Distress among cancer patients: some considerations in identification and treatment

Abstract

This article discusses the merits of screening for distress in the context of a chronic illness, such as cancer, with a particular focus on resource-constrained health care systems such as those in low-and-middle income countries. Despite calls for distress to be considered a vital sign, like pain it is not objectively verifiable as it relies solely on the person’s subjective appraisal. Accordingly, the Distress Thermometer has limited validity considering its concordance with the Hospital Anxiety and Depression Scale, which itself has limitations in terms of its psychometric properties. Indeed, an elevated score on a self-report measure does not indicate caseness for a mental health condition. Distress is often self-limiting and transient, whereas common mental disorders require evidence-informed treatment. In the context of scarce resources as is the case in low-and-middle income countries, efforts should instead be directed at identifying common mental disorders among persons living with cancer and others who have serious health threats. Such an approach will increase the likelihood of resources being directed at those who are most likely to benefit from psychological interventions. Where persons living with cancer indicate the need for psychosocial services, ways to manage distress include problem-solving therapy, motivational interviewing, and mindfulness-based stress reduction.

High rates of prostate, breast, cervical, and other cancers exist in many sub-Saharan African countries and are expected to more than double over the next two decades (CANSA, 2021; Rebbeck, 2020). A cancer diagnosis is often accompanied by the emotional response of distress as the diagnosis signals a threat to the person’s health, well-being, and lifestyle. A cancer diagnosis introduces the likelihood of receiving chronic or acute treatment, a change in lifestyle, including diet and exercise regimens, the possibility of pain and discomfort from either the condition or its treatment or both, and the possibility of disability or death that may result from the condition. As a result, emotional distress is a normative response to receipt of a cancer diagnosis.

Several writers have argued that distress should be regarded as a vital sign that medical practitioners should address (Bultz et al., 2015; Carlson et al., 2012; Holland & Bultz, 2007; Howell & Olsen, 2011). Vital signs refer to a cluster of indicators about the status of the body’s life-sustaining functions (Lockwood et al., 2004). They are used to assess the health of a person and provide the attending clinician with data to arrive at diagnostic impressions and identify possible diseases. The four usual vital signs that are assessed during a medical examination are body temperature, blood pressure, heart rate, and respiratory rate (Hornstein, 2011). Other measures often included are weight, height, and waist circumference. There are normal ranges within which most people’s readings fall and scores that fall outside of these ranges are referred to, for example, as hypothermia in the case of temperature and hypertension in the case of blood pressure (Hornstein, 2011).

In the late 1990s, the Joint Commission of Healthcare Organisations in the United States indicated the need for screening for pain as a fifth vital sign in medical examinations (Berry et al., 2001). This suggestion was due to the view that pain was invisible, under-recognised, and thus under-treated. The development was heralded as a paradigm shift in the monitoring and management of chronic pain. After pain was included as a vital sign, emotional distress was then proposed as a sixth indicator of health, in keeping with a biopsychosocial view of chronic illness (Bultz et al., 2015). Distress is a non-stigmatising term that refers to the psychological, social, nonphysical aspects of well-being. Examples of distressing emotions include normal fears, worry, and sadness. The extent to which distress leads to a mental health condition such as major depressive, generalised anxiety, or panic disorder is a complex matter. Nonetheless, a study of service users at 55 cancer centres in North America found that 46% had elevated distress (Carlson et al., 2019). Among Chinese persons with cancer, 20% reported elevated psychological distress (Rao et al., 2019). Among individuals attending a cancer rehabilitation facility in Singapore, approximately 30% reported distress in the elevated range (Kuo, 2022). Among South African persons with cancer, more than 33% of study participants had high scores on the Center for Epidemiological Studies – Depression Scale (CESD) and the Hopkins Symptom Checklist, which correlated with perceived social support and body change stress (Kagee et al., 2018).

In the field of oncology, clinicians have been exhorted to include distress as a vital sign (LeBlanc & Kamal, 2017). The International Psycho-Oncology Society (2014) has advocated for the integration of distress screening into routine care. In most low-and-middle income countries (LMICs), much of the care in public health systems is the responsibility of nurses. Thus, a brief screener for psychosocial distress can help streamline the process by quickly identifying service users in need of referral.

In 2003, the National Comprehensive Cancer Network (NCCN) Distress Panel recommended that distress can be identified and treated at each point of the disease trajectory in cancer care; that screening should be applied to all service users at the initial visit and at appropriate intervals such as remission, recurrence, disease progression; and that distress should be treated in accordance with clinical practice guidelines (Holland & Bultz, 2007).

