Chapter 14: Descriptive Research

Descriptive research is designed to document the factors that describe characteristics, behaviors and conditions of individuals and groups. For example, researchers have used this approach to describe a sample of individuals with spinal cord injuries with respect to gender, age, and cause and severity of injury to see whether these properties were similar to those described in the past.¹ Descriptive studies have documented the biomechanical parameters of wheelchair propulsion,² and the clinical characteristics of stroke.³ As our diagram of the continuum of research shows, descriptive and exploratory elements are commonly combined, depending on how the investigator conceptualizes the research question.

Descriptive studies document the nature of existing phenomena and describe how variables change over time. They will generally be structured around a set of guiding questions or research objectives to generate data or characterize a situation of interest. Often this information can be used as a basis for formulation of research hypotheses that can be tested using exploratory or experimental techniques. The descriptive data supply the foundation for classifying individuals, for identifying relevant variables, and for asking new research questions.

Descriptive studies may involve prospective or retrospective data collection, and may be designed using longitudinal or cross-sectional methods (see Chapter 13). Surveys and secondary analysis of clinical databases are often used as sources of data for descriptive analysis. Several types of research can be categorized as descriptive, including developmental research, normative research, qualitative research and case studies. The purpose of this chapter is to describe these approaches.

Concepts of human development, whether they are related to cognition, perceptual-motor control, communication, physiological change, or psychological processes, are important elements of a clinical knowledge base. Valid interpretation of clinical outcomes depends on our ability to develop a clear picture of those we treat, their characteristics and performance expectations under different conditions. Developmental research involves the description of developmental change and the sequencing of behaviors in people over time. Developmental studies have contributed to the theoretical foundations of clinical practice in many ways. For example, the classic descriptive studies of Gesell and Amatruda⁴ and McGraw⁵ provide the basis for much of the research on sequencing of motor development in infants and children. Erikson's studies of life span development have contributed to an understanding of psychological growth through old age.⁶

Developmental studies can be characterized by the method used to document change. The longitudinal method involves collecting data over an extended period, to document behaviors as they vary over time. Because the same individuals are tested throughout the study, personal characteristics remain relatively constant, and differences observed over time can be interpreted as developmental change. With the cross-sectional method, the researcher studies various developmental levels (usually age levels) within a particular cohort of subjects and describes differences among those levels as they exist at a single point in time.

One of the earliest developmental studies was actually a longitudinal case report of data collected between 1759 and 1777, chronicling the physical growth of a child at 6-month intervals, from birth to 18 years. These data still represent one of the most famous records of human growth.⁷ Intellectual growth has been the subject of many longitudinal studies, in children⁸ and adults.⁹ Changes that occur in psychological and physiological processes with aging are also best described using longitudinal methods. For example, research has documented the development of personality through late adulthood¹⁰ and cognitive effects of aging.¹¹ At the other end of the spectrum, researchers have described longitudinal patterns of development in infant heart transplant recipients, demonstrating mild motor delays and age-dependent variability in cognitive skills.¹²

Marsala and VanSant¹³ used the cross-sectional approach to study toddlers as they rose to a standing position from the floor in a sample of 60 children aged 15 to 47 months. They classified movement patterns of the upper and lower extremities and trunk across different age groups. In this study, the investigators chose to examine the characteristics of a broad sample at one time, rather than follow a group over several years.

Developmental research provides an invaluable source of information for the production of correlational and experimental hypotheses. A foundation of descriptive data is needed for the generation of developmental theories and determination of which variables are most important for studying treatment effects. Cross-sectional data are most effective when the primary interest in a developmental study is the description of typical individuals at various stages of life or the description of existing groups in contemporary society. The cross-sectional method will provide a greater possibility of sampling large representative groups for such assessments. If, however, the primary interest is the study of patterns of change, the longitudinal method is preferred, as only this method can establish the validity of temporal sequencing of behaviors and characteristics.

Natural History

Longitudinal studies may focus on the natural history of disease states. This type of information is important in the future design of clinical trials and the generation of hypotheses about the etiology and progression of disease. For example, Godbolt and associates¹⁴ documented the 10-year progression of symptoms of Alzheimer's disease in a familial cohort of 23 individuals, demonstrating the challenge of recording subtle early deficits and the later deterioration of spelling and naming. Howieson and coworkers¹¹ prospectively examined the occurrence and outcome of cognitive decline in healthy, community-dwelling elders. Ninety-five elders (mean age 84 years) who at entry had no cognitive impairments were followed for up to 13 years. Outcomes of aging were characterized as intact cognition, persistent cognitive decline without progression to dementia, and dementia.

