Master the Vocabulary: A Single-Case Design Terminology Guide

Welcome to the world of Single-Case Designs (SCD), a unique family of research methodologies with a singular focus – the individual. These designs revolve around the collection of repeated, systematic outcome measurements (e.g. pain) over time, allowing for conclusions tailored specifically to each participant. You might encounter single-case designs under various names, such as single-case studies, single-case designs, single-participant studies, or single-subject designs. It's crucial to note that all types of single-case designs go beyond the realms of "qualitative" or "descriptive" case studies. Instead, they represent prospectively designed quantitative studies, rigorously collecting data to unveil valuable insights about each individual participant. In this blog we explore the intricacies and significance of single-case designs.

Observational vs. Experimental Designs

Within the realm of single-case designs, there exist two main categories: Single-Case Experimental Designs (SCEDs) and Single-Case Observational Designs (SCODs). We will take a look at some of the main types of SCEDs first.

N of 1 Trials

N of 1 trials are the most rigorous type of single-case experimental design. They involve randomly allocating different time periods within an individual to repeated intervention and no intervention (or an alternative intervention) periods and then comparing outcomes across the different time periods. Where possible, blinding is incorporated into the design. N of 1 trials are also known as N-of-1 RCTs, individualised medication effectiveness tests (IMETs), personalized trials, and single-patient trials. In clinical practice, an unblinded and usually nonrandomised variation called a single-patient open trial (SPOT), can be valuable as it is logistically easier to conduct.

Figure 2. Simulated data illustrating an A-B-A-B design.1

Alternating Treatments Design

Alternating treatment designs consist of rapid and random (or semi-random) alternation of two or more conditions such that each has an approximately equal probability of being present during each outcome measurement. These types of studies provide information about how an individual responds to one treatment relative to another.

Figure 3. Simulated data demonstrating an alternating treatment design to compare two treatments. The treatment phase may be preceded by a baseline phase (1)

Changing-Criterion Design

The changing-criterion design can represent an experimental or a quasi-experimental design depending on the amount of experimental control there is. Gradual changes in performance over time demonstrate experimental control: stepwise changes in criteria predict concurrent changes in performance.

Figure 4. Simulated data demonstrating a changing-criterion design. Vertical solid lines indicate phase changes. Dashed horizontal lines indicate performance criterion level for a given treatment phase. Phase change occurs when the criterion has been met, in this instance on three consecutive occasions (1)

Multiple Baseline Design

In a multiple baseline design, treatment commencement is staggered (i.e., started at different times) across individuals. This strengthens the conclusion that any changes observed are due to the treatment as opposed to external factors. We start by measuring an outcome of interest repeatedly, then applying a treatment when the “baseline” is stable, and then repeatedly measuring that outcome again. If outcomes change in all participants after they receive treatment, this suggests the treatment is effective.

Figure 5. Simulated data demonstrating a MBD across three different behaviours.1

Single-Case Observational Designs (SCODs)

SCODs (also known as N-of-1 observational studies), involve repeated measurement of an outcome in an individual without applying an intervention. Therefore, SCODs can help us draw conclusions about naturally occurring patterns and predictors of outcomes. The data from a SCOD can be used to tailor treatments and interventions so that they target unique predictors of outcomes for individual patients.

Figure 6. Graphed SCOD data showing fluctuations in body pain and physical activity over time experienced by an individual with a chronic condition (2)

Self-Study Designs

Both experimental and observational SCD studies can have a ‘self-study’ design, where an individual conducts the study on themselves, to answer research questions they have generated themselves (“citizen scientist”). Technology now enables informed self-measurement for “citizen scientists”. The Quantified Self is an international community of users and makers of self-tracking tools (usually digital) who share an interest in “self-knowledge through numbers.”

Figure 7. Lines show finger tapping results (right and left hand) from 12th (day 1) and 13th (day 2) of March 2012. Bars show medication intake times.3

Pooled (aggregated) single-case studies

In a pooled (also known as aggregated) single-case design study, data from multiple studies are aggregated to provide a population-based effect estimate. Each participant contributes multiple data points to the analysis, rapidly increasing statistical power. Research has shown that pooled single-case designs achieve statistical power equivalent to a traditional group-based parallel randomised controlled trial, but with far fewer participants. Various statistical approaches can be used to analyse pooled single-case studies including Bayesian modelling, hierarchical linear modelling, and multi-level modelling.

Figure 8. Pooled N-of-1 trials contributing multiple datasets to a virtual RCT. Effectively, each participant provides multiple datasets to each side of an RCT (4)

We hope this brief overview has made the different terminologies describing this powerful and flexible design clear.

N-of-1 Hub is a consulting company that specialises in designing, conducting and analysing personalised clinical studies using single-case designs. N-of-1 Hub provides consulting and collaborative services to companies, clinicians, researchers, and healthcare consumers who wish to conduct personalised clinical studies using single-case designs. N-of-1 Hub also offers data management and analysis services and customised workshops and courses. Additionally, our training programs are tailored to equip researchers and clinicians with the skills needed to excel in personalised clinical studies.

For more information on how N-of-1 trials can revolutionise your clinical research and improve patient outcomes, we invite you to contact us for a free consultation. Don’t forget to follow us on LinkedIn for the latest updates and insights in the field of personalised clinical studies.

References

(1) Tate, Robyn L. and Perdices M. N-of-1 Trials in the Behavioral Sciences. In Jane Nikles and Geoffrey Mitchell (Ed.), The essential guide to N-of-1 trials in health (pp. 19-41). Dordecht, Netherlands: Springer.

(2) McDonald, S., Vieira, R., & Johnston, D. W. (2020). Analysing N-of-1 observational data in health psychology and behavioural medicine: a 10-step SPSS tutorial for beginners. Health Psychology and Behavioral Medicine, 8, 32-54.

(3) Reproduced from Sara Riggare and Maria Hägglund. Precision Medicine in Parkinson’s Disease – Exploring Patient-Initiated Self-Tracking. J Parkinsons Dis. 2018; 8(3): 441–446.

(4) Mitchell, Geoffrey (2015). In Jane Nikles and Geoffrey Mitchell (Ed.), The essential guide to N-of-1 trials in health (pp. 57-66) Dordecht, Netherlands: Springer.

(5) Nikles J, Onghena P, Vlaeyen JWS, Wicksell RK, Simons LE, McGree JM & McDonald S (2021). Establishment of an International Collaborative Network for N-of-1 Trials and Single-Case Designs. Contemp Clin Trials Commun, 2, 23:100826.