Considerations for Planning and Conducting Clinic-Based Research in Physical Therapy.

Recently, the American Physical Therapy Association (APTA) developed its Clinical Research Agenda (CRA), which is designed to "support, explain, and enhance physical therapy clinical practice by facilitating research that is useful primarily to clinicians."[1(p499)] The CRA is supposed to serve as a mechanism to direct and prioritize the profession's research effort in response to changes in the health care environment, particularly the increased demands for evidence to support interventions.[1] The CRA can become the focal point for research programs within APTA as well as a major influence on a principal funding source for physical therapists--the Foundation for Physical Therapy.

The CRA calls for research to be done on patients in clinical settings (eg, outpatient clinics, hospitals, facilities) and for the purpose of: (1) documenting effectiveness of interventions, (2) validating diagnostic or classification systems, or (3) accurately identifying prognostic indicators of functional limitations or disability. Implementation of the agenda should result in a shift in the types of research funded by the Foundation for Physical Therapy. It should also change what is published within our profession's principal journal--Physical Therapy. Many of the research reports published in Physical Therapy have not been clinically based, they have not been done on patients, and their principal goals are not the goals we have mentioned above. For example, when we reviewed the research reports published in Physical Therapy during 1998 and 1999, only 6% (6/95) were studies that included patients as subjects, were performed in clinical settings, and addressed the effectiveness of physical therapy treatments. Moreover, only 27% (26/95) were studies that included patients as subjects, were performed in clinical settings, and addressed issues related to diagnostic testing or measurement characteristics of tools used by physical therapists in clinical practice.

Some authors have argued that implementing evidence-based practice in physical therapy will most likely require clinicians to upgrade their current skills or even acquire new skills.[2] By the same token, implementing a research agenda to support evidence-based practice may require researchers to upgrade and acquire new knowledge and skills related to conducting this type of research (ie, with patients in clinical settings). Two factors lead us to believe that a retooling may be needed. First, in response to their own investigation that identified several disturbing trends in clinical research, the Department of Health and Human Services announced several additional steps to improve the safety of subjects in clinical trials, including: (1) an aggressive education and training effort that now results in required credentialing of investigators, (2) more specific guidelines on informed consent with full expectation that institutions will now increase surveillance along with routine auditing of records to ensure full compliance, (3) requirements for investigators to supply monitoring plans for small-scale clinical trials (eg, phase I and II trials) along with data safety and monitoring boards' requirements for larger-scale trials, (4) clarification of regulations related to conflict of interest, and (5) pursuing legislation to enable the Food and Drug Administration to levy civil monetary penalties (up to $250,000 per clinical investigator and up to $1 million per institution) for violations of informed consent and other research practices.[3] Second, there has been a relatively limited volume of physical therapy research that has patients as subjects and has been performed in clinical settings.

In our opinion, when designing research studies that involve patients as subjects and are conducted in a clinical setting, some special considerations need to be addressed to ensure the successful completion of these projects. These considerations include facility or institutional resources; patient management issues; availability of target patient populations; and acquiring and maintaining support from therapists, physicians, and institutional review boards (IRB). There are also issues about the responsibilities of investigators when research-related injury or illness occur to subjects during the course of a clinical trial as well as ethical considerations that need to be addressed when determining whether a protocol needs to be modified or a project needs to be terminated. The purposes of this perspective are to discuss these considerations and to offer suggestions for addressing them when designing and carrying out research projects in a clinical setting.

Facility or Institutional Resources

In order to conduct research that is consistent with the CRA, that has patients as subjects, and that is conducted in clinical settings, a researcher must have adequate access to clinical settings and the resources and support necessary to carry out this research, including:

* Adequate access to targeted patient populations and effective mechanisms for identifying and recruiting potential subjects from within the facility, institution, or community

* A network of professionals from other disciplines (eg, physicians, surgeons, epidemiologists), all of whom may play important roles in supporting the research, especially for large-scale clinical trials

* Institutional review boards and medical-legal support to deal with the complexity of clinical research, including confidentiality and other monitoring

Academic medical centers typically have all of these resources, and a clinical researcher's task is certainly more manageable if they work within these environments. We are not suggesting that academic medical centers are the only places where meaningful clinical research can be accomplished. We do suggest, however, that researchers outside of an academic medical center should consider seeking support and resources when determining the feasibility of carrying out clinical research. The inability to acquire such support may create barriers that prevent the completion of the research.

