According to the authors of the latest study claiming to demonstrate effectiveness of homeopathic remedies, colds are common in the pediatric population. They further explain that colds and cough symptoms are a frequent impetus for parents to seek pediatric medical care. Finally, they add that evidence in support of decongestants, antihistamines and cough suppressants for the treatment of pediatric cold symptoms is lacking and that there are significant potential risks with their use in young children.
All of this is true and information I give to medical learners and patient caregivers all the time. I only wish they had quit while they were ahead. Sadly, the authors of “A randomized controlled trial of a homeopathic syrup in the treatment of cold symptoms in young children” continued:
One option for treating cold symptoms in young children is with homeopathy. Because the concentrations of active ingredients in homeopathic medications are extremely dilute, they are generally considered to be safe. However, there is a widely held belief that any efficacy related to use of homeopathic remedies is related to a placebo effect.
The study authors apparently stared into the face of homeopathy’s utter lack of scientific plausibility, and the fact that homeopathy is nothing but trumped up placebo, and laughed into the void. This actually isn’t surprising considering that the lead author, Jennifer Jacobs, is a long-time proponent and former employee of Standard Homeopathic Company, also known as Hyland’s, the makers of bogus teething remedies currently being investigated by the FDA. Those paying close attention to the unethical involvement of pediatric patients in homeopathy studies over the years may recognize Jacobs.
The study particulars
The study took place at the University of Washington Medical Center-Roosevelt Pediatric Care Center, a seemingly legitimate facility where pediatric residents learn primary care and parents probably don’t expect to be deceived by medical professionals. Children aged 2 to 5 years diagnosed with an upper respiratory infection and having less than a week of symptoms were targeted, and those with medical complications such as asthma or who were being prescribed a medication were excluded. Also excluded were children exposed to a homeopathic remedy within the 48 hours prior to recruitment, because multi-nothing interactions can be a real confounder.
Upon recruitment, parents of study subjects rated the severity of their child’s cold over the prior 24 hours based on 4 core cold complaints: runny nose, congestion, cough, and sneezing. Each complaint could be ranked from 0 to 3 in severity, with a maximum overall score of 12 points. This is an entirely arbitrary scoring system that is 100% subjective.
Study subjects were assigned an ID number and their parent given a bottle of either the homeopathic syrup or a placebo syrup, both essentially being simple licorice extract in water. This was done randomly with the parent, child, and research team unaware of what each bottle contained. The specific homeopathic syrup used in the study, which “contains” allium cepa 6X, hepar sulf calc 12X, natrum muriaticum 6X, phosphorous 12X, pulsatilla 6X, sulphur 12X, and hydrastasis 6X, was Hyland’s 4 Kids Cold ‘n Cough. Hyland’s, in the form of Standard Homeopathic Company, also funded the study.
Advanced data mining
Parents were instructed to give their child a dose of whatever was in their bottle every 4 to 6 hours as needed for cold symptoms over a three day period. If a dose was given, parents were asked to document changes in the 4 core symptoms using a 7-point Likert scale (from “much worse” to “much improved”). They were also charged with tracking non-specific symptoms such as lethargy, fussiness, and appetite as well as any observed side effects. This is homeopathy after all and they do like their provings.
Twice each day over the three day study period, parents completed a symptom diary which rated overall severity of the four core symptoms on the aforementioned 4 point scale. Again, these scores were added together to make a composite score with 12 being the maximum. Parents noted missed days of school and/or work and recorded any other medications or treatments given. Once each day, parents also were asked to use a 5-item functional health scale to assess their children on various activities such as being read to, eating, and sleeping. Finally, they also rated their child’s overall health once daily using a 1-10 scale with 1 being “very well” and 10 being “very sick.”
5 to 10 days after initiation, the caregiver was contacted and asked to assess the subjects’ current cold symptom severity, overall health, and functional status.
The primary outcome being assessed in this study was the effect of the homeopathic syrup on the 4 core colds symptoms an hour after being administered. All the other secondary outcomes described above were just bet hedging. If a parent assessed their child as even just slightly improved an hour after a dose, it was counted in the win column during statistical analysis.
The authors chose to perform a “preliminary, interim analysis” after the first 63 children had been enrolled in order to “assess for possible efficacy of the homeopathic syrup, with the option of discontinuing the study if there was no indication of efficacy.” They evidently found what they were looking for. Call me a cynic, but that part seemed fishy to me.
They recruited 261 subjects, collecting 154 dosage logs amounting to 957 doses, 497 being of the homeopathic syrup. Parents also returned 162 symptom diaries. This fell far below their predetermined goal of 1,425 total doses of study medication.
The results speak for themselves…I think?
