9.3 Definition and assessment of chronic pain in advanced disease

The International Association for the Study of Pain (IASP, 2011) defines pain as an ‘unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage’. Chronic pain is defined as ‘pain which persists beyond the usual course of healing or is associated with chronic pathological illness which causes continuous pain or pain which recurs at intervals for months of years’ (Bonica, 1990). Pain is always subjective, corresponding to what the patient describes. This point, which is emphasized by the IASP, underscores the observation that correlation between observed behaviour and patient reports of pain may be poor as individual response to pain varies considerably. All pain-related behaviour, including verbal report, is influenced by its perceived meaning, and an individual’s culture, ethnicity, and mood (Cleeland et al., 1996; Twycross, 1997). Verbal report is the simplest to interpret clinically, but other behaviours are relevant to a broader understanding of the pain. In those patients unable to communicate effectively, such as infants or those who are cognitively impaired, changes in behaviour may be the key clinical finding.

The concept of ‘total pain’ acknowledges the physical, psychological, social, and spiritual influence on a patient’s perception of pain, and the multidimensional effects it has on a person’s life (Saunders, 1967; Portenoy and Lesage, 1999; Zaza and Baine, 2002; Foley, 2004; Ferreira et al., 2008; Portenoy, 2011). Without attention to all these areas, pain relief is unlikely to be optimal (Saunders, 1967).

Defining the likely pathophysiology of the pain is an essential part of the broader understanding which may help enable appropriate treatment. Nociceptive pain is defined as ‘pain that arises from actual or threatened damage to non-neural tissue and is due to the activation of nociceptors’ (IASP, 2011). In contrast, neuropathic pain is defined by the Neuropathic Pain Special Interest Group (NeuPSIG) of the IASP as ‘pain caused by a lesion or disease of the somatosensory nervous system’. It is often then further classified as peripheral or central and by anatomical site and disease. Both these terms are descriptions of pain, not diagnoses.

Breakthrough pain is the term most widely used to describe variations in quality, intensity, and timing of pain on a background of stable pain control (Haugen et al., 2010). It can be predictable, unpredictable, spontaneous, or evoked with the same quality or a different quality to the baseline pain (Portenoy et al., 2006; Hagen et al., 2008). Patients with uncontrolled breakthrough pain are more likely to have a poorer quality of life and be depressed (Portenoy et al., 1999).

Pain occurs in more than 50% of patients with advanced disease, and although the ‘analgesic ladder’ approach to treatment promulgated by the World Health Organization is both accepted as the mainstay therapy and widely disseminated, pain is undertreated (Solano et al., 2006; van den Beuken-van Everdingen et al., 2007a; DeAndrea et al., 2008; Breivik et al., 2009). Unrelieved pain interferes with daily functioning and quality of life. It may have multiple negative effects, interfering with sleep, activity, and interaction with others, and causing psychological and existential distress (Sheinfeld et al., 2012).

Most of the empirical evidence about pain in advanced disease has been related to patients with cancer, but a review by Solano et al. (2006) shows that pain is a common feature of advanced disease in all patients regardless of the underlying pathology. Prevalence of pain in cancer patients was 35–96%, in those with heart disease, 41–77%, AIDS, 63–80%, chronic obstructive pulmonary disease (COPD), 34–77%, and renal disease, 47–50% (Solano et al., 2006). However, what is not clear is whether this pain is a result of the underlying advanced disease or associated co-morbidities. For example, a patient with COPD may have severe back pain but this may be due to osteoporotic fractures as a result of long-term steroid use.

In cancer patients, pain is more common as the disease progresses. A systematic review of 52 articles focusing on patients with cancer found that the prevalence was 64–74% in those with advanced disease, 59% of those on anti-cancer treatment, and 33% in those that had been cured of their cancer (Hearn and Higginson, 2003; van den Beuken-van Everdingen et al., 2007b). These findings are similar to those in a recent European Pain in Cancer survey of over 5000 patients in 11 European countries, which showed an overall pain prevalence of 72% of patients with cancer in the community (Breivik et al., 2009). Prevalence rates of pain associated with different types of cancer vary in the literature, but overall rates all remain over 50%. Breivik et al. (2009) found that the highest prevalence rates, of over 85%, were associated with cancers of the pancreas, bone, brain, lymphoma, lung, and head and neck.

Although confirming that pain is an extremely common symptom among diverse populations with serious illness, the specific prevalence rates must be interpreted with caution. Pain is heterogeneous and prevalence rates are affected by the setting of the patient, such as a specialist pain or a general respiratory clinic.