Preliminary Work

Preliminary work is usually necessary in the design phase of a clinical trial to determine the researchers' ability to complete a project successfully. Investigators engaged in all types of research commonly conduct pilot studies for a variety of purposes, including determining their measurement procedures, determining whether modifications are necessary to ensure the quality of data collection, generating data that can be used in power analyses and estimations of sample sizes, and convincing themselves and external funding agencies that they have the capabilities to conduct and complete the study. However, other pragmatic factors related to conducting clinical trials, which are not necessarily addressed through pilot studies, are equally important in determining if a clinical trial can be completed successfully. These factors include issues related to patient management and safety, availability of adequate numbers of subjects from the targeted patient population, obtaining support and participation from therapists and physicians on staff, and identifying whether other personnel are needed. These factors may impede the progress of a clinical trial or prevent the completion of the project. Preliminary work can help investigators identify potential obstacles and help determine mechanisms to prevent or minimize their occurrence during the course of the project.

Patient Management Issues

When investigators are designing a clinical trial to determine the effectiveness of interventions, we believe they should address a number of issues related to the clinical management of subjects in the study. These may include application of interventions, the frequency and duration of the intervention, appropriate outcome measures that will likely identity an intervention effect, the time frames in which outcome measurements should be taken, and potential risks that may be associated with the examination procedures and interventions.

In some instances, previous research or other publications may be useful in providing information concerning patient management issues that can guide the design of a clinical trial. As part of the preliminary work, investigators may find it helpful to examine and apply interventions, as described in the literature, to a few people before proceeding with the design of the clinical trial. Doing so allows the investigators to determine the feasibility of using these procedures in their own clinical settings or whether modifications may be needed in order to avoid potential constraints specific to the investigators' setting. For example, modifications may be required if the investigator is planning to apply the interventions to different patient populations, if he or she is using slightly different equipment or instrumentation to examine patients or provide interventions, or if he or she needs to use a different frequency and duration of interventions than those previously described in the literature. The investigators could also determine whether the interventions, when performed by the investigators, are likely to result in effect sizes similar to those reported in the literature. This information may assist the investigator in estimating the sample sizes needed to conduct the clinical trial.

In some instances, such as when new interventions are being developed, the literature may provide little information on patient management issues, and the investigators may have to rely on their clinical observations to guide their decisions in designing the clinical trial. In such instances, a series of case studies conducted as part of the preliminary work could be an important component of designing a clinical trial. The case studies may help the investigators determine the best method for applying the interventions in order to improve the likelihood that the interventions will produce some desired effect. The case studies may also help the investigators determine the type of outcome measures and the time frames in which these measurements should be taken in order to detect an intervention effect. Unknown or unreported risks that may be associated with the intervention may also be discovered when case studies are conducted before a clinical trial is designed. Case studies also can serve as a framework from which a broader clinical trial can be designed.

Estimating the Availability of an Adequate Number of Patients

Perhaps the biggest obstacle encountered in conducting a clinical trial is the inability to recruit adequate numbers of patients as subjects. Even in the best settings for conducting clinical trials, the need to find specific types of patients as subjects can slow the progress of a study. Preliminary work can be useful in determining whether the investigator's setting will provide an adequate number of patients who meet the inclusion criteria. Patient scheduling databases or chart reviews can be used to predict the available number of patients who may meet the inclusion criteria over specified periods of time. For example, we were designing a study in which we wanted to compare the effectiveness of neuromuscular electrical stimulation for quadriceps strengthening in patients with and without a persistent knee extensor lag following anterior cruciate ligament (ACL) reconstruction. We wanted to enroll the necessary number of subjects for the study within a 1-year period. We used our facility's patient scheduling database to identify the number of patients who underwent ACL reconstruction by our surgeons over a 6-month period. We then used the database to determine how many of these patients were referred to our department for postoperative rehabilitation. Next, we performed a review of the physical therapy charts for these patients to determine the proportion of subjects who either did or did not exhibit a knee extensor lag after 1 week of physical therapy. A knee extensor lag was considered present if the therapist's notes indicated that the patient's active range of motion in knee extension was less than the available passive range of motion in knee extension. From this preliminary work, we learned that there would be an adequate number of these patients referred to our facility and that there was a fairly even distribution of those who either did or did not exhibit an extensor lag. Our preliminary work convinced us that we had a reasonable probability of being able to recruit an adequate number of subjects within our target time period. Had the results of this review been insufficient, we then would have had to decide whether we were able to do the study at our facility, or perhaps we would have had to determine if including other sites would provide an adequate number of subjects.