In regards to the primary outcome being studied, the effect of the homeopathic syrup on 4 core cold symptoms (runny nose, sneezing, cough, congestion) an hour after dosing, this study was a complete and utter failure. There were no differences in any of the 4 core symptoms compared to placebo. There were also no differences in the non-specific secondary outcomes an hour after each dose. The homeopathic syrup was found to be indistinguishable from placebo. End of story.
Wait, there’s more. There is always more. Why is there always more? Let the torturing of data and convoluted post-hoc reasoning begin!
Next up for analysis were the twice daily symptom diaries. They found that improvement from baseline was significantly better in the homeopathic syrup group…for runny nose, cough, and sneezing (not congestion???)…and only at the first and second assessments. The composite cold scores were statistically better in the homeopathic group as well, also only on day one. On day 2, 3 and at the follow-up 5-10 days later, there were no differences in cold symptoms whatsoever.
Children randomized to the homeopathic syrup arm also received more ibuprofen and chest rubs (good old fashioned TLC) on day one. When taken into account, the supposed benefit of homeopathy was no longer seen in cough scores. Regarding the other secondary functional outcomes, there were no differences between the two groups at any point in the study. Furthermore, children in the homeopathic syrup group had worse cough at the 5-10 day follow-up and there were statistically more side effects in this group as well, although none were considered to be serious.
The authors attempt to salvage this shipwreck of a study
Jacobs and Taylor describe the results of this study as “provocative.” I disagree. After admitting their complete failure to find any benefit in the primary outcome being measured, or in any functional outcome, they focus on the finding that children randomized to the homeopathic syrup had statistically significant improvement in 2 of 4 of the arbitrarily chosen core symptoms, but only on the first day of treatment. This is extremely weak sauce.
The authors attempt to shrug off alternative explanations for their non-findings:
Because homeopathy is not consistent with the current understanding of chemistry and physics, it is crucial to consider alternative explanations for our findings.
As one astute follower of mine on Twitter (FuturePedsNP) exclaimed:
@skepticpedi I love that "current". Like any day now, results are going to come out of CERN that show how sugar pills can treat pneumonia.
— FuturePedsNP (@cantab_biker) November 14, 2016
They list some potential biases in the study which may have affected the results, such as they were. 38% of the symptom diaries were not returned, for example. The authors claim, however, that a higher response rate would have likely increased the chance of positive findings rather than revealing their thin gruel to be even more meager. They blame a lack of additional funding (NCCIH hasn’t funded homeopathy since 2008) and a lower than predicted number of kids with colds, the latter being almost as ludicrous as homeopathy itself.
They then go on to discount the possibility that poor blinding, inadequate randomization, or statistical chicanery could have biased their results, which I agree with. They aren’t needed and just serve as a distraction. They are essentially just straw man arguments. Random noise is a good enough explanation for me.
The study authors conclude with a few howlers:
The worsening of cough in the homeopathic group at the 5-10 day follow-up is difficult to explain but might be due to a rebound effect after stopping the medication.
No, just more random noise.
Clinicians should consider recommending it for symptomatic relief, since there are no conventional medications that have been shown safe and effective in treating URIs in children.
Since nothing works, try syrup of nothing? I prefer a tincture of time and a soupçon of reassurance myself.
There is presently no definitive explanation for the possible biological activity of such highly diluted homeopathic preparations. Current [there is that word again] speculation about a mechanism of action centers on the field of nanomedicine with the possibility of immune system modulation.
Why seek a definitive explanation for something that doesn’t exist?
Conclusion: Yet another unethical homeopathy study
This study is yet another in a long unethical line of studies boasting cargo cult methodologies designed to answer a question that nobody has asked since 1865. We know that homeopathy shouldn’t work, and it truly toes the line of can’t work. But let’s humor this study a bit and ask ourselves if the results might be enough cause doubt in our joyless skeptic hearts.
The study authors have presented a treatment that fails to reduce cold symptoms in any way a full hour after taking it, while somehow still causing an overall improvement in those same cold symptoms but only for one day. This improvement comes in the form of having slightly less runny nose and sneezing as determined by an outside observer, who I am assuming is not measuring nasal runoff in milliliters per hour. After one day of improved arbitrarily determined symptoms, the treatment stops reducing said symptoms and then has no effect on the overall severity or duration of the cold. Furthermore, despite reducing symptoms on the first day, it doesn’t lead to any functional improvement at any point.
Kids with the “statistically significant” improvement in composite cold symptom scores on day one didn’t have better appetites, more energy, or less fussiness, and they weren’t more likely to go back to school. And their cough was worse a week later! But it’s all just noise. And nanomedicine. Something something nanomedicine.