Advertisement

Abstract

Background Increasing challenges to health care systems and the prominence of patient-centered care and evidence-based practice have fostered the application of qualitative approaches in health care settings, prompting discussions of associated ethical issues in a range of disciplines.

Objectives The purposes of this work were to identify and describe the application and value of qualitative health research for physical therapy and to identify ethical considerations in a qualitative research study.

Design This was a qualitative interview study with telephone follow-ups.

Methods Forty-six participants were interviewed about their early experiences with rheumatoid arthritis. They also were asked what motivated them to volunteer for the study. To inform the discussion of ethics in qualitative health research, this study drew on the in-depth interviews, took a descriptive approach to the data, and applied the traditional ethical principles of autonomy, justice, and beneficence to the study process.

Results Ethical issues emerged in this qualitative health research study that were both similar to and different from those that exist in a positivist paradigm (eg, clinical research). With flexibility and latitude, the traditional principle approach can be applied usefully to qualitative health research.

Conclusions These findings build on previous research and discussion in physical therapy and other disciplines that urge a flexible approach to qualitative research ethics and recognize that ethics are embedded in an unfolding research process involving the role of the subjective researcher and an active participant. We suggest reflexivity as a way to recognize ethical moments throughout qualitative research and to help build methodological and ethical rigor in research relevant to physical therapist practice.

Increasing challenges to health care systems and the prominence of patient-centered care and evidence-based practice (EBP) have fostered the application of qualitative approaches in health and illness concepts,112 prompting discussions of associated ethical issues in a range of disciplines (eg, medicine,1 ethics,1315 social science,16 health care17,18). Although there is no unified definition or agreed-upon way of doing qualitative research, it has been noted that the qualitative approach has salience for physical therapy, given its efforts toward EBP,19 a patient-centered approach, and the call to focus on the ethics of care.20 According to Jensen, “Qualitative methods provide researchers with the tools to examine social settings and human behavior. The methods are well suited to studying the complex, multidimensional environments present in physical therapy practice and education.”21(p492)

This article contributes to the discussion on the utility and value of qualitative health research and associated ethical concerns. Drawing on a qualitative interview study, we apply the traditional ethical principles of autonomy, justice, and beneficence22 to the qualitative health research process and outline: (1) the development of medical research and governance; (2) the application and value of qualitative health research; (3) research strategies in a qualitative health research study; and (4) emerging ethical considerations in the qualitative research process. In our discussion, we suggest the process of reflexivity as a way to foster ethical and methodological rigor in applied qualitative health research and, ultimately, to offer enhanced care. The concept of reflexivity has been used in a variety of ways in a range of disciplines; for the purposes of this article, it is a researcher self-awareness that frames actions and interactions during the research process. Being reflexive engenders attending to participant priorities, respecting participant experiences in the context of their daily lives, and building relationships based on mutual respect and shared information in the health research process.

Background: Medical Research and Governance

Health research is inherently a moral enterprise, characterized by asymmetrical relationships of trust and power, underpinned by ethical tensions between means (eg, potential for risks posed to volunteers in the research process) and ends (the quest for knowledge for the greater good).23 Guiding moral principles in research ethics consider the overall potential benefits accrued against the possible harms to volunteers, and formalized structures of research governance exist to ensure that research is conducted in an ethical manner and that researchers act with integrity.24 The advancement of medical knowledge and associated research scandals involving abuses of power by health care professionals and gross harm to research participants25 spawned the development of systematic ethical guidance in medical research.26 The protective measures introduced rested on basic and shared moral principles. Research volunteers were to be fully informed of the nature of the research, decide freely about participation, and be assured that their participation would not affect their health care and treatment. Regulatory safeguards and guidelines were declared in the Nuremberg code (1949),27 the World Medical Association Declaration of Helsinki (1964; last revised in 2008)28 and the Belmont Report (1979).29

Increased federal accountability for human subjects research is further illustrated in updated ethics guidelines, such as the Council for International Organizations of Medical Sciences−World Health Organization's International Ethical Guidelines for Biomedical Research Involving Human Subjects (1993; updated 2002),30 the Canadian Tri-Council Policy Statement (1998; amendments in 2000, 2002, and 2005),31 and the Statement of Ethical Practice for the British Sociological Association (2002; appendix updated 2004).32 In the United States, the Office for Human Research Protections details research regulation, offers education (is currently running a campaign to inform the general public about research participation), and recently updated its document on compliance oversight.33 Historically, guidelines for human subject protection are anchored in the biomedical model,1 with clinical trials acting as the benchmark of research governance.34

Broadly, the biomedical model rests in the positivist paradigm, which typically (although not always) generates quantitative data. The positivist approach is associated with a hypothetico-deductive model of science—a systematic process in which observable “facts” are collected, variables are scrutinized, hypotheses are tested, reliability and validity are measured, and statistical generalizations are made (Fig. 1). The researcher objectively observes the data and collects the results. In contrast, qualitative research is associated with the interpretivist paradigm, an inductive approach to studying naturally occurring phenomena and understanding multiple realities; it applies different criteria to validity and reliability, and an assessment is made of how transferable the findings are to different settings (Fig. 1). Qualitative researchers attempt to interpret the meaning people attach to their experiences and investigate the complexity, context, and process of “lived experience.” Generally, ethnography, although its definition is debated, is a favored interpretive approach that typically involves participant observation with interviews. Often researchers are unable to immerse themselves in the lives of those being studied, but conduct a series of interviews (often termed “ethnographic”) in an attempt to gain an in-depth understanding of individual experience. The researcher co-constructs the data generated, acting as the research instrument. Samples can range from single cases to large-scale studies, depending on the research aim.35 Quantitative and qualitative approaches typically ask different questions and gain different types of evidence, all of which are needed to fully inform an evidence base.20 It should be noted that positivist and interpretivist approaches are not always tied to particular methods and that researchers often are involved in mixed-methods initiatives and make decisions about research design based on pragmatic factors rather than philosophical preferences.36

Figure 1.

Model illustrating examples of positivist and interpretivist approaches to research. Based on: Shepard KF, Jensen GM, School BJ, et al. Alternative approaches to research in physical therapy: positivism and phenomenology. Phys Ther. 1993:73:88–97.

Although qualitative research is increasingly addressed in research compliance guidelines for the protection of human subjects,31 the principle-based approach to conducting ethical research is commonly cited as the most appropriate framework for ensuring human subject protection: “Respect for autonomy, beneficence and non-maleficence has become a key component of any discussion of the researcher-researched relationship, which, it is argued, will ensure that the end objective in qualitative research does nor override the rights, health, well-being and care of research participants.”37(pp1151–1152) However, evidence indicates that medical research ethics committees encounter difficulties when assessing qualitative research, which subsequently has hampered research of sensitive topics.1,38

The Application and Value of Qualitative Health Research

The increasing relevance of qualitative research to health care practice has been documented.3,35 Typically designed to reveal a range of experiences and identify commonalities and differences between groups or individuals, methodological approaches include grounded theory, phenomenology, and narrative. Investigation reveals how outcomes are achieved and situations unfold and highlights the interactions and minutiae of daily life. Qualitative approaches have increased our understanding of medication use (eg, revealing patient ambivalence and the shortcomings of the compliance model9) and highlighted contradictions and tensions in practicing self-management not identified by a variable-based approach.8 Qualitative studies have offered explanations for unexpected or contradictory findings revealed by quantitative studies and shortcomings of a rational-choice decision-making model in seeking care and have identified unintended consequences of service developments.35 In physical therapy, grounded theory studies20,39,40 have revealed the complex ways in which health care practitioner experts make decisions. Such studies contribute knowledge to an evidence base that cannot be generated by a quantitative approach.41 Overall, qualitative research can offer useful stand-alone projects, evidence alongside quantitative methods, or a stage of a mixed methods project and can contribute to a body of work for synthesis.42

Here we describe an in-depth interview study and then focus on ethical issues that arose throughout the study process. We use the traditional principles of autonomy, justice, and beneficence/nonmaleficence as an organizing framework22,34 and comment on their utility in qualitative research.

Research Strategies in a Qualitative Health Research Study: The Early Rheumatoid Arthritis Help-Seeking Experience (ERAHSE) Project

Our research aim was to gain an in-depth understanding of the meanings people applied to early symptoms of rheumatoid arthritis (RA) and their impact and the actions individuals took in the context of their daily lives. We wanted to investigate the “lived experience” of illness.4 We draw on interview extracts and detailed field notes (from both the pilot study and the main study) to explore and illustrate the links between method and ethics. Because this is a discussion article that draws on data, we use basic description,43 remaining close to the data, and do not make highly conceptual or abstract statements. As Sandelowski noted, qualitative description is a valuable method, but often neglected as a stand-alone approach to presenting qualitative data.43 We compare the quantitative and qualitative paradigms (Fig. 1) and identify the ways in which traditional ethical principles can be applied (Fig. 2), focusing on the qualitative research context.44

Figure 2.

Model illustrating theoretical approaches in research and associated ethical issues.

