The Use of Suffering in Pediatric Bioethics and Clinical Literature: A Qualitative Content Analysis

Annie B Friedrich, Kirsten M Dempsey, Erica K Salter

METHODS

Our primary aim was to first determine how often the term “suffering” is used in the pediatric bioethics and clinical literature and to identify how authors define and operationalize suffering when referring to issues in pediatric treatment and care. We conducted a qualitative content analysis in order to identify patterns in the textual data as well as to interpret why these patterns exist. [11,12]

In order to first identify bioethics articles that address pediatric care, one author (ABF) searched prominent bioethics and pediatric clinical journals using the following key words: pediatric(s), child(ren), young adult, teen(ager), and kid(s) (see below, Table 1).  Only English-language articles published between 2007 and 2017 specifically pertaining to the care or treatment of children were included in the content analysis. We then used a computer search function to find all occurrences of the word “suffering” in the articles that fit inclusion criteria: first, each article had to be primarily a pediatrics article, then it had to include at last one use of the term “suffering.” The verb form “to suffer” or “suffering” used as a verb (i.e. “suffering from”) were not included in the analysis because of their common use in medical literature pertaining to physical ailment (for example, “the patient suffers from high blood pressure”). We also excluded all uses of the word “suffering” as a person (i.e. “to care for the suffering”) (see below, Figure 1).

We used a combination of inductive and deductive coding strategies, both to identify well-established definitions and uses of the word “suffering” and to identify instances in which “suffering” is used uncritically or ambiguously. All authors developed the initial codebook based on a review of the literature, and two authors (ABF and KMD) conducted initial coding. We coded at three levels: 1) the one who experiences suffering, 2) the cause of suffering, and 3) the experience of suffering (i.e. what the experience of suffering entails). Our first code aimed to identify who is suffering: the patient, the care team, the parents or family, or a combination of parties. Then, we assessed what is causing the suffering. For the patient, suffering could be caused by treatment or physical intervention, the patient’s underlying condition, or an ambiguous or other specified cause of suffering. For the care team and family, the cause of suffering could be ambiguous or related to making a difficult decision. Our third level of coding explored the experience of suffering as either physical pain or distress, a negative emotional, psychological, or spiritual experience, a combination of both physical and emotional distress, or an ambiguous experience of suffering (see below, Table 2).

After the codebook was developed and the data were analyzed, the authors established inter-rater reliability by checking 10% of the codes and reconciling all differences until >80% agreement was reached. In order to achieve consistency in our coding process, the authors made several initial qualifications: first, whenever parent or family suffering was identified, we characterized the experience of suffering as a negative emotional or psychological reaction. Although it is certainly possible that parents and family members could be experiencing physical distress, authors seemed to most often draw on this concept to highlight emotional turmoil rather than physical discomfort, as the care team would not be treating the family’s physical symptoms.

Also, whenever the phrase “pain and suffering” was used, we characterized suffering in this context as something in addition to pain, therefore as something emotional or psychological; however, it is possible that the phrase “pain and suffering” is used to refer to a unitary concept. Lastly, there were several occurrences of the concept of suffering as it relates to general “human suffering.” We coded these excerpts as an example of suffering among all parties—patients, families, and care teams, as it seemed likely that the rhetorical intent of the term was to evoke a sense of a more general shared suffering present in many difficult cases.

As all of the data are publicly available, research ethics approval for this study was not necessary.

RESULTS

Description of Findings

We identified 651 unique instances of the concept of “suffering” in 121 different articles. Fifty-nine articles (48.8%) discuss general pediatric issues such as pediatric euthanasia (n=9), end-of-life issues apart from euthanasia (n=19), or other pediatric care topics (n=31). Forty-five articles (37.2%) address neonatal issues such as the management of congenital anomalies (n=18) or end-of-life care in neonates (n=8).  The remaining seventeen articles discuss perinatal care or other issues such as supporting the grieving family and clinician moral distress (see below, Table 3).