In another example, Munsat and colleagues¹⁵ followed 50 patients with amyotrophic lateral sclerosis (ALS) over 6 years, to document the rate and pattern of motor deterioration. They described a linear and symmetric rate of motor neuron loss (see Figure 14.1). Understanding the developmental nature of disease states is an important part of moving toward study of their management. For instance, the natural history data on ALS was used to generate hypotheses about specific drug interventions to slow the progression of the disease.^16,17,18 Without the natural history data, it would not have been possible to determine whether or not the drugs were effective.

The utility of evaluative findings in the assessment of a patient's condition is based on the comparison of those findings with known standards of performance. For example, joint motion is compared with ranges "within normal limits,"¹⁹ and nerve conduction velocities are interpreted with reference to "normal values."²⁰ Clinicians need these standards as a basis for documenting the existence and severity of weakness, limitations and patient problems as a guide for setting goals.

The purpose of normative research is to describe typical or standard values for characteristics of a given population. Normative studies are often directed toward a specific age group, gender, occupation, culture or disability. For example, researchers have established normative values for speech perception capacity in children aged 5 to 10 years;²¹ the diagnosis of osteoporosis is based on "young normal" and "age-matched" norms;²² age and sex-matched Constant scores have been developed as norms for shoulder function.^23,24

Norms are usually expressed as an average, or mean, within a range of acceptable values. Therefore, the normal nerve conduction velocity of the ulnar nerve is expressed as 57.5 meters/sec, with a normal range of 49.5 to 63.6 m/s,²⁵ with corrections needed for limb temperature,²⁶ age and height.²⁷ The normal cadence of women walking in high heels is given as 117 steps/minute, with a range from 100 to 133 steps/minute.²⁸ Average values are often given with a standard deviation (see Chapter 17). Therefore, we can describe normal knee ranges for healthy adults during free speed walking as 60 ± 7 degrees.²⁹ Norms can also represent standardized scores that allow interpretation of responses with reference to an arbitrary "normal" value. For example, the Wechsler Intelligence Scales scores are "normed" against a mean of 100 and a standard deviation of 15.³⁰

The importance of establishing the validity of normative values is obvious. The estimation of "normal" behavior or performance is often used as a basis for prescribing corrective intervention or for predicting future performance. If the interpretation of assessments and the consequent treatment plan are based on the extent of deviation from normal, the standard values must be valid reflections of this norm. Because no characteristics of a population can be adequately described by a single value, normal values are often established with reference to concomitant factors. For instance, several studies have established normative values for grip strength in children based on age and hand dominance.^31,32,33 Hager-Ross and Rosblad³⁴ studied 530 boys and girls aged 4 through 16, and found that there was no difference in grip strength between the genders until age 10, after which the boys were significantly stronger than the girls (see Figure 14.2). Another study established norms for 5-15 year olds based on age, gender and body composition.³⁵ Normative data for grip strength in adults³⁶ and older persons³⁷ is also distinguished by gender and level of fitness activity. These data provide a reference for making clinical decisions about the management of hand problems and for setting goals after hand injury or surgery.

There is still a substantial need for normative research in health-related sciences. As new measurement tools are developed, research is needed to establish standards for interpretation of their output. This is especially true in areas where a variety of instruments is used to measure the same clinical variables, such as balance, function and health status assessments. For example, norms have been published for health status instruments such as the SF-36 that help health care providers and clients assess health in several domains³⁸ and in different geographic populations.^39,40,41 It is also essential that these norms be established for a variety of diagnostic and age groups, so that appropriate standards can be applied in different clinical situations. For example, the functional reach test has been used extensively as a measure of stability in the elderly.⁴² It has also been tested as a balance assessment for patients with spinal cord injury⁴³ and age-related normative values have been determined for children and adults without disabilities.^44,45

Researchers should be aware of the great potential for sampling bias when striving to establish standard values. Samples for normative studies must be large, random and representative of the population's heterogeneity. The specific population of interest should be delineated as accurately as possible. Replication is essential to this form of research, to demonstrate consistency and, thereby, validate findings.

Qualitative research paradigms offer a perspective to explore and understand human behavior that arises from a different philosophy than quantitative research designs. Quantitative methodology is linked to the philosophy of logical positivism, in which human experience is assumed to be limited to logical and controlled relationships between specific measurable variables. The rationale for studying these relationships can be defined in advance, based on hypotheses that guide the methods of inquiry. Accordingly, variables can be operationalized and assigned numerical values, independent of historical, cultural or social contexts within which performance is observed.⁴⁶ For example, many quality of life assessments, by virtue of their list of questions, are based on assumptions about measurable behaviors that reflect health status. By using a single rating scale for all subjects, investigators demonstrate the reductionist premise of quantitative research: that experience and clinical phenomena can be reduced to a set of specific questions and variables predetermined by the researcher.

