case-control study
TODAY IS ‘WORLD HOMEOPATHY DAY!!!
Let’s celebrate it by looking at the latest ‘cutting edge’ research on the world’s most commercially successful homeopathic remedy, Oscillococcinum®, a preparation of duck organs that are so highly diluted that not one molecule per universe is present in the end-product. It is therefore surprising to read that this new investigation finds it to be effective.
According to its authors, the goal of this controlled observational study was “to investigate the role of the homeopathic medicine in preventing respiratory tract infections (RTIs)”. The ‘study’ was not actually a study but a retrospective analysis of patients’ medical records. It examined 459 patients who were referred to a respiratory diseases specialist in Italy. Subjects who had taken any form of flu vaccine or any other type of vaccine (immuno-stimulants, bacterial lysates, or similar) were excluded from the study. 248 patients were treated with the homeopathic medicine, while 211 were, according to one statement by the authors, not treated. The latter group was deemed to be the control group. All patients were followed-up for at least 1 year, and up to a maximum of 10 years.
A significant reduction in the frequency of onset of RTIs was found in both the homeopathic medicine and untreated groups. The reduction in the mean number of RTI episodes during the period of observation vs. the year before inclusion in the study was significantly greater in the homeopathic-treated group than in untreated patients (-4.76 ± 1.45 vs. -3.36 ± 1.30; p = 0.001). The beneficial effect of the homeopathic medicine was not significantly related to gender, age, smoking habits or concomitant respiratory diseases when compared to the effect observed in untreated patients. The number of infections during the follow-up period is plotted in the graph.
The authors concluded that these results suggest that homeopathic medicine may have a positive effect in preventing RTIs. However, randomized studies are needed before any firm conclusion can be reached.
This could well be the worst study of homeopathy, an area where there is no shortage of poor research, that I have seen for a long time. Here are some of its most obvious problems:
- The aim was to investigate the role of homeopathic medicine – why then do the authors draw conclusions about its effectiveness?
- The ‘control group’ was not ‘not treated’ as the authors claim, but these patients were prescribed the homeopathic remedy and did not comply. They were neither untreated – most would have self-medicated something else) nor a proper control group. This is what the authors state about it: “The physician initially instructed all 459 patients to take 1 dose of homeopathic medicine…A total of 211 patients were found to be non-compliant (i.e., they did not take the homeopathic medicine as recommended by the medical doctor), and these formed the control group.”
- Why were vaccinated patients excluded? This would skew the sample towards anti-vaxers who tend to be homeopathy-fans.
- A follow-up between 1 and 10 years? Are they serious? The authors tell us that “a total of 21 (4.6 %) patients ended the follow-up before 2012…” Did all the others die of homeopathic over-dose?
- Before the start of the ‘study’, patients had more than 5 infections per year. This is way beyond the normal average of 1-2.
- The authors inform us that “the primary outcome measure for assessing the effectiveness of the preventive treatment with homeopathic medicine was the reduction in the average number of RTI episodes per year versus the year before inclusion in the study.” This begs the question as to how the primary endpoint was assessed. The answer is by asking the patients or phoning them. This method is wide open to recall-bias and therefore not suited as a primary outcome measure.
My favourite alternative explanation for the reported findings – and there are many that have nothing to do with homeopathy – goes like this: some patients did not comply because their condition did not respond to homeopathy. These were the ones who were, on average, more severely ill. They needed something better than a homeopathic placebo and they therefore became the ‘control group’. As the differed systematically from the verum group, it would be most extraordinary, if they did not show different findings during follow-up.
So, is there nothing interesting here at all?
Not really…hold on, here is something: “The corresponding author thanks the Scientific Department of Laboratories Boiron S.r.l. (Milan, Italy) for funding the independent statistical analysis made at the Department of Medical and Surgical Sciences of the Alma Mater Studiorum-University of Bologna (Bologna, Italy).”
BOIRON SEEMS TO DECIDE ON FUNDING AFTER HOMEOPATHICALLY DILUTING SCIENTIFIC RIGOR AND COMMON SENSE.