Acquiring and Maintaining Support From Physical Therapists

Clinical trials in physical therapy often require some level of participation by physical therapists employed in the facility where the research will be conducted. The level of therapist participation may range from identifying potential subjects for a study to carrying out examinations and interventions as part of the study. The quality of data and the ability to successfully complete a clinical trial will largely depend on the willingness of physical therapists to participate. In our experience, how the therapists perceive the investigator's attitude makes a substantial difference in their willingness to participate in a clinical trial. Therapists may be less likely to participate in a clinical trial if they believe that an investigator views them simply as a convenient work force to collect data. Therapists may be more likely to participate if they view the investigator as an important member of their clinical team, whose research may have a direct impact on their practice environment. Therapists may also be more likely to participate if they believe that the investigator views them as important members of the research team.

We strongly believe that an essential factor in obtaining support from therapists is to design the research study to provide information that the therapists perceive as relevant to their clinical practice. The investigator may find it beneficial to begin a dialogue with the therapists as the idea for a clinical trial is being developed. This dialogue may guide the investigator in tailoring the research design to match the interests and questions that are important to both the investigator and the therapists. In addition, therapists often can provide perspectives that may enhance the clinical relevance of the study. For example, when we were considering our trial on neuromuscular electrical stimulation in postoperative ACL rehabilitation programs, we met with the therapists to discuss our idea. During this meeting, we learned that some therapists believed that neuromuscular electrical stimulation could be an effective adjunct to strength training, but they did not believe it was necessary for all patients. They informed us that they typically reserved this treatment for those patients who exhibited a persistent knee extensor lag after 1 to 2 weeks of voluntary exercise and that they were interested in determining if this criterion for treatment selection was effective. This information prompted us to design our study in a manner so that we could provide an answer to this question. By doing so, we increased the therapists' interest in our research because they could see that their participation led to a study more pertinent to their clinical practice.

Another factor concerning therapist participation that investigators should consider is the productivity demands that therapists must meet in their daily practice. Many therapists may be willing to participate in clinical trials, but their jobs may depend on their ability to maintain a daily patient visit quota. If the burden of testing and documentation for a clinical trial limits therapists' ability to complete their patient management and clinical documentation responsibilities, the investigator may find that data collection may be incomplete or that therapists may withdraw from the study. In our experience, this obstacle can be overcome if the investigators determine what are reasonable time commitments from therapists before the study is initiated. This may mean that the investigator will have to share some of the testing and documentation responsibilities related to the research or recruit other personnel (research assistants) who can carry out these responsibilities. Another alternative may be to obtain funding to cover a proportion of the therapist's salary for his or her time devoted to the research activities.

We dealt with the therapist productivity issue in designing our neuromuscular electrical stimulation trial by establishing a patient caseload at the study clinic site for a 3-month period. We specifically maintained the required daily patient visit quota to determine what would be a reasonable level of participation to ask of therapists in our study. We learned fairly quickly that it would be reasonable to ask the therapists to carry out the interventions, record patient progress, and administer the self-report surveys that would be used in the study. We also learned that, although the therapists had the expertise to carry out the study's physical performance testing procedures, their schedules would not always allow them the time to collect the data. In addition, we learned that, because of their daily clinic schedules and responsibilities, the administrative responsibilities for the study (eg, obtaining informed consent, maintaining and managing data files) would be far too time-consuming for therapists to consistently perform. We realized that we would need to recruit assistance from other personnel to carry out the testing procedures and that we would need to assume the administrative responsibilities for the study.