The essence of the qualitative method, on the other hand, obliges the researcher to understand the person's perspective first. Qualitative research seeks to describe the complex nature of humans and how individuals perceive their own experiences within a specific social context. Qualitative methodology uses the subject's own words and narrative summaries of observable behavior to express data, rather than numerical data derived from predetermined rating systems. The qualitative approach emphasizes an understanding of human experience, exploring the nature of people's transactions with themselves, others and their surroundings. Qualitative designs and methods also allow the study of many simultaneous variables contained in a phenomenon. Questions that lend themselves to qualitative inquiry are generally broad, seeking to understand why something occurs, what certain experiences mean to a patient or client, or how the dynamics of an experience influence subsequent behaviors or decisions.⁴⁷

For example, Carpenter⁴⁸ explored the experience of a spinal cord injury (SCI) with individuals who had sustained such an injury. This work demonstrated that the education about living with SCI provided by health care professionals did not match the lived experience of those with the injury, suggesting the need to transform educational approaches. The qualitative investigation into this phenomenon helped to uncover the meaning of SCI to those who experience it, and how it affects their behavior, emotions, body image, self-esteem and interactions. The purpose of qualitative inquiry is to examine such experiences using a holistic approach that is concerned with the true nature of "reality" as the participants understand it. Qualitative methodology has been a cornerstone of research in sociology and anthropology and has more recently received attention by clinical researchers.^49,50

The need to understand the patient's view of the world is particularly important with the widespread adoption of principles of evidence-based practice, which compel the practitioner to consider the patient's values and circumstances in combination with clinical judgment and evidence from the literature. Qualitative designs are well suited to explore patients' preferences, giving practitioners the opportunity to understand the concepts of health, illness and disability from the direct perspective of the person who lives it.^51,52,53

Combining Qualitative and Quantitative Approaches

Qualitative research is not just a description of a particular situation. To qualify as a research method, such inquiry must be tied to understanding, explaining, or developing theory about an observed phenomenon. From such insightful description, relevant variables can be uncovered, and questions can then be posed to study quantitative aspects of those variables in controlled settings. Qualitative and quantitative aspects can also be combined within one study to measure certain components of behavior and to see how such measurements relate to the nature of the actual experience.⁵⁴ Use of both qualitative and quantitative methods in the same study can increase the validity of the findings.⁵⁵

For example, Talion and colleagues⁵⁶ used a focus group of patients with osteoarthritis of the knee to design a questionnaire that exposed a mismatch between treatment priorities of patients and those of health care practitioners. Hayes et al⁵⁷ used both a descriptive questionnaire and qualitative interviews to better understand clinical instructors' perspectives of problematic student behaviors. Paterson et al⁵⁸ used both semi-structured interviews and standardized outcome tools to measure the outcomes of a program of massage in patients with Parkinson disease. The two approaches were used together to identify difficulties with the standardized tools and specific perceptions of the participants not available in those tools. Because of the richness of the combined data, the authors were able to recommend several very specific features for future research.

Perspectives in Qualitative Research

There are a number of different approaches one may take when using a qualitative research design. These include phenomenology, ethnography and grounded theory.⁵⁹ These methods are considered naturalistic inquiry because they require substantial observation and interaction with subjects in their own natural environment.

Phenomenology

The tradition known as phenomenology seeks to draw meaning from complex realities through careful analysis of first-person narrative materials.⁶⁰ The researcher begins this type of inquiry by identifying the clinical phenomenon to be studied. Illness, physical disability and childbirth are examples of phenomena that have been explored by health professionals. Within the phenomenological perspective, experience is constructed within the individual's social context and is, therefore, intersubjective.⁶¹ As an example of this approach, DeGrace⁶² studied the meaning of a family's experience of daily life with a child with severe autism. Her results showed that the family had difficulty engaging in daily activities that held positive meaning for them, and that they relied on stringent routines that revolved around the child to meet daily life demands.

In a phenomenological study in the workplace, Blau and colleagues⁶³ describe the responses of physical therapists to major changes in the healthcare system in which they practiced. They identified four common themes related to stress and discontentment, but also identified that participants were able to find positive affirmation in their work. In a similar study, Dale et al⁶⁴ investigated the influences of cost containment constraints on occupational therapists in a hand therapy clinic. They found that the therapists modified their skills, their professional settings, and implemented innovative interventions to function effectively.