Recruitment

We recruited individuals who had been diagnosed with RA in the 12 months prior to their recruitment and who were English speaking and lived in British Columbia, Canada. In our pilot study,45 recruitment packages were mailed to 163 family physicians and 4 rheumatologists for forwarding to patients who fulfilled the inclusion criteria. Out of a target of 10 participants, we recruited 8 participants (4 recruited through family physicians and 4 recruited through rheumatologists). Follow-up contact was attempted with the health care professionals, but no further participants were recruited. This paucity of participants reflects the low incidence of new cases of RA (33 per 1,000 in North America), but also reveals problems, common to both qualitative and quantitative research,46,47 with recruiting participants through health care professionals.4850 In most cases, we were unable to speak to family physicians to discuss recruitment but spoke to assistants, several of whom reported physicians did not take part in this type of research. The pilot recruitment strategy was considered adequate for the pilot phase (we gained in-depth data that warranted further investigation) but insufficient for the main study, in which we wanted to gain a range of experiences based on diverse social and illness factors in a range of settings. We also wanted to identify similarities and differences among individuals.

For the main study, we recruited participants through patient organization Web sites, newsletters, and information leaflets at local arthritis centers, as well as 4 rheumatologists’ offices. We gained 38 participants (37 female and 1 male). Although we originally had sought 36 participants, we extended recruitment (and adapted information leaflets) in the hope of gaining more male participants; however, this attempt was unsuccessful, and recruitment ended for practical reasons.

Although in the main study we provided more opportunity to “self-select” via patient organizations and arthritis centers, we encountered recruitment problems. Despite requiring an equal number of male and female participants to gain insight into gendered experiences, recruiting only 1 male participant limited the findings. Also, the majority of participants who contacted us did so via the recruitment leaflets, not their health care professionals. Participants indicated (in their initial contact and during the interview) that a major reason for taking part in the study was the hope that they would gain treatment or care benefits or that they could share their story to assist others. This participant self-selection may have biased the sample. It also was possible that those volunteers with few treatment or illness problems saw no reason to participate and that those who were among the most disadvantaged, with multiple conditions, may not have had access to the recruitment information or could have had other priorities. This possibility does not dilute the significance of the findings, but it does highlight that caution and clarity are needed when making data-driven claims, for instance, about transferability of the findings to other settings.

Interviews

The guide was organized around 3 broad, overlapping areas: (1) early symptoms, including impact and illness actions; (2) interactions with health care professionals and gaining a diagnosis; and (3) post-diagnosis experiences. We also asked individuals their motivations for participating in the research. Informed consent was obtained, and all participants agreed to a follow-up telephone interview for elaboration and clarification and for the interviews to be audio-recorded. Detailed field notes were taken. The interviews were transcribed verbatim, and identifying information was removed from the transcripts. Pseudonyms, chosen by the participants, have been used for all interview data. The structure, content, and form of the guide were designed to elicit open and detailed responses from the study participants, giving them opportunities to discuss their priorities. To avoid, as far as possible, researcher bias and medical model preconceived ideas about treatments and decision making, consumer collaborators (volunteers on the research team with a diagnosis of arthritis) contributed to the interview guide, offering their perspective on what topics should be included, the type of language or phrasing to be used, and the order of the questions asked. Thus, the language was anchored in the everyday world rather than in a medical, therapeutic, or theoretical paradigm. The interview was designed to help build rapport (mutual trust and emotional affinity) and aid validity (gain an in-depth account as close as possible to participants’ experiences). Practical issues included arranging the interview for the convenience of the participant in terms of venue, timing, and comfort (eg, stretch breaks were offered).

Knowledge Translation

We define knowledge translation as an exchange of knowledge and perspective among an interdisciplinary team (including practitioners and consumers), as well as strategies for dissemination to a range of stakeholders and decision-makers in applied settings. We also include the transfer of information between the research team and participants. Communication strategies were built into the research design to facilitate team discussions and negotiate different perspectives and their application in answering the research question. Patient groups and health practitioners contributed key perspectives to the interview guide and dissemination process, which was designed to offer a range of relevant data formats and outlets encouraging best use of the collective findings (peer-reviewed publications; specialized and plain language/information documents, conferences, round-tables, workshops, education initiatives, and updates and results regularly published on arthritis Web sites). Dissemination activities continue to be aimed at different stakeholders to build a bank of knowledge leading to actionable research/action. Regular (quarterly) progress reports were sent out to all participants and the research team, updating them on all dissemination activities.

The analysis to date has informed an interactive educational initiative called the Animated Self-serve Web-based Research Tool (ANSWER), which is near completion, to enhance the patient experience, underpinned by a shared decision-making approach and recognition of the role of partnership in health care relationships. This initiative operationalizes the concepts of integrated knowledge translation and end-of-project knowledge translation.51 The findings also have led to a longitudinal help-seeking study in another health care setting (the United States) and a Canada-wide survey to suggest hypotheses about help-seeking to add to the knowledge base.

Emerging Ethical Considerations in the Qualitative Research Process

We apply the concepts of autonomy, justice, and beneficence/nonmaleficence to discuss ethical concerns that emerged in our study. We use this principle-based approach because it typically is used as a framework in assessing human protection in biomedical research. However, it increasingly is regarded as unsuitable in assessing ethical issues that arise in a qualitative paradigm1 (Fig. 2). Some claim it is insufficient and neglects the context of research,37 whereas others defend its utility if the principles are used flexibly and context is addressed.52

Autonomy

Autonomy has been defined as the capacity to think, decide, and act on the basis of a freely made decision.52 Beauchamp and Childress22 identified 2 fundamental components of autonomy: (1) liberty (independence from controlling influences) and (2) agency (capacity for intentional action). Hewitt stated, “In the context of research, informed consent is an explicit agreement by participants to participate in the research after receiving and comprehending information regarding the nature of the research. Such consent is given without threat or inducement and requires that participants have the mental capacity to give consent and voluntariness”37(p1152) From a qualitative perspective, informed consent typically is considered an ongoing process,15 of which recruitment is a part. In our pilot study, participants were recruited through the offices of their health care professionals, provoking potential ethical concerns regarding decision making and research participation.

Our study was designed to ensure, as far as possible, that health care professionals had no personal contact with patients and that participation remained confidential. Communications between the researcher and health care professionals were limited to mailing recruitment documents and associated general follow-ups and inquiries. Recruitment documents were to be mailed to relevant patients from health care professionals’ offices. However, in the pilot study, several participants reported interpersonal contact with their physician regarding participation (Box 1, quotes 1, 2, and 3). It has been noted that potential issues of inducement and consent emerge in the health care setting if a study participant is recruited through his or her physician.49 Issues of vulnerability and power may influence decisions to participate, and being recruited by an individual's health care professional may pose obstacles to the process of free and informed consent.

Box 1.

Autonomy

Typically, participants in our pilot study described their physicians as pivotal resources and, in some cases, as allies as they faced early-stage RA and accessed a range of medical services and information at a particularly vulnerable time in their illness trajectory. This relationship has a potentially coercive influence on study participation. Study information from a trusted health care professional (particularly when gained in the consultation) may put implicit pressure on the patient to participate. It would be paternalistic to assume that the patient feels obliged or coerced into volunteering, but safeguards are needed to ensure the decision-making process to take part in health research is fully informed and unencumbered.15 Also, if physicians select patients beyond the inclusion criteria, they may deny access to others by declining to pass on study information, perhaps for paternalistic reasons to “protect” patients from an assumed burden of participating or for deeming that the research and patient are “not well matched.” Any interpersonal communication between patient and health professional regarding participation raises potential ethical concerns. Although patients may not necessarily be perceived as forming a vulnerable group, in our study, we found that individuals moved through vulnerable moments in their illness trajectory, and this vulnerability may have implications for how to recruit patients for research while offering them every opportunity to volunteer without being coercive in subtle ways. Throughout this process a reflexive self-awareness when interacting with potential patient participants may assist those involved in recruitment to navigate a path between being overly paternalistic and practicing a subtle or nuanced coercion.

Respect for autonomy encompasses an acknowledgment of agency and respect for the participants’ priorities, experiences, and motivations. An interview situation may leave the participant vulnerable to an implicit agenda of the researcher, who attempts to balance listening to detailed accounts, against the aims of the research and practical considerations such as time. There is the potential to “mute” the participant by categorizing some aspects of the conversation as “going off topic” and keeping to an agenda of topics to be covered. Considering many of the participants expressed they had taken part in the study to share their experiences, with several commenting “I wanted to tell my story” or “I have waited a long time for this” and with others coming to the interview with logs of their experiences, this is of particular significance. Built into the topic guide were opportunities for participants to ask questions and give feedback about the interview, as well as introduce or elaborate on their own priorities. When applying an ethics lens to the interview, the strategic process of “active listening” becomes suffused with “ethical moments.” Negotiating the requirements of the research aim and the autonomy of the participant in the reality of the interview situation brings to the surface the “at odds” relationship of researcher and researched, which often is muted by descriptions of rapport building. This process illustrates the ongoing ethical moments that researchers face and the need for a reflexive approach; that is, the researcher reflects on the details of the research process and her or his role and interactions with participants and takes actions accordingly.

Participants are vulnerable to being misrepresented. Because the researcher co-constructs the interview with the participant and analyzes and interprets the talk, the potential to misrepresent the individual and associated groups is always present. In our study, interdisciplinary discussions highlighted different aspects of the transcripts as significant, and this was a constant reminder of the need to be “ethically vigilant,” staying as true as possible to the reported experiences of the participant by, for example, drawing on the preinterview communications and field notes to contextualize the interviews when interpreting the data. This process extends beyond being transparent when making data-driven claims and offering participants opportunities to comment on the emerging findings; it is underpinned by a reflexivity that involves a continual self-awareness of the researchers’ personal responses to the data.

Justice

Hewitt stated, “Concepts of justice are explained in terms of what is deserved by each individual, and to what each individual is entitled, without partiality and with the aim of delivering equitable treatment”37(p1153) Attending to this principle means offering participants a fair and equitable distribution of burdens and benefits. Although other forms of justice, such as procedural justice, may be applicable here, it is beyond the scope of this article to explore the range of ways justice may be conceptualized and utilized in discussions of ethical aspects of the qualitative paradigm.

Many of the volunteers who made contact conveyed how they hoped to benefit in some way from study participation. Some hoped for information to help them manage their symptoms or for advice about a medical system that they found difficult to navigate (Box 2, quote 1). We offered a resource sheet to all participants who contacted us, with details of Web sites and patient- and arthritis-related organizations that offered advice and information. Some participants reported receiving the resource sheet was a benefit of taking part in the study.

Box 2.

Justice

Some participants described the opportunity to share their experiences as a benefit and stated that taking part in the study helped them face and manage a new diagnosis of a debilitating, painful chronic illness. In such circumstances, being denied access to research can be perceived as a justice issue, for example, when a gatekeeper (eg, a health care professional) offers study information to some patients for reasons beyond the formal inclusion criteria, denying others the opportunity for potential benefits. Patients may interpret being denied access to research as being denied a voice, which may be particularly significant for those who feel they are not being heard and that their experiences were “invalidated” in the medical consultation. The context in which individuals volunteer to participate in health research has implications for the ways in which researchers communicate nonaccess to potential participants who may not fulfill the study criteria and who may see subsequent nonparticipation as being denied a form of care.

The most disadvantaged individuals may not have access to research participation for a number of reasons; they may be preoccupied with negotiating multiple conditions (social, medical, and personal).53 Some of our participants volunteered but did not take part due to a reported inability to negotiate their illness, daily life activities, and participation. Others may not have access to a gate-keeping professional (who may or may not provide them with study information); they may live in remote communities or may not have access to recruitment materials from the Internet or patient organizations.

Giving participants the opportunity to express their experiences is a justice issue because the interview contributes to our understanding of human experience. Particular meanings, processes, and complexities might remain hidden or misunderstood if research does not generate in-depth knowledge of a range of experiences under investigation. Reflexive construction of an interview schedule or topic guide means researchers being aware of the potential to impose their perspectives in subtle ways on the interview, through their use of language and choice of content and providing “spaces” for participants to prioritize their concerns (in keeping with the research's broad aims). Analysis and interpretation of interviews also are justice issues if a qualitative study designed to explore the complex messiness of the lived experience of illness is explained in terms of variables and cause and effect more suitable to the positivist paradigm. It also is important to apply theory to qualitative accounts in order to respect the experiences of individual accounts, while recognizing the broader context in which personal experiences take place, and to assess the feasibility of transferring the findings to other settings. Careful researcher attention to the complexity, multidimensionality, and nuances of context and to the relevance of theory work to guard against a distortion of findings, which may stigmatize groups, while neglecting fundamental social problems.

Dissemination of the qualitative findings also can be framed as a justice concern if participants give their time and share their experiences (which may be burdensome), with the understanding that they are contributing to a knowledge base to improve practice (which they may perceive as a benefit). Participants in our study hoped to contribute to medical knowledge and improve patient experience of illness, treatment, and care; they conveyed helping others as a benefit of taking part in the research, even if they felt they personally would not benefit directly from research participation (Box 2, quotes 2 and 3).

Despite increasing applicability and visibility, qualitative research still is neglected in some prominent medical journals and can be seen as anecdotal, or criticized on the basis of quantitative measures of reliability, validity, and generalizability.54 Efforts need to be made to publish qualitative research widely and make it accessible to different communities and disciplines. It is a waste of funding resources if findings are not considered useful, worthy of publication, or disseminated to suitable stakeholders.

Most participants in our study reported an interest in knowing the results of the project. Some identified how they perceived learning about the research outcomes as a benefit of taking part (Box 2, quote 4). The accounts make explicit the responsibility the researcher has to participants55 to attempt to disseminate the research findings and improve awareness of the results among all relevant stakeholders, including participants themselves. A reflexive self-awareness about the details and implications of the research process invites researchers to be conscious of the justice issues from study inception to dissemination.

Beneficence/ Nonmaleficence

The principles of beneficence and nonmaleficence involve an obligation to provide benefits for the patient and to balance such benefits against risks22 and require that the researcher should do the patient no harm and should prevent harm and remove existing sources of harm.56 This concern highlights a potential conflict for health researchers who fulfill multiple roles, such as nurse research coordinators who advocate on behalf of the researcher (eg, principal investigator) and the patient, as well as the potential for misunderstanding when the patient places trust in the health researcher. Such situations raise ethical issues and have been identified as warranting a sort of “trust wariness” on behalf of the participant.15

Risks include overburdening individuals due to physician estimations of patient suitability beyond the inclusion criteria when not engaging fully with risks (or problems in assessing risks) that are associated with qualitative research rather than those in clinical trials. When obtaining consent for an interview study, it is never possible to accurately estimate risk, in terms of emotional upset, but we can anticipate its potential and take appropriate measures to prevent harm. In our study, respect for the priorities of the participant and the aims of the project were negotiated throughout the interview. We attempted to recognize when it was inappropriate to probe further, despite the interview guide, as the risk to a person's emotional well-being may outweigh the hope of gaining rich data. For example, seemingly straightforward questions (from the perspective of the interviewer) could elicit an emotional response (Box 3, quotes 1 and 2).

Box 3.

Beneficence/Nonmaleficence

Ongoing decisions about how the interview unfolds need to be made in vivo. Some participants talked in emotional terms when describing what prompted their first appointment, as it symbolized a loss in their lives and the start of what was, in some cases, a debilitating illness and an uncertain future. In another example, a seemingly innocuous question caused emotional upset for the participant, who was close to tears as she responded to a probe to gain further detail about family life and her children (Box 3, quote 2). This participant was able to continue in paid employment, but a routine task for her symbolized being a mother, and she conveyed suffering and loss beyond her functional debility.

Participants may offer personal and sensitive information spontaneously, with no apparent upset, or they may become distressed unexpectedly (Box 3, quotes 1 and 2). The observation here is that aspects of risk are subtle and ongoing and emotional risk in an interview situation cannot be anticipated in advance, but the potential for it can be, and researchers need to attempt “ethical listening”—a blend of flexibility, structure, sensitivity, and pragmatism. The resource sheet we provided also included counseling service details (a free service offered locally).

There also is a risk of confusing the research interview with the therapeutic interview, so the participant may have an inaccurate estimation of risks and benefits. One participant noted that she was so frustrated with the health care system that she wanted to “dump” on somebody (Box 2, quote 1). Another participant reported severe illness, debilitating symptoms, acute reactions to medication, and depression and anticipated that she would find it helpful “just to talk to someone.” Although participants may gain benefits from the interview, the potential for psychological distress—during or after the interview—should not be underestimated. Also, the line between building rapport in a research interview and offering a therapeutic interview encounter needs to be navigated with care.

One participant related how, in a previous interview in a hospital setting, she had shared sensitive details and found it difficult to adjust to the sudden termination of the interview and being back out “on the street.” Another participant described discussing help-seeking and had later spent some time reflecting on her previous actions and wondering whether she could have consulted her family physician earlier and prevented symptom deterioration. Participants also reported more practical burdens and inconvenience. Some conveyed negotiating participation with their medication regimens, appointments, daily life, and symptoms (fatigue, pain, and discomfort). Others described the impact of the interview. One participant noted how she had been surprised by feeling physically tired on returning to work after her lunch-time interview and had not anticipated any negative impact, despite her employer advising of this possibility. It is difficult to assess these potential risks prior to an interview study, but this difficulty underscores the emerging ethics in qualitative research and the need for a reflexive approach, anticipating ethical moments and how to manage them throughout the process, as well as building measures into the study design that will minimize harm.

Discussion

We identified the value of qualitative health research57 and ethical issues that emerged in our in-depth interview study, applying the principle-based approach of autonomy, justice, and beneficence/nonmaleficence, which traditionally is associated with human subject protection in medical research, although increasingly criticized for its suitability to qualitative research. Our findings mirrored ethical concerns that arise in quantitative research (eg, issues of coercion and recruitment through health care professionals). We also identified ethical considerations in the research process (Fig. 2). There are fundamental differences between the positivist and interpretivist paradigms (Fig. 1) that are core to the different protections required for participants. The interpersonal researcher-researched relationship, which involves ongoing interactions, building mutual trust and respect, the active role of the participant and researcher in co-constructing the data generation, and the qualitative interpretation of the findings, is in direct contrast to the ideal of the passive research participant and the “arms-length” scientific researcher who observes the phenomenon and collects the results. The unfolding nature of qualitative research and the intersubjectivity of researcher and researched shifts the notion of informed consent as “a moment in time” and the lynchpin of human protection to a process embedded with ethical moments58 inextricably linked to research design and methodology inviting reflexivity throughout.59

Our findings are limited to issues that arose in our interview study. We did not design the project to analyze ethical aspects of research, but have descriptively drawn on field notes and interview extracts to illustrate ethical issues that emerged. We use the traditional principle-based approach as an exploratory exercise; we do not claim this framework is the most effective way of considering human subject protection. Applying the traditional approach to our qualitative study extends our knowledge of the utility and value of this approach in context and is a strength of this article.

In line with our study, recruitment through gate-keeping has been seen as a practical obstacle to research in the primary care setting.48,50 Some authors have suggested that physicians may not be suitably familiar with qualitative research and its potential contribution to EBP,60,61 and other authors have identified how this lack of familiarity is paralleled in quantitative research.62 Other authors broach ethical aspects and ask whether family physicians should be encouraged to recruit patients by offering both reimbursements and payments46 and whether clinician recruitment of patients can be assumed to be coercive or whether this assumption is another form of paternalism.49 Delaney suggested that research should be prioritized and made “directly relevant” to the training and development of the nonacademic clinician.50 Because ethical issues can emerge throughout the research process and are not confined to particular stages, such as gaining informed consent, a reflexive approach characterized by a thoughtful and ongoing self-awareness, attention to detail, and sensitivity to the individual's role in the research enterprise (eg, health care professional as research worker/recruiter) highlights relationship dynamics and enhances ethical oversight. Education and training that delineate the ethical differences between the quantitative and qualitative approaches seem key if health care professionals continue to play a role in research, including recruitment.

Gate-keeping can diminish autonomy in both the positivist and interpretivist paradigms. Our findings support other patient-participant reports of being recruited or influenced by health care professionals to participate in research such as in clinical trials.15 Other authors have called for an investigation into barriers to recruitment of groups perceived as vulnerable because of misguided ethical guidelines.63 Such barriers raise questions that also were prompted in our study: When does an individual's illness status induce vulnerability, and what is the impact on decisions to participate and giving informed consent?38 Gate-keeping also can threaten the validity of the findings; if the research fails to generate useful data (eg, due to a compromised sample),45 the subsequent wasted resources and production of research of limited validity are ethical issues.49 As in our study, other types of recruitment may be more appropriate to offer individuals the opportunity to exercise self-selection. This opportunity to exercise self-selection seems more in keeping with patient-centered care and collaborative research. However, sampling problems still arose in our study, illustrating that particular groups may be under-represented in research and their experiences neglected in EBP. In Canada, effective strategies need to be put in place to ensure groups and individuals have access to participation. These strategies may involve building trust and reaching remote communities and disadvantaged groups.

Reflecting our findings, the interview has been described as a moral endeavor,64 and 3 types of potential ethical problems in qualitative interviews have been identified: (1) the design itself, (2) the research relationship, and (3) the process involved in interpreting qualitative data, and particularly the role of the researcher in co-constructing the findings (results).65 Cox also noted the importance of ethical listening and being sensitive to when, and when not, to probe.66 Interviewer awareness of insensitive probing is indicative of a reflexive approach; sometimes the offer of “good data” may need to be relinquished in favor of “good ethics.” The potential for psychological harm when sharing sensitive information in in-depth interviews and the impact of taking part have been considered,67 as well as the potential to distort findings.68

Our findings illustrated how interviews are an effective way of gaining experience of suffering69 and how ethics emerge in interpersonal relationships in the interview situation.70 The researcher has a moral obligation to ensure that the research is worthwhile when people share time and suffering,71 and because there is scope to interpret and co-construct the realities that are then part of the knowledge base, acting reflexively is crucial to ethical practice.37 Our experiences revealed that it was difficult to know when ethical issues would arise, or how nuanced (constructing the interview guide) or unique (psychological harms) they would be.

Ethical issues emerged in the construction of the interview guide and the practical circumstances of the interview. Skillful researchers try to avoid imposing their own structures and assumptions upon interviewees’ view of the world.3 Careful recording of field notes and analytic reflection are required. Field notes serve as the written account of what the researcher sees, hears, experiences, and thinks and supplement the interview data. The interview is an ethical process, from the creation of the interview guide to the analysis of the accounts. This approach to ethics in research is informed by a process model of research ethics, whereby considerations of autonomy and respect for the perspectives and experiences of consumers are central from the early stages of study design.58

Regarding issues of knowledge translation, our interdisciplinary team faced challenges in discussions that highlighted our disciplinary assumptions and diverse perspectives. To facilitate mutual respect and shared decision making, measures were built into the research process to ensure inclusion of all perspectives; ongoing team communications were crucial in order to clarify meaning, recognize subtle findings, and communicate complexity effectively.

There also are ethical implications for knowledge dissemination and the development, application, and visibility of qualitative methods in health research. Although there has been seminal qualitative health research for more than 50 years (eg, about the patient-practitioner relationship72 and help-seeking73), disagreement exists about both the value and acceptance of the utility of the qualitative paradigm. Although previously ignored or viewed as a weak source of evidence, Sandelowski43 noted that qualitative research is now considered essential to the EBP goal of improving health care. Atkin54 described how patient accounts are still seen as “meaningless” and not regarded as evidence, but rather as anecdotal. Thus, despite the apparent shift beyond the oppositional quantitative and qualitative paradigms, opposition still characterizes many evaluative discussions in an interdisciplinary research context. Atkin noted that an emphasis on positivist-based priorities and quality appraisal may discourage the more reflexive engagement that characterizes qualitative methods and asked: “Is there a danger that by pressing for the inclusion of qualitative accounts in more applied research, often dominated by quantitative methods, we risk producing analysis that is descriptive, uncontextualised and little more than a token gesture, paying lip service to individual experience?”54 This question illustrates the inextricable link between ethics and methodology: if lip service is paid to investigating the “lived experience” of illness in all its messiness, there is a danger that we will not do justice to participant accounts.

Conclusion

We have identified: (1) that the traditional principle-based approach can be usefully applied to qualitative research ethics, if we use the concepts flexibly and prioritize context and process, and (2) that the realities of doing “everyday ethics” highlight ethical moments and the need for a reflexive stance at every level and stage of research. Reflexivity involves the researcher always being aware of the role of self in the process, a sensitivity to the ethical concerns that may arise, and taking measures to prevent harm (eg, anticipating that participants may experience psychological distress). An ongoing awareness of the ethical moments that may arise throughout the research process and one's role in the process increases ethical rigor at all stages of research, such as when designing the study (offering counseling services or a resource sheet with useful and valid Web sites) and during the interview (how to listen sensitively and when and when not to probe).

Ethical, rigorous, and useful qualitative research involves being accountable to research participants and echoes the ethical concerns in physical therapy. An ethos of patient-centered care and issues of empathy, autonomy, and respect are highlighted in our account of qualitative research from the study's inception to the dissemination of results. An ethical lens in qualitative health research brings particular challenges and insights to methodological concerns about recruitment procedures, the interview or other data-generation processes, and knowledge translation, and, ultimately, the usefulness of the findings to bring benefits to patients/consumers (and health care professionals) and to enhance health care.

Footnotes

  • Dr Townsend conceptualized the article and wrote all drafts of the manuscript, conducted the majority of the interviews, led the analysis, and was the principal investigator during the pilot phase of the project. Dr Cox consulted on the ethical issues emerging from the project and data analysis and assisted with the writing of the manuscript by reading the manuscript, making comments, and clarifying content. Dr Li was the principal investigator of the ERAHSE Project. As the senior author, she provided guidance on drafts and contributed to the preparation of the manuscript.

  • The authors thank the participants, who shared their time and experiences, and the members of the ethics review board who reviewed the study protocol.

  • Ethical approval was obtained from the University of British Columbia Behavioral Research Ethics Board and the Vancouver Health Research Institute.

  • This article was developed from a paper given at the CARE V International Conference; April 23–25, 2008; Oslo, Norway.

  • This research was funded by the Canadian Institutes of Health Research (CIHR). Dr Cox was supported through a career award from the Michael Smith Foundation for Health Research during the early part of this research. Dr Li was supported by a CIHR New Investigator Award and an American College of Rheumatology Research and Education Foundation Health Professional New Investigator Award.

  • Received December 4, 2008.
  • Accepted December 28, 2009.

References

View Abstract