Clinical guidelines tell doctors what the best treatments are, based on the latest and most reliable research. The guidelines are based on the best available evidence from medical research.
A strong recommendation must be based on reliable evidence.
This is called evidence-based medicine. It means that doctors look at research to help with treatment decisions, rather than relying on judgement, experience, and medical training alone.
Decisions needs to be based on the best available research.
Some studies can be wrong or misleading, possibly due to poor scientific design. Poor evidence is not a reliable guide for clinical decision makers.
To establish which evidence is trustworthy, studies are graded according to their reliability. Study design methods are graded from poor to high quality.
Expert bodies such as the World Health Organization (WHO) and the Centers for Disease Control and Prevention (CDC) make decisions about the best available medical interventions. They produce guidelines for health care professionals to use.
These bodies use a system of grading known as the “GRADE” system, developed by the Grading of Recommendations, Assessment, Development and Evaluations (GRADE).
The decisions are endorsed by panels of leading doctors from the relevant field. The panels produce sets of recommendations after reviewing all the research studies.
The doctors review the best health care programs and medical treatments, they look at each piece of medical research, and they give them approval ratings.
Doctors want to be confident in the treatment they offer. Stronger evidence makes the guidance more influential. The ratings reflect how good the evidence is that underlies each recommendation.
The GRADE system is widely used in the United States and internationally. It is one way of setting ratings when reviewing the strength of medical research evidence.
It works on the basis that good scientific study design gives strong medical evidence. There are three grades:
- High-grade evidence is based on randomized controlled trials (RCTs) and systematic reviews of RCTs
- Low-grade evidence is taken from observational studies
- Very low-grade evidence is any other evidence.
Some types of study design are more robust than others. They may involve more people, or they may incorporate findings from a wide range of studies.
The following types of scientific study design are listed in order of increasing robustness. A case study is the least robust, a meta-analysis, the most.
Following the links will take you to other articles for more detail about each one.
A case report is rather like a story. It may give an account of an individual case. It is normally used for an unusual situation, such as an unusual disease or complication. It tells people about something new that the practitioner has learned in clinical practice. It could also describe a novel approach to managing a common condition.
A case report is considered low-grade evidence, but it can trigger new ideas and new research.
A case-control study is an observational study that looks back and compares a group of people who have had a disease with people who have not had the disease. Those with the condition represent the “case.” Those without are the “control.”
A case-control study may look at how often a risk factor occurs in each group, in a search for links between the risk factor and the disease.
It is useful for looking at rare conditions, but since it relies on memory, people may make mistakes in recalling details.
A cohort is a large group of people. A cohort study looks at factors in a “group of people with defined characteristics who are followed up to determine incidence of, or mortality from, some specific disease, all causes of death, or some other outcome.”
Researchers first identify an outcome, say cardiovascular disease. Then, they follow a specific population over time, and they note when the disease occurs.
The long time frame enables them to assess possible causes of the condition. It can provide strong scientific evidence, but it can take a long time to gather results.
Randomized controlled trial
A Randomized controlled trial (RCT), according to the BMJ, is “The most rigorous way of determining whether a cause-effect relation exists between treatment and outcome and for assessing the cost effectiveness of a treatment.”
In an RCT, participants are randomly allocated to groups, and they are normally unaware of which treatment they have received until after the trial.
Apart from the intervention under study, conditions must be the same for all participants.
The aim is to measure how the groups differ, in terms of the predefined outcome.
One drawback is that it may be unethical to treat any patient in a way that could be inferior to a method that is already in use and known to be effective.
Most studies review previous research, but they can only select a few items. This can lead to bias and skewed results.
It uses explicit and systematic methods to establish what can reliably be said about the issue. The findings are then synthesized to make them easily for policymakers to access. The aim is to minimize bias.
The review may include judgments and recommendations.
A meta-analysis combines the findings of a range of studies, both qualitative and quantitative, to produce a single conclusion.
The conclusion of a meta-analysis is more powerful than that of a single study, because it involves large numbers of diverse participants and an accumulation of effects and results.
It can incorporate findings from smaller studies, that might not be significant on their own, into a bigger, more complete picture. It can give a more precise analysis of, for example, the safety and harms of an intervention.
Good medicine is not just about research.
Medical research is important, but it is not always applicable. People are individuals, and what is true for most people may not be true for everyone.
Patients’ preferences also play a role in the effectiveness of medicine.
An experienced clinician will use guidelines based on reliable evidence to help a patient choose the best solution within a given context.