The time spent in the clinic also allowed us to develop the interventions used in the study with feedback from the therapists. We were able to train therapists who were unfamiliar with the neuromuscular electrical stimulation protocol, and we were able to discuss and obtain consensus on other interventions and the progression for the study. This was done in an effort to ensure that the therapists were comfortable with and agreed to the patient management procedures and that consistency in patient care would be maintained during the clinical trial. Most importantly, the time spent in the clinic gave us an opportunity to establish ourselves as a members of their clinical team before the study was initiated.

Once a clinical trial is initiated, an investigator may need to take a proactive role in maintaining therapist interest and participation. Based on our experience, frequent meetings or discussions with therapists regarding the progress of the study may reinforce a belief that their efforts are important and will eventually contribute to the literature. Investigators may also consider providing assistance in patient coverage from time to time for those therapists who overextend themselves because of the study. We believe there should be ongoing dialogue between the investigator and the therapists to eliminate obstacles that therapists may encounter and to consider the need for modifications as the study progresses.

Acquiring Support From Physicians

Clinical trials conducted by physical therapy researchers most often will require support from physicians. The need for support from physicians extends beyond their role as a source of patients for the study. Physicians can play an important role in broadening the investigator's perspective on the safety and patient management issues related to a clinical trial. Physicians may provide additional information to the investigator concerning medical conditions or risks associated with a given patient population that could assist the investigator in developing sample inclusion or exclusion criteria or in determining whether testing or intervention procedures need to be modified. Physicians may also assist in identifying medical cointerventions that may be potential confounding factors, so that the investigator can determine whether these cointerventions should be either controlled or accounted for during the course of the study. Like physical therapists, physicians may be more willing to support a clinical trial if they are convinced that the study will address questions that are relevant to managing patient care. Initiating a dialogue with physicians early in the design phase of a clinical trial can help investigators plan a study with minimal patient risks and with relevant questions. In addition, if they have played a role in the design of the study, physicians may be less likely to withdraw their participation and may be more likely to support the investigator if adverse events occur.

Just as support may be gained from physical therapists through their involvement, investigators may be more successful in gaining support from physicians if the investigators are recognized as members of the clinical team. There are a number of ways an investigator can foster a working relationship with physicians to gain their trust and support. Managing their patients, attending their clinical rounds, and participation in their journal clubs as well as assisting with physicians' research projects can be helpful. We have found that providing educational programs on rehabilitation issues are ways that physical therapist investigators can establish a collegial relationship with and gain trust from physicians. Investigators can build the physicians' support during the course of a clinical trial and for future projects by providing them with periodic updates on the progress of the trial and their patient's response to the interventions related to the trial. Informing them of adverse events when they occur and involving them in the decision process to either modify or terminate procedures in response to an adverse event will also encourage physicians to provide continued support for an investigator's research agenda.

Although preliminary work during the design phase may place a considerable time burden on the investigator and may delay implementation of a clinical trial, without this effort, unanticipated obstacles can occur, including some that would not allow completion of the study. The reader only needs to consider a scenario where researchers have obtained a large amount of funding, with several people on the project payroll, but the trial was terminated because safety issues required design modifications in the middle of the trial or because inadequate subject recruitment prevented the project aims from being attained. The importance of preliminary work in the design phase of a clinical trial cannot be overemphasized.

Support From the Institutional Review Board (IRB)

The institutional review board (IRB) is an administrative body of the institution where the research is being conducted that protects the rights and welfare of human subjects recruited to participate in research activities.[4-7] The IRB has the authority to approve or disapprove and to require modifications in all research activities in its jurisdiction involving human subjects, most often in accordance with regulations of the Department of Health and Human Services and the Food and Drug Administration.[4] Federal regulations apply to those institutions where the research is either conducted or supported by a federal agency. Physical therapists who want to conduct their work in a setting that does not have an IRB would be obligated to find out how they may become associated with one, based on the professional Code of Ethics and the Guide for Professional Conduct adopted by APTA.[8] The Code of Ethics is binding on all physical therapists who are members of APTA. Principle 6 of the Code of Ethics states that physical therapists "shall maintain and promote high standards for physical therapy practice, education, and research."[8(p697)] The Guide for Professional Conduct states, in part, that physical therapists engaged in research shall ensure the consent of subjects, confidentiality of the data on individual subjects and the personal identities of the subjects, and the wen-being of all subjects in compliance with facility regulations and laws of the jurisdiction in which the research is conducted.[8(pp697-700]) Promoting high standards in research would include a review of a research protocol by an IRB prior to conducting the research. In some cases, the IRB from a local academic institution or medical center may be willing to review and approve a study that is being conducted at a facility outside of the institution, but this situation may require that either a faculty member or an employee of the IRB's institution become a member of the investigative team.

