Correlation, causation or confounder

After drinking three bottles of beer, I came home sober at around 3 am, to find my wife who had been drinking tea, awake and waiting for me. That night, I realised tea is dangerous. It brews violence.

Whilst I lay in bed, not uttering a word, she kept abusing me and shouting on top of her voice till I finally slept and had some peace of mind. Tea is dangerous. Avoid it.

The first paragraph is a paraphrased joke I read, which on the face value, sounds absurd.

A woman had a BP of 240/140 and was visited by a Doctor who gave her chilli, 1 raw banana, rock salt and tap water. After 5 minutes, her headache was gone, and she walked without the support she previously required. So, when BP is rising all you need is chilli.

This sounds more believable, doesn’t it?

Due to the proliferation of social media and the patronage of WhatsApp especially by both young and old, educated and not educated, (by this I mean certain information will sound differently to different people depending on their educational bias. E.g. In matters of religion, I am more likely to believe what I read because I am naïve or uneducated), a lot of information will be shared, by a lot of people to groups and individuals, without recourse to verifying the message. And yes, adding ’shared as received’, doesn’t absolve you of blame if your post does harm to a follower of your message’s dictum in the same way as leaving DDT in close proximity to a toddler, with ‘TOXIC’ inscribed in red ink on the bottle, together with a picture of a skull and two femurs also in red, doesn’t make you less blameworthy if the child ingests its.

Now to medicine, which I studied. There is a concept referred to as, evidence-based medicine, which originated as far back as Aristotle’s days and is currently gaining prominence again. It requires practitioners to use the best available medical evidence in making clinical judgements. In the absence of proof, anything which sounds right will be imbibed and put into practice. This may prove to be fatal!

In the practice of medicine, which is a science, and because human life is precious, we go to great lengths in testing hypotheses before putting them into practice. This means that when a revolutionary idea such as stopping antihypertensive medication and adopting chilli to control very high blood pressure is postulated, it would have gone through a clinical trial.

A very good trial involves human volunteers and evaluates safety and treatment benefit of a new drug- the larger and more diverse the people involved, the better. It will involve clients who have high blood pressure (not dangerously high) and those who don’t. Clients are however not denied the benefits of their usual treatment. In other words, no client is harmed during the trial. It will involve giving some volunteers chilli and others a placebo. Only when there is significant benefit from using chilli, will it be recommended for hypertensive patients.

In other words, the mere fact that some remedy has worked for someone does not guarantee it will work for another. Otherwise, I would recommend telenovelas for headaches, because I am sure when Abi finally marries Pragya and all other villains get their due recompense and the painfully long series comes to an end, a lot of people will be relieved of their headache.

Another instance where this concept is worth mentioning is when someone tests positive for hepatitis b, receives treatment, prayers or herbal preparations, and later tests negative. Naturally, all praise will be due to intervention. However, as WHO puts it, ‘less than 5% of healthy persons who are infected as adults will develop chronic infection’. Therefore, the question to ask is, did the virus vanish on its own or the drug worked? From what WHO says, in the acute phase for most adults, the virus will go on its own.

The fact that two things are related doesn’t mean that one resulted in the other- correlation or association is not causation. The case of taking chilli for hypertension is like attributing having a dog to reducing the risk of heart attacks. In reality, however, the real reason for the reduction in heart attacks is that having a dog increases your chances of exercise which in turns reduces your risk of heart disease. It will be misleading to attribute the reduction in risk of heart attacks to dogs.

Also sometimes, certain experiences are caused by confounders which if not accounted for in our analysis, will lead to wrong conclusions- another reason why we can’t trust all WhatsApp doctrines.

If a Doctor says I have diabetes and I stop attending hospital, start taking bitter leaf, according to alternative medical practitioner’s advice and on top of that, based on an alluring TV advert, start taking a herbal preparation and when I go back to the hospital in 3 months, my blood sugar is normal, do I simply say that the bitter leaf or herbal mixture is the reason?

Firstly, healthy lifestyle including diets rich in vegetables is recommended even before commencing drug therapy. In other words, Doctors recognise that lifestyle change alone can solve the problem of diabetes and hypertension (e.g. DASH diet). Secondly, people who are diagnosed with such chronic conditions automatically adopt healthier lifestyles, including an exercise regime.

Based on the above, many will say and perhaps even recommend the herbal preparation to all diabetics, and further dissuade them from taking antidiabetics because they were healed through other than orthodox medicine.

Such a conclusion, based on aforementioned reasons will be fallacious.

Before we adopt new therapeutic options, we must be guided by evidence, and the onus of providing proof must rest squarely on the person forwarding the message. Other ways of identifying such flawed messages include author’s name not written, no citation and no useful link for further reading.

Orthodox medicine has its own problems regarding researchers’ interests, which sometimes lead to a desire to produce data that is favourable and not inimical to big companies’ interests, especially those in the pharmaceutical sector, who usually fund research projects. Perhaps, such a discussion will be appropriate for another occasion.

Till we find evidence to controvert large, randomised, double-blind control trials, systematic reviews or meta-analyses of randomised controlled trials, let’s stick to them. If chilli works for BP of 240/140 without proper evidence to back it, so does Kumkum Bagya relieve a headache.