Often overlooked in primary care research is the single subject research design. A single subject research design can be used to study the time course, variability, or effect of an intervention or treatment on a single patient. This experimental research design involving multiple measurements over time on a single subject (n = 1) has been labelled as a clinical trial of n = 1, a randomised clinical trial in a single patient, a within subject design, patient care study, A-B, or a single subject design.1–9
When conducting patient care, a physician is most probably treating the patient using a single subject paradigm. Typically, a patient presents to a physician with a complaint or symptom. As a result of the history, examination, and medical testing by the physician, for the patient’s presenting complaint, an intervention or treatment might be started. Evaluation is then made of the effectiveness of the treatment in eliminating or modifying the complaint or underlying identified condition. When conducting a single subject research design as compared with patient care, sound research methodology would need to be included. This would include standardisation of the intervention including the assessment of compliance, objective measurement of the outcomes, procedures to minimise bias such as blinding or second observers, and so forth. With the implementations, it can be argued that patient care is a special case of a single subject research design.
The randomised parallel group clinical trial design is usually considered as the gold standard in clinical research. The results of these trials provide information on the average treatment effects for the studied populations and serve as guidelines for evidence based medicine. At times, the results from these trials might not always be applicable in determining the most effective treatment for an individual patient: (1) patients might be different clinically from those that participated in the trial, (2) the heterogeneity of the patients in the trial may limit the generalisability of the results to particular patient populations, and (3) the strict criteria for trial participation might be of limited applicability to a general clinic patient populations. To examine the applicability of the results of a clinical trial to a specific patient, a single subject research design could be used.
An Institute of Medicine report10 has provided some guidelines as to the use of these small clinical trials. Specifically, warranted situations might include rare diseases, unique study populations, individually tailored treatments, isolated environments, emergency situations, and public health urgency.10 Practice based primary care research, commonly, encompasses individually tailored treatments (for example, glycaemic control), isolated environments (for example, rural health), and unique study populations (for example, HIV infected pregnant adolescent).
A patient care or an experimental single subject research design can have multiple periods of measurement and multiple times for intervention or treatment. At the onset of each of these designs, a series of baseline (A) observations are taken to assess the patient or subject in the initial or diseased state. A course of treatment is prescribed, denoted by the treatment or (B) phase. The patient or subject continues with the treatment, while assessing the effectiveness across time, using the same outcome variables that were used during the baseline phase. This describes the primary A-B single subject design. Other designs might include removal of the intervention or treatment referred to as washouts, the second baseline (A), a readministration of the treatment (B), or different treatments (C). The research question of interest should guide the single subject design used in terms of combinations of baselines and treatments (A-B, A-B-A, A-B-A-B, A-B-C, etc). Depending upon the research question of interest, varying combinations of phases of observation consisting of baseline, washout, intervention, and so forth.11 If more than one treatment is planned, the order of the treatments may be randomised; and, where feasible, the researcher and patient are blinded to the order of the treatments. Within each of these phases, multiple across time observations are obtained. If there are multiple repeating phases, these are considered as periods. One, or more than one, outcome variable(s) may be measured.
A WORKING EXAMPLE
As an example of the use of a single subject research design in clinical practice, a person with diabetes might present to their family physician for additional disease and treatment management. In the language of a single subject design, the initial reporting of the patient to their family physician, and the physician’s evaluation of the patient is considered the baseline phase. This phase, of a single subject design, encompasses an evaluation of the initial state of the patient. For this diabetic patient, this phase would consist of review of the past blood glucose readings and some initial testing (for example, HbA1C). For a single subject design, the period of time when the patient is receiving an intervention is referred to as the treatment phase. For this diabetic patient, that could include an insulin adjustment, a knowledge based diabetic education course, a behavioural intervention, and others. The assessment of the effectiveness of the treatment is then made by a comparison of the initial with the treatment phase.12 If the first treatment was not effective, a choice might be made to stop the initial treatment or change to a new treatment, or both, thus introducing another baseline or treatment phase into the design. During each of these phases, one or more measurements might be taken to assess the course or variability in the response, which entails multiple measurements of HbA1C at spaced visits. In a single subject design, the typical number of measurements within a phase is seven.9
Presented in figure 1 is an illustration of an implemented A-B single subject research design. The research question of interest for this study was whether a comprehensive intervention for diabetes management would be effective in lowering the fasting blood glucose values. In choosing the patient to participate in this single subject study, careful consideration would be given to identify a patient that would be considered as:
An illustration of an A-B single subject research design.