Seeking IRB approval to conduct a clinical trial may be viewed by some investigators as a necessary evil that must be dealt with in order to get their project off the ground. This attitude can be harmful to the investigator, particularly if he or she does not provide information in the proposal in order to expedite approval from the IRB. The obvious scenario occurs when potential risks are not disclosed and then a subject is injured during the course of the study. In such cases, the investigator may be subjected to a number of disciplinary actions, which may include termination of the project; limitations on a further role as an investigator at the institution; probation, suspension, or termination of employment at the institution; loss of funding; litigation; and assessment of fines. However, other aspects may not be so obvious. For example, investigators may modify the experimental design without realizing that they have compromised the scientific merit of the study. In such instances, the investigators may be accused of unethical behavior after the fact, because subjects enrolled in the study may have been placed at risk unnecessarily. Rather than viewing the IRB as a necessary evil, we believe that investigators ought to consider the IRB approval process as an additional opportunity for peer review of their work to ensure that the study will be conducted in a safe and ethical manner.

Subjects have the right to give informed consent to participate in a research study.[9] In general, informed consent means that the subject understands the purpose of the research, the procedures and activities in which he or she will be involved, and the potential risks and benefits that may be incurred through participation in the study.[4] Required elements of the informed consent form may vary among institutions. There are a number of references that provide detailed descriptions and explanations of important elements in the informed consent process.[11-13] Investigators should be familiar with the guidelines for informed consent of their respective institutions.

Considerations for Modifying or Terminating a Trial

There are several reasons why a research protocol may need to be modified or terminated during the course of a clinical trial. In our experience, the most common reason is when patient safety issues arise and the risk/ benefit ratio needs to be re-evaluated. This situation may occur when new knowledge is acquired regarding risks and, therefore, safety becomes an issue. For example, investigators may learn from a published report that a treatment or procedure they are currently using has resulted in a previously unknown risk of injury or side effect. The investigators, in this instance, will have to consider whether their own procedures will need to be modified to maintain a reasonable level of subject risk.

Another instance when the risk/benefit ratio needs to be re-evaluated is when a subject in the investigator's clinical trial receives a research-related injury or illness while participating in the study. Research-related injuries or illnesses are commonly referred to as "adverse effects" and are categorized as either expected or unexpected adverse effects. Expected adverse effects are research-related injuries or illnesses in which the potential for occurrence is known and the risk of such occurrence has been acknowledged in the IRB proposal, investigators are obligated to acknowledge the expected risks or research-related injury or illness in the risk section of the IRB proposal. When acknowledging expected risks, the investigator should state the nature and potential severity and frequency of the injury or illness in the proposal. The IRB will usually require the investigator to explain steps that will be taken to minimize the risks. During the design phase of a clinical trial, investigators should evaluate all aspects of the study to identify potential risks. The use of preliminary work or thorough examination of the literature can help investigators identify the potential for research-related risks.

When research-related injury or illness occurs, the investigator's first responsibility is to make certain that the subject receives appropriate medical attention. A physician should be notified as soon as possible. If the injury or illness is minor, the investigator may administer first aid and then refer the subject to appropriate medical personnel for examination. If the injury or illness is a medical emergency, the investigator should be aware of and follow all procedures consistent with the facility's policies for medical emergencies. The second responsibility of the investigator is to report the adverse event to the IRB. The IRB will have specific policies and procedures for reporting adverse events, and investigators should be familiar with these policies at their respective institutions.

