Introduction
As clinical landscapes evolve and technology advances at an unprecedented pace, clinicians find themselves at crossroads. They face the pressing dilemma of adhering to traditional, established practices or adapting and innovating in response to the rapidly changing world. This chapter explores the challenges and considerations in balancing these approaches, offering insights into how healthcare professionals can navigate this complex terrain.
Clinicians are increasingly at the mercy of the marketplace, facing constant pressure from sales representatives to adopt the latest products and technologies. This relentless influence presents a formidable challenge, forcing them to weigh evidence-based practices against emerging clinical advancements’ allure, and sometimes the urgency. Navigating this tension requires a careful balance of professional judgement, integrity and adaptability. Furthermore, the scarcity of high-quality, evidence-based information on certain new advancements or products often makes it easier for clinicians to inadvertently accept low-quality, non-evidence-based interventions or treatments. This reliance on unsubstantiated information poses a serious risk, as such interventions may not only prove ineffective but, in some cases, could potentially harm patients. The importance of rigorous evidence in clinical decision-making cannot be overstated, especially as the healthcare field continues to evolve rapidly.
It is essential for clinicians to be able to address two critical questions before initiating any treatment: ‘Why am I doing this for this patient?’ and ‘Is there evidence supporting a more effective approach?’ These questions serve as guiding principles, prompting clinicians to reflect on the rationale and efficacy of their chosen interventions, thereby reinforcing a commitment to patient-centred, evidence-based care.
This chapter emphasises the importance of evidence-based decision-making in the face of marketplace pressures and the need for clinicians to critically evaluate the rationale and effectiveness of their treatments. Figure 99.1 illustrates the different hierarchical levels of evidence in a pyramid shape, starting from lowest level of evidence and it increases towards the apex. The following key points are discussed to enhance the foundation of evidence-based medicine and dentistry.
-
i.
The two main types of study designs in evidence-based orthodontics: observational studies and experimental studies.
-
ii.
Key concepts such as bias and confounding factors, along with strategies to mitigate them. These strategies include randomization, strict inclusion and exclusion criteria, and statistical adjustments.
-
iii.
The importance of blinding in clinical trials to reduce performance and detection bias.
-
iv.
The role of systematic reviews in synthesizing existing literature to inform clinical practice.
The different levels of evidence are represented in a hierarchical pyramid, with study designs at the base considered the lowest level of evidence.
The level of evidence increases towards the apex, with studies at the top regarded as providing the highest quality of evidence.
Based on the concept of Forrest JL, Miller SA. EBDM in Action: Developing Competence in EB Practice. Colbert, Washington: ebdLibrary LLC; 2016 .
Study design
Study designs are broadly categorised into two types: observational and experimental. Observational studies focus on gathering data from participants or a sample group without introducing any active intervention. In contrast, experimental studies involve the application of an active intervention, which is typically evaluated against a control group for comparison ( Fig. 99.2 ).
Flowchart representing types of study design.
Importance of study design in evidence-based orthodontics
The foundation of evidence-based orthodontics lies in comprehending study designs. A well-structured study must be guided by a clear, concise and impactful research question, alongside a well-defined approach to participant recruitment, data collection and analysis.Collectively, these elements constitute the study design, ensuring the validity and reliability of the results, thereby enabling their application to clinical practice.
Types of study designs
Observational studies
This type of study design merely observes a phenomenon or a sample with no intervention, that is, to study people as we find them. Observational studies can involve data collected at a single point in time (cross-sectional) or data gathered over an extended period to track changes or outcomes (longitudinal) ( Table 99.1 ). Observational studies can be broadly categorised into descriptive and analytical types.
TABLE 99.1
Advantages and disadvantages of observational studies
| Advantages | Disadvantages |
|---|---|
|
|
Descriptive studies
Descriptive studies are akin to storytelling, as they provide detailed accounts of specific cases or situations. For instance, when a single case is described and interpreted, it is referred to as a case report. If similar cases are reported collectively, it is termed a case series.
However, a significant limitation of descriptive studies is the potential for selection bias. Often, cases may be selectively chosen to showcase the most successfully treated outcomes, which might not accurately reflect the broader clinical reality. This selective reporting can distort the findings, limiting the generalisability of the results and potentially creating a misleading impression of the treatment’s efficacy. Consequently, descriptive studies occupy the lowest level of evidence in the hierarchical pyramid of research evidence ( Fig. 99.1 ). Table 99.2 summarises advantages and disadvantages of descriptive studies.
TABLE 99.2
Advantages and disadvantages of descriptive studies
| Advantages | Disadvantages |
|---|---|
|
|
Analytical studies
Analytical studies involve a deeper understanding of a problem, requiring critical thinking to generate hypotheses and investigate processes to understand how and why certain phenomena occur. They aim to identify the relationships between different aspects of a disease and health conditions. These studies often compare different groups, such as healthy individuals and those with diseases, to explore causes and effects and to generate data. The data is analysed using appropriate statistical methods to draw logical conclusions. These are further classified into observational and experimental.
Cross-sectional studies.
