How Can I Tell If Diabetes Research Is Valid?
New findings from medical research on diabetes are published nearly every week. Keeping up with the latest studies can help us maintain optimal health—if we read them with a critical eye.
Clinical Trials
Much of the research we come across involves clinical trials that test whether an intervention might be helpful and safe for patients to use:
- Some clinical trials are pilot studies that explore the potential efficacy and safety of a new drug or treatment.
- Other research involves well-planned, complicated, and lengthy clinical trials that measure an intervention’s effectiveness, and determine possible side effects.
Pilot studies are typically small, but positive results often lead to the larger, more complex clinical trials.
Questions To Ask
Whether reading about a pilot study, or an in-depth clinical trial, there are several questions we should be asking to assess the study’s validity. Seeking answers to these questions helps us become informed and discriminating healthcare consumers:
- How big was the study? The accuracy of a study is generally increased when many people participate. In smaller studies, there is a greater probability that observed effects are owed to chance.
- Was the clinical trial controlled? In typical controlled clinical trials, experimental treatments are compared to standard treatments with known effects, or to placebos. For instance, one group of participants will receive an experimental retinopathy treatment, and their outcomes are compared to the outcomes of a control group receiving either the standard retinopathy treatment, or a sham treatment (placebo).
- Was bias minimized? It’s difficult to avoid bias when running clinical trials, especially if patients, and (or) the investigators know which treatment - experimental, standard, or placebo - each participant is receiving.
- Ideally, study participants are randomly assigned to either the experimental or the control group, and are “blind” or unaware which group they are in. In rigorous double-blind studies, neither the participants, nor the researchers know which volunteers are getting the experimental intervention.
- Conflicts of interest? A study’s credibility is suspect when the investigators or research sponsors have a monetary or reputational “stake” in certain outcomes. Be attentive to who funds a study, whom the investigators work for, and look for corroborating research evidence from independent scientific sources.
- Any Replication? No single study can provide conclusive answers to scientists’ questions. Research results need to be replicated by different investigators using the same scientific methods, but with a new set of volunteers. Once studies are replicated, we can be more confident the outcomes are valid.
- Significant results? When research findings are “statistically significant” it means the different outcomes between experimental and control groups were probably not owed to chance. “Clinically significant” findings have to do with the size or intensity of a treatment’s observed effects.
- For example, a diabetes drug trial may report statistically significant differences (not owed to chance) between the experimental and control groups, but the differences may be clinically insignificant, or too small to consider the new drug useful.
It’s also important to look at the date of each study. If it’s more than a few years old, newer research might have altered scientists’ view of the drug or treatment. Curious people might utilize the National Library of Medicine’s database of published studies at the website PubMed (link below).
Sources: National Center for Complementary and Integrative Health; PubMed
Photo credit: Petras Gagilas