The third responsibility of the investigator is to determine whether procedures need to be modified to reduce the risk of future events or, in some cases, to determine whether the study needs to be suspended or terminated. If the event was expected and its occurrence has not changed the risk/benefit ratio of the study, modifications may not be required. If the event is unexpected, moderate in nature, and simple modifications can clearly reduce the risk of future occurrence, the investigator can request these modifications when reporting the adverse event to the IRB. The investigator, however, will need to suspend further subject recruitment until the IRB approves the modifications. If the investigator is uncertain about the modification needed or if the investigator believes that major modifications in the experimental design will need to be considered, he or she should suspend the study, pending further evaluation. Depending on the nature and consequences of the adverse event, the IRB may also require the investigator to suspend or terminate the study.

When modifications of research procedures are required, the investigator will have to justify them to the IRB. The investigator will have to demonstrate that not only will the risks be minimized, but that the modifications will not compromise the scientific merits of the study. Further review of the literature and, in some cases, further pilot study data may be needed to help justify the modifications and demonstrate that there will still be a benefit. If modifications are made to address safety issues and these modifications alter the ability of the study to address the specific aims, the investigators may decide that the study is no longer worth conducting and may terminate it. Investigators may be able to salvage the study by considering a modification of the study's aims so that its scientific merit is maintained. However, if the investigator determines that modifications in the research protocol have been so extensive that a viable research question cannot be addressed, he or she will have to consider terminating the study. For example, required modifications to a test procedure to improve subject safety might actually alter the ability of the test to measure the primary outcome variable of the study. If there were no other alternative test procedures available to address the primary outcome variable, the investigator would have to terminate the study. In addition, investigators should be aware that a change in the specific aims of a funded study will usually have to be re-evaluated by the funding agency. The funding agency may withdraw funds from the project if the changes are no longer compatible with the funding.

Modifying or terminating a research protocol may also need to be considered in cases where the investigators have either overestimated or underestimated the intervention effect that was anticipated at the beginning of the study. In the case where overestimation of the intervention effect has occurred, the investigators may have to modify the protocol to maintain sufficient statistical power by increasing the sample size. When the anticipated intervention effect is underestimated, a difference in outcomes between the experimental and control groups may be achieved before the projected number of subjects in the original design have been managed and tested. This could raise ethical concerns regarding the undertreatment of subsequent subjects who were assigned to the control group. This dilemma can be dealt with if the investigators analyze the data continually during the course of the study. By doing so, the investigators would be able to monitor if or when an intervention effect became clear enough for the investigator to make definitive conclusions regarding intervention effectiveness. This could, therefore, justify an early termination of the study.

For example, an investigator may be studying the effect of supplementing a voluntary exercise program with a joint mobilization technique to improve shoulder range of motion. The experimental group receives the voluntary exercise program combined with the joint mobilization technique, and the control group receives only the voluntary exercise program. An a priori power analysis indicates to the investigator that 60 subjects in each group would be needed to detect a difference in shoulder range of motion between groups, based on the anticipated effect that adding the joint mobilization treatment may have in improving range of motion. However, after 30 subjects in each group have been treated and tested, the investigator finds that the experimental group is already demonstrating greater improvements in should range of motion than the group that receives only voluntary exercise. The addition of the joint mobilization technique to the exercise program had resulted in a greater than anticipated benefit to the treatment program. At this point, it would not be ethical to continue withholding the joint mobilization technique from additional control subjects because the data indicate that this treatment is clearly beneficial. Ongoing data analysis and monitoring can be used to circumvent any ethical concerns about withholding an intervention shown to be beneficial to participants in the study.

Recently, the National Institutes of Health (NIH) has developed a policy that all NIH-sponsored or conducted clinical trials should have a system for oversight and monitoring in place to ensure the safety of participants and the validity of the data.[14] The people or groups involved in monitoring data and safety of trials review research protocols, with an emphasis placed on data integrity and patient safety. The data and safety monitoring boards are composed of people who are considered to be content experts in areas related to the research and who are from institutions other than the investigator's institution. The activities of the data and safety monitoring boards include monitoring adverse events and making recommendations to funding agencies and the principal investigators. These boards also oversee protection of the confidentiality of the trial data. Although data and safety monitoring boards are currently only required for federally funded projects, it may be a good idea for investigators who are involved in projects that are not federally funded to assemble a group of experts who could function as a data safety and monitoring board. These people may not necessarily have to be from other institutions, but they could be content experts from other academic or clinical departments who could provide an additional opportunity for continuing review of the protocol as the study progresses. These groups could help investigators conduct their research in a safe and ethical manner, and provide guidance in situations where the modification or termination of their studies needs to be considered.