Cause and Experience of Suffering

In over 90% of the excerpts (n=588), “suffering” referred to the suffering of the patient, as opposed to the suffering of the care team or family. Three excerpts (0.45%) reference the suffering of the care team, twenty-eight excerpts (4.3%) reference the suffering of the parents or family, and thirty-two excerpts (4.95%) refer to some combination of suffering among the patient, family, and care team.

In seventy-three percent of the excerpts, the cause of the patient’s suffering was ambiguous. When the cause of the patient’s suffering was identified, authors identified specific treatments and physical intervention (9.4%) or the patient’s underlying condition (7.4%) as the cause of suffering. Other specific causes of suffering among patients refer to societal attitudes regarding disability [13] or impaired bonding between child and parent. [14]

The cause of suffering of the care team was most often related to making a difficult decision (67%) and for the majority of instances of family suffering, the cause of suffering was ambiguous (79%), although some authors did identify inability to protect one’s child as a cause of suffering for parents. [10] (see below, Table 4)

Next, we analyzed each occurrence according to the specific type of experience referred to by the term “suffering”: (1) physical pain or distress; (2) negative emotional, psychological, or spiritual experience; (3) a combination of both physical and emotional distress; or (4) ambiguous with regard to experience. A vast majority (74%) of the excerpts were ambiguous about the experience of suffering (n=484). When the experience of suffering was specified, the most commonly used characterization referred to a negative existential, spiritual, or emotional experience (17%) (see below, Table 5).

Comparing the Use of Suffering in Bioethical and Clinical Contexts

When comparing the use of “suffering” in the bioethics literature and the clinical literature as represented by the AAP’s journal Pediatrics, clinical articles were more likely than bioethics articles to identify specific treatments or physical interventions as the cause of suffering (15% vs. 6%) and were more likely to characterize suffering as physical pain or distress (11% vs. 2%). Furthermore, clinical articles were less likely to posit an ambiguous characterization of suffering than bioethics articles (65% vs. 78%).

Use of Suffering in the Context of Decision-Making

We hypothesized that the use of the concept “suffering” is often used to influence decision-making and identified four specific ways in which the concept of suffering might impact decision-making: (1) suffering used as a reason to withhold/withdraw treatment (or offer palliative care); (2) suffering used as a reason to offer or continue curative interventions or treatment; (3) suffering explicitly rejected as a valid concept to influence decision-making; or (4) suffering not used in connection with decision-making (or no specific position articulated with regard to treatment). The authors used surrounding textual context to identify instances where the term was invoked to influence decision-making.

Fifty-two percent of the articles (n=63) included in the study invoked the term “suffering” as a part of a specific medical decision, such as whether to withhold or withdraw aggressive medical intervention, whether to terminate a pregnancy because of certain genetic diagnosis, or whether to euthanize a non-terminally ill child because of concerns of present or future suffering. In thirty-two percent (n=208) of the excerpts, suffering was used as a justification for withholding or withdrawing treatment. In twenty percent of the usages, suffering was used as a reason to offer treatment or curative therapy (10%, n=67) or was explicitly rejected as a valid marker to influence decision-making (10%, n=66). The remaining forty-eight percent (n=31) of the occurrences did not draw on suffering as a factor in decision-making or did not articulate a particular decision with regard to treatment (see below, Table 6).

DISCUSSION

Differences between Bioethicists and Clinicians

Because bioethicists and clinicians have different perspectives and priorities when treating ill children, their use of the term “suffering” differs. Overall, the clinical literature more often characterizes suffering as physical pain related to specific treatments or interventions compared to the bioethics literature. Furthermore, clinicians are less likely to be ambiguous with their use of the concept of “suffering.” These differences may, in part, stem from the different relationships and responsibilities of parents, providers, and bioethicists. For example, de Weerd and colleagues [15] show that while parents’, physicians’, and nurses’ perceptions of suffering in pediatric patients overlap some, they also differ in important ways based on the relation to and kind of responsibility owed toward the child. While parents describe the suffering of their child in relation to physical signs and future perspective of the child, nurses have a more short-term perspective of suffering in terms of the child’s current level of pain and comfort, and physicians describe suffering as discomfort, pain, anxiety, and dyspnea related to the intensity and impact of therapy. Likewise, bioethicists may be more focused on generating general themes and claims when discussing pediatric suffering than making actual decisions for particular patients, which may lead to more ambiguous characterizations of suffering overall.

Characteristics of Suffering

Given these differences between bioethicists and clinicians, the authors were interested in identifying specific definitions of pediatric suffering in the excerpts from which to compare other uses. However, of the articles examined, suffering is never given an explicit definition. Rather, only defining characteristics of suffering are discussed. An exploration of the data yielded three common characteristics: the subjective nature of suffering, suffering as an experienced phenomenon, and the immeasurability of suffering. First, one clinician writing in a bioethics journal identifies suffering as something that “is wholly subjective and therefore can be gauged only by the individual.” [16^p.457] He distinguishes this from the quantifiable features of pain, nausea, and other noxious sensations that may be described as anatomic or physiologic processes. Meanwhile, he asserts that the “experience” of these physiologic processes (like the experience of pain or the experience of nausea) is part of what constitutes the experience of suffering, and thus is purely subjective. Because suffering involves an individual person’s experience, Kon deduces that “any quantification of an individual’s suffering that is not based on the personal report of that individual is unreliable.” [16^p.457] Thus, using Kon’s definition, any second-hand claim of patient suffering (which constitutes much of pediatric suffering) is contestable and therefore, may not belong in a decision-making calculus.

Two other bioethicists also see suffering as relative from person to person, claiming that the way “one might quantify suffering may vary from individual to individual.” [[17^p.43] One clinical article further asserts that because suffering is relative from individual to individual, it is a subjective experience that cannot be measured. [18]Bioethicists Lindemann and Verkerk apply a slight variance to the subjective characteristic of suffering. They claim suffering is a relative concept not just from person to person but rather in relation to pain. To clarify this, they state: “It has become a bioethical commonplace to distinguish suffering, which is subjective, from pain, which is not.” [6^p.47] We witnessed Lindemann and Verkerk’s claims to some degree in our analysis, where suffering language is set in contrast to pain.

Thus, while no concrete definitions of suffering were found in our data, the few common characteristics of suffering explicated by the authors above include: the subjectivity of suffering, the phenomenological experience of suffering, and the immeasurability of suffering. Both bioethicists and clinicians identify suffering as a subjective or relative term, varying from individual to individual. Implied in the understanding of subjectivity and expressed clearly by Kon is the idea that suffering is an experience of physiologic processes such as pain, but it is distinct from pain itself. This understanding reflects Cassell’s characterization of suffering as something that goes beyond mere physical pain and may threaten one’s personhood and sense of wholeness. [2] Because suffering is subjective and experiential, Lindemann and Verkerk (as well as Ottenhoff et al.) call out the inability to quantify suffering. [6,18]

However, this subjective, phenomenological, and immeasurable nature of suffering raises questions about who can suffer. It has been suggested that neonates and infants could not suffer because they lacked the capacity to construct an integrated sense of self. [19] Yet over one-third of the articles analyzed in this content analysis (37%) address neonatal suffering, and both clinicians and parents claim that children do indeed suffer. [7, 20] Thus, the theoretical characterizations of suffering that we identified may not comport with understandings of suffering in a neonatal context, and more research is needed in order to 1) clarify what parents and providers mean when they talk about neonatal suffering, and 2) develop more robust theoretical understandings that can account for the experience of neonatal suffering.

Ambiguity of the Concept

Although we discovered several consistently expressed characteristics of suffering, no concrete definitions were provided, and thus what we found in most articles was an overwhelming amount of ambiguity in both the use and possible understandings of suffering. In almost three-quarters of the excerpts, both the cause and the experience of suffering were ambiguous. Many of these ambiguous uses of the term “suffering” were related to death or dying or were characterized by the emotional and physical distress that may be associated with dying. However, since parents and providers are often the ones assessing a dying child’s suffering, these second-hand accounts and descriptions may not accurately portray the child’s experience of dying. Further, parents and providers may mistakenly associate their own emotional suffering with the patient’s suffering, and thus may make decisions about treatment based on their own assumptions about the patient’s quality of life, rather than the patient’s actual experience. Thus further research is needed in order to specify these ambiguous uses of the term, classify them according to their rhetorical use, and assess the role that a particular child’s suffering should play in decision-making.

Decision-Making

Whether and how claims of patient suffering should enter into specific medical decisions for children is a question that has received some, but not much, attention in the bioethics literature. For example, the Groningen Protocol, a set of Dutch guidelines allowing physicians to actively terminate the life of a newborn that is not terminally ill, gives a special place to suffering in the decision-making process: among the protocol’s requirements is the presence of “hopeless and unbearable suffering” in the child, as well as confirmation of that suffering by at least two physicians. [9] This protocol has been widely criticized as giving too much weight to second-hand assessments of current and future patient suffering, assessments that are necessarily merely speculation about the first-hand experience of the newborn. [21]

Alternatively, the Baby Doe Regulations, an amendment to the U.S. Child Abuse Law passed in 1984 after the deaths of two disabled newborns, explicitly rejects estimations of a child’s quality of life or suffering as acceptable reason to limit life-prolonging treatment. [22]

This content analysis reveals that despite the fact that “suffering” is rarely defined or characterized, the concept is still regularly used as a factor in medical decisions for children. And, notably, the concept is three times more likely to be invoked to support a life-ending decision (withholding or withdrawing life-prolonging interventions or active euthanasia, in the case of the Netherlands) than it is to support a life-extending decision (32% vs. 10%, respectively). This raises several issues related to pediatric decision-making. First is the likelihood that if “suffering” is referred to broadly and without a definition, then the specific concerns of the speaker—whether primarily physical symptoms (air hunger, pain, agitation) or the existential or emotional experience of the patient (hopelessness, loneliness) or something else entirely—are left without response.   In many cases, there are ways to respond to and mitigate these specific concerns, aside from the extreme option of making a life-ending decision. However, when claims of suffering remain ambiguous, these concerns cannot be addressed or remain obscured under the broader decision to withhold or withdraw treatment.

A second concern is that invocations of the concept “suffering” may allow for the either intentional or unintentional “smuggling in” of subjective and untested notions of quality of life. When a speaker makes a second-hand claim that a patient is suffering, what he or she may actually be saying is, “I wouldn’t want to live that way,” or, “This situation is causing me distress,” which are not equivalent to the patient saying, on behalf of him or herself, “I am experiencing unbearable suffering and it’s not worth it to me to continue treatment.” Finally, we must acknowledge that “suffering” is a powerful word that may wield great rhetorical power, swaying decision-making toward “dramatic solutions” and inadvertently obscuring the realities of the patient’s actual experience in the process.

Strengths and Limitations of the Study

To our knowledge, this is the first qualitative content analysis of the term “suffering” in pediatric bioethics and clinical literature. It is important to characterize how suffering is currently being used in order to validly critique the concept and offer suggestions for ways to be more precise with the terminology. However, because the authors only reviewed select bioethics journals and clinical literature, it is possible that the concept of suffering is used more judiciously in other journals or by other authors. Also, due to the nature of qualitative research, it is possible that articles were inadvertently overlooked or left out of this comprehensive content analysis. However, the authors believe that substantive conclusions can still be drawn despite potential omissions.

Conclusions and Implications for Practice

As evidenced by this content analysis, the concept of “suffering” is often used uncritically in the pediatric bioethics and clinical literature to convey a negative experience of pain and distress, which leads to confusion about the authors’ goals of invoking the concept. Although several theoretical characterizations of suffering were identified, these did not neatly apply to the neonatal experience, even though parents and providers make claims about neonate and infant suffering. Thus, further work is needed in order to better understand what parents and providers mean when they talk about infant suffering and to develop more robust theoretical understandings that can account for the experience of neonatal suffering.

Last, this content analysis shows that authors draw on the ambiguous yet negatively characterized experience of “suffering” in order to sway parental and provider decision-making without giving a substantive account of what suffering is and why it should lead to a particular kind of decision. We conclude that authors—both bioethicists and clinicians—need to be more precise with their language about what is distressing the patient, family, or care team in order to better address and alleviate that specific experience of distress.