Pain prevalence may be related to aetiology. Pain may be due to a direct effect of underlying advanced disease or its treatment; an indirect effect of the disease, such as pressure sores from immobility; or a disorder unrelated to the primary disease, such as osteoarthritis. One survey observed that 76% of patients with cancer have pain directly related to the cancer, 11% have pain related to cancer treatment, 5% have pain due to an indirect effect, and 8% have pain due to unrelated conditions (Grond et al., 1996).

A study of patients with chronic cancer pain observed that somatic pains were more common (71%) than either neuropathic (39%) or visceral (34%) pains (Caraceni and Portenoy, 1999). Many patients experience both nociceptive and neuropathic pain, and the origin of pain must be considered when interpreting prevalence rates. In neuropathic cancer pain, for example, although the majority of pains are caused by cancer (64%), a significantly higher proportion is caused by cancer treatment (20%) in comparison with all cancer patients (Bennett et al., 2012).

The pathophysiology of pain affects its impact on patients and treatment. Patients with neuropathic pain have been shown to have greater pain intensity, a worse quality of life, and a greater negative impact on their daily living than patients with nociceptive pain (Torrance et al., 2006; Smith et al., 2007; Bouhassira et al., 2008). Similarly, those with neuropathic cancer pain have been shown to have a worse quality of life, poorer performance status, and a need for both higher opioid doses and a longer time to achieve pain control than those with nociceptive pain (Fainsinger et al., 2010; Rayment et al., 2013).

Further clouding interpretation of prevalence figures is the observation that patients often report more than one pain (Twycross and Fairfield, 1982). For example, a systematic review found that, on average, patients with cancer have two pains, 20% of which are neuropathic in origin; these patients with multiple types of pain contribute to a 40% prevalence rate for neuropathic pain overall (Bennett et al., 2012).

Finally, pain prevalence must be interpreted in light of severity and other factors. Pain severity varies, but as a minimum, more than one-third of patients with cancer pain grade their pain as either moderate or severe (van den Beuken-van Everdingen et al., 2007b; deAndrea et al., 2008). In hospitalized patients with cancer, patients reported a pain score of 3.7 on a 0–10 numerical rating scale for average pain; two-thirds of patients reported that their worst pain was higher than 5 (Klepstad et al., 2002; Yates et al., 2002). In the community, 93% of those patients who experienced cancer pain at least several times a month rated the severity as moderate to severe, 44% described it as severe, and 3% regarded it as the worst pain imaginable (Breivik et al., 2009).

Like pain itself, under-treatment has many influencing factors. Discrepancies between patient and physician rating of pain severity are one such predictor (deAndrea et al., 2008). Under-treatment also is more likely when patients have a better performance status or are in the early stages of their diseases, findings that may suggest a reluctance to believe that those with early disease can have severe pain or an increased likelihood that those with advanced disease are managed by specialists (deAndrea et al., 2008).

It is not clear if older age has an effect on pain management. Evidence from the cancer population suggests that there are no differences in pain intensity but older people may require less or more analgesia than the younger population, perhaps due to altered physiology (Vigano et al., 1998; Mercadante and Arcuri, 2007; Hall et al., 2003; Mercadante et al., 2006; van den Beuken-van Everdingen et al., 2007b; Bennett et al., 2009). Older patients also do not have more adverse effects or more need for opioid switching (Mercadante et al., 2006). However, they may have different attitudes towards opioids and therefore may be more reluctant to use these drugs (Yeager et al., 1997; Closs et al., 2009). Gender does not alter pain prevalence rates, but education level and ethnicity may (van den Beuken-van Everdingen et al., 2007b; deAndrea et al., 2008).

Assessing a patient’s pain should be a continual process supported by communication between patient, carer, and clinician. The aims are to characterize the complaint, correctly identify the underlying pathophysiology, and determine the effect of pain on the patient’s life. All of this information informs the management plan. Without this focus on a comprehensive assessment, pain treatment becomes less targeted and less effective.

A history and clinical examination are vital for pain assessment and radiological and laboratory tests may be indicated. Inadequate pain control is most often due to a poor history and examination (Grossman et al., 1992; Von Roenn et al., 1993). Establishing the effect of pain on the person’s quality of life and activities of daily living, recording previous effects of analgesic treatments, and exploring the patient’s goals and expectations are crucial in being able to effectively treat all aspects of pain’s impact on the person. Creating a trusting, open relationship encourages communication and allows patients to be involved in the management of their pain. Concerns generated by the experience of fear, such as worry about cancer progression or about the risks of addiction or tolerance to opioids, may affect the information given to the clinician.

The acronym SOCRATES is useful in prompting systematic assessment of pain characteristics: Site, Onset, Character, Radiation, Associated factors, Timing, Exacerbating/relieving factors, and Severity (Box 9.3.1). As previously mentioned, patients often report more than one pain and it is important to ascertain a pain history for each pain a patient reports, as its cause and therefore treatment may vary. Answers to these questions provide clues as to the likely underlying cause. For example, a patient who describes unpredictable severe shooting and stabbing pains in both toes and feet, with no radiation, may well be describing a peripheral neuropathic pain perhaps related to previous chemotherapy. In contrast, a patient describing severe shooting and stabbing pains radiating down both legs to their feet, which is worse on movement, may be describing impending cauda equina/spinal cord compression. Management of these two syndromes is substantially different and requires physicians’ knowledge of the underlying disease and likely prognosis.

The pain assessment also must clarify the history of the underlying disease process, including previous treatments and current status. This information informs both treatment selection and prognostication. Similarly, a history of other symptoms is needed to both predict the impact of pain therapy and assist in the development of a broader plan of care for symptom control.

For those patients who are cognitively impaired it can be harder to ascertain symptoms and there must be increased reliance on non-verbal cues, as well as proxy reports from families and other carers who can highlight abnormal behaviours. A recent Norwegian cluster randomized controlled trial of patients in nursing homes with moderate to severe Alzheimer’s dementia and significant behavioural disturbances showed significant benefit in agitation in patients who received analgesia, the vast majority receiving just regular paracetamol (Husebo et al., 2011).

English, which can express the thoughts of Hamlet and the tragedy of Lear, has no words for the shiver and the headache. . . . The merest schoolgirl, when she falls in love, has Shakespeare or Keats to speak her mind for her; but let a sufferer try to describe a pain in his head to a doctor and language at once runs dry. (Virginia Woolf, On Being Ill)

Although patients often find it difficult to articulate the character of their pains, it is important to elicit descriptors as a step in diagnosing the underlying pathophysiology. Nociceptive pain is divided into somatic pain, which arises from injury to the soft tissues and bone, and visceral pain, which arises from injury to internal organs. Somatic pain is usually well localized and described as aching, sharp, or throbbing. In contrast, visceral pain due to obstruction of a hollow viscus is poorly localized and can be described as gnawing or cramping; injury to other tissues, such deep fascia, or organs such as the heart, may be described as pressure or a deep pain. Visceral pain is commonly referred, such as shoulder tip pain from diaphragmatic irritation. Patients often find visceral pain harder to describe than somatic pain (Bennett et al., 2005; Osta and Bruera, 2008).

Neuropathic pain is often described as numb or burning with sharp, shooting pains on movement. Patients may describe spontaneous pains in the absence of any stimuli or evoked pains such as allodynia (painful response to a non-painful stimulus), hyperalgesia (increased painful response to a painful stimulus), and hyperpathia (delayed and prolonged response to painful stimulus). Due to cancer’s natural history, neuropathic pain can evolve as masses increase in size; over time, many patients with tumour-related neuropathic pain experience the emergence of deep aching related to the underlying lesion. See Box 9.3.2 for an example of a patient’s description of neuropathic pain from head and neck cancer.

Neurological signs may precede pain and reflect the distribution and severity of nerve injury (Twycross, 1995). Sensory changes, weakness, and altered tendon reflexes can localize the site of injury and help clarify the aetiology of the pain.

However, distinguishing between neuropathic and nociceptive pain on the basis of pain descriptors alone is not easy. A study in non-cancer patients found that while patients with peripheral neuropathic pain reported more descriptors such as hot, cold, sensitive, itchy, and significantly less dull and deep than those with nociceptive pain, they still used nociceptive descriptors, although to a lesser extent (Dworkin et al., 2007). A study in lung cancer patients found that several words traditionally thought to describe neuropathic pain such as burning, shooting, tingling, and cold did not distinguish between the two pains (Wilkie et al., 2001).

As noted, individual pains also can have more than one mechanism. A patient describing pain from a vertebral metastasis that is compressing a nerve root is likely to describe features of both nociceptive and neuropathic pain. This would be classified as mixed pain or a more useful term to aid management may be whether the pain is more or less neuropathic in nature (Bennett et al., 2006).