Cross-sectional studies collect data from a specific sample at a single point in time, without any follow-up. These studies are particularly useful for identifying the prevalence of a condition, disease or risk factor within a population. While they are relatively easy to conduct, quick (the authors often refer them as ‘one-day studies’) and cost-effective, they are prone to significant biases and occupy a lower position on the evidence hierarchy. In orthodontics, many questionnaire-based surveys are examples of cross-sectional study designs.
Example: Ericson S, Kurol PJ: The authors evaluated 107 children with ectopic maxillary canines diagnosed in 2D and 3D (CT) to report on the incidence and severity of lateral incisor root resorption with impacted canines.
Case-control studies.
Case control studies are commonly used to assess factors that are associated with the disease or outcomes. In orthodontic research, they are often used to evaluate whether a particular treatment is more effective compared to a control group (e.g. no treatment) for a specific outcome. These studies do not involve any intervention; instead, they retrospectively assess risk factors or treatment success by comparing individuals who received treatment with those in the control group.
Example: Handelman et al.: The authors retrospectively evaluated the effects of RME in 47 adults, 47 children and compared them with 52 untreated controls. The paper evaluated non-surgical RME success in adult population.
Cohort studies.
Cohort studies are non-interventional in nature, involving the follow-up of patients either prospectively or retrospectively over a defined period to explore cause-and-effect relationships. The longitudinal design of these studies allows for a comprehensive understanding of the natural progression of a condition. This approach generates valuable data that contributes to evidence on risk factors, treatment effectiveness and the long-term impact of various exposures on outcomes.
Example: The Cardiff Study : This prospective cohort study evaluates the impact of orthodontics on quality of life (QoL) of individuals over the long term. Extensive data on QoL was collected around 11–12 years of age and at 30 years. The authors evaluated the effect of orthodontics on QoL at 30 years of age.
Experimental studies
Experimental studies involve administering an intervention to one group and comparing its effects with a control group (either untreated, receiving conventional treatment or subjected to another intervention). Unlike observational studies, where participants are studied as they are, experimental studies require the investigator to determine which participants receive the intervention and which do not. In orthodontic research, it is common to compare a new appliance (intervention) with either no treatment, a conventional appliance (control), or another type of appliance (alternative intervention). These studies are less prone to confounding factors. When the intervention is assigned randomly, the study is referred to as a randomised controlled trial (RCT). Conversely, if the allocation is based on the investigator’s discretion, it is termed a controlled clinical trial (CCT). The advantages and disadvantages of experimental studies are tabulated in Table 99.3 .
TABLE 99.3
Advantages and disadvantages of experimental studies
| Advantages | Disadvantages |
|---|---|
|
|
Randomised controlled trials.
RCTs are prospective studies designed to assess the effectiveness of an intervention or treatment. They hold a prominent position near the top of the evidence hierarchy, second only to systematic reviews and meta-analyses, due to their ability to minimise bias. Through randomisation, RCTs ensure that confounding factors are evenly distributed across groups, making them a powerful tool for examining cause-and-effect relationships between interventions and outcomes. A detailed exploration of RCTs will follow in the next section.
RCT are prospective studies that compare the effectiveness of different treatments or interventions with participants being randomly assigned to one of the two or more groups.
Controlled clinical trials.
CCTs share similarities with RCTs in that they are prospective studies, but the key difference lies in how participants are allocated to intervention groups. In CCTs, the allocation of the intervention is determined at the discretion of the investigator, rather than through randomisation. Like RCTs, CCTs include both an intervention group and a control group (either no treatment or placebo), with both groups being followed concurrently. However, because the allocation is not random, CCTs are susceptible to selection bias. While randomisation is generally preferred, there are situations where it may not be possible or ethical to randomise participants. For instance, testing the effect of smoking on oral hygiene would be unethical (or even impossible) to randomise participants to smoking groups.
Example: In a study by Lund and Sandler, the authors prospectively evaluated the effects of twin block appliance and compared the results with no treatment controls. The study included 73 adolescent patients requiring twin block and treated 36 of them with an intervention (Twin block), and 27 of them acted as no treatment controls. The study pointed out the side effects of twin block appliance, that is, proclination of lower incisors.
Bias in research
Bias refers to a systematic error or flaw in the study design, methodology or conduct that can influence the results of a study. It can arise at any stage, including study design, execution, analysis or reporting. Bias impacts the validity of the findings and can often be identified by researchers, regardless of their expertise in the field.
For instance, selection bias occurs when participants are selectively recruited to include only the most successful cases while excluding failures, thereby skewing the results. Common types of bias include selection bias, detection bias, performance bias, attrition bias and reporting bias.
Eliminating bias requires careful planning, including a robust study design and a well-written protocol developed before the study begins. Importantly, once bias is introduced into a study, it cannot be corrected through statistical methods. In the following sections, we will explore various types of bias in greater detail and discuss strategies to minimise them.
Confounding factors in research
A confounder, or confounding factor, refers to a variable other than the intervention under study that can influence the study’s outcome. Identifying confounders often requires expertise in the relevant clinical field. For instance, when evaluating the effect of electric versus manual brushing on oral hygiene during orthodontic treatment, smoking could significantly influence the results. If not accounted for, this variable may distort the findings and lead to incorrect conclusions.
Strategies to manage confounders
Stay updated, free dental videos. Join our Telegram channel
Leave a Reply
VIDEdental - Online dental courses