typical in terms of the practice demographics,
typical for the disease presentation and progression,
in need of lower fasting blood glucose values,
anticipated to be compliant for the treatment changes,
anticipated to be compliant for the necessary outcome assessments, and
would give informed consent.
For this particular single subject design, there are two phases to this design: a baseline (A) and a treatment (B). The baseline and treatment are both administered one time. There are measurements across time, within both the baseline and treatment phases, the patient was seen every two months for a fasting blood glucose measurement.
Presented in figure 1 are the data for the fasting blood glucose values. The patient was a 57 year old white woman, with an onset of insulin dependent diabetes in 1992. Presented are four measurements, every two months, before and after the intervention. A total of eight observations are presented, and each observation is measured and reported (for example, laboratory used, time of day, etc) in the same manner. The intervention consisted of a prescribed exercise regimen, weight management, and participation in a counselling session that was goaled to present diabetes management. As can be seen, it appears that the intervention was effective in lowering the measured fasting blood glucose in this subject. As in all single subject designs, the research question of interest should guide the single subject design used.
RECENT EXAMPLES FROM LITERATURE
Fisher, although most often associated with multiple subject designs, first introduced a single subject experimental paradigm in 1945.1 Since the introduction, the single subject design has been commonly used within the social and educational sciences.6 This design has recently been used as a means of investigation in medicine involving such areas as drug therapy,9 gastroenterology,13–15 internal medicine,15 paediatrics,16 family medicine,17 cardiology,18 nutrition,19 and others.9,16,20–22 During the 1980s, McMaster University established a service to direct and collaborate with physicians in planning and conducting n-of-1 trials.20 It is reported20 that of the 57 completed trials, 50 of those trials provided a definite clinical answer and in 15 cases the study results of the trial resulted in the physician changing the treatment of the patient.
A recent literature example by Langer et al16 reports on a single subject randomised trial to assess the effect of cisapride on symptoms arising from gastro-oesophageal reflux in paediatric patients. A placebo and cisapride phase were studied, with three study periods (A-B-A-B-A-B). The outcome variables of interest were number of episodes per five days for vomiting, gagging, and stools. In addition, Guyatt et al22 report on a single subject study of a randomised controlled investigation of theophylline. Two study periods of drug and placebo phases were used (A-B-A-B). The symptoms of shortness of breath, the need for an inhaler, and sleep disturbance as patient reported on a seven point scale were used as the outcomes.
At times, single subject research designs have advantages over more traditional group based designs. Some advantages outlined by Franklin et al25 that are especially applicable to practice based family medicine research, include: (1) research situations where research funds are scarce, especially for professionals working in private practice or small clinical settings; (2) research questions that aim to study the process of change; and (3) research questions that are driven by clinical work with the crucial question as to whether a treatment would work for a particular patient. The utilisation of the single subject design for the example above showed these (that is, the data presented in fig 1). This design permitted the investigation of the process of change, both within the baseline and after intervention phases and across these phases, with limited research funds in a private physician office, with the overall intention of answering the research question as to whether this treatment would work for this particular patient.
Limitations of the single subject research design are generalisability of the study conclusions and the methodological and statistical assumptions that are typically needed for inferential statistical tests.4,23 A single subject design provides limited support for conclusions regarding populations of subjects. The results of a single subject design may provide positive findings of the effectiveness of an intervention for a particular subject, however the portion of the population that would show this effectiveness and the size of the benefit remain unknown. The non-violation of the methodological and statistical assumptions that are typically needed for inferential tests are difficult to evaluate and test when using a single subject design, because of the limited available data. The evaluation of the validity of the assumptions is more difficult to assess in these small samples. Even with these limitations, estimates and the tests of the effectiveness or intervention effect on the studied subject can be accurately and validly tested using a single subject design.