Pain is associated with many features that may be clinically important or predict treatment response. For example, interference is the extent to which pain has affected aspects of a patient’s life. It is important to analyse because the extent to which pain interferes in various functional domains may enable better classification of its cause or predict those patients who are likely to have a poor response to management and a longer time to pain control. Although pain interference is strongly related to pain intensity at a group level (Serlin et al., 1995), there are significant inter-individual differences. In addition there are differences in reported levels of pain-related interference between patients with and without advanced medical illness: at any level of pain intensity, those without advanced illness report greater interference (Fayers et al., 2011).

Other associated features also are important. A recent longitudinal study by Knudsen et al. (2012) confirmed that pain intensity, incident pain, and younger patients predicted more complex pain and longer time to achieve pain control. In an international multicentre study, Fainsinger et al. (2010) showed that these three domains and psychological distress and neuropathic pain were associated with a longer time to pain control. Those with addictive behaviours, a history of drug or alcohol abuse, did not take longer to reach stable pain control but required higher opioid doses. Alcoholism is a major predictor of rapid escalation of opioids and opioid-related neurotoxicity (Bruera et al., 1995).

In addition, sleep disturbance, lack of social support, and concerns over finances and family all affect the impact pain has on the patient. Given this complexity, the treatment of pain with analgesics alone is unlikely to optimize outcomes. Working within a multiprofessional team to target and help each domain while expertly providing analgesics is more likely to result in better pain management for the patient.

It is important to assess the severity of pain, as the greater its severity the more impact it has on daily functioning and quality of life (Serlin et al., 1995). Various pain measurement tools can be applied to this task, ranging from simple unidimensional scales such as the visual analogue scale (VAS), to more complex pain questionnaires (see following sections). For some patients, facial scales may be more useful in establishing the severity of pain.

In the clinical setting, pain severity usually is measured by asking patients to rate their pains on a 0–10 scale or as mild, moderate, or severe. The response to this question may provide enough information to help initiate appropriate treatment, as well as its urgency. It may also help to establish the underlying cause of the pain; for example, severe pain on movement may support a diagnosis of a fracture, whereas mild pain may suggest osteoarthritis. Ongoing assessment of pain severity enables evaluation of interventions, and monitoring of variation over hours or days can help establish factors which worsen or ease the pain. All this information may be useful when there is a large team caring for the patient (Twycross, 1995).

Visual, categorical numerical, and categorical verbal rating scales are validated and used as tools to measure intensity of pain (Caraceni et al., 2002; Hjermstad et al., 2011) (see also Chapter 7.2). The VAS is a 10 cm line labelled at each end with the minimum or maximum extremes of what is being measured. In measuring pain, these anchors often are ‘no pain’ and ‘pain as bad as you can imagine’. Patients indicate their pain severity on the line. Categorical numerical scales ask the patient to rate their pain on an 11-point numeric scale (e.g. ‘0 to 10, where 0 is no pain and 10 is pain as bad as you can imagine’) or on a 5-point numeric scale (e.g., a 0–4 scale with the same anchors), and categorical verbal scales provide verbal descriptors, such as none, mild, moderate, or severe. They are reliable and can be used clinically and in research. Expert consensus recommended an 11-point numerical rating scale with ‘no pain’ and ‘pain as bad as you can imagine’ as anchors for the assessment of chronic pain including cancer pain (Dworkin et al., 2005; Kaasa et al., 2011).

The Brief Pain Inventory (BPI) (Daut et al., 1983; Cleeland and Rayn, 1994) has long (15 minutes) and short versions; either a health-care professional or patient can complete it. The short version assesses pain severity from four frames of reference—‘pain right now’, ‘pain at its best’, ‘pain at its worst’, and ‘pain on average’—and also measures pain relief using a VAS. A pain schematic shows the localization of pain, and pain interference during the last 24 hours is determined in terms of various domains, for example, general activity, mood, walking ability, normal work, relations with other people, sleep, mood, and enjoyment of life. The longer version of the BPI also records information about medical history, has pain descriptors, and asks about interference with daily activities in the last month and last week. The BPI has been translated and validated in many languages. The short form has been recommended for cancer pain assessment (Caraceni et al., 2002).

The McGill Pain Questionnaire (MPQ) is another self-report tool validated in cancer patients. It assesses intensity, quality, temporal pattern, relieving and exacerbating factors, and site of pain. Intensity is assessed on a categorical verbal rating scale, which uses the descriptors mild, discomforting, distressing, horrible, and excruciating (Melzack, 1987). It has been validated in the cancer pain population (Graham et al., 1980; Dudgeon et al., 1993).