Summary

In our opinion, APTA's Clinical Research Agenda will result in increased funding opportunities and support from the profession for clinic-based research. We believe it is important that our profession makes certain that our researchers are prepared and ready to meet the needs and challenges of the new agenda successfully. In this perspective, we have identified obstacles that we believe may create problems in designing and implementing clinic-based research programs. We offer some suggestions for dealing with these obstacles in both the design and implementation phases of clinical research studies. Our thoughts and comments have been based on our own experiences in conducting clinic-based research and most likely do not include all of the issues that clinical researchers in physical therapy may need to consider. We hope our perspective has provided some direction for improving the readiness of those who want to pursue a clinic-based research agenda and that it may provide some insight for reassessing research training programs, so that future investigators will be better prepared to meet the challenges of the new agenda.

References

[1] Clinical Research Agenda for Physical Therapy. Phys Ther. 2000;80: 499-513.

[2] Thompson-O'Brien M, Moreland J. Evidence-based physiotherapy practice. In: Sackett DL, Strauss SE, Richardson WS, et al, eds. Evidence-based Medicine: How to Practice and Teach EBM. 2nd ed. New York, NY: Churchill Livingstone Inc; 2000 (chapter is located in the CD-ROM accompanying the book under "Discipline-Specific Chapters").

[3] Shalala D. Protecting research subjects: what must be done. N Engl J Med. 2000;343:808-810.

[4] Reference Manual for the Use of Human Subjects in Research. Pittsburgh, Pa: Institutional Review Board, University of Pittsburgh; 2000.

[5] Levine RJ. The institutional review board. In: Levine RJ, ed. Ethics and Regulation of Clinical Research. Baltimore, Md: Urban and Schwarzenberg; 1981:207-243.

[6] Loue S. Ethical issues before the study begins. In: Loue S, ed. Textbook of Research Ethics: Theory and Practice. New York, NY: Kluwer Academic/ Plenum Publishers; 2000:71-112.

[7] Smith T. Medical research ethics committees: protecting patients, researchers, and institutions. In: Smith T, ed. Ethics in Medical Research: A Handbook of Good Practice.

[8] Guide to Physical Therapist Practice. 2nd ed. Phys Ther. 2001;81: 9-744.

[9] The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research. Bethesda, Md: National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, National Institutes of Health; 1979.

[10] Spilker B. Informed consent and ethics committee/institutional review board processes. In: Spilker B, ed. Guide to Clinical Trials. Philadelphia, Pa: Lippincott-Raven; 2000:202-218.

[11] Levine RJ. Informed consent. In: Levine RJ, ed. Ethics and Regulation of Clinical Research. Baltimore, Md: Urban and Schwarzenberg; 1981: 69-115.

[12] Loue S. Ethical issues during and after the study. In: Loue S, ed. Textbook of Research Ethics: Theory and Practice. New York, NY: Kluwer Academic/Plenum Publishers; 2000:113-170.

[13] Smith T. Importance of informed consent. In: Smith T, ed. Ethics in Medical Research. Cambridge, United Kingdom: Cambridge University Press; 1999:57-86.

[14] National Institutes of Health Policy for Data Safety and Monitoring [NIH Guide]. Bethesda, Md: National Institutes of Health; June 12, 1998.

GK Fitzgerald, PT, PhD, OCS, is Assistant Professor, Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh, 6035 Forbes Tower, Pittsburgh, PA 15260 (USA) ([e-mail]). Address all correspondence to Dr Fitzgerald.

A Delitto, PT, PhD, FAPTA, is Associate Professor and Chair, Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh.

Both authors provided concept/idea and writing. Julie M Fritz, PT, PhD, ATC, and M Kathleen Kelly, PT, MS, provided consultation (including review of manuscript before submission).