POSSIBLE RESEARCH QUESTIONS FOR THE PRIMARY CARE PHYSICIAN
Depending upon the interest of their clinical and patient care needs, the number of research questions that can be addressed by a primary care physician in the practice setting is quite broad. A few examples of possible research questions that can be answered using a single subject research design include:
How does the introduction of a specific antihypertensive drug affect the systolic blood pressure for this patient?
Is there a change in compliance for this patient with twice daily blood glucose readings with the introduction of a reminder system?
What is the optimal dose of a specific antidepressant drug for the control of this patient’s self reported symptoms and obtained scores on the Beck depression scale?
Based upon the reported low back and leg pain scores for this patient, does the introduction of an exercise regimen have an impact on those scores?
Does the introduction of an email contact reduce the missed appointments for a particular patient?
Is the level of clinical effectiveness obtained for this particular patient comparable to the results of a recently published clinical trial for the effectiveness of acid suppressive therapy for the control of dyspepsia?
With group based research designs, internal and external validity issues need to be considered and balanced.26 As the primary questions for a single subject research design concern the investigation of the process of change and whether a treatment would work for a particular patient, internal validity (elimination of bias) issues are paramount. Unfortunately, because of the nature of these designs, external validity (generalisability) when balanced against internal validity is typically left with limited control.
The generalisability of the results from a single subject research study is limited. Possible means of increasing the external validity, generalisability of the results, is by: (1) choosing a subject that is representative of the general type of patients for which this intervention would be used6 and (2) by conducting replication studies involving variation in researchers, subjects, or practices.27
When planning to implement a single subject design, the specifications necessary to conduct a multiple subject randomised clinical trial must also be followed. The planning phase must incorporate forethought in the choice of the:
implementation of the treatment,
number of phases,
number of periods, and
number of observations.
Some guidelines have been prepared for the design of single subject research studies.4,8,9,16,24 Given the research question of interest, a single subject design should be considered as a means of investigation for practice based primary care research.
In design of experiments, single-subject design or single-case research design is a research design most often used in applied fields of psychology, education, and human behavior in which the subject serves as his/her own control, rather than using another individual/group. Researchers use single-subject design because these designs are sensitive to individual organism differences vs group designs which are sensitive to averages of groups. Often there will be large numbers of subjects in a research study using single-subject design, however—because the subject serves as their own control, this is still a single-subject design. These designs are used primarily to evaluate the effect of a variety of interventions in applied research.
The following are requirements of single-subject designs:
- Continuous assessment: The behavior of the individual is observed repeatedly over the course of the intervention. This ensures that any treatment effects are observed long enough to convince the scientist that the treatment produces a lasting effect.
- Baseline assessment: Before the treatment is implemented, the researcher is to look for behavioral trends. If a treatment reverses a baseline trend (e.g., things were getting worse as time went on in the baseline but the treatment reversed this trend) then this is powerful evidence suggesting (though not proving) a treatment effect.
- Variability in data: Because behavior is assessed repeatedly, the single-subject design allows the researcher to see how consistently the treatment changes behavior over time. Large-group statistical designs do not typically provide this information because repeated assessments are not usually taken and the behavior of individuals in the groups are not scrutinized; instead, group means are reported.
Phases within single-subject design
- Baseline: this phase is one in which the researcher collects data on the dependent variable without any intervention in place.
- Intervention: this phase is one in which the researcher introduces an independent variable (the intervention) and then collects data on the dependent variable.
- Reversal: this phase is one in which the researcher removes the independent variable (reversal) and then collects data on the dependent variable.
It is important that the data are stable (steady trend and low variability) before the researcher moves to the next phase. Single-subject designs produce or approximate three levels of knowledge: (1) descriptive, (2) correlational, and (3) causal.
Flexibility of the design
Single-subject designs are preferred because they are highly flexible and highlight individual differences in response to intervention effects. In general, single-subject designs have been shown to reduce interpretation bias for counselors when doing therapy.
Interpretation of data
In order to determine the effect of the independent variable on the dependent variable, the researcher will graph the data collected and visually inspect the differences between phases. If there is a clear distinction between baseline and intervention, and then the data returns to the same trends/level during reversal, a functional relation between the variables is inferred. Sometimes, visual inspection of the data demonstrates results that statistical tests fail to find.
Researchers utilizing single-subject design begin with graphic analysis. During the baseline, data are repeatedly collected and then graphed on the behavior of interest. This provides a visual representation of the subject's behavior before application of the intervention. It is critical that several (three to five is often recommended) data points are collected during baseline to allow the researcher to describe the effects on the target behavior during intervention.
In interpreting, the general strategy of all single-subject research is to use the subject as their own control. Experimental logic argues that the subject's baseline behavior would match its behavior in the intervention phase unless the intervention does something to change it. This logic then holds to rule out confound, one needs to replicate. It is the within-subject replication and allows for the determination of functional relationships. Thus the goal is:
Research designs are traditionally preplanned so that most of the details about to whom and when the intervention will be introduced are decided prior to the beginning of the study. However, in single-subject designs, these decisions are often made as the data are collected. In addition, there are no widely agreed-upon rules for altering phases, so—this could lead to conflicting ideas as to how a research experiment should be conducted in single-subject design.
The major criticism of single-subject designs are:
- Carry-over effects: Results from the previous phase carry-over into the next phase.
- Order effects: The ordering (sequence) of the intervention or treatment affects what results.
- Irreversibility: In some withdrawal designs, once a change in the independent variable occurs, the dependent variable is affected. This cannot be undone by simply removing the independent variable.
- Ethical problems: Withdrawal of treatment in the withdrawal design can at times present ethical and feasibility problems.
Historically, single-subject designs have been closely tied to the experimental analysis of behavior and applied behavior analysis.
- Kazdin, Alan (1982). Single-Case Research Designs. New York: Oxford University Press. ISBN 0-19-503021-4.
- ^Cooper, J.O.; Heron, T.E.; Heward, W.L. (2007). Applied Behavior Analysis (2nd ed.). Prentice Hall. ISBN 0-13-142113-1.
- ^Kazdin p. 191
- ^Kazdin, pp. 103–10
- ^Tripodi, T. (1998). A Primer on Single-Subject Design for Clinical Social Workers. Washington, DC: National Association of Social Workers (NASW) Press
- ^Thompson, C.K. (1986). Flexibility of Single-subject Experimental Designs. Part III: Using Flexibility to Design or Modify Experiments. The Journal of Speech and Hearing Disorders, 51(3), 214–25
- ^Moran, D.J. & Tai, W. (2001). Reducing Biases in Clinical Judgment with Single-Subject Treatment Design. The Behavior Analyst Today, 2(3), 196–206 BAO
- ^Backman, C.L. & Harris, S.R. (1999). Case Studies, Single-Subject Research, and N of 1 Randomized Trials. Comparisons and Contrasts. American Journal of Physical Medicine & Rehabilitation, 78(2), 170–6.
- ^Bobrovitz, C.D. & Ottenbacher, K.J. (1998). Comparison of Visual Inspection and Statistical Analysis of Single-Subject Data in Rehabilitation Research. Journal of Engineering and Applied Science, 77(2), 94–102.
- ^Nishith, P.; Hearst, D.E.; Mueser, K.T. & Foa, E. (1995). PTSD and Major Depression: Methodological and Treatment Considerations in a Single-Case Design. Behavior Therapy, 26(2), 297–9
- ^Alberto, P.A. & Troutman, A.C. (2006). Applied behavior analysis for teachers (7th ed.). Upper Saddle River, NJ: Pearson Education, Inc.
- ^Kazdin, p. 284
- ^Kazdin, p. 291