Various neuropathic screening tools exist (Bennett et al., 2007). The Leeds Assessment of Neuropathic Signs and Symptoms has five self-report pain items and two clinical examination items (Bennett 2001). PainDETECT is another self-report questionnaire consisting of seven questions related to the quality of the pain, a body map, and two further questions on radiation and timing (Freynhagen et al., 2006). The Douleur Neuropathique en 4 questions or DN4 has seven symptom and three clinical examination questions (Bouhassira et al., 2005), while the Neuropathic Pain Questionnaire has ten questions on sensation and two on affect (Krause and Backonja, 2003). All have been validated in populations without cancer and emerging evidence suggests that for screening tools in cancer populations these tools may be less reliable or require adapting (Rayment et al., 2013).

The Alberta Breakthrough Pain Assessment Tool was developed for use in clinical studies. There are 15 patient answered questions and questions on aetiology, pathophysiology, medications, and descriptions of the pain for the health-care professional.

Clinical examination is essential to ensure accurate diagnosis of the pathophysiology of pain. It not only aids accurate diagnosis but also allows for assessment of comorbidities and the patient’s overall physical state. This information is essential when considering management strategies.

A neurological examination often provides valuable information and is essential if neuropathic pain is suspected. The recent IASP guidelines for neuropathic pain underline the importance of clinical examination for an accurate diagnosis (Haanpää et al., 2011), and the recent NeuPSIG guidance (Haanpää et al., 2011) proposed a four-criterion grading system for non-cancer neuropathic pain highlighting the importance of examination in this subgroup. This guidance indicates that the examination of the patient with presumed neuropathic pain should demonstrate that the pain is neuroanatomically plausible and the patient has a history suggestive of a relevant lesion or disease of the central somatosensory or peripheral nervous system. With these criteria, neuropathic pain may be considered possible if there is at least one confirmatory bedside measurement, and probable if there is at least one confirmatory test. If all four of these criteria are met, then neuropathic pain may be considered definite.

Confirmatory imaging will help in the assessment of pain if there is doubt as to its cause or if it may lead to further directed treatment, such as radiotherapy (Twycross, 1995). Analgesia should not be withheld while a diagnosis is being obtained. Plain radiography and bone scintigraphy may help in determining the appropriateness of radiotherapy or orthopaedic intervention. Computed tomography and magnetic resonance imaging help determine the causes of intrathoracic or intra-abdominal and pelvic pain, and nerve conduction studies may help localize the lesion.

Patients with advanced disease face many challenges. Pain is often only one symptom of an illness that may have been associated with many losses—normality, independence, health, and the future. The significance of pain varies among individuals; a patient with peripheral neuropathic pain due to diabetes may be very distressed but a patient with ischaemic heart disease may believe chest pain means they are dying. It is still common for people to believe that severe pain is unavoidable in advancing illness, especially cancer, and this may place extra suffering on patients. It therefore is a clinical imperative to understand a patient’s understanding of the cause of their pain, and their concerns about management and what may happen in the future. This inevitably needs to be done with consideration for other symptoms. Other factors are important to help predict response to treatment. For example, organ failure, cachexia, and age affect pharmacokinetics and response to treatment, and should be considered within any management plan.

It is essential when considering a management plan to consider both the potential benefits and burdens. Performance status and likely prognosis may alter what treatment is offered. For example, a patient with pain from a femoral fracture due to metastatic disease who has been gradually deteriorating for many months, was bed bound before the fracture and is anorectic, cachetic, and fatigued with a prognosis of days will not benefit from radiotherapy, while a patient with a similar pain due to a femoral fracture who was walking and independent prior to the fracture, and has a prognosis of many months, should be considered for both orthopaedic and oncological intervention.

Explanation of the cause of pain and its management needs to be made not only to the patients but to any carers/family that the patient identifies as helping in their day-to-day life. Their understanding may be crucial in designing and implementing management. Consideration also needs to be made as to whether initial assessment and management is better done in an inpatient unit or on an outpatient basis.

Continuing assessment of pain and its response to management is vital in ensuring that patients get the treatment they deserve. Clinicians working within a multidisciplinary team with expertise in nursing, social work, physiotherapy, occupational therapy, spiritual care, and complementary therapies ensure a holistic assessment is made of a patient’s pain and a holistic management plan is implemented.