Ethnography

A second common perspective, called ethnography, is the study of attitudes, beliefs and behaviors of a specific group of people within their own cultural milieu.⁶⁵ In ethnographic studies, the researcher becomes immersed in the subjects' way of life to understand the cultural forces that shape behavior and feelings. Questions often emerge as data are collected. The ethnographer begins this type of inquiry by identifying the setting or culture to be studied and may specify the types of phenomena that will be observed. Classic examples of ethnographic research are found in the well known anthropological works of Margaret Mead.^66,67

This approach has been used to study the traditional beliefs and practices related to pregnancy and childbirth among Native American women.⁶⁸ Swigart and Kolb⁶⁹ interviewed sheltered and street-dwelling homeless persons to describe factors that influence their decisions to utilize or reject a public health disease-detection program. Wingate et al⁷⁰ studied the perceptions of activity and vocational status in women with cardiac illness. As these examples illustrate, the concept of culture in ethnographic research is taken broadly.

Qualitative research has also expanded to include the concept of research synthesis through the use of meta-ethnography. This approach uses analysis of multiple sources to develop new insights into the phenomenon being studied. For example, Smith et al⁷¹ studied factors that delay a person's willingness to seek help for a potential cancer. Their synthesis of 32 papers was able to identify several factors that delay seeking help, including lack of recognition of the meaning of symptoms, fear, and gender of the patient.

Grounded Theory

One of the unique features of qualitative methodology is that it allows the researcher to develop theory to explain what is observed. This approach is called grounded theory research, in which the researcher collects, codes, and analyzes data simultaneously. This facilitates identification of relevant variables, and using an inductive process, identification of theoretical concepts that are "grounded" in the observations.⁷² These concepts are not based on preconceived hypotheses, but instead grow out of an ongoing constant comparative analysis of each set of data collected. As data are gathered and coded, each idea or theme is compared to others to determine where they agree or conflict. At any point in the study, if data do not support the theory, the data are not discarded, but the theory is refined so that it fits the existing data; that is, the theory must come from the data.

As this process continues, interrelationships emerge that lead to the development of a theoretical framework. Data collection and analysis continues until data being collected become repetitious, affirming what has already been identified and no new concepts or relationships emerge. This method requires a sophisticated approach to coding and categorizing data.

A wonderful example of grounded theory research is found in the work of Jensen and colleagues,^73,74 who collected and analyzed data over a 10-year period to develop a theory of what constitutes expert practice in physical therapy. Working with recognized "experts" in a variety of specialty areas, they formulated a theoretical model with four dimensions, as shown in Figure 14.3. The theory suggests that these dimensions may exist in the novice practitioner, but not in an integrated manner. They propose that these elements become increasingly integrated as a therapist's competence and expertise grow, moving toward a well-developed philosophy of practice.

Methods of Qualitative Data Collection

Because qualitative data can come from a wide variety of sources and can take many different forms, the methods of data collection are also quite varied. The most common forms of data collection are observation and interviews.

Observation

Field observation of the phenomenon being studied is often conducted prior to interviews. The purpose of this observation is to identify people, interactions, the influence of sociocultural context and even artifacts that might be studied in depth to acquire relevant data to answer the research question. As a "nonparticipant" in the activities being monitored, the researcher quietly and as inconspicuously as possible simply observes. Immediately following the observation, the researcher records field notes about what was observed and uses memos to capture possible questions for follow-up interviews. Nonparticipant observation sessions can also be videotaped so that later the participants can discuss with the researcher what they were thinking while they observe their own behavior. This technique is used to capture the participant's reality while diminishing the sometimes distorting effects of recall.

The essence of qualitative research is that the individual's experience should be described as it is lived by that individual. Therefore, the researcher can also become embedded within the group, using the technique of participant observation. With this method, the researcher actually becomes a participant in the activities of the group being studied, so that observation of behaviors can be appreciated from the standpoint of those who are being observed. While this technique, as with other research techniques, is inherently biased by the researcher's own preconceptions, it does provide a mechanism to describe the interactions of individuals within a social context and to analyze behaviors as a function of the subjects' personal realities. The researcher is in a position to recognize feelings and thoughts that emerge from the subjects' frame of reference. For example, Hasselkus⁷⁵ studied the meaning of daily routines and activities at a day-care center for persons with Alzheimer disease, as experienced by the staff. Data collection included interviews and participant-observation, by working directly with the staff. Through this experience, the researcher determined that the foremost guiding principle for all activities during the day was prevention, that is, to prevent participant behavior that would be harmful to self or others. She was also able to identify characteristics of the staff's perception of a "good day" versus a "rough day." Participant-observer is a complex role, but one that is believed to enhance the validity of qualitative observations.⁷⁶

Interviews

Interviews involve a form of direct contact between the researcher and the subjects within their natural environment. Interviews are used to gather information, with the researcher asking questions that probe the subject's experiences and perceptions. For example, Monninkhof and colleagues⁷⁷ interviewed patients with chronic obstructive pulmonary disease (COPD) in their homes. The data obtained was used to explain how and why standardized health related quality of life scales failed to capture accurately the patient's experience. Potter et al⁷⁸ used structured interviews and a nominal group technique to identify the kinds of patients perceived as difficult by physical therapists, as well as strategies for dealing with them. They identified behavioral problems and patient expectations as leading to the greatest difficulties, with improved communication as the primary strategy. Fogarty⁷⁹ used focus group interviews to identify the benefits of exercise for patients with schizophrenia who resided in a community care facility. In addition to reported gains in physical status, both staff and participants reported improvement in cohesion among residents and between residents and staff.

Interviews should be approached with a broadly structured script that will guide the discussion and provide a basis for comparing responses; however, they must also be flexible enough to allow the interviewer to probe and ask follow-up questions that are relevant to the specific individual's circumstances. For example, an interviewer might simply ask, "Tell me what it has been like for you to have a spinal cord injury? Such a question is usually followed by additional probes, such as "Can you tell me more about that?" and "Can you give me an example of___?" to elicit the full richness and truth value of the data. This process requires that the interviewer have expertise in the subject matter that will be discussed. This should include prior observations of the participant in his or her natural setting so that the appropriate and relevant follow-up questions will be asked.

The quality of the data collected will depend on the knowledge and skill of the interviewer. Therefore, interviewers should be trained in both interviewing and observation skills, and must be sensitive to the issues that will be raised by respondents.⁸⁰ While interviews for qualitative research may seem similar to the clinical interview, it is important to remember a fundamental difference. Clinical interviews have as their primary focus to arrive at a diagnostic decision; qualitative interviews are designed to bring about a better understanding of the phenomenon from the participant's perspective.⁸¹

Data Analysis and Interpretation

Qualitative data analysis is primarily an inductive process, with a constant interplay between data that represent the reality of the study participants and theoretical conceptualization of that reality. Therefore, the process of analysis is ongoing as data are collected. Because observational and interview responses are recorded as narratives, qualitative data are typically voluminous. Data will usually be recorded through written memos or transcribed from audio or videotapes. The specific techniques of data analysis can vary from purely narrative descriptions of observations to creating a coding system from which categories can be developed, and in a systematic way, patterns or themes develop from the mass of information.⁸²

Analysis of qualitative data by hand involves many hours of sifting through narratives, coding and organizing. There are computer programs that help the qualitative researcher manage the large amounts of data that are typically gathered. The available programs are user friendly and highly interactive and are designed to assist the researcher to record, store, index, cross-index, code, sort and interconnect text-based material.⁸³ There are also numerous programs available which are referred to as computer-assisted software. These are categorized into five main software families which reflect their primary function: text base managers, code programs, retrieval programs, code-based theory builders, and conceptual network builders.⁸⁴

There is considerable debate in the field of qualitative research about the use of computer programs for data analysis. The debate focuses on the loss of intimacy with the data as well as issues of confidentiality and security of the participants. The key to researchers using these programs is to keep in mind that they are designed only to assist the researcher in managing the data, but not to analyze data, develop theory or draw conclusions about findings. It is of utmost importance that researchers not allow the computer program to direct the interpretive activity.⁸⁴

Reliability and Validity

Reliability and validity are issues of concern in qualitative research just as they are in all types of research. For qualitative study, the concept of "measurement error" must be examined in terms of judgments rather than numerical equivalency.⁸⁵ Because the data sources are words rather than numbers, different terms and techniques are used to describe and determine the trustworthiness of the data. Lincoln and Guba⁸⁶ have suggested the terms "credibility" and "truth" to refer to internal validity, "transferability" to refer to external validity, and "consistency" and "dependability" to refer to reliability. They describe a number of techniques that can be used to increase credibility and consistency of qualitative data. Similar to techniques used in quantitative research, these approaches reflect a need to consider the rigor of data analysis and the potential for investigator bias.

