Pain: A Descriptive Study in Patients With Cancer

Abstract

Ninety percent of patients with advanced cancer are reported to experience pain at some time during the course of their illness. This pain usually is improperly evaluated because it varies by patient and is difficult to control. Pain often cannot be conceptualized because of the individual nature of patient experience; however, patterns of pain occur in those suffering from similar conditions. This secondary analysis was conducted to examine the characteristics of pain in individuals with advanced cancer. The sample consisted of 717 patients with cancer who were receiving hospice home care in the southeastern United States. An understanding of the total pain experience, including type of pain and pain location, is important for addressing pain in patients with cancer in hospice care. More effective strategies are needed for pain management in this patient population. In addition, steps need to be taken to ensure that hospice care is understood by and made available to minorities.

Pain, “an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage,” is experienced by everyone at some point (International Association for the Study of Pain, 1994, p. 2010). For patients with cancer, pain is a significant problem. Cancer pain is unique, multidimensional, unpredictable, and affects many aspects of a patient’s life (Brawley, Smith, & Kirch, 2009; Herr et al., 2010; Knudsen et al., 2009; Mehta & Chan, 2008).

In 2012, 13.7 million people had a history of cancer and 1,665,540 new cases are projected in 2014 (American Cancer Society, 2014). Of those with advanced disease, two-thirds have severe pain that is highly variable and thus difficult to control (Ngamkham, Holden, & Wilkie, 2011). Therefore, the frequency and severity of chronic pain in these individuals requires careful assessment using the concept of total pain, an individualized patient experience characterized by physical, psychological, social, emotional, and spiritual components (Mehta & Chan, 2008; Saunders, 1993). Despite its importance, pain measurement does not occur regularly in clinical cancer practice (Herr et al., 2010; Hjermstad, Fainsinger, & Kaasa, 2009), and pain is often undertreated (Levy, Chwistek, & Mehta, 2008; Owens, Simmons, Gibson, & Weeks, 2001; Strassels, Blough, Hazlet, Veenstra, & Sullivan, 2006). A lack of understanding about patients’ pain experiences and the terms they used to describe their pain may result in ineffective pain management (Mehta & Chan, 2008).

Studies have evaluated general pain assessment and management in patients with cancer receiving hospice care or pain in patients in hospice in general (Bisht, Bist, Dhasmana, & Saini, 2011; Marinangeli et al., 2004; Owens et al., 2001; Sanders et al., 2013; Strassels et al., 2006; Tang, 2010; Tse et al., 2012). A few have assessed the pain experience of patients with cancer in hospice; however, most involved a broad appraisal of the pain experience (Dobratz, 2001, 2009; Dumitrescu, van den Heuvel-Olaroiu, & van den Heuvel, 2007; McMillan & Small, 2002; Mercadante et al., 2012; Mystakidou et al., 2007). A thorough evaluation of symptomatology in patients with cancer in hospice is necessary for understanding the pain experience and addressing the pervasive issue of pain.

The current study was conducted to describe patterns of pain in patients with cancer in hospice in the southeastern United States. A detailed explanation of the pain characteristics in this secondary analysis complements the original study by McMillan, Small, and Haley (2011), which involved systematic assessment and feedback to hospice interdisciplinary teams. Information regarding pain characteristics adds to this knowledge and affects interventions intended to improve the patient experience.

The current study aimed to answer the following questions: What are the characteristics of pain in individuals with advanced cancer in hospice care? Does a relationship exist between pain severity and descriptors used by patients? Is cancer type related to descriptors used by patients? Does an association exist between level of education and descriptors used by patients?

Methods

The current study was a descriptive, secondary analysis of a larger National Institute of Nursing Research–funded study involving patients with cancer receiving hospice care. Although it was a longitudinal study, only baseline data were used for this analysis.

Primary Study

A randomized clinical trial of 717 patients with cancer was conducted to evaluate the effectiveness of using systematic feedback from standardized assessment tools in improving hospice outcomes for patients and their caregivers. Interdisciplinary teams caring for the patient and caregiver dyads were randomly assigned to experimental or control groups. Data were collected at hospice admission and one week after each of the first two interdisciplinary team meetings (McMillan et al., 2011).

