This article discusses single-case research designs as a feasible and appropriate method for clinicians to incorporate research into their practice. Three designs are introduced and described including comments on their strengths and weaknesses.

This paper centers on the multiple baseline across subjects design for the authors feel it has the most promise for field based chiropractic research. The discussion clearly outlines consideration for applying this approach to research and an example of such a study is plainly described.

This article proposes replicated AB research design to be the best method for single-subject research in the chiropractic practice. The author describes the design and points out how the usual threats to validity in single-subjects research are ruled out in the replicated AB design. Also mentioned is the use of visual analysis or graphed data using four suggested criteria: change in mean level and trend and latency of change.

This article presents a good overview of single-subject design. The alternating-treatment design is emphasized as a tool for clinical decision making. Also provided are a brief examples of design options.

Offers guidelines for conducting an N-of-1 randomized controlled trial. The guidelines are set in a question format and in an easy to follow table.

This article supports replication of an N-of-1 randomized controlled trial to increase the generalization of the information gained. Also provides discussions on prerequisites, statistical issues and validity assessments.

Written from the standpoint of drug trials, this article discusses the advantage of single-subject over group designs and factors that influence the appropriateness of controlled single-subject trials. Heavy emphasis is placed on the control aspect in this design.

This article emphasizes the responsibility of the chiropractic profession to develop an experimental data-base. Time-series designs are discussed as a viable experimental strategy. The methods of time-series designs (TSD's) and variations of them are explained with an emphasis on chiropractic application. The discussions and examples provided made this article essential for the understanding of TSD's and their place in chiropractic.

Dr. Meeker nicely summarizes the concepts concerning single subject time-series designs. Not only is the reader introduced to the design, but he is also guided step-by-step through the process of conducting a study and how to write it. This article is an excellent starting point for learning about single subject design

.

This article offers a "to the point" description of a single-case study. The author points out their statistical short comings but concludes that single-case studies may be a valuable tool in medical practice.

This monograph discusses various designs for single-case studies and is meant to be a source of guidelines for their use in applied settings. Many examples of graphs and tables which simplify results are included. This book is indexed by subject and by author and has an extensive list of references.

The authors have organized their monograph into three sections. The first, concerned with the introduction of the scientist- practitioner, provides a historical perspective and discusses the role of the model in today's practice. The second section concentrates on measures while section three focuses on strategies and designs. Issues, ethics and clinical replications are also discussed. Author and subject indexes are provided.

This monograph emphasizes methodology. Single-subject research design options are presented to exemplify the wide range of uses in research. Also presented are the limitations and problems of this type of design. Included are two appendixes concerning visual and statistical analysis and two indexes by author and by subject.

Although this article was written for the osteopathic practitioner, its relevance to the chiropractor is significant. The authors clearly describe singe-subject research designs and their place in research. An explanation of the statistical short comings as well as the ethical considerations involved with single-subject designs are presented. Aldeman's five rival hypotheses that compete with the experimental hypothesis are presented with a discussion on how to minimize their competition with rigorous research methods.

The authors use a conversational style to explain time-series research design in a simplified manor. The A/B/A/B and multiple-baseline designs are described and hypothetical examples are provided. The chapter concludes with a brief discussion on the importance of an investigative attitude in promoting time- series clinical research.

This book was written for the "scholar of research methodology"; not necessarily for the clinician interested in incorporating N-of-1 studies within their practice. The language used throughout the text is entrenched with research terminology that may be unfamiliar to many.
The first chapter provides the standard overview of single-case research design and analysis. However, the author summarized much of the information and put it into tables which are understandable. One table compares low-inference strategies with high-inference strategies of 12 research characteristics. High- inference strategies increase validity.

This monograph not only identifies the need of research for professional development but emphasizes, "the integration of empirical findings and methods into clinical practice." The author hopes to demonstrate how research and clinical practice can be combined. A general overview of research is offered an introduction before and in-depth discussion on single-subject research design ensues; including advanced designs, visual analysis and statistical analysis. The author kindly provides a glossary as well as an author and subject index.

EXAMPLES:

A supplement to the first paper that provides examples of many studies that used the single case report format. Citations from the fields of psychology and neurology.

This article offers a five step breakdown in the implementation of a single-subject study. Although the article is written in terms of occupational therapy, the issues discussed are of concern to any single-subject design. Example studies are presented.

Following a brief discussion of the clinical presentation of spondylolysis, the authors present an example single-subject study. The results were graphed in a typical N-of-1 format which concludes that chiropractic care may offer some relief for patients presenting with this condition.

Presented here is an example of a single-case time-series design. A case of a 14 year old male with primary nocturnal enuresis was used to demonstrate the method of an A-B single-subject design.

With the use of an AB-AB single-case experimental design, the authors of this article evaluate the effects of flaxseed oil has an ameliorative effect on the patient's symptoms.

The authors of this article provide yet another example of an N-of- 1 study. The methods were well documented including difficulties encountered in the study. In so doing, future studies of chiropractic's effect on autonomic function will benefit by resolving these difficulties.

The author examines the effectiveness of Maitland's concept of manipulation in treating chronic cervical dysfunction by employing an AB single-case design. The situation where the treatment resolves a condition making an AB-AB design not clinically feasible.

This article presents an example of a study performed on a patient who complained chiefly of cervical pain and stiffness but also experienced dull left sacroiliac joint pain, loss of grip strength of her right hand, left arch pain and constipation.

Provided here is an example of a multiple-interrupted, time-series design on a patient with bilateral patellofemoral pain syndrome. After a brief discussion on patellofemoral pain syndrome, a detailed description of the case and treatment protocol were provided. The graphed results of the pain diary were also provided, demonstrating the effectiveness of the conservative care. Recommendations for future research were provided by the authors.

GENERAL:

This article presents information relevant to physiotherapy concerning single-case studies. The author stresses the importance of objective and reliable outcome measures in such studies and provides a brief discussion on visual and statistical analysis of data from single-case studies.

Bithell raises design issues that threaten the validity and reliability of an N-of-1 study. One of the issues discussed was how the use of frequent measures (a fundamental aspect of an N- of-1 study) may increase the effects of error. The article's conclusion states that single subject design is inappropriate for demonstrating the general effectiveness of a treatment.

. In 1985, the authors of this article designed an N-of-1 service for clinicians involved with N-of-1 drug therapy efficiency studies. The purpose of the article is to discuss the feasibility and effectiveness of N-of-1 randomized controlled trials based on their 3 years experience with the service. They consider an N-of-1 trial to be clinically relevant if the results of the trial allowed high levels of physician confidence in a management plan. Also discussed were statistical criteria for assessing the results of an N- of-1 trial and the use of a 7-point scale questionnaire as an out come measure. The authors, based on the results of 57 trials consider N-of-1 randomized controlled trials to be both feasible and useful.

A debate is presented on the use of single subject trials for drug development. Johannessen concludes that such trials provide a valid means for new insight in poorly defined conditions and improve therapeutic decisions. Lewis discusses the limitation of such trials; mainly the lack of general inferences derived from the results of a N-of-1 trial.

This article describes the format of an N-of-1 randomized clinical trial (RCT), presents some issues in methodology and plainly lists out terminology in N-of-1 RCT. The main purpose of the article is to introduce an N-of-1 service which will either run a trial on a referred patient or set up a trial for a clinician to run him/her self.

. This article clearly demonstrates the benefit of single-subject research designs for the practicing doctor. Examples of three designs are presented, including advantages and disadvantages of each. The discussion on data analysis in terms of visual inspection and statistical tests provides an insight to the controversy of single-subject designs. The authors suggest guidelines in the application of these designs.

Sites clinical situation that would be good candidates for N-of-1 RCT. The author warns clinicians planning single-subject studies should become familiar with such studies before hand as to prevent repetition of common errors. The ability of an N-of-1 trial to ascertain treatment protocols for the general population is questioned since they only determine effective treatment for an individual patient.

The author discusses the short comings of group studies for the evaluation of a treatment protocol from the standpoint of rehabilitation. Group studies have mixed success in predicting the effects of a particular treatment on an individual. Single-case experimental studies are presented an alternative with examples of alternating phase designs and multiple baseline designs.
The results of the trial were favorable in showing a positive effect of chiropractic on enuresis; however, many design short comings were discovered. Future studies can avoid these same short comings.

This article is a response to Bithell's paper on single subject design. The authors defend N-of-1 studies as a tool to evaluate practice, especially when done in series.

The authors argue the legitimacy of single subject designs as a means for the clinician to contribute to the documentation, evaluation and improvement of clinical practice. Single subject designs are described as the simplest form of replication-based research when a procedure is repeated on the same subject in the hopes to reproduce a past result. A discussion on baselines and various designs within the single subject design is provided.

This article describes and compares two types of single subject research designs; the case study and the single-case experiment. Also discussed is criteria for evaluating the results. The author points out tow issues to consider when engaging in single subject research: replication and evaluating change.

The author states the importance of research to validate clinical practice and feels time-series designs are an effective tool to do so. Essential criteria, advantages and disadvantages, and ethical considerations of several designs are discussed.

Provided by Alan Adams, D.C.
Chairman of the Research Division
Los Angeles College of Chiropractic

For more information contact:
Douglas W. Smith, D.C.
Santa Barbara Health Center
Santa Barbara, CA 93105
e-mail-dwsmith@silcom.com