Techniques for Ensuring Trustworthiness of Qualitative Data

Triangulation refers to a process whereby concepts are confirmed using more than one source of data, more than one data collection method, or more than one set of researchers. The concept actually originated as a technical term in surveying, to demonstrate how two visible points could be used to locate a third point. In social sciences, the concept has been adopted to reflect multiple methods or data sources to substantiate an outcome. For instance, a researcher may identify a specific concept through an interview, by direct observation of group performance, and by analysis of written materials. If comparable conclusions are drawn from each method, the internal validity or credibility of the interpretation is considerably strengthened. For example, Galantino and colleagues⁸⁷ demonstrated that use of exercise groups resulted in positive physical changes, enhanced coping and improved social interactions for a group of people living with HIV/AIDS. They supported the validity of their findings by showing common themes using focus groups, nonparticipant observation and journals.

The validity of findings can also be supported by a clear description and documentation of the thought processes used to interpret data. This process is referred to as an audit trail, allowing those who read the research to follow the investigator's logic. This provides an opportunity for others to agree or disagree with conclusions, and to reconstruct categorizations. In their study of therapists' reactions to changes in the health care system, Blau and colleagues⁶³ transcribed initial and follow-up interviews, used process notes, and other documentation of data reconstruction to carefully document their process of analysis.

Other strategies for improving accuracy include the involvement of more than one investigator to confirm ideas, confirmation of conclusions with the subject of the study through member checks, and analysis until data saturation (no new themes identified) is reached. For example, Jensen et al⁷³ used three non-physical therapist consultants to support their theoretical formulations. Blau et al⁶³ summarized the themes they extracted from their data and mailed them to participants several months later, interviewing them to validate their interpretations. These strategies are important to control for the potential bias in qualitative analysis.

Sampling

In qualitative research, subject selection proceeds in a purposeful way, as the investigator must locate subjects who will be effective informants, and who will provide a rich source of information.⁸⁸ Depending on the research question, the researcher may select one of many types of sampling strategies including typical, maximum variation or extreme. Another type of sampling, termed theoretical sampling, is based on the need to collect data to examine emerging categories and their relationships, and not on identifying specific age, gender or other characteristics of subjects.⁸⁹ In this type of sampling, a few subjects are initially chosen because they belong to a certain group, but further subjects are recruited based on their fit with theory that emerges from the initial data.⁹⁰

A common misconception about sampling in qualitative research is that all samples are small. Sample size remains an important consideration. Samples that are too small will not support claims of having reached a point of saturation in the data. Samples that are too large will not permit the in-depth analysis that is the essence of qualitative inquiry. Sandelowski⁹¹ suggests that determining adequate sample size in qualitative research is a matter of judgment and experience in evaluating the quality of the information collected and the purpose of the research.

We recognize that this brief introduction to qualitative analysis is by no means sufficient to demonstrate the scope of data collection and analysis methods that have been developed. This approach has great promise for generating understanding of health and how it is evaluated. Those interested in pursuing qualitative research are urged to read the references cited in this chapter. We also suggest reading professional literature, such as the journal Qualitative Health Research, to gain an appreciation for the breadth of qualitative research and to develop familiarity with the techniques and terminology of qualitative methodology.

Surveys are often used as a source of data to collect information about a specific group, to describe their characteristics or risk factors for disease or dysfunction. These types of studies are generally focused on a particular issue or aspect of the group's behaviors or attitudes. The purpose of this approach is to provide an overall picture of the group's characteristics, but may involve some correlational interpretations regarding association among certain variables. As an example, Jensen and co-workers⁹² used a survey to examine the nature and scope of pain in persons with neuromuscular disorders. They demonstrated that, while pain is a common problem in this population, there are some important differences between diagnostic groups in the nature and scope of pain and its impact. In another example, Miller et al⁹³ described the most prevalent disorders of older adults that were seen by physical therapists, and compared these across several regions of the United States. Giesbrecht⁹⁴ surveyed occupational therapists in 75 facilities across Canada to describe risk assessment tools, referral patterns and interventions related to management of pressure ulcers. We will discuss the design of surveys in Chapter 15.

Clinicians and researchers have long recognized the importance of the case study or case report for developing a clinical knowledge base. A description of interesting, new and unique cases is necessary to build a foundation for clinical science and as a means of sharing special information among professional colleagues. Typically, case studies involve the in-depth description of an individual's condition or response to treatment; however, case studies can also focus on a group, institution, or other social unit, such as a particular school, healthcare setting, community or family. A case series is an expansion of a case study involving observations in several similar cases.

Purposes of Case Studies

Perhaps the greatest advantage of the case study as a form of clinical investigation is that it provides an opportunity for understanding the totality of an individual's condition outcomes of care. Rather than simply recording behavior, the depth of the case study allows the researcher to explore the subject's condition, emotions, thoughts, and past and present activities as they relate to the focus of the study. The researcher tries to determine which variables might be important to the subject's development or behavior, in an effort to understand why the subject responds or changes in a particular way, or why certain outcomes were achieved and others were not. The case study becomes a valuable source of information for evidence-based practice because it shares important aspects of clinicians' experiences with different patients. Case studies can serve several purposes.

Understanding Unusual Patient Conditions

Clinical case studies often emphasize unusual patient problems or diagnoses that present interesting clinical challenges. For instance, Gann and Nalty⁹⁵ highlighted the unusual diagnosis of a patient with a vertical patellar dislocation. The case was remarkable because the injury occurred without trauma. The authors documented their findings as a contribution to understanding mechanisms of patellar injury. As another example, Herrera and Stubblefield⁹⁶ reported on a case series of eight patients with rotator cuff tendonitis as an unusual complication of lymphedema, and discussed the possible etiology and treatment options.

Sometimes unique situations offer perspectives to support the implementation of guidelines for patient care. Laursen et al⁹⁷ report on a case of a 27-year-old male who suffered from acute delusional psychotic disorder and severe physical agitation. Following a period of physical restraint, the patient was diagnosed with bilateral deep venous thrombosis (DVT). The case report suggests an association between the occurrence of DVT and the immobilization by physical restraint in the absence of pre-existing risk factors. The authors used this case report as a forum to recommend that medical guidelines for prevention of DVT be considered when physical restraint is necessary in this patient population.

Providing Examples of Innovative or Creative Therapies

Case studies may focus on innovative approaches to treatment. For example, Sullivan and Hedman⁹⁸ reported on the use of a home program with electrical stimulation as an effective therapy for upper extremity tasks in the case of a patient 5 years after stroke. MacLachan et al⁹⁹ described the first case of the successful use of "mirror treatment" in a person with a lower limb amputation to treat phantom limb pain. They demonstrated that this approach, which had been successful with upper extremity amputation, was equally effective for improving lower extremity phantom pain.

Changes in technology provide an excellent opportunity to demonstrate resourceful approaches to patient care. For instance, Gillen¹⁰⁰ described a combination of occupational therapy interventions to improve mobility of a man with multiple sclerosis whose impairments included ataxia, and decreased strength and endurance. Through the use of assistive technology, positioning, orthotic prescription and adaptation of movement patterns, she observed improved postural stability and independent control of a power wheelchair. Case studies like this one allow clinicians to disseminate creative ideas that can contribute to successful patient outcomes.

Schuman and Abrahm¹⁰¹ illustrate the use of the case study approach in describing implementation of educational and consensus-building strategies to engage hospital staff in providing palliative sedation for patients in intractable pain at the end of life. In this case, the unit of concern was not an individual patient, but a team of clinicians responsible for making decisions and administering pain management, including nurses, physicians and pharmacists, as well as family members. The focus of the case study was on the course of institutional change, using several patient histories to demonstrate the process. Descriptive case studies can be extremely useful for understanding institutional culture and change.

Generating and Testing Theory

Treatment and diagnostic decisions frequently require reference to theory in the absence of more direct evidence. Case studies are especially helpful for demonstrating how clinical theories can be applied. For example, Rosenbaum and co-workers¹⁰² describe the case of a man who experienced widespread brain damage in a motor cycle accident 20 years earlier, and who has been followed because of his unusual memory impairment. The researchers have documented how this patient's experiences contribute to several aspects of memory theory, including understanding episodic and semantic memory, and the distinction between implicit and explicit memory.

Case studies can also offer insight into existing theories, providing opportunities to confirm or challenge specific hypotheses. For instance, Lott et al¹⁰³ used the physical stress theory (described in Chapter 2) to explain the effect of weight-bearing activities on healing of foot ulcers in a 66-year-old man with a history of diabetes. They accounted for the recurrence of ulcers based on cumulative plantar stresses that occurred with a rapid change in the patient's walking patterns after initial healing.

Providing Future Research Directives

Although case studies focus on the details of individual patients, the results can often highlight issues that require further inquiry, providing a rich source of research questions. One major contribution of the case study to research is its ability to provide information that can be used to generate inductive hypotheses. Because the case study allows for a thorough analysis of a single situation, it often leads to the discovery of relationships that were not obvious before. As more and more cases are reported, a form of "case law" gradually develops, whereby empirical findings are considered reasonable within the realm of accepted knowledge and professional experience.¹⁰⁴ Eventually, with successive documented cases, a conceptual framework forms, providing a basis for categorizing patients and for generating hypotheses that can be tested using exploratory or experimental methods.

As an example, Reinthal et al¹⁰⁵ examined the effectiveness of a postural control program in an adolescent who had problems walking and talking simultaneously following traumatic brain injury. They hypothesized that her difficulty was due to excessive co-activation of trunk, extremity and oral musculature. Following 2 years of speech and physical therapy, the patient was able to significantly improve her ability to communicate intelligibly while walking. The results of treatment suggested that improved postural control allowed less rigid compensation of oral musculature, allowing her to speak. This hypothesis can now be tested in other patients or under different conditions.

Format of a Case Study

A clinical case study is an intensive investigation designed to analyze and understand those factors important to the etiology, care and outcome of the subject's problems. It is a comprehensive description of the subject's background, present status, and responses to intervention.

A case study may begin with an introduction that describes background literature on the patient problem. Theoretical or epidemiologic information is often helpful to understand the scope and context of the disorder. A full history of the patient is then provided, delineating problems, symptoms, and prior treatments, as well as demographic and social factors that are pertinent to the subject's care and prognosis. Authors sometimes include photographs (with the patient's permission) to illustrate a patient's condition.

A section on methods presents elements of the treatment plan. Expectations should be presented as justification for the treatment approach chosen.¹⁰⁴ Where relevant, literature should be cited to support the rationale for treatment and interpretation of outcomes. If special assessments are used, they should be described in functional detail. Some case studies are actually geared to describing the applicability of new or unusual assessment instruments for diagnosing certain problems.

A results section will include the subject's responses and any follow-up data. When appropriate, tables or graphic presentation of data may be useful to explain the patient's progress. The discussion section of the report provides interpretation of outcomes and conclusions, including reference to research and other clinical information that may support or challenge the current findings. This section should include discussion of unique or special considerations in the patient's condition or responses that may explain unexpected outcomes, or that may distinguish the patient from others with the same condition. Finally, the author of a case study should present questions for further study, suggesting where the case study's findings may lead.

Generalizability of Case Studies

The case study is probably the most practical approach to research because of its direct applicability to patient care, but it is also the least rigorous approach because of its inherent lack of control and limited generalizability. The interaction of environmental and personal characteristics and the effect of multiple interventions make the case study weak in internal validity. Generalization from one case to a larger population is also limited because the responses of one individual or social unit may bear little resemblance to those of others in similar circumstances. In addition, case studies are often concerned with exceptional situations or rare disorders, and subjects are generally not representative of the "typical" patient seen in the clinic. Therefore, external validity is also limited.

The validity of inferences from a case study can be enhanced, however, by taking steps to objectify treatment effects and to demonstrate them under different conditions.^106,107 For instance, interpretations can be made stronger by direct quantified observation and by taking repeated measurements over the course of treatment. Treatment effects can be further supported by using multiple dependent variables and by choosing outcome measures that show large and immediate changes. Generalization can be enhanced by documenting the subject's behavior in more than one setting and by including information from the follow-up visit to establish the long-range success of treatment. Literature should be used to demonstrate how results support a particular theoretical approach to treatment. When validity is an important issue to demonstrate the effect of intervention, a single-subject design should be considered.

Case Reports in Epidemiology

An epidemiologic case report is a description of one or more individuals, documenting a unique or unusual occurrence or medical condition. The purpose of the case report is to present as complete a picture as possible about the characteristics of, and exposures faced by, that individual, often resulting in the presentation of a hypothesis about the causal factors that might account for the observed outcome. Many notable examples of this approach exist, such as the original report of a patient with unique dementia characteristics by Alzheimer in 1905;¹⁰⁸ the single case report in 1961 of a 40-year-old pre-menopausal woman who developed a pulmonary embolism 5 weeks after starting to use oral contraceptives;¹⁰⁹ and a series of reports documenting the first cases of AIDS in five young previously healthy homosexual males in Los Angeles.¹¹⁰ These cases led to the formulation of important analytic hypotheses that have since been tested and supported.

In a more recent example, Jeong and coworkers¹¹¹ documented the first report of a "floating total knee," which resulted from fractures of the distal femur and proximal tibia after a total knee arthroplasty. The patient was an 80-year-old man with rheumatoid arthritis who fell in his home 3 years after the original procedure. The report described surgical revision of the joint replacement and postoperative follow-up. Given that periprosthetic fractures are rare following knee replacement, the authors used this case to discuss risk factors related to the patient and surgical technique. They described their choice of treatment options and successful short-term outcomes. This type of case report provides important information that can be used to minimize future risks and to prompt further study.

As is true with clinical case reports, case studies in epidemiology do not provide sufficient control to allow for generalizations or conclusions about causality. They can only act as a catalyst for further study; however, as the preceding examples illustrate, case reports can be vitally important for identifying new health hazards and facilitating further analytic research.