On this blog, I have repeatedly tried to alert consumers and patients to the risks of herbal medicine. The risks include:
- toxicity of one or more ingredients of the plant,
- interactions with other medicines,
- contamination with toxic non-herbal substances such as heavy metals,
- adulteration with synthetic drugs such as steroids,
- ineffectiveness in treating the target disease
- reduction of adherence to prescribed medicines.
A new paper throws more light on the latter issue which has been not well-studies so far.
The objective of this study was to investigate the relationship between the use of medicinal plants and medication adherence in elderly people. The authors conducted an observational, cross-sectional study of elderly residents in Cuité-PB, Northeastern Brazil, through a household survey. A stratified proportional and systematic random sample of 240 elders was interviewed in their homes and the use of pharmaceutical medicines and of medicinal plants was assessed by direct examination. The association of medication adherence with socio-demographic, clinical, medication and use of medicinal plants was analysed with multiple logistic regression.
The results showed that medication non-adherence increases with use of herbal medicines (adjusted odds ratio 2.022, 95% CI 1.059–3.862, p = 0.03), as well as with the number of different medicinal plants used (adjusted odds ratio 1.937, 95% CI 1.265–2.965, p = 0.002).
The authors concluded that this study provides first-hand evidence that the use of herbal medicines is associated with poor medication adherence. Given the high frequency of the use of herbal medicines, further research into the mechanisms of this association is justified.
This conclusion is well-put, I think. If these findings are confirmed in other populations, we are confronted with a somewhat paradoxical situation: combining herbal and synthetic medicines can reduce adherence to the synthetic drugs and, in cases where adherence is not affected, it could increase the risk of herb/drug interactions.
Anthroposophic medicine is based on Rudolf Steiner’s mystical ideas. It is popular in Germany and is slowly also spreading to other countries. Anthroposophic drugs are prepared according to ancient notions of alchemy and are fly in the face of modern pharmacology. Anthroposophic doctors treat all sorts of diseases, and their treatments include anthroposophic medications, and a range of other modalities.
A recent paper reported a secondary analysis from an observational study of 529 children with respiratory or ear infections (RTI/OM) <18 years from Europe and the USA. Their caregivers had chosen to consult physicians offering either anthroposophic (A-) or conventional (C-) treatment for RTI/OM.
During the 28-day follow-up antibiotics were prescribed to 5.5% of A-patients and 25.6% of C-patients (P < 0.001); the unadjusted odds ratio for non-prescription in A- versus C-patients was 6.58 (95%-CI 3.45-12.56); after adjustment for demographics and morbidity it was 6.33 (3.17-12.64). Antibiotic prescription rates in recent observational studies with similar patients in similar settings, ranged from 31.0% to 84.1%. Compared to C-patients, A-patients also had much lower use of analgesics, somewhat quicker symptom resolution, and higher caregiver satisfaction. Adverse drug reactions were infrequent (2.3% in both groups) and not serious.
What can we conclude from these data?
Not a lot, I fear!
The authors of the study are a little more optimistic than I; they conclude that this analysis from a prospective observational study under routine primary care conditions showed a very low use of antibiotics and analgesics/antipyretics in children treated for RTI/OM by physicians offering AM therapy, compared to current practice in conventional therapy settings (antibiotics prescribed to 5% versus 26% of A- and C-patients, respectively, during days 0–28; antipyretics prescribed to 3% versus 26%). The AM treatment entailed no safety problem and was not associated with delayed short-term recovery. These differences could not explained by differences in demographics or baseline morbidity. The low antibiotic use is consistent with findings from other studies of paediatric RTI/OM in AM settings.
They are clearly careful to avoid causal inferences; but are they implying them? I would like to know what you think.
Few subjects lead to such heated debate as the risk of stroke after chiropractic manipulations (if you think this is an exaggeration, look at the comment sections of previous posts on this subject). Almost invariably, one comes to the conclusion that more evidence would be helpful for arriving at firmer conclusions. Before this background, this new publication by researchers (mostly chiropractors) from the US ‘Dartmouth Institute for Health Policy & Clinical Practice’ is noteworthy.