The various instruments that have been marshalled in the service of such screening include the Profile of Mood States (POMS; Kim & Smith, 2017), the Hospital Anxiety and Depression Scale (HADS; Annunziata et al., 2020), and the Distress Thermometer (DT; Ownby, 2019). The move to recognise distress as the sixth vital sign has garnered considerable support and it is commonly regarded that such an approach will improve treatment outcomes for oncology service users around the world (Bultz et al., 2015). Relatedly, health-related quality of life (HQOL) is an accepted multi-dimensional approach to understanding how the illness affects the patient’s well-being and functioning across a number of domains and can be used in the co-ordination and integration of care. A full discussion of HQOL is outside of the scope of this article (for further reading please see Bakas et al., 2012; The EuroQol Group, 1990).

Objective and subjective vital signs

There is an important difference between the traditional vital signs such as temperature, blood pressure, respiration rate, and heart rate, on one hand, and the newer ones of pain and distress. The former are objectively verifiable and require no cognitive reflection on the part of the service user. For example, data provided by thermometer, blood pressure cuff, or stethoscope reading are objectively correct and are not vulnerable to being challenged by the person being examined, unless these instruments themselves are flawed. If a health service user reports a fever, for example, their temperature can be taken, and if it is no higher than 37.2°, then it can be said that that they do not have a fever. Similarly, if someone reports a rapid heartbeat but objectively their heart beats at 80 beats per minute, then no tachycardia is present.

While signs can be objectively observed by another person, symptoms refer to subjective experiences (Hornstein, 2011). By this account, reports of headache, itching, and vertigo are symptoms as they cannot be objectively corroborated. Phenomena such as cough, skin rash, or inner ear abnormalities, on the other hand, are signs as these can be objectively observed. As Hornstein observes, ‘if someone other than the patient can’t see, hear, palpate, percuss, or measure it, it’s a symptom; anything that can be perceived by someone else is a sign’ (Hornstein, 2011). Of course, when someone reports headache, vertigo, itching, or pain, this does not mean it is not real. It may indeed be a real experience for that person. The point is that these are not signs and therefore cannot be vital signs in the sense of the four existing vital signs.

Pain and distress require a person to reflect on their subjective experience, appraise a measurement instrument such as the DT, and rate their subjective experience in accordance with the scale (Coyne, 2014). There are no objective ways for the attending clinician to verify or corroborate the assessment that the person makes. Thus, while pain and tissue damage may be expected to concur, there may also be a mismatch between the person’s report of pain and actual tissue damage, in either direction. For example, a person may report no pain while tissue damage is apparent or report considerable pain in the absence of tissue damage. Indeed, pain behaviours exist, such as grimacing or holding the body part where the pain is reported to be situated. Yet, the absence of such behaviours does not mean no pain, and conversely someone may engage in such behaviours even in the absence of pain. For these reasons it has been argued that pain cannot be regarded as a vital sign (Coyne, 2014). A rating on a measure of pain, which has been advocated (Cutillo et al., 2017), is not an objective measurement in the same way that other vital signs are. The same argument applies to distress.

The Distress Thermometer

The NCCN created the DT in 1997 to guide the management of distress among persons living with cancer (Cutillo et al., 2017). The DT, pictorially presented as a thermometer, is a 1-item screening instrument with response options that range from 0 to 10. Zero refers to the absence of distress and 10 refers to extreme distress. The respondent is then asked to indicate the presence or absence of problems on an accompanying problem list consisting of 40 items that commonly occur among persons with cancer. These problems include (1) practical issues such as childcare, finances, housing, transport, and occupation; (2) family problems such as childcare or relationship difficulties; (3) emotional problems such as depression, anxiety, sadness, worry, and loss of interest activities that were formerly pleasurable; (4) spiritual concerns; and (5) a range of physical problems that include appearance, bathing, dressing, breathing, and constipation. It has been suggested that asking service users to rate their distress should be commonplace for clinicians working in oncology, and that distress then should become a sixth vital sign alongside pulse rate, respiratory rate, temperature, blood pressure, and pain (Holland & Bultz, 2007).