The study sample included patients aged 18 years or older who were receiving hospice home care from one of two large not-for-profit hospices in the southeastern United States. Hospice administrators and the university’s institutional review board approved the project. Within 24–72 hours of admission to hospice, patients were contacted by phone and invited to participate. Patients were excluded if they were confused, excessively debilitated, comatose, or actively dying.

Demographic data collected included gender, age, ethnicity, educational level, marital status, religion, length of time since diagnosis, presence of pain, medication use, and cancer type. Patients were screened using the Palliative Performance Scale (PPS) and the Short Portable Mental Status Questionnaire (SPM-SQ). The PPS, a variation of the Karnofsky Performance Status Scale, measured physical status of patients receiving palliative care (scaled from 0–100) (Anderson, Downing, Hill, Casorso, & Lerch, 1996). The SPMSQ assessed for cognitive impairment (scaled from 0–10) (Pfeiffer, 1975). The Brief Pain Inventory (BPI) assessed pain (scaled from 0–10) throughout the study and measured pain at its worst, least, average, and current levels. Previous studies suggested that the BPI is a valid and reliable instrument (Daut, Cleeland, & Flaner, 1983; McCormack, Li, Zarowny, & Singer, 1993). In addition, patients were asked to describe their pain; the descriptors were recorded because the percentage of individuals who chose a given descriptor may be useful information. The Memorial Symptom Assessment Scale (MSAS) measured symptoms related to cancer according to severity, frequency, and associated distress. Likert-type scales were used to evaluate pain severity and pain distress. The items in each subscale (physical and psychological) were summed to obtain a score; the higher the score was, the more severe, frequent, or distressing the symptom cluster was for the patient (Portenoy et al., 1994). The validity and reliability of the MSAS have been supported by analyses in previous studies (McMillan & Small, 2002; Portenoy et al., 1994).

Secondary Study

Conceptual framework:

The Theory of Unpleasant Symptoms (Lenz, Suppe, Gift, Pugh, & Milligan, 1995) served as a guide for secondary analysis regarding symptom quality and patient experience. This theory suggests that a person’s susceptibility to or experience of a given unpleasant symptom is affected by physiologic, psychological, and situational factors. The dimensions (i.e., duration, intensity, quality, and distress) of the symptoms, in turn, affect the person’s level of cognitive functioning and physical performance (Lenz et al, 1995). The theory highlights three major components of the patient’s multidimensional symptom experience (i.e., symptoms, influencing factors, and consequences) (Lenz, Pugh, Milligan, Gift, & Suppe, 1997). In the current study, the quality of one symptom, pain, was evaluated using descriptive terms associated with the dimensions in the theory.

Procedures:

Baseline patient data related to the pain experience, a subset of the primary data set, were used for this study. The data were evaluated for value and data entry errors using descriptive statistics and frequencies.

Data analysis:

Deidentified data from the original study were provided by the principal investigator. Analysis of the data was conducted using descriptive statistical methods within SPSS^®, version 21.0. In the original study, patients were asked to describe their pain experience, and their descriptors were coded into two main categories: patient experience and inferred pain mechanism. These categories coincide with various classification systems and tools, such as the McGill Pain Questionnaire, that have been used to describe or evaluate pain in patients with cancer and were useful in coding the data for this study (Knudsen et al., 2009; Melzack, 1975).

Results

The study included 717 participants; the majority were Caucasian, male, aged 73 years, and married (see Table 1). Cancers seen most often included lung, pancreas, prostate, and colon. Months since diagnosis ranged from 0–444 ($\overline{\text{X}}$ = 26.8, median = 11). About 65% of the participants were taking medications; however, medication type was not recorded. The patients’ SPMSQ scores ranged from 7–10 ($\overline{\text{X}}$ = 9.23), and 51% had a perfect score of 10. All participants met the performance status criterion with scores of 40 or higher. All patients had advanced cancer and were in the final stages of their illness.

