Taking a risk over assessing outcomes in clinical research

At first glance, it would seem that the most important part of clinical research is making sure that you gather the right data to accurately measure the outcome you are testing. In some ways, it is the crucial stage, as it is not possible to properly perform the subsequent analyses without it.

However, you can develop a brilliant hypothesis, identify the correct cohorts in which to test it, and gather all the raw data you could possibly need, but your efforts will come to nothing if you don’t analyze and present the information in the right way. After all, clinical research showing that 30 out of 100 people with condition X have a mutation in gene A compared with 3 out of 100 healthy individuals is, in terms of its practical implications and clinical applicability, meaningless. It is not possible to infer a significant correlation, causality, predictive status or any other conclusion on such data. And such inferences are essential if your clinical research is going to have the impact it merits.

For a regulatory body to approve, a local health authority to sanction, a guideline body to recommend, and a clinician to prescribe a medication or medical device, they need to be able to understand how to apply your clinical research findings to a wider population. More specifically, they need to understand the likelihood of key outcomes to be able determine the absolute and relative effects of interventions or exposures.

There are a number of measures designed to aid this process, and include:

  • Absolute risk – the probability of a particular outcome occurring over a defined period in a given population
  • Absolute risk reduction – the difference in absolute risk between exposed and unexposed groups
  • Relative risk – the probability of an outcome in an exposed group divided by the probability of the same outcome in a unexposed group
  • Relative risk reduction – the absolute risk reduction as a percentage of the risk in the unexposed group
  • Odds ratio – the odds of an event in an exposed group divided by the odds of the same event in a unexposed group
  • Hazard ratio – the rate at which events happen in the exposed group divided by the rate at which events happen in the unexposed group. Hazard rations are frequently used when presenting results from clinical trials involving survival data.
  • Number needed to treat – an estimate of how many people would have to be exposed to a treatment in order for one person to avoid experiencing a particular outcome.

The preferred measures for summarizing clinical research data that compares treatment interventions are the relative risk, relative risk reduction, and hazard ratio. However, they should always be considered in relation to absolute differences, as ratios may appear to exaggerate clinical effects when the baseline occurrence of an event is low.

The odds ratio should also be used with caution when presenting clinical research, as it offers only an approximated estimate of the effect being studied, and it is more difficult to understand than other measures.

As an example of calculating the relative risk and relative risk reduction in clinical research, a recent study examined whether kidney cancer patients had a greater risk of non-cancer-related mortality than disease progression and cancer-related death [1]. Among 6655 renal carcinoma patients with a first lifetime cancer diagnosis, 7.0% of patients died of kidney cancer and 21.2% died of other causes over a median follow-up of 43 months.

Here, the absolute risk of death from kidney cancer is 7.0% and that of non-cancer-related death is 21.2%, giving an absolute risk reduction of 14.2%. The relative risk of kidney cancer death versus non-cancer-related death is 7.0% ÷ 21.2% = 0.33. Therefore, the relative risk reduction is 100 x (1 – 0.33) = 67%. As can be appreciated, the relative risk reduction, on its own, can give a different interpretation than when given with the absolute risk.

Contact us to find out more about our medical writing services. We can help you identify the key statistics that will reinforce your arguments and support your recommendations, in a clear, accessible manner.

 

References

  1. Kutikov A, Egleston BL, Canter D, Smaldone MC, Wong YN, Uzzo RG. Competing Risks of Death in Patients with Localized Renal Cell Carcinoma: A Comorbidity Based Model. J Urol 2012; 188: 2077–2083.

 

About Wesley Portegies

Wesley has over 10 years' experience as a marketing manager in the medical industry. He has successfully launched several products in the medical device market and has a great passion for sales and marketing.
0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *