Using Qualitative Research for Complex Interventions

What is already known?

While qualitative research has been recognized as a means to conduct evaluation of complex interventions, most existing literature only describes this in general. In our article, we provide examples of a complex intervention and detail how hermeneutics was used as a research methodology to further understand this intervention.

What this paper adds?

This article provides a specific example of evaluation of complex interventions using hermeneutic methodology. We believe this specific detail will be helpful to further understand this type of qualitative research and how it may be applied, more generally, to other types of complex interventions.

Complex interventions are seldom described comprehensively or their effects explained adequately (Medical Research Council [MRC], 2008; Pawson & Tilley, 1997; Petticrew, 2011; Shepperd et al., 2009). Consequently, there is a need for better methodologies to understand and evaluate complex interventions (Clark, 2013; Clark, Briffa, Thirsk, Neubeck, & Redfern, 2012; Lewin, Glenton, & Oxman, 2009; MRC, 2008; Mohler, Bartoszek, Köpke, & Meyer, 2012; Moren & Blom, 2003). Qualitative research methods have been proposed for undertaking research into complex interventions by increasing knowledge of intervention assumptions (MRC, 2008), components (Clark, 2013), and active ingredients or mechanisms (Pawson & Tilley, 1997); these methods to research complex interventions are underdeveloped. Despite the existence of frameworks to set global objectives for evaluation (MRC, 2008; Moore et al., 2015), practical and detailed guidance on how to use qualitative research in complex intervention research is scant. Methodological guidance is therefore needed that provides specific guidance into how to examine complex interventions (Blackwood, O’Halloran, & Porter, 2010; Clark, 2013; Paley & Lilford, 2011).

The current article addresses how to use qualitative research methods to study complex interventions. In particular, this article shows how hermeneutics can be used to evaluate complex interventions. An overview of complex interventions and methodological issues will be given, followed by an outline of Hans Georg Gadamer’s (2004) philosophical hermeneutics. We then describe how this qualitative research can be used to understand complex interventions using examples from family intervention research.

Complex Interventions

Complex interventions are defined as “interventions with several interacting components” (MRC, 2008, p. 6). For the intervention to work, a large number of parts must come together in certain ways (Byrne, 2005; Clark, 2013; Pawson & Tilley, 1997). These parts can be “material, human, theoretical, social, or procedural in nature” (Clark, 2013, p. 8). The MRC (2008) lists several characteristics of complex interventions: there can be a number of different components to the interventions; there can be a range of possible outcomes, varying by population; outcomes are influenced by the behaviors of those delivering and receiving the intervention; interventions can target a number of groups or organizations and may incorporate varying degrees of flexibility or tailoring. Common complex interventions in health include chronic disease management programs, tobacco cessation programs, or health policy interventions, like a public smoking ban.

The complexities of these health interventions are not always well acknowledged, represented, or incorporated into intervention design and evaluation (Clark, 2013; Glasziou, Meats, Heneghan, & Shepperd, 2008; McCleary, Duncan, Stewart, & Francis, 2013). Current methodology and reporting practices make it difficult to replicate complex interventions, understand their effects (Pawson & Tilley, 1997), understand which components of interventions contribute most to effects (Clark et al., 2012), translate research into practice and policy (Glasziou et al., 2008), and translate the benefits of effective interventions from research into real-world contexts (Pawson, 2006). Furthermore, interventions often have inconsistent effects (Clark, MacIntyre, & Cruickshank, 2007). In short: Most nonpharmacological interventions are poorly described (McCleary et al., 2013) and this decreases the applicability and usefulness of findings about these interventions to practice (Glasgow, 2008). Often only a small number of components that are most easily observed and/or measured are studied—but these may not be the components that are actually influential (Clark et al., 2012).

The MRC (2008) framework encourages developers and evaluators of interventions to have a sound theoretical understanding of how the intervention causes change, to evaluate its effectiveness in everyday practice, and to ascertain how the intervention works. Guidelines for process evaluation of complex intervention outline the need to understand the context of the intervention, implementation, and mechanisms of impact (Moore et al., 2015). Context is the preexisting conditions, or anything external to the intervention, that may affect the outcome. Implementation includes how the intervention was delivered, for example, training of staff, timing, and dosing. The mechanism of impact refers to the causal mechanism, or what created change, and needs to be understood to facilitate transfer of the intervention to different contexts or populations.

Better methodologies are needed to address important questions about how complex interventions work, for whom, under what circumstances, and in which contexts (Pawson & Tilley, 1997). This involves using methods to understand aspects of the intervention context, implementation, and mechanisms, including study of the people who provide and receive the intervention, elements of the interventions that are observable and readily measurable, and those which are more hidden and tacit (Clark et al., 2012).

The recent guidance published by the MRC (Moore et al., 2015) has described how to carry out process evaluations to include consideration of context, implementation, and mechanisms of impact. They encourage the use of quantitative and qualitative research and process as well as outcome evaluations. In terms of process evaluation, qualitative research is suggested for two purposes: to explain quantitative findings and generate hypothesis to then be tested quantitatively (Moore et al., 2015). Although the MRC provides guidance on general use of qualitative methods, there is limited detail on the specific qualitative methods, and readers of those guidelines are encouraged to consult other sources. Several recent writings have attempted to address the process evaluation of complex interventions. For example, May (2013; Johnson & May, 2015) focused on theories of implementation, in particular changing the behavior of health-care providers who are delivering interventions. We suggest an interpretive methodology, such as hermeneutics, can be a useful part of process evaluation for complex interventions to not only understand issues of implementation but also theorize mechanisms of actions and understand context and acceptability of the interventions.

An interpretive methodology for qualitative research is appropriate for complex interventions. Pawson (2006) has stated that social science research has strong interpretive dimensions and that meta-analyses often write out nuance and interpretation in their processes. In this way, health services interventions are represented in journals as consistent, static, pills but actually involve interpretation in provision, receiving (e.g., from patient) and then in the publications in writing and interpretation of the research. In other words, complex interventions and the research and evaluation of them are actually interpretive from start to finish. In our following examples, we will show how the results from hermeneutic research can both offer explanations and generate hypotheses about the context, implementation, and mechanisms of complex interventions.

Use of Experience

One major difference between Gadamer’s (2004) hermeneutics and how other qualitative research commonly operates is the positioning of human experience. Qualitative research is often focused on eliciting participants’ accounts of their experience of a phenomenon or how they make sense of their experiences (Paley, 1998) as a psychological state or subjective experience (Paley, 2014). In hermeneutic research guided by Gadamer’s philosophy, a person’s experience of the phenomenon is seen as an experience of some thing. This thing is the topic, or phenomenon, of interest, rather than the interior, subjective, psychological meaning of the participants (Gadamer, 2004). If the topic is a complex intervention, researchers would speak with participants, who may be delivering or receiving the intervention, about their experience of the intervention as one possible way to further understand the intervention—keeping in mind that these participant’s recall and bias will also influence how they have experienced and report what they believe to be useful about the intervention. The purpose of interviews in hermeneutic inquiry is therefore not to “give full expression to the emotional lives” of the participants (Nunkoosing, 2005, p. 705).

To provide a simplified illustration of the focus on the topic rather than the experience, consider a group of people seated in a room around a table. If researchers were interested in learning about the subjective meaning or experience of these people, they would ask them what it is like for them to be seated around the table. Hermeneutics asks what their experience of sitting around the table can tell us about the table. What might each of their perspectives provide to help us better understand the table? It is this shedding light on the topic that is valued in hermeneutic research, more so than how people experienced the topic. Applied to an intervention, the purpose of hermeneutical dialogue about the intervention is to learn something about the complex intervention rather than gain insight into the subjective, psychological experience, or meaning that the intervention has for the participants. Hence, participants’ experience of complex interventions provides a means to increase understanding of the intervention as opposed to uncovering the essence, meaning, or subjective experience of the intervention.

An example of this might be a heart failure home management program. Rather than focusing on how the participants experienced the program, the focus is more on what their experience can tell us about the program itself. This may be about the materials or technology, the timing, the context, the relationships, or the providers of the program, and so on, but it is less about the subjective meaning that the participants give to participating in the program. Similarly, in relation to complex interventions, providers or recipients of the intervention can be asked about its components, mid-range theory can be invoked about what the likely effects of particular components will be, behavioral research can help to explain the success of particular responses, and metaphor may be used to facilitate understanding of several interacting components.

Participants’ perspectives are not the only way to learn about the topic. While “people’s understanding of their ‘world’ is necessary for a social scientific account of that world…it does not have to be, and usually is not, sufficient” (Paley, 1998, p. 820). Other literature or theory can also be used to help understand what is going on (Jardine, 2006). For instance, linguistic and etymological analysis, whether it is transcripts of interviews or other text, may also be used for analysis. Metaphors may help to explain the complexities of topics that are difficult to understand—providing something familiar in the context of the unknown (Moules, Simonson, Prins, Angus, & Bell, 2004). It may include research from other disciplines and from other methods. The possibilities for what diverse sources may be required to understand the topic is numerous and this searching out becomes a part of the research process. Returning to the example of the heart failure management, information about the intervention could also be sought from family and caregivers, from theories of change, writings about technology interface, and so on. Researchers may look at how different programs worked in the same contexts, current political or geographical circumstances, and education programs of the intervention providers. All of this information will provide information about the intervention, rather than a person’s experience or subjective meaning.

If interviewing participants is chosen as part of the data collection, subsequent analysis in hermeneutic research is not about the accurate representation of the experience of the participants, but rather what it is they have to say about the topic that points to a new or different understanding. In this way, participants’ language signifies, or points to, something of significance in how the phenomenon is understood. This would include information such as the acceptableness of the intervention.

Language is a significant component of analysis in hermeneutics. Language reveals social and personality processes, extending beyond literal meaning and semantic context, and reflecting particular worldviews (Pennebaker, Matthia, & Neiderhoffer, 2003). In qualitative research, language is the measurement tool, reflecting beliefs and attitudes using a “calibrated linguistic scale” (Paley & Lilford, 2011, p. 957). Rather than reflecting the participant’s access to some internal psychological state, qualitative data usually more accurately reflect the typical and common manner in which the topic is talked about. “It [the interview] should be construed, instead, as an example of what ‘they’ say about events of the kind being narrated. It is merely the ‘script’—incorporating a series of stereotypical assessment, roles and imputed motives” (Paley, 2014, p. 9). These factors are important to consider in the analysis of the participants’ accounts.

Through the work of reconciling the researcher’s previous understanding of a topic with the new understanding that is achieved through analysis, potentially using a multitude of resources to develop this analysis, there is a fusion of horizons (Gadamer, 2004). This explains why it is not necessarily important to return to participants to verify if the interpretation that the researcher arrived at was accurate. In the table example, it is easily demonstrated that what the researcher learns from the carpenter may not make sense to the person seated at the table, but that does not make it invalid. Likewise, providers and recipients of a complex intervention may well differ in their perceptions of its components and indeed are likely to because of their different backgrounds, qualifications, socializations, biases, and roles. Rather than making the findings of this research completely relative, however, these differences actually uncover further complexities.

Because hermeneutics is not about understanding the subjective experience of the participants, it avoids the trap of relativity. Gadamer (2004) was critical of the relativism of postmodernism, stating that knowledge cannot be merely someone’s opinion.

For postmodernism, criticism of the infatuation with method and with scientific certainty ends in questioning all truth, all notion of the adequacy of consciousness and of reality. For the postmoderns, there are therefore only interpretations or perspectives, interchangeable and equivalent. (Grondin, 2003, p. 3)

Contradictory Findings

Gadamer (2004) famously argued that truth is not solely reached through scientific method or via interpretations and perspectives (Grondin, 2003). Truth is considered in the sense of aletheia—the opening of something that was closed, the remembering of something that was forgotten, revealing of something concealed, enlivening something that was dead (Jardine, 2000; Moules, 2003). Truth in the sense of hermeneutic philosophy is aletheia (Jardine, 2000), not the complete unconcealment of things, but the simultaneous exposure and concealment (Gadamer, 2004).

It is this type of hermeneutical understanding that offers a way to understand contradictions in findings related to complex health interventions. This addresses the thorny of issue of how discrepancies in accounts of interventions should be handled and reconciled—for example, when providers and recipients present conflicting accounts of an intervention’s components, mechanisms, or powers. While current frameworks, notably from the MRC (2008) and realist evaluation (Pawson & Tilley, 1997), leave this vital issue ignored, hermeneutics provides researchers with a means to allow for and begin to reconcile these differences and to practically handle data about complex interventions. Indeed, as Gadamer (2004) stated, there will never be an end to the ways in which we understand a topic, because “there is no absolute progress and no final exhaustion” (Gadamer, 2004, p. 86) of the topic of interest.

Differing accounts of the most useful parts of a complex intervention does not necessarily mean that one participant’s account was correct, and the other’s was incorrect, but rather reveals the nature of complexity. Indeed, there may be a variety of components, mechanisms, and processes within a complex intervention, and because it is complex, there may be differing reasons why the intervention was helpful or at least different perceptions of why it worked. Different accounts of what worked in the intervention may exist because for different people, at different times, and in different contexts, different mechanisms contribute to change. Working out these discrepancies, moving from the parts that may be revealed in interviews, to a whole picture of the intervention is part of the work of analysis in hermeneutic research.

Preunderstanding

The rigor of qualitative research is particularly vulnerable when it lacks some of the devices that have been employed in quantitative research to ensure that what is produced is not just well-composed rhetoric of a well-meaning, but biased, researcher’s opinion. Morse, Barrett, Mayan, Olson, and Spiers (2008) claim that rigor in qualitative research also needs to be reclaimed as a component of the research process, rather than only the post hoc evaluation of whether the research is written up in a convincing and plausible manner.

Prejudice and bias has been written about extensively related to hermeneutics, reflecting on Gadamer’s (2004) original description of this phenomenon. Jardine (2006) and Moules (2002), for example, stated that the researcher’s experience with the topic is what allows the topic to be understood in a certain manner and thus does not require the researcher to forget or bracket everything they know about the topic to conduct research. Gadamer did not consider these preunderstandings to be negative, but as something that is actually necessary and useful for understanding (Maggs-Rapport, 2001). Moreover, prejudices allow researchers to understand the topic in ways that would not have been otherwise possible (Moules, 2003). It is not possible to cleanly remove and suspend all of our prejudices and biases as researchers (Jardine, 2000; Tversky & Kahneman, 1975).

In qualitative research, every attempt must be made to be open to what might be different, new, or even surprising and to overcome inherent human cognitive biases or illusions (Paley, 2005). Gadamer (2004) wrote about this in terms of experience—which means to also be open to experience things differently. In all types of research, interpretive decisions influence the research.

Understanding a complex intervention requires researchers to incorporate their personal reflections in analysis of data—drawing on their hunches, disciplinary background, and theoretical and substantive knowledge rather than dismissing such contributions as being biased. Previous experience and understanding of a topic does not prevent a researcher from being open to new understanding of the topic but is an asset that enables the researcher to be better prepared for understanding. Using interview data uncovers the stereotypical ways in which the intervention may be typically thought about. Researchers, therefore, need to be able to move beyond their own preconceived notions of the intervention, to challenge the rhetoric that may be reported by participants. Otherwise, the results of the qualitative research will only be what the researcher already knew about the topic (Paley, 2014), or a reflection of common discourse, rather than a contribution of new knowledge.

Analysis and Rigor

The story of La Loba, as written by Clarrisa Pinkola-Estes, explains the process of hermeneutic research metaphorically (as cited by Jardine, 2006). It particularly reflects how the work of the researcher is collecting information and putting it together in a way that makes coherent sense as a whole and affords a better understanding of the parts. In this legend, La Loba’s work is to collect bones and then assemble a skeleton; her specialty is wolves. It is then said that she sings over this skeleton, and as she sings, the creature grows skin and fur, begins to breathe, and eventually runs away (Jardine, 2006).

The way in which hermeneutic research on a complex intervention can be judged as being of sound quality, or having rigor, is the same way in which we know that La Loba has constructed a wolf. The interpretive process of data collection about the intervention involves the collecting of bones; the process of analysis and writing is the fleshing out—to come to a more complete understanding of the intervention. In the writing, however, it still needs to be recognized as an intervention and be an accurate and comprehensive account.

To relate this metaphor to the study of complex interventions, it would be obviously insufficient to understand a wolf simply by counting the number of bones. This is something that is certainly measurable and perhaps meaningful, but to understand how something works as a whole (i.e., the entirety of the wolf or the entirety of the intervention) is different. Because of the fallacies of self-reports from participants (Paley, 1998, 2013), the researcher needs to take what is found (e.g., the bones) and work to flesh out the research by moving beyond typical understanding of an intervention to a more detailed and sufficient account of what is going on in the intervention.

Gadamer (2004) wrote at length about the fecundity of the individual case. Whereas some qualitative research projects gather data from or with participants until there is saturation—that is, nothing novel is emerging in themes—the number of participants is irrelevant in hermeneutic inquiry and is not a valid assurance of rigor (Jardine, 1992). What is more important is the amount of new information that the researcher can gather—from a variety of sources—to further understand the topic in general. Gadamer stated that this is how judgment operates, as exemplified in jurisprudence. How law is applied affects the individual case, but the individual case also contributes inductively to the understanding of the law (Gadamer, 2004). Likewise, even a single, specific experience of a complex intervention may shed new and transformative light on an aspect of the intervention that could further an understanding of the whole intervention. It is this type of analysis that can help with explaining quantitative findings and also generating hypothesis to be further tested.

Timing With a Metaphor

One of these particular aspects that stood out to the researchers was how time relates to grief (Thirsk & Moules, 2012). At the time of the study (data collected 2007–2008), there was debate about including complicated grief, or prolonged grief response, in the Diagnostic and Statistical Manual, fifth edition (DSM-V). The proposed criteria implied that most problematic symptoms of grief should be resolved within 6 months (Prigerson & Maciejewski, 2005). Other research claimed that grief was a life-long, life-changing experience (Moules, 1997, 1998). In addition, there are numerous colloquial expressions related to grief such as “time heals all wounds” or “you’ll get over it in time”—this represented the way in which grief was typically thought about. Being knowledgeable about the literature on grief, the researchers were attuned to these issues of time. This previous knowledge helped in understanding the current thinking about grief but was not a prejudice that needed to be suspended or forgotten.

In an interview with one of the family members, the interviewee stated that the therapeutic conversations were not that helpful because the timing was not right (Thirsk, 2009). The nurse who worked with this family member offered further insight into the timing. She suggested that perhaps the intervention was not helpful because the woman was trying to keep an even keel on her family’s ship. This phrase is a colloquialism but was originally a nautical term meaning to keep a ship steady or a state of balance, not tilting. This was further explored in the analysis, thinking of when it might be difficult to keep an even keel, such as in a storm. This metaphor of a ship in a storm was taken up by the researchers and further explored (Thirsk & Moules, 2012). For example, what is needed for intervention in the middle of a storm (e.g., battening the hatches, trimming the sails) is different than what is needed for intervention when the storm has passed and it is realized that the ship has veered off course. The use of this metaphor seemed much more useful than aligning what type of interventions a grieving family might need with the amount of calendar time that had passed.

Investigating this topic of time further revealed information about how human beings perceive time. Not surprisingly, most people view and mark time in terms of significant events, these are called eras (Vessey, 2007). It is less prevalent that people think or experience life according to clockwork or calendar time. These ideas are incongruous with the time line originally proposed for the DSM-V diagnosis of complicated grief and question the validity of this clock or calendar-based conception of time. An alternative conception would be, as with this storm metaphor, to align grief interventions with the ebb and flow of family needs and adjustment (Thirsk & Moules, 2012).

The metaphorical framing that resulted from this hermeneutic research offered a different insight into what was going on in the intervention and the timing in grief. It offers both practical guidelines around timing of interventions and a different conception of why some interventions might be more appropriate than others at certain times. Although this metaphor and analysis was based on information from only a few participants or cases, it helps to inform a larger transformative reframing or understanding of this complex intervention. What was discovered here was not significant because of saturation of the data, or numerous recurrent themes, but rather because the analysis led to new understanding of the topic of interventions with families experiencing grief.

Uncovering this component helps in understanding and explaining the complex intervention and also in generating further hypotheses, particularly around the timing of the implementation of the intervention. For example, if families are offered more instrumental assistance when they are in the immediate crisis of bereavement, and then more reflective, meaning-making interventions once their lives have settled somewhat, does that lead to more effective interventions for grief? This provides an opportunity to integrate the qualitative research findings with designing quantitative research which can help to confirm or refute if this was indeed a significant component of the complex intervention. Not only would this new information be helpful to theorize when designing an intervention but also in the evaluation and improvement of existing interventions.