The purpose of this study was to quantify the risk of stroke after chiropractic spinal manipulation, as compared to evaluation by a primary care physician, for Medicare beneficiaries aged 66 to 99 years with neck pain.
The researchers conducted a retrospective cohort analysis of a 100% sample of annualized Medicare claims data on 1 157 475 beneficiaries aged 66 to 99 years with an office visit to either a chiropractor or to a primary care physician for neck pain. They compared hazard of vertebrobasilar stroke and any stroke at 7 and 30 days after office visit using a Cox proportional hazards model. We used direct adjusted survival curves to estimate cumulative probability of stroke up to 30 days for the 2 cohorts.
The findings indicate that the proportion of subjects with a stroke of any type in the chiropractic cohort was 1.2 per 1000 at 7 days and 5.1 per 1000 at 30 days. In the primary care cohort, the proportion of subjects with a stroke of any type was 1.4 per 1000 at 7 days and 2.8 per 1000 at 30 days. In the chiropractic cohort, the adjusted risk of stroke was significantly lower at 7 days as compared to the primary care cohort (hazard ratio, 0.39; 95% confidence interval, 0.33-0.45), but at 30 days, a slight elevation in risk was observed for the chiropractic cohort (hazard ratio, 1.10; 95% confidence interval, 1.01-1.19).
The authors conclude that, among Medicare B beneficiaries aged 66 to 99 years with neck pain, incidence of vertebrobasilar stroke was extremely low. Small differences in risk between patients who saw a chiropractor and those who saw a primary care physician are probably not clinically significant.
I do, of course, applaud any new evidence on this rather ‘hot’ topic – but is it just me, or are the above conclusions a bit odd? Five strokes per 1000 patients is definitely not “extremely low” in my book; and furthermore I do wonder whether all experts would agree that a doubling of risk at 30 days in the chiropractic cohort is “probably not clinically significant” – particularly, if we consider that chiropractic spinal manipulation has so very little proven benefit.
My message to (chiropractic) researchers is simple: PLEASE REMEMBER THAT SCIENCE IS NOT A TOOL FOR CONFIRMING BUT FOR TESTING HYPOTHESES.
The very first article on a subject related to alternative medicine with a 2015 date that I came across is a case-report. I am afraid it will not delight our chiropractic friends who tend to deny that their main therapy can cause serious problems.
In this paper, US doctors tell the story of a young woman who developed headache, vomiting, diplopia, dizziness, and ataxia following a neck manipulation by her chiropractor. A computed tomography scan of the head was ordered and it revealed an infarct in the inferior half of the left cerebellar hemisphere and compression of the fourth ventricle causing moderately severe, acute obstructive hydrocephalus. Magnetic resonance angiography showed severe narrowing and low flow in the intracranial segment of the left distal vertebral artery. The patient was treated with mannitol and a ventriculostomy. Following these interventions, she made an excellent functional recovery.
The authors of the case-report draw the following conclusions: This report illustrates the potential hazards associated with neck trauma, including chiropractic manipulation. The vertebral arteries are at risk for aneurysm formation and/or dissection, which can cause acute stroke.
I can already hear the counter-arguments: this is not evidence, it’s an anecdote; the evidence from the Cassidy study shows there is no such risk!
Indeed the Cassidy study concluded that vertebral artery accident (VBA) stroke is a very rare event in the population. The increased risks of VBA stroke associated with chiropractic and primary care physician visits is likely due to patients with headache and neck pain from VBA dissection seeking care before their stroke. We found no evidence of excess risk of VBA stroke associated chiropractic care compared to primary care. That, of course, was what chiropractors longed to hear (and it is the main basis for their denial of risk) – so much so that Cassidy et al published the same results a second time (most experts feel that this is a violation of publication ethics).
But repeating arguments does not make them more true. What we should not forget is that the Cassidy study was but one of several case-control studies investigating this subject. And the totality of all such studies does not deny an association between neck manipulation and stroke.
Much more important is the fact that a re-analysis of the Cassidy data found that prior studies grossly misclassified cases of cervical dissection and mistakenly dismissed a causal association with manipulation. The authors of this new paper found a classification error of cases by Cassidy et al and they re-analysed the Cassidy data, which reported no association between spinal manipulation and cervical artery dissection (odds ratio [OR] 5 1.12, 95% CI .77-1.63). These re-calculated results reveal an odds ratio of 2.15 (95% CI.98-4.69). For patients less than 45 years of age, the OR was 6.91 (95% CI 2.59-13.74). The authors of the re-analysis conclude as follows: If our estimates of case misclassification are applicable outside the VA population, ORs for the association between SMT exposure and CAD are likely to be higher than those reported using the Rothwell/Cassidy strategy, particularly among younger populations. Future epidemiologic studies of this association should prioritize the accurate classification of cases and SMT exposure.
I think they are correct; but my conclusion of all this would be more pragmatic and much simpler: UNTIL WE HAVE CONVINCING EVIDENCE TO THE CONTRARY, WE HAVE TO ASSUME THAT CHIROPRACTIC NECK MANIPULATION CAN CAUSE A STROKE.
How many times have we heard from practitioners of alternative medicine, particularly chiropractors, that their patients are more severely ill than those of conventional clinicians. The claim is usually that they have tried all that conventional medicine can offer and eventually, as a last resort, they turn to the alternatives.
But is this true? If so, it would explain why these patients do no better or even worse than those treated conventionally.
Here is a new article that goes some way in addressing these issues.
For this study, Danish chiropractors and general practitioners recruited adult patients seeking care for low back pain (LBP). Extensive baseline questionnaires were obtained and descriptive analyses were performed to define the differences between the two populations.
Questionnaires were returned from 934 patients in chiropractic practice and 319 patients from general practice. Four out of five patients had previous episodes, one-fourth were on sick leave, and the LBP considerably limited daily activities. The general practice patients were slightly older and less educated, more often female, and generally worse on all disease-related parameters than chiropractic patients. All the disease specific parameters showed a statistically significant difference between general and chiropractic practice. Patients in general practice were generally more severely affected. They had higher pain intensity (mainly for leg pain), longer pain duration, more previous episodes, more sick leave, more activity limitation on the disability scale, slightly higher level of depression, slightly more fear-avoidance beliefs, and a poorer self-reported general health. All these differences were statistically significant.
The authors concluded that LBP in primary care was recurrent, causing sick leave and activity limitations. There were clear differences between the chiropractic and general practice populations in this study.
I know, I know: these findings are from Denmark and therefore they cannot be generalised to other countries. However, the authors point out that similar findings have been reported from the US. Furthermore the observations relate to chiropractors and must not be applied to other alternative practitioners. Nevertheless they do show that, in this specific scenario, patients opting for the alternative are not more but less severely ill.
The next time an alternative practitioner claims ‘my patients have worse outcomes because they are sicker’, I will insist on seeing the evidence before I believe it.
For this blog, I am constantly on the lookout for ‘positive news’ about alternative medicine. Admittedly, I rarely find any.
All the more delighted I was when I found this new study aimed to analyse the association between dietary long-chain n-3 polyunsaturated fatty acids (PUFAs) and incidence of rheumatoid arthritis (RA) in middle-aged and older women.
Data on diet were collected in 1987 and 1997 via a self-administered food-frequency questionnaire (FFQ). The risk of RA associated with dietary long-chain n-3 PUFAs and fish intake was estimated using Cox proportional hazard regression models, adjusted for age, cigarette smoking, alcohol intake, use of aspirin and energy intake.
The results show that, among 32 232 women born 1914–1948, 205 RA cases were identified during a mean follow-up of 7.5 years. An intake of dietary long-chain n-3 PUFAs (FFQ1997) of more than 0.21 g/day (lowest quintile) was associated with a 35% decreased risk of developing RA compared with a lower intake. Long-term intake consistently higher than 0.21 g/day (according to both FFQ1987 and FFQ1997) was associated with a 52% decreased risk. Consistent long-term consumption (FFQ1987 and FFQ1997) of fish ≥1 serving per week compared with<1 was associated with a 29% decrease in risk.
The authors concluded that this prospective study of women supports the hypothesis that dietary intake of long-chain n-3 PUFAs may play a role in aetiology of RA.
These are interesting findings which originate from a good investigation and which are interpreted with the necessary caution. As all epidemiological data, this study is open to a number of confounding factors, and it is therefore impossible to make firm causal inferences. The results thus do not led themselves to clinical recommendation, but they are an indication that more definitive research is warranted, all the more so since we have plausible mechanisms to explain the observed findings.
A most encouraging development for alternative medicine, one could conclude. But is this really true? Most experts would be surprised, I think, to find that PUFA-consumption should fall under the umbrella of alternative medicine. Remember: What do we call alternative medicine that works? It is called MEDICINE!
Linus Carl Pauling (1901 – 1994), the American scientist, peace activist, author, and educator who won two Nobel prizes, was one of the most influential chemists in history and ranks among the most important scientists of the 20th century. Linus Pauling’s work on vitamin C, however, generated considerable controversy. Pauling wrote many papers and a popular book, Cancer and Vitamin C. Vitamin C, we know today, protects cells from oxidative DNA damage and might thereby block carcinogenesis. Pauling popularised the regular intake of vitamin C; eventually he published two studies of end-stage cancer patients; their results apparently showed that vitamin C quadrupled survival times. A re-evaluation, however, found that the vitamin C groups were less sick on entry to the study. Later clinical trials concluded that there was no benefit to high-dose vitamin C. Since then, the established opinion is that the best evidence does not support a role for high dose vitamin C in the treatment of cancer. Despite all this, high dose IV vitamin C is in unexpectedly wide use by CAM practitioners.
Yesterday, new evidence has been published in the highly respected journal ‘Nature’; does it vindicate Pauling and his followers?
Chinese oncologists conducted a meta-analysis to assess the association between vitamin C intake and the risk to acquire lung cancer. Pertinent studies were identified by a searches of several electronic databases through December of 2013. Random-effect model was used to combine the data for analysis. Publication bias was estimated using Begg’s funnel plot and Egger’s regression asymmetry test.
Eighteen articles reporting 21 studies involving 8938 lung cancer cases were included in this meta-analysis. Pooled results suggested that highest vitamin C intake level versus lowest level was significantly associated with the risk of lung cancer. The effect was largest in investigations from the United States and in prospective studies. A linear dose-response relationship was found, with the risk of lung cancer decreasing by 7% for every 100 mg/day increase in the intake of vitamin C . No publication bias was found.
The authors conclude that their analysis suggested that the higher intake of vitamin C might have a protective effect against lung cancer, especially in the United States, although this conclusion needs to be confirmed.
Does this finding vindicate Pauling’s theory? Not really.
Even though the above-quoted conclusions seem to suggest a causal link, we are, in fact, far from having established one. The meta-analysis pooled mainly epidemiological data from various studies. Such investigations are doubtlessly valuable but they are fraught with uncertainties and cannot prove causality. For instance, there could be dozens of factors that have confounded these data in such a way that they produce a misleading result. The simplest explanation of the meta-analytic results might be that people who have a very high vitamin C intake tend to have generally healthier life-styles than those who take less vitamin C. When conducting a meta-analysis, one does, of course, try to account for such factors; but in many cases the necessary information to do that is not available, and therefore uncertainty persists.
In other words, the authors were certainly correct when stating that their findings needed to be confirmed. Pauling’s theory cannot be vindicated by such reports – in fact, the authors do not even mention Pauling with one word.
Tai Chi has been suggested to have many health benefits. Might it even prolong life? There are many enthusiasts who claim just that, but is there any evidence?
This study is a retrospective cross-sectional investigation to compare the rejuvenating and anti-ageing effects among a Tai Chi group (TCC) and a brisk walking group (BW) and a no exercise habit group (NEH) of volunteers. Thirty-two participants were separated into three groups: the TCC group (practicing TC for more than 1 year), the BW group (practicing BW for more than 1 year), and the NEH group. The CD34+ cell counts in peripheral blood of the participants was determined, and the Kruskal‐Wallis test was used to evaluate and compare the antiaging effects of the three groups. The results show that the participants in the TCC group (N = 10) outperformed the NEH group (N = 12) with respect to the number of CD34+ progenitor cells. No significant difference was found between the TCC group and the BW group. The authors of this study conclude that TCC practice sustained for more than 1 year may be an intervention against aging as effective as BW in terms of its benefits on the improvement of CD34+ number.
I was alerted to this new paper by several rather sensational headlines in the daily press which stated that Tai chi (TC) had anti-aging effects. So I searched for the press release about the article where I found the following quotes:
“It is possible that Tai Chi may prompt vasodilation and increase blood flow,” said Lin. “Considering that BW may require a larger space or more equipment, Tai Chi seems to be an easier and more convenient choice of anti-aging exercise.” “This study provides the first step into providing scientific evidence for the possible health benefits of Tai Chi.” said Dr. Paul R. Sanberg, distinguished professor at the Center of Excellence for Aging and Brain Repair, Morsani College of Medicine, University of South Florida, Tampa, FL. “Further study of how Tai Chi can elicit benefit in different populations and on different parameters of aging are necessary to determine its full impact.”
Personally, I find both the press release and the original conclusions of the authors quite amazing. If anyone wanted to write a textbook on how not to do such things, he/she could use them as excellent examples.
Seen with just a tinge of critical thinking the paper reports a flimsy case-control study comparing three obviously self-selected groups of people who had chosen to follow different exercise regimen for several months. In all likelihood they also differed in terms of life-style, nutrition, sleeping pattern, alcohol intake, smoking habits and a million other things. These rather tiny groups were then compared according to a surrogate measure for ageing and some differences were identified.
BIG DEAL!
To conclude from this, or even to imply, that TC has anti-ageing effects is as far-fetched as claiming the tooth fairy has money problems.
This story could be just funny or trivial or boring – however, I think, it is also a bit worrying. It shows, I fear, how uncritical researchers in conjunction with some naïve press officer are able to induce silly journalists and headline-writers to mislead the public.
After a traumatic brain injury (TBI) the risk of stroke is significantly increased. Taiwanese researchers conducted a study to find out whether acupuncture can help to protect TBI patients from stroke. They used Taiwan’s National Health Insurance Research Database to conduct a retrospective cohort study of 7409 TBI patients receiving acupuncture treatment and 29,636 propensity-score-matched TBI patients without acupuncture treatment as controls. Both TBI cohorts were followed for up to two years and adjusted for immortal time to measure the incidence and adjusted hazard ratios (HRs) of new-onset stroke.
TBI patients with acupuncture treatment (4.9 per 1000 person-years) had a lower incidence of stroke compared with those without acupuncture treatment (7.5 per 1000 person-years), with a HR of 0.59 (95% CI = 0.50-0.69) after adjustment for sociodemographics, coexisting medical conditions and medications. The association between acupuncture treatment and stroke risk was investigated by sex and age group (20-44, 45-64, and ≥65 years). The probability curve with log-rank test showed that TBI patients receiving acupuncture treatment had a lower probability of stroke than those without acupuncture treatment during the follow-up period (p<0.0001).
The authors conclude that patients with TBI receiving acupuncture treatment show decreased risk of stroke compared with those without acupuncture treatment. However, this study was limited by lack of information regarding lifestyles, biochemical profiles, TBI severity, and acupuncture points used in treatments.
I want to congratulate the authors for adding the last sentence to their conclusions. There is no plausible mechanism that I can think of by which acupuncture might bring about the observed effect. This does not mean that an effect does not exist; it means, however, that it is wise to be cautious and to not jump to conclusions which later need to be revised. The simplest interpretation, by far, of the observed phenomenon is that those patients opting to have acupuncture were, on average, less ill and therefore had a lower risk of stroke.
Having said that, the findings are, I think, intriguing enough to conduct further investigations – provided they are rigorous and eliminate the confounders that prevented this study from arriving at more definitive conclusions.
