-0 vote, would provide immunity from criminal charges for people under the age of 21 who are seeking emergency medical assistance for themselves or friends as a result of a prescription drug overdose.


The Next (Bigger) Step Forward

Rep. Sam Singh has introduced House Bill 4931 that will expand the Good Samaritan protections to people of all ages and illegal drugs.

The bill is in the House Criminal Justice Committee.

Now is the time to contact your representative and/or the members of the Criminal Justice Committee.


The Facts

Keep these facts in mind:

  • Overdose is now the leading cause of accidental death in the U.S. Most of these overdose deaths are due to opioids.
  • If help arrives in time, overdoses can be safely and quickly reversed with a squirt of naloxone up the patient’s nose.
  • More than half of all overdoses occur in the presence of other people, usually other drug users.
  • Too often, people do not call 911 in a timely manner due to fear of arrest for possession of drugs.

A Start

At the urging of parents who have lost children to overdose, the Michigan legislature made some good first steps last year. They enacted laws that increased access to naloxone, the drug that reverses overdoses.

On October 13, 2014, Public Acts 311, 312, 313 and 314 of 2014 were signed into law.

These acts will:

  • Allow Narcan to be prescribed to friends and family of heroin addicts, so it’s readily available in the event of an overdose.
  • Protect a person administering Narcan in good faith to be immune from criminal prosecution or professional sanctions.
  • Require emergency medical personnel to carry the drug in their vehicles and be trained in how to administer it.
  • Require the state Department of Community Health to complete annual reports of opioid-related overdose deaths.

911 Good Samaritan Laws

NARCAN-KITThe Drug Policy Alliance provides a really good summary of 911 Good Samaritan laws:

Accidental overdose deaths are now the leading cause of accidental death in the United States, exceeding even motor vehicle accidents among people ages 25 to 64. Many of these deaths are preventable if emergency medical assistance is summoned, but people using drugs or alcohol illegally often fear arrest if they call 911,  even in cases where they need emergency medical assistance for a friend or family member at the scene of a suspected overdose.The best way to encourage overdose witnesses to seek medical helpis to exempt them from arrest and prosecution for minor drug and alcohol law violations, an approach often referred to as Good Samaritan 911.

The chance of surviving an overdose, like that of surviving a heart attack, depends greatly on how fast one receives medical assistance. Witnesses to heart attacks rarely think twice about calling 911, but witnesses to an overdose often hesitate to call for help or, in many cases, simply don’t make the call. In fact, research confirms the most common reason people cite for not calling 911 is fear of police involvement.

It’s important to know that this is not a liberal vs. conservative or Republican vs. Democrat issue. Some of the reddest and the bluest states in the country have passed 911 Good Samaritan laws.

Twenty states and the District of Columbia have enacted policies to provide limited immunity from arrest or prosecution for minor drug law violations for people who summon help at the scene of an overdose. New Mexico was the first state to pass such a policy and has been joined in recent years by Alaska, California, ColoradoConnecticut, Delaware, Florida, Georgia, Illinois, LouisianaMaryland, Massachusetts, Minnesota, New Jersey, New York, North Carolina, Rhode Island, Vermont, Washington and Wisconsin.

Further, these laws don’t protect dangerous or predatory criminals.

Good Samaritan laws do not protect people from arrest for other offenses, such as selling or trafficking drugs, or driving while drugged. These policies protect only the caller and overdose victim from arrest and/or prosecution for simple drug possession, possession of paraphernalia, and/or being under the influence.

The Bad News

An overdose is a major medical crisis, right? It’s not unlike a heart attack.

Here’s what happens when someone has a heart attack41KSA2GA12L._SX300_

  1. A person has a heart attack at the grocery store and . . .
  2.  . . . thank goodness, the store has an automatic defibrillator.
  3. Someone has been trained to use the defibrillator and performs the rescue.
  4. Someone else calls 911 to make sure the patient gets all the care they need.
  5. The patient is taken to the emergency department and medically stabilized.
  6. Once stabilized, the patient gets transferred to care that will address the cause of the heart attack and/or care that will prevent future heart attacks.
  7. The patient’s treatment plan will generally include lifestyle changes. (Diet, exercise, etc.)
  8. Then, the patient gets follow-up care that might include:
    • follow up appointments with specialists,
    • periodic tests to monitor for indicators of a recurrence,
    • self-monitoring (blood pressure), and
    • monitoring by the patient’ primary care physician.
  9. If problems recur or there are indications of a potential recurrence, the care plan will be re-evaluated and the patient will get whatever care they need.

Here’s what happens when someone ODs and is rescuednarcan

  1. A person overdoses and . . .
  2.  . . . thank goodness, the someone has naloxone.
  3. The person has been trained to use naloxone and performs the rescue.

Maybe, if they are lucky, these steps happen.

  1. Someone else calls 911 to make sure the patient gets all the care they need.
  2. The patient is taken to the emergency department and medically stabilized.

Naloxone is not enough.

We’d never tolerate cardiac patients being sent home without the proper care. Why should people with an addiction be treated any differently?

The good news

The good news is that there are models that work.

The Gold Standard

A male doctor writes on a patients chart.The best example of what should happen is the the kind of care that opioid addicted doctors, nurses, pilots and lawyers get. They all have low relapse rates and return to work at very high rates.

Here’s what would happen if one of them overdosed at work (or if it was known to their employer):

First, the recovery planning begins with some important assumptions:

  • abstinence is the goal;
  • full recovery with a return to full functioning is the expectation;
  • addiction is a chronic illness and recovery requires long term treatment, support and monitoring; and
  • for recovery to be durable, the addict must be an active participant in treatment and recovery maintenance.

Signpost along the road to recovery.The recovery plan is likely to include the following:

  • Formal treatment. The first phase of formal addiction treatment for most of these professionals is residential care ranging from 30 to 90 days.
  • Supportive services. Supportive services used by these professionals includes AA or NA 12-step groups, aftercare groups from their formal treatment programs, and follow-up from case managers.
  • Long-term support and monitoring. After completion of initial formal addiction treatment, they develop a continuing care contract consisting of support, counseling, and monitoring for usually 2 to 5 years.
  • Drug testing. Regular testing for 2 to 5 years, usually with more frequent testing at the start and reduced testing following periods of stable negative drug test results.
  • Dealing with relapse. Relapses are usually addressed by a combination of increased intensity of care and monitoring and by immediately informing family and colleagues of the physician to enlist their support.

Other options

Buprenorphine (Suboxone) and methadone have been shown to reduce drug use, overdose risk, criminal activity and disease transmission.

Some people are able to stabilize and live normal lives on these medications but, at this point, there is no research demonstrating its effectiveness with quality of life indicators like employment.

Many people hope to use these drugs as an interim step toward abstinence. However, there is no established model for successfully transitioning buprenorphine and methadone patients to abstinence. A large federally-funded study attempting to do this reported, “near universal relapse.”

Another option is an injectable drug called Vivitrol. It is injected once per month and can protect against overdose. Unfortunately, it’s very expensive.

Which to choose?

Hope Traffic SignThe gold standard offers a path to full recovery, but it does demand a lot of structure, effort and lifestyle changes.

Many professionals prefer drug maintenance as a goal. It’s an easier plan to implement and many professionals are not confident that their patients are capable of drug-free recovery. (Look for professionals that are optimistic and believe in you ability to achieve full recovery.)

However, most patients and families, for a wide variety of reasons, prefer abstinence as a goal–the most common reason is that they want their life back the way it was before they became addicted.

Patients not sticking with the treatment plan is the biggest barrier to success with both approaches.

It’s harder than it should be

Getting the gold standard for yourself or a family member is likely to be very difficult. But, there are steps you can take to improve the odds.

Comments Off on One step taken. More to go. 911 Good Samaritan Legislation in Michigan

Filed under Uncategorized

“I would have welcomed the help, and it would have changed my life.”

“VERY SCARED”: Tory Schlier accidentally suffocated her baby. At her sentencing, she said she had been “very scared to bring a helpless human being into the world.” In a letter from prison, Schlier writes that she needed help. REUTERS/Handout

A new Reuters story on newborns affected by the opioid epidemic highlights the inadequate treatment provided to addicted pregnant women.

In America, a baby is born dependent on opioids every 19 minutes. But doctors aren’t alerting social services to thousands of these infants, many of whom come to harm in families shattered by narcotics.

The article does a nice job of highlighting the heart-breaking systemic failures after the baby is born.

Reuters identified 110 cases since 2010 that are similar to Brayden’s: babies and toddlers whose mothers used opioids during pregnancy and who later died preventable deaths.

Being born drug-dependent didn’t kill these children. Each recovered enough to be discharged from the hospital. What sealed their fates was being sent home to families ill-equipped to care for them.

. . .

The cases illustrate fatal flaws in the attempts to address what President Barack Obama has called America’s “epidemic” of opioid addiction, a crisis fed by the ready availability of prescription painkillers and cheap heroin.

In 2003, when Congress passed the Keeping Children and Families Safe Act, about 5,000 drug-dependent babies were born in the United States. That number has grown dramatically in the years since. Using hospital discharge records, Reuters tallied more than 27,000 diagnosed cases of drug-dependent newborns in 2013, the latest year for which data are available. On average, one baby was born dependent on opioids every 19 minutes.

The federal law calls on states to protect each of these babies, regardless of whether the drugs their mothers took were illicit or prescribed. Health care providers aren’t simply expected to treat the infants in the hospital. They are supposed to alert child protection authorities so that social workers can ensure the newborn’s safety after the hospital sends the child home.

But most states are ignoring the federal provisions, leaving thousands of newborns at risk every year.

The article makes it clear that access to maintenance medications is not the problem. (These are frequently touted as the “treatment that actually worksmade inaccessible by anti-science zealots.)

Too often, addicts of all sorts are given a passive referral to treatment of inadequate quality, intensity and duration. (Whether it’s medication assisted or abstinence-based.)

What a perfect population to try a treatment model based on the gold standard. Why not provide these women with years of treatment, recovery support and recovery monitoring.  If there are ways in which the gold standard is not an ideal fit for their circumstances, what modifications and adaptations could be applied?


Filed under Uncategorized