Clinical Judgment

Another interpretation was developed regarding clinical judgment during therapeutic conversation (Thirsk, 2009). This analysis revealed the complexity of these types of interventions in terms of resolving seemingly conflicting accounts in data. In this instance, the family members and the nurse gave different accounts during the interviews about whether or not the intervention was helpful (Thirsk, 2009). On review of the videotapes of the sessions, the final session concluded with the family stating that the intervention had been helpful and that the clinical work could be concluded. When the nurse was interviewed later for the research, she stated she had a sense that the intervention had been helpful, while the family, when interviewed, stated they had not found it to be useful.

This divergence is important because, for example, both the MRC (2008) framework and other approaches (Pawson & Tilley, 1997) suggest that it is important to understand the powerful components of interventions but do not address how to resolve differences in perception about what is powerful. A hermeneutical approach to complex interventions in this instance leads to an exploration of sense (Thirsk & Moules, 2013) that involved an exploration of clinical judgment and expertise (Cioffi, 1997) and description of aesthetic judgment (Gadamer, 2004). The concept of a change in beliefs being the foundation for sustained change in families is the main tenant of the Illness Beliefs Model (Wright et al., 1996). Overall, an initial interpretation that perhaps because some peripheral beliefs, rather than core beliefs, had been changed, the nurse judged that the clinical work had been useful, but the change from the intervention was temporary and not sustained with the family (Thirsk, 2009). In other words, some issues had been addressed but the underlying concern had not been shifted sufficiently. The family thus reported that things were improving, which the nurse interpreted as sufficient, but as time passed, the unresolved reasons for this family’s suffering resurfaced. This is one way in which the work of analysis leads to an interpretation that accounted for the discrepancy and created a coherent whole that would allow for both accounts to be true.

Another interpretation might suggest the family, at the conclusion of the sessions, agreed that the intervention had been helpful, when in fact, it had not been. This desire to please, perhaps particularly in relationships with a power differential, such as with a health-care provider, has been demonstrated in social psychology research (Higgins, 1992) as an act of tuning the communication for the audience. To date, there has been minimal exploration of cognitive bias such as these in family nursing practice (Thirsk, Moore, & Keyko, 2014), but this hermeneutic research on grief interventions offers a clue that this interaction could be a component of the intervention. This is another example of how the work of analysis can offer explanations of discrepancies and contradictions.

This example provides important insight into the nature of complex interventions. Professionals and lay participants (such as patients, caregivers, and family) may offer conflicting accounts of the same intervention in terms of effective components, mechanisms, and the moderating role of contextual factors as a result of markedly different agendas, history, life position, and values. There are numerous other interpretations that could be made of these data, using a wide variety of resources to help understand the intervention. For example, because participants were interviewed up to 10 years after their loss and up to 7 years after the intervention, memory, recall, and shifts in narrative perspective all become issues.

As such, the seemingly simple task of collecting data on the influential components of interventions, as recommended, for example, by the MRC (2008) framework, is not straightforward. While data can be collected from one perspective only (e.g., the recipient of the intervention), it raises the issue of whether the providers of intervention are equally well placed to identify how it works—a stance argued for by Pawson and Tilley (1997). Yet, by collecting data from multiple parties, researchers have to grapple with what to do when accounts compete and conflict. This often occurs in the context of the need to identify specific recommendations of how to improve the intervention. As such, mere description or documentation of different perspectives and equivocal conclusions do not suffice.

Hermeneutical perspectives on complex interventions can successfully embrace the multiple interpretations that can be made of the intervention. This type of research requires the researcher work out the various possibilities of what was happening in the intervention. Discovering the nuances of mechanisms and components can then provide a framework for further research and testing, as it would not be possible to say conclusively which interpretation, or hypothesis, was the correct one. Specifically in this example, further quantitative study design could help to determine the significance of memory/recall, or a desire to please, by using nonself-report indicators. Further studies could also be designed that examine core beliefs and whether or not it is a change in core beliefs that leads to sustained change. This provides an example of how hermeneutic research can help to explain other findings (that the intervention was or was not helpful) and also generate further hypothesis to understand implementation, context, or mechanisms in complex interventions.

Being a Stranger

A final example offers further understanding of how important mechanisms of complex interventions are discovered using hermeneutics. Prior to researching family nursing interventions with grief, the first author was curious about why a thing such as grief interventions exist (Thirsk, 2009). When, where, and why did it happen that people started seeking help from professionals when a loved one died? This led to an exploration of the evolution of grief in the professional literature throughout the 20th century. Several pivotal and historical documents about grief were reviewed—beginning with Mourning and Melancholia (Freud, 1957/1917) and reading further about Bowlby (1980), Worden (1982), Kubler-Ross1 (1969), Attig (1996), Silverman and Klass (1996), and Neimeyer (2001). Seeking grief intervention from a professional is a relatively new phenomenon in Western culture (Jordan & Neimeyer, 2007; Walter, 2000). From a sociological perspective, Walter (2000) suggested this shift occurred because the bereaved no longer have a support system of family, friends, and neighbors nearby to talk with or to engage with in mourning rituals. A further explanation was the decreased involvement of people in organized religion (Walter, 1996). This background information helped to guide the research to uncover what, if anything, was significant about family members seeking help for grief from a professional.

One family member attributed a significant part of the helpfulness of the intervention to the nurse (Thirsk & Moules, 2013). Talking with a professional, as opposed to a family member or friend, meant that this woman felt no burden of protecting anybody, and thus she was able to hear things from the nurse that she did not hear when said by family and friends. Although this idea aligns with previous research about nurses being “compassionate strangers” (Robinson, 1996) who are involved in the process of healing, but not the problem, this interpretation offered new knowledge to specifically explain a mechanism for grief interventions by going beyond previously documented explanations for why bereaved people seek help from professionals. So rather than professionals being needed because of a decrease in organized religion, or the globalization and movement of family and friends, this intervention’s effectiveness could be in part due to the nurse being someone unrelated, who the family member does not need to worry about (Thirsk & Moules, 2013).

This analysis again exemplifies how the knowledge/experience the researchers had prior to the research enabled a different understanding of the topic, rather than confirming a bias that was already held. The goal of this analysis was not to better understand the experience of the family member, but rather furthering understanding of the intervention through the experience of the family. Whereas the limited description of qualitative research found in the MRC (2008) framework might lead researchers to conclude that professionals are a key component of interventions, expanding on this interpretation invited further analysis about why it was helpful to talk with a professional—namely, because of a belief that the bereaved would burden family and friends by speaking with them. It can be seen here that when components and mechanisms of complex interventions (e.g., how, when) are understood, it is easier to implement, modify, and adapt to different settings. For example, if someone who is experiencing grief cannot access a professional, but talking about their grief is helpful, then who might they be able to talk to without feeling they are a burden? On the other hand, can some of the beliefs about being a burden be challenged, to enable the person to speak with family and friends about their grief? This latter option may be helpful when considering the experience of one couple’s grief who stated that 7 years after their daughter’s death, the research interview was the first time they had heard about what each other’s grief had been like, and it was one of the most helpful conversations they had had (Thirsk, 2009). Perhaps their worry about burdening each other, and thus seeking help from numerous professionals, prevented this helpful conversation from happening earlier.

Summary

Hermeneutics can be used to increase understanding of complex interventions—providing explanations of quantitative findings and generating hypothesis for further investigation. Our examples reveal that hermeneutics offers a useful and sound understanding of context, implementation, and mechanisms in complex interventions. Judging these interpretations of data to be valid, reliable, or rigorous cannot be done by ensuring that method was adhered to, but rather that the work of the research is a careful analysis and interpretation of the possible mechanisms, parts, and components of complex interventions. Other researchers may have arrived at differing interpretations in a hermeneutic study, but their interpretations would likely add to our knowledge and do not necessarily mean others’ interpretations were wrong. This does not mean that all interpretations are equal, and thus everything is relative. Judgments should be made about the interpretations usefulness as an explanation and, when viewed as a mode of discovery, further research can explore whether these explanations of the complex intervention hold up. Most importantly, this allows researchers to address whether the contribution added to the understanding of the complex intervention in a new way that allows for new and different understanding.