6 Tips for Finding Quality Cancer Research
The Internet has become a rich source of information for cancer patients. It exposes a plethora of research and opinions that would otherwise only be available to medical professionals allowing patients to play a more active role in their treatment.
Unfortunately, the gold nuggets that could extend, and possibly save our lives, are hidden in a quagmire of dubious, dangerous and misleading information. So how can we use this fantastic, but unregulated resource to its maximum value and spot the clickbait, charlatans, crooks and flat-earthers?
In this blog, I will be giving you a six-point checklist that will help you assess the reliability and validity of any claim or opinion you find on the Internet.
1. Identify the Source
Hardly a week goes past without a sensational headline espousing a new cancer cure or miracle treatment. If these headlines were actually true, we’d have cured cancer thousands of times over.
Often there is some scientific basis for the headline, but in the search for a compelling headline, the science can become manipulated and simplified so that it no longer accurately represents the findings of the original study. Good robust science gets morphed into a gimmicky and misleading article that sells papers whilst raising false hopes. Take for example the Daily Mail’s 2013 claim that “Chocolate and fizzy drinks could be used as cancer detectors because malignant tumours feed off sugar”. The article is an inaccurate extrapolation of a research paper published by UCL which demonstrated the use of glucose as an alternative to current contrast agents used in MRI scans. Cancer Research UK were quick to bust the hype by explaining that the experiment would require participants to fast before their scans and then drink a salt and glucose solution rather than a Mars Bar and a can of Coke. Furthermore, the claim that tumours feed off sugar is an oversimplification of a very complex process. Whilst good for our overall health, science is a long way from claiming that cutting down on our sugar intake affects the progression of our cancer.
Luckily, newspapers and other non-specialist publications are good at citing sources on which an article is based. These can be accessed (usually without charge) on research database sites such as PubMed, NCBI, NCRI or NCI. Even if the sources aren’t listed in the original article, a simple Google search should be enough to find the basis of their claims. Of course, if there are no sources cited and you’re unable to find anything using a Google search, the article is almost certainly ‘fake news’.
2. Look for Large Studies
Findings based on studies with a small number of subjects generally hold less weight than those based on a larger population size. There are two reasons for this:
Firstly, larger studies reduce the possibility of chance. This can be illustrated by a heads/tails analogy. The chances of flipping 4 coins and them all landing on heads is about 6%. Double the number of coins, and the chance of them all landing on heads is just 0.3%. As the number of coins increases, the element of chance is reduced.
Secondly, larger studies reduce the possibility of other factors influencing the results. For example, a study of 10 randomly-selected people who reported drinking 3 or more litres of water per day and who reported less sickness-based absence compared with the general population may conclude that increased water consumption has a positive effect on the immune system. But what if all 10 of these people also ate a vegan-based diet or were keen runners? This is quite possible, given the small study size and it would be impossible to isolate the effect of their water consumption from other lifestyle factors they had in common. Now increase the study sample size to 1000 - it is virtually impossible that all the subjects observed the same diet, exercised regularly or shared any other habit that might influence results.
3. Controls and Structure
Well-designed studies are well structured and include carefully designed control measures so that the researchers can more accurately isolate cause and effect. A good quality study will detail their methods in the study report. Methods that suggest a high level of control include:
Control Group - Where possible, a control group should be formed containing individuals who do not receive the treatment or intervention on trial. Observations recorded in the control group are used as a benchmark to measure differences between them and the experimental group.
Double-blind - To eliminate the risk of researcher bias in a study containing a control group, subjects are assigned to either the control arm or experimental arm without the knowledge of either the researchers or the subjects.
Another important distinction is that between statistical significance and clinical significance. A study may be statistically significant, but it may lack the controls, structure or scale to be clinically significant. In other words, the results are not robust enough for it to change clinical practice.
4. Conflicts of Interest
Clinical research is expensive. Even a small trial involving 100 patients can cost $6million, with a trial of 1000 patients costing an average of $77million. These costs are generally beyond the reach of public bodies and universities. However, for multi-million pound drug companies, these costs are very manageable and represent a tiny fraction of the overall drug development cost. This is why many larger studies are sponsored by pharmaceutical companies who have a vested interest in their results. In fact, it is often claimed that drug companies would conduct multiple trials of the same drug and only publish the results of those trials that aligned with the business goals of the organisation. This practice is now illegal, but the fact remains that many large-scale trials are sponsored by the pharmaceutical industry.
Most medical journals will declare sources of funding for the studies they publish. Studies funded by the pharmaceutical industry are often highly relevant and informative, but as a consumer, you should be aware that they may be influenced by the prospect of a huge financial return for shareholders.
5. Spot Patterns
Perhaps the strongest evidence supporting a piece of research is the presence of other published research with similar results. The absence of similar research, should invite scepticism.
Strongly debated hypotheses such as the link between sugar and cancer will also be the subject of meta-analyses and systemic reviews. These are research papers that attempt to collate research from a large number of studies to form an overall conclusion. They are usually independent, funded by public bodies and universities rather than pharmaceutical companies so are less prone to bias.
6. Age of Study
Science, especially in the oncology field evolves quickly. As our knowledge increases and new theories are developed, historical papers lose relevance and accuracy. When reading older papers, it is important to search for more recent publications to verify that the conclusions are still relevant in today’s context.