Similar to pain, distress in the context of physical illness is vulnerable to the criticism that it has limited concordance with illness severity. In as much as there have been calls for routine screening for distress in oncological care, it has also been argued that such calls should be tempered by existing data. A systematic review of 14 eligible intervention studies to ameliorate distress found that pharmacological and psychotherapeutic care reduced distress with small or moderate effects (Meijer et al., 2013). In addition, a high score on a distress measure is not necessarily an indication for psychosocial services beyond what people may currently be receiving (Meijer et al., 2013). Thus, an elevated score on a screener for distress may not indicate interest in or uptake of psychosocial services. Conversely, persons who score below a cut-off for distress may not be referred to services even though they may indicate such an interest. In a study of 970 persons living with cancer, 423 scored in the distressed range on the Hopkins Symptom Checklist – 25, of whom 215 expressed a desire for psychosocial services (van Scheppingen et al., 2014). Of the original sample, 4% agreed to participate in an intervention study to ameliorate distress, leading the authors to conclude that 27 individuals needed to be screened to recruit one person into the study and that up to 17 hours were required to recruit each patient (van Scheppingen et al., 2014). Although some studies have suggested that the DT is useful for assessment of patients (e.g., Klingenstein et al., 2020), a systematic review found little evidence to support its validity, as well as uncertainty about what the tool indeed measures (Stewart-Knight et al., 2012). Stewart-Knight et al. suggest that future research should compare the DT with qualitative interviews in which service users are asked how distressed they feel and to identify concerns that contribute to their distress.

There is an assumption that an elevated score on the DT means that the person requires counselling and should thus be referred for mental health services. However, among 277 Dutch persons with cancer who were assessed by the DT, the rate of interest in referral for counselling was low. Of the total sample, over 70% did not wish to be referred, and of those who scored above 5 on the DT, 14% stated they wished to be referred, 29% indicated they ‘maybe’ wanted a referral, and 57% indicated they did not wish to be referred (Tuinman et al., 2008). It is thus unclear that an elevated score on the DT is concordant with a wish to receive counselling or whether indeed counselling would be beneficial for such persons. It is entirely possible that mental health services were not required by participants in this study or was not what they preferred. In the face of these data, Coyne (2014) has argued that making distress a vital sign conveys a false objectivity and insufficient validity that does not yield positive psychosocial outcomes. The evidence for the benefit of screening persons with cancer for distress is thus minimal.

What does an elevated score on a distress measure mean?

Correct identification of common mental disorders can facilitate treatment referral, potentially resulting in the amelioration of symptoms of the disorder, and consequently in improved quality of health. Diagnostic interviews such as the Structured Clinical Interview for the DSM (SCID), the Composite International Diagnostic Interview (Wittchen, 1994), and the Mini-International Neuropsychiatric Interview (Sheehan et al., 1998) are usually considered the gold standard to establish caseness for a mental disorder. However, a diagnostic interview requires resources such as a trained assessor, usually a psychologist or psychiatrist, a private office where the clinician can meet the person, and sufficient time to conduct the assessment. Furthermore, the clinician then needs to be able to offer persons living with cancer an evidence-informed intervention to ameliorate symptoms of their disorder which in many resource-constrained contexts can be difficult. To save time and resources, screening for mental disorders has been considered a useful alternative. Screening instruments are easy to administer as service users may respond to a questionnaire in a clinic waiting room. An administrative assistant may then calculate the total score and determine the severity of the symptoms. However, an elevated score on a self-report measure does not necessarily indicate caseness for a mental health condition and indeed self-report measures usually yield higher prevalence rates than diagnostic interviews (Coyne, et al., 2000). Important concepts in determining the concordance between a positive screen and true caseness are sensitivity (the proportion of people correctly identified with the disorder), specificity (the proportion of people correctly identified without the disorder), positive predictive value (the probability that those who score above the optimal cutpoint on a screening instrument actually have the condition), and negative predictive value (the probability that those who score below the optimal cutpoint on a screening instrument indeed do not have the condition) (Zou et al., 2011). A large number of false positives means that many people will be incorrectly identified as having a mental health condition and inappropriately referred for treatment. A large number of false negatives means that people who actually require treatment will not be identified as such and may thus go untreated. Among South African samples, the concordance between elevated scores on self-report measures and caseness as determined by a diagnostic interview has been poor. For example, among 500 HIV test-seekers, 3.4% met the diagnostic criteria for generalised anxiety disorder as assessed by the SCID. A cutpoint of 22 on the Beck Anxiety Inventory found 82% sensitivity and 80% specificity, with positive and negative predictive values of 13% and 99%, respectively (Saal et al., 2019). These findings indicate a 13.0% probability that those who scored 22 and above would indeed have generalised anxiety and a 99.2% probability that those who scored below this cutpoint would not. Among the same sample, 14.4% met the criteria for major depression against which the Beck Depression Inventory (BDI) was 67% sensitive and 67% specific. The BDI had a 25% probability that those who scored 20 and above would meet the criteria for major depression and a 92% probability that those who scored below this cutpoint would not (Saal et al., 2018). Clearly, these measures were superior in identifying non-cases than cases of these mental disorders. These data as well as others (Coyne, Thompson, et al., 2000; Kagee et al., 2013) indicate that self-report screening instruments are more likely to accurately identify non-cases but mis-identify cases of persons with a mental health condition.

When such an analysis is applied to the DT, further problems arise. A meta-analysis that included 42 studies with a total of 14,808 people with cancer showed that the DT displayed a good balance between pooled sensitivity and pooled specificity of 0.79 (95% CI = [0.76, 0.81]) and 0.80 (95% CI = [0.78, 0.82]), respectively, against the HADS as a reference standard, using a cutpoint of 4 (Ma et al., 2014). By this logic, there is a case to be made for using the DT to determine which individuals may benefit from a referral to manage their distress. However, such reasoning rests on the assumption that the HADS, as the reference standard, is indeed a valid and reliable measure of distress.

According to Coyne and van Sonderen (2012), there are several reasons to question this assumption. First, it has been demonstrated that the factor structure of the HADS lacks consistency across samples. A systematic review of 50 reviewed studies showed that 25 of the studies found a two-factor structure, 5 studies found one factor, 17 studies found a three-factor structure, and 2 studies found a four-factor structure. One study found both a two- and three-factor structure. The number of factors found also depended on whether exploratory factor analysis, confirmatory factor analysis, or item response theory was used in the analysis (Cosco et al., 2012). An unstable factor structure thus limits assumptions about the validity of the HADS. Second is the problem of response options. The HADS has 15 items, 6 of which are reverse scored so that a low score indicates low distress and a high score indicates high distress. For example, a distressing symptom such as feel-ing tense or wound up is scored 0 as not at all and 3 for most of the time. The other 9 items are scored so that a high score indicates greater distress, and a low score indicates less distress. In addition, the response key for the item ‘I can sit at ease and feel relaxed’ has different anchors from the rest of the items (1 corresponds to definitely; 4 corresponds to not at all, versus 1 corresponds to not at all; 4 corresponds to most of the time). Presumably, these differences were deliberately intended by the test constructors to discourage a uniform response style. It has been pointed out that such discrepancies in response options can disorient the respondent, to the extent that they would not be alert to the changes in direction of the scoring of these items as they respond to the items (Coyne & van Sonderen, 2012).

A third problem with the HADS raised by Coyne and van Sonderen is the inclusion of the item ‘I get a sort of frightened feeling like “butterflies” in the stomach’. This item may be poorly comprehended in cultures where such an expression is not commonly used. A fourth concern expressed by Coyne and van Sonderen is the emphasis on items related to anhedonia. These are as follows: ‘I still enjoy the things I used to enjoy’, ‘I have lost interest in my appearance’, ‘I look forward with enjoyment to things’, and ‘I can enjoy a good book or radio or TV program’. Depressed mood and sadness are only represented by the items ‘I feel cheerful’ (scored negatively) and are thus underrepresented. Finally, concern has been expressed about the varying cutpoints associated with the HADS that validation studies have yielded (Coyne & van Sonderen, 2012).

These concerns bring into focus the limitations inherent in any attempt to regard the HADS as a reference standard on which to base the DT. Following this reasoning, it is fair to question what a high score on the DT may actually mean (Stewart-Knight et al., 2012). Distress is not validated except by correlating scores on the DT with other measures of distress which, as is evidenced by the HADS, is susceptible to misinterpretation.

Whom to refer for treatment?

The number of mental health professionals in LMICs is insufficient to serve their populations (Lund et al., 2010), thus there is a need for accurate case identification so that interventions can be provided that yield optimal benefits. If the purpose of screening is to identify a common mental disorder, it may be useful for public health services to adopt a two-stage approach to screening (Eack et al., 2008). Thus, only those persons who screen above a commonly used cutpoint on a valid self-report measure such as the CESD would undergo a diagnostic interview to determine the presence of a common mental disorder. In such a two-stage approach, persons who are true cases may then be referred for treatment where these are available, such as anti-depressive medicines or psychological counselling. People who are non-cases but report elevated symptoms of distress can similarly be referred for appropriate support.

By all accounts, the various aspects of the cancer experience, diagnosis, remission, recurrence, disease progression can be extremely distressing. However, as argued above, it may be inappropriate to assume that elevated distress equals a mental health condition. Indeed, for most people, being confronted with one’s mortality, impending physical decline, pain and physical disability, all of which may accompany the cancer experience, can be highly distressing. In this sense, distress is a natural response to a health condition.

The question then is how to assist people living with cancer who experience elevated distress but do not have a mental disorder. Perhaps distress in and of itself is not an indicator of the need for mental health services. Instead, when a person living with cancer indicates a need for such services or is offered a referral and accepts it, such a referral should be made. Antidepressant medication is improperly prescribed to treat distress. In the absence of sufficient mental health professionals in the public health systems of LMICs, task-shifting of functions such as case identification to less specialised personnel by means of a two-phase screening and diagnosis process holds promise.

Various interventions such a problem-solving therapy, motivational interviewing (MI), and mindfulness-based stress reduction can be effective to assist persons with cancer to manage their distress. The goals of problem-solving therapy (Nezu et al., 2019) include helping the person living with cancer adapt to their experience and other problems in living, and to learn skills such as emotional regulation and problem-solving in a planful manner. Problem-solving skills can assist persons with cancer to develop effective ways of coping with their problems. Such problems are framed as life situations that necessitate the person to develop a response that is adaptive so that negative consequences can be prevented. Situations that create problems include novelty (such a diagnosis with cancer), ambiguity (such as uncertainty about treatment options or paying for health care), unpredictability (such as uncertainty about the trajectory of illness), conflicting goals (such as weighing up treatment benefits with side effects), performance skills deficits (such as communicating with family or health care providers), and lack of resources (such as insufficient funds to pay for treatment, transport to the hospital, or to pay for medication) (Nezu et al., 2019). Distress, where it occurs, can be managed by regarding problems as challenges rather than threats, by assuming that problems are solvable, by having self-efficacy about one’s ability to cope with the problem, and accepting distress as part of the problem-solving process that can help one cope when the problem is especially stressful.

Another intervention is MI. MI facilitates and develops the person’s motivation for behaviour change by means of open-ended discussions (Miller & Rollnick, 2012). In terms of facilitating behaviour change, including behaviours such as having a healthy diet, engaging in physical activity, and stopping smoking, MI has shown considerable promise. A systematic review has found that MI can be useful in effecting changes in behaviour related to diet, exercise, and smoking (Bennett et al., 2007).

Finally, mindfulness-based stress reduction, a mental training practice that instructs the person to slow down racing thoughts, relinquish thoughts of negativity, and engage in mental calming activities (Kristeller, 2007), has received much attention from researchers. A meta-analysis of mindfulness interventions for pain related to cancer indicated improved pain severity, symptoms of anxiety, depression, and quality of life (Ngamkham et al., 2019). A major concern is that most mindfulness studies have been conducted in the global north, with very few involving participants in LMICs.

Concluding comments

Part of the problem with distress is that it is not a technical term and is thus open to many uses in many contexts. If distress refers to the unpleasant feelings or emotions that people have when they feel overwhelmed, then such experiences would apply to almost all people in the world. Thus, distress is likely to follow receipt of a diagnosis of cancer, news that the person’s condition is worsening, physical pain, or the side effects of medication. Distress therefore is a natural response to such experiences. At least some of the time distress ameliorates spontaneously or may be the result of personal challenges that are not related to their physical illness (Walters et al., 2008). For those individuals who indicate the need for a referral for mental health services or, when offered such a referral accept it, such services are certainly indicated.

While efforts to ameliorate distress are laudable, it is unclear that distress should be considered a vital sign as it is not objectively verifiable. Distress in and of itself for many people without a mental health condition may be transient and self-limiting. Also, as has been argued above, there is a lack of concordance between elevated scores on self-reported measures of distress and caseness for a mental health condition. In the context of scarce resources in LMICs, efforts should instead be directed at identifying common disorders among persons living with cancer and other serious health threats. Such an approach will increase the likelihood of resources being directed at those who are most likely to benefit from psychological interventions.