TABLE 1.

Sample Characteristics (N = 717)

Characteristic	n	%
Gender
Male	400	56
Female	311	43
Missing	6	1
Age (years)
20–35	6	1
36–50	26	4
51–65	159	22
66–80	313	44
81–95	204	28
Missing	9	1
Ethnicity
Caucasian	690	96
African American	11	2
Hispanic	7	1
Asian or Pacific Islander	1	< 1
Missing	5	1
Other	3	< 1
Education level (years)
0–12	416	58
13–16	239	33
17–28	52	7
Missing	10	1
Marital status
Married	451	63
Widowed	131	18
Divorced	82	11
Never married	37	5
Separated	8	1
Missing	8	1
Religion
Christian (not Catholic)	408	57
Catholic	204	28
Agnostic	2	< 1
Buddhist	1	< 1
Shamanism	1	< 1
Spiritual	1	< 1
None	93	13
Missing	7	1
Cancer type
Lung	241	34
Abdomen (pancreas, stomach, liver, gallbladder, appendix)	112	16
Genitourinary (prostate, kidney, bladder, duodenal, intestinal)	80	11
Colon, rectal, anal	53	7
Gynecologic (cervical, ovarian, uterine, endometrial, vulvar)	40	6
Breast	40	6
Head, heck and throat (larynx, trachea)	26	4
Lymphoma, Hodgkin lymphoma, mycosis fungoides	23	3
Skin (basal cell carcinoma, melanoma)	18	3
Acute myeloid leukemia, leukemia, myelodysplasia	13	2
Nervous system, brain	13	2
Other	42	6
Unknown	7	< 1
Missing	9	1

Note. Because of rounding, not all percentages total 100.

Pain

On the MSAS, pain presence during the week prior was assessed. Patients rated the pain severity (1–4, slight to very severe) and distress caused by the pain (0–4, not at all to very much). About 71% of the participants reported pain. Average ratings were 2.28 (SD = 1.01) for severity and 2.3 (SD = 1.16) for distress.

Pain Sites

On the BPI, patients were asked whether they ever had pain, and up to three pain sites could be identified (site A, B, or C). Pain was reported by 76% of the participants, and 94% identified at least one site. The average pain rating (scaled from 0–10) at each site ranged from 2.13 (SD = 2.31) to 2.6 (SD = 2.74). About 66% of participants had mild pain (1–3), 22% had moderate pain (4–6), and 12% had severe pain (7–10). The average for the worst pain in the past 24 hours ranged from 3.71 (SD = 4.22) to 5.19 (SD = 2.97) (see Table 2).

TABLE 2.

Average Pain and Worst Pain

Category	n	$\overline{\text{X}}$	SD
Average pain (N = 694)
Site A	497	2.601	2.743
Site B	148	2.13	2.316
Site C	49	2.35	2.377
Worst pain (N = 237)
Site A	159	5.19	2.97
Site B	61	4.75	4.05
Site C	17	3.71	4.22

Note. Sites A, B, and C represent three independent pain sites.

Pain Descriptors

About 51% of the participants described their pain using a term from the affective category (see Table 3). Pain was least frequently (1%) described in terms of severity with words such as “severe/intense” and “bad”. Descriptive terms also were delineated into categories based on inferred mechanism (see Table 4). Terms that did not correspond with one of these categories were assigned to a miscellaneous category. Most patients (67%) used somatic terms to describe their pain. Use of neuropathic terms was less frequent at 18%; these terms included “burning,” “tingling,” and “shooting.” Visceral terms were used least frequently (14%) and included “pressure,” “squeezing,” and “deep”. Some patients (6%) could not isolate a pain location, which likely reflects characteristics of visceral pain and often is generalized and difficult to locate (Levy et al., 2008).

TABLE 3.

Frequency and Percentage of Descriptive Terms According to Patient Perception (N = 449)

Category	Definition	Descriptorsa	n	%
Affective	Physiologic and psychological elements that are part of the pain experience	Aching, exhausting, sickening, dull, sore, hurts, touchy, hard, sensitive, deep	238	51
Sensory	Physical or sensory qualities of the pain experience	Throbbing, cramping, burning, gnawing, stabbing, jabbing, pinching, shooting, sharp, piercing, tingling, pulling, prying, wringing, twisting, bruising, shock-like, like being hit or kicked, pounding, squeezing, tightness, pressure, gassy, bloated, stiffness, spasm, catch, friction, irritation, grabbing, radiating	204	47
Timing	Pain frequency	Intermittent, constant, steady	12	3
Evaluative	Subjective description of pain intensity	Miserable, nagging, indescribable, annoying, aggravating	8	2
Severity	Pain strength or intensity	Severe, intense, discomfort, bad	5	1

^aListed in order of frequency of participant use

Note. Participants could use more than one category of descriptor. Because of rounding, not all percentages total 100.

Note. Based on information from Melzack, 1975.

TABLE 4.

Frequency of Descriptive Terms According to Inferred Pain Mechanism (N = 449)

Category	Definition	Descriptorsa	n	%
Somatic	Associated with musculoskeletal tissues that is localized and is often considered sharp	Aching, dull, sharp/like a knife, throbbing, spasm, pounding, splitting, stiffness, sore, bruising, pulling, hurts, prying, wringing/twisting, like being hit/kicked, tension, hard, friction, irritating, grabbing	313	67
Neuropathic	Related to nerves and is often described as itching, burning, or pins and needles	Burning, tingling, shooting, stabbing, jabbing, shock-like, piercing, radiating, gnawing, pinching, touchy, sensitive	82	18
Visceral	Related to the internal organs that is unclear, is not localized, and is experienced as an ache	Pressure, squeezing, deep, dull, cramping, sickening, constant, steady, tightness, gassy/bloated	64	14
Miscellaneous	–	Exhausting, miserable, nagging, indescribable, severe/intense, discomfort, catch, bad, intermittent, annoying, aggravating	20	4

^aListed in order of frequency of participant use

Note. Participants could use more than one category of descriptor. Because of rounding, not all percentages total 100.

Note. Based on information from Levy et al., 2008.

A standardized, composite score of two pain measures, pain severity from the MSAS and worst pain in the 24 hours from the BPI, was created to further evaluate descriptors used in varying degrees. Pain was delineated into three categories (i.e., low, medium, and high) based on the composite score. A total of 449 patients indicated a pain intensity score and identified an adjective that described their pain. The majority of the patients (43%) used “aching” to describe their pain, with 72% reporting medium (44%) or high pain (28%). The most common descriptors (i.e., “aching,” “sharp/like a knife,” “stabbing,” and “throbbing”) were used equally frequently to describe all pain categories. Pain severity also was evaluated according to cancer diagnoses. However, no significant differences were noted among the means for pain severity for five common cancers (i.e., lung, pancreas, colon, prostate, and breast), and “aching” was used most frequently by all groups. In addition, education was coded into four levels according to years of school (1–8, 9–12, 13–16, and 17 or more). Patients also used the term “aching” most frequently regardless of education. Patients with more years of education reported lower pain scores. No significant correlations were noted among education level, patient age, and pain category.

Discussion

Study results are similar to previous studies that reported a prevalence of 64%–80% for pain experienced in advanced disease (Bisht, Bist, Dhasmana, & Saini, 2008; Herr et al., 2010; Hjermstad et al., 2009; Levy et al., 2008; Luger, Mach, Sevcik, & Mantyh, 2005; Stark, Tofthagen, Visovsky, & McMillan, 2012). About 71% of participants reported pain on the MSAS, and 76% reported pain on the BPI. The difference in wording on each instrument may have contributed to varied reports. Nonetheless, pain was noticeably prevalent. Patients identified up to three pain sites and average pain intensity varied by site, ranging from 2.13–2.6 (scaled from 0–10). However, this rating is not an exact representation of pain intensity because the presence of extreme scores affected the mean. Evaluation of the worst pain in the 24 hours prior was more informative; the mean ranged from 3.71–5.19 (scaled from 0–10). The upper end of this range was consistent with previous studies that reported the prevalence of moderate to severe pain in patients with cancer (Chow et al., 2006; Hjermstad et al., 2009; Owens et al., 2001; van den Beuken-van Everdingen et al., 2007); however, it should be noted that these patients already were receiving some pain medication before the start of the study, potentially lowering scores. According to the MSAS, the average rating for pain severity was 2.28 (scaled from 1–4), and the average for distress was 2.3 (scaled from 0–4), which appear low; however, these findings should be evaluated in the context of the scales used. These ratings equate to higher scores of 5.7 for severity and 5.75 for distress when scaled from 0–10.

Patients most often described the physical and psychological properties (affective domain) of their pain. Previous studies also identified words from these categories as the most frequently used descriptors (Dobratz, 2001). Therefore, physical and psychological processes are closely related and need to be understood to address pain management approaches.

Pain often is described as somatic, visceral, or neuropathic (Knudsen et al., 2009). The prevalent use of somatic terms in the current study supports previous studies that identified somatic pain as the most common type of pain in individuals with advanced cancer (Dobratz, 2009; Potter & Higginson, 2004). This finding may suggest somatic metastasis and has implications for pain management given that somatic, visceral, and neuropathic pain are all treated differently. For example, anti-inflammatories may work best for low to moderate somatic pain; however, visceral pain may respond better to opioids, and neuropathic pain to anticonvulsants or tricyclic antidepressants (Dobratz, 2009; Levy et al., 2008). Patients, however, may have mixed types of pain requiring multiple pharmacologic approaches.

No association existed between pain severity and descriptors, between cancer type and descriptors, or between education level and descriptors; however, “aching” was used more frequently regardless of pain level, cancer diagnosis, or education level. In addition, individuals with higher education reported lower pain scores. Also, individuals with higher education levels are likely to use more words than others; however, this was not evaluated in this secondary study.

This study provides data that apply to the standardized use of descriptors in pain assessment. Pain was least frequently described in terms of severity; however, those are the terms clinicians often use to understand patients’ pain. This finding suggests a possible disconnect between provider and patient perceptions of pain. Although pain is frequently encountered in the clinical setting, clinicians typically are not aware of literature regarding adjectives used by patients to describe pain. Study findings may help to inform pain assessment approaches, as well as improve communication between patients and providers.

Limitations

The original study consisted of a majority of Caucasian patients from one geographic area, decreasing the generalizability of the findings. Future studies should include more diverse samples. Patients were accrued to the study within 24–72 hours of admission to the hospice. Therefore, a drawback was that pain management by the hospice team might have been initiated prior to patient enrollment in the study and could have affected patient reports about pain severity. Various cancer types also were studied; therefore, patients had different pain issues, which sometimes require varied pain management strategies. Information regarding specific pain locations was not collected. Future studies should collect this information to better understand patients’ total pain experience. In addition, some data were missing, which may be attributed to patient fatigue during data collection.

Clinical and Research Implications

Nurse educators should include instruction about pain types, evaluation, and management approaches in schools of nursing and continuing education settings. Sources for patient education and resources include the National Cancer Institute (https://pubs.cancer.gov/ncipl/detail.aspx?prodid=P258) and the American Cancer Society (www.cancer.org/treatment/treatmentsandsideeffects/physicalsideeffects/pain/index). Nurses in clinical settings need to conduct more thorough assessment of pain so that they understand how best to manage it. Future research using more diverse samples of patients at all stages of the disease is needed.

The high prevalence of pain in the current study suggests that more effective strategies are necessary for pain management in patients with cancer receiving hospice care. Both physical and psychological processes are involved in pain; therefore, pain management approaches should address the total pain experience. In addition, the presence of moderate to severe somatic pain should be evaluated further because it will influence the types of modalities used to achieve pain management. More research identifying specific pain locations in patients with cancer receiving hospice care would add to an understanