'One of the first duties of the physician is to educate the masses not to take medicines' William Osler

“One of the first duties of the physician is to educate the masses not to take medicines’ Sir William Osler said just over 100 years ago. How much has changed since then?

The use of prescription medication has been rising dramatically in recent years. See CDC.gov US prescription statistics.  48% of patients in the US are on regular prescription medication. The number of patients taking 5 or more drugs rose from 6 to 11 percent between 2000 and 2008. In the over 60’s 37% take 5 or more different prescription drugs regularly.  Figures from the UK are similar.

Although there are some over 75’s who take no long term medication, they are rare. The majority take medication for a combination of chronic conditions like high blood pressure, reflux symptoms, joint pain, low bone density, angina, irregular heart rhythm, cholesterol, depression, peeing too much or too little, etc.

Some of this medication is intended to control symptoms, but the majority of drugs are taken in an effort to prevent future illness such as a stroke. The evidence for the effectiveness of this strategy is generally derived from research trials and evaluated by organisations such as the National Institute of clinical excellence, and similar organisations around the world using complex statistical techniques. This process is called meta-analysis (translation ."further analysis"), and the result is supposed to be so called Evidence Based Medicine in which we no longer rely mainly on the opinions of eminent experts, who may be biased, but are guided by an objective view of the available research conducted by a dispassionate and numerate expert.

Guidelines on preventative treatments are also produced by other organisations such as Royal Colleges of a particular specialty, and also by special interest groups such the British Heart Association.

The guidelines pretend that patients have only one long term condition, and take no account of multiple illnesses, even though patients with only one long term problem are a small minority.

Many patients will have for example, high blood pressure and diabetes and heart disease and joint pain and acid reflux, but the guidelines say nothing about how the treatment of any of these problems should be influenced by the others. As a result patients with terminal cancer are commonly still taking drugs designed to reduce a fairly small risk of a heart attack.

Guidelines calling for increases in the use of a particular drug treatment are widely promoted by the pharmaceutical industry, who advertise in journals, sponsor conferences on the relevant disease, finance health charities which increase awareness of the condition in the population, and lobby politicians. In recent years there has been a lot of publicity about dementia. It is unlikely that this would have happened without the potential for drug makers to gain huge amounts from the sale of new drugs for this problem. The fact that the drugs are of very doubtful benefit, has made the promotion even more vital to the makers profits. Pfizer issued a legal challenge to NICE and forced them to alter the advice that limited the use of the new drugs to only a few patients.

As well as the promotion of guidelines being heavily biased, the evidence on which they are based is also tilted in favour of the use of medications. The original research is mostly done by the drug firms, who own the data, do the sums, and then get the 'authors' to help write it up. However the conclusions are biased in various ways, as shown by ben goldacre in his book Bad Pharma. See wikipedia.org/wiki/Bad_Pharma The main techniques are:

 1 to do several studies and only report those with the best results, as has happened with Roche and trials of the flu drug called tamiflu.

2 to do lots of different statistical tests on the results, and only report the ones that come out best, pretending that that was the so called primary outcome you had set out to test.

3 losing inconvenient patients to follow up so their outcomes are not included. Merck did this with Vioxx which worked for arthritis but caused heart attacks. 

4 doing repeat analyses during the course of the trial and reporting the results when they are best, pretending that the trial was shorter than it actually was. Pfizer did this with Celebrex, where the trial lasted a year but they only published six months.

5 Giving all patients a pre trial course of the drug and excluding those who had side effects from the main trial.

The result of the above problems is that according to many doctors, Evidenced based medicine is largely broken and its conclusions should be treated with suspicion.

Another major problem with the guideline approach is that the patients in the trials are almost always very different from the patients who end up taking the drug. Trial patients are much younger and fitter, with more effective livers and kidneys to break down the drugs, and much less likely to have significant other conditions. They are also usually banned from taking significant amounts of other medications.

Licensing authorities around the world are aware of these problems and often ask producers to do so called post marketing studies, but these requests are usually ignored. Sometimes it happens that another organisation does some research into a new drug looking for other uses and discovers  harms. For example researchers investigating whether the newer selective arthritis painkillers might prevent bowel cancer found a large increase in heart attacks. In general there is very little systematic research into drug side effects and harms.  Most well known drug harms that we are aware of today such as kidney damage from anti inflammatory drugs, or diabetes from drugs for mental illness went unreported in the initial trials, and were only discovered after several years of widespread use. Other potentially serious harms are not being researched. For example we suspect that the main drugs for brittle bones may cause cancer of the oesophagus, that blood pressure drugs cause falls and broken hips, and that newer diabetes drugs cause deaths through pancreatitis and possibly pancreatic cancer, but the relevant companies are not forced to research these issues. Most glaringly, after 30 years of the use of Statin drugs by hundreds of millions of people, we are still very uncertain about the frequency of muscle pain and memory problems, that are reported by many patients but dismissed as nocebo (negative placebo) effects by those who advocate their even wider use. Similarly almost as many people take so powerful drugs called PPIs to reduce stomach acid and relieve heartburn. We know that these drugs increase broken bones, and hugely increase the chance of contracting the hospital superbug C. Diff, but we have very little idea of the scale of the problem. 

Even when we are aware of harms these are rarely communicated to the patient. One example of  a very 'dirty' drug category is the so called 'anticholinergic' class which has been around for a long time and is most often used to reduce urine frequency by inhibiting the main bladder muscle. These drugs have a long list of side effects which are almost universal, including constipation, irregular heart rhythm, impaired memory, blurred vision, dry mouth, but they are often prescribed by specialists who focus on the bladder and ignore the rest.

Similarly patients are discharged from hospital on a long list of medication that is seldom explained in detail even when there are likely to be significant harms. For example many patients who have an acute illness have a temporary disturbance of heart rhythm called atrial fibrillation. The hospital doctors have no idea whether this a long existing condition or a temporary one. There is an increase stroke risk with this condition of roughly 5% per year, and the doctors want to do anything they can to reduce that so they prescribe blood thinning treatment. The patients will have to have frequent blood tests, will be in increased danger if they have an accident, will not easily be able to have operations, and will be unable to take many other drugs which affect the way the blood thinner works. On top of this there is a significant risk of roughly 1/1000 per year of bleeding to death. This surely requires detailed discussion and yet this rarely occurs. Once it has been started it is a brave GP who stops it. If the patient has a stroke it will be his fault.

Another factor increasing drug use is the so called Quality and Outcomes framework (QOF), or 'points mean prizes' for GPs. This system has been going for 10 years or so and rewards the achievement of targets which can be for example control of blood sugar in diabetes or simply the prescription of certain drugs such as blood thinners for patients with Atrial Fibrillation. GPs can exempt patients for various reasons but have to be prepared to justify it. This so called ‘exception reporting’ is closely monitored with the possible sanction of payments being withdrawn and a bad report from the Health inspectors. Not surprisingly people find that it is easier simply to prescribe the drug. Patients who are included in QOF are usually reviewed by nurses, who often lack detailed pharmaceutical knowledge.

Young doctors in particular find it hard not to follow guidelines even when they have a good reason to ignore them. All the guidelines state in their preamble that they are to be interpreted in the context of a particular patient. But in practice, when complaints are made, or inspections look at particular case reports there is a heavy onus on the doctor or nurse to explain why they have not followed a particular guideline.

Lots of eminent doctors are aware of the harms of the present degree of polypharmacy, especially in the frail elderly who are much more vulnerable to harm. Several trials have been done of stopping multiple medication in the frail elderly, often with great benefit both in terms of quality of life and also survival. But in practice very little of this is done. It is rare to find specialists stopping long term treatment, and although practices have annual drug budgets, these are ignored as there is no sanction if they are exceeded. Patients are supposed to have medication reviews at least every year, but it is common to see patients who have been taking a medication for many years despite not knowing what it is for and despite the original reason having long since resolved. Doctors are busy and it is easier not to raise the subject when going through a list of 9 or 10 drugs would take far longer than the 10 minutes allocated to the consultation.

So what have the overall results of this huge increase in resources devoted to long term medication.? The short answer is that we do not really know. Long term trends in the number of heart attacks and deaths from various conditions have continued without for example showing any evidence that long term outcomes have been improved by QOF.

So benefits zero to low. What about harms?

Money that could have been used elsewhere has been diverted, both to the direct cost of the drugs but more significantly to a huge increase in the number of medical appointments for long term conditions. Patients have certainly suffered adverse effects but they are impossible to quantify.

Life expectancy in the uk has shown the first downwards blip in a long upward trend which has been going on for generations. Perhaps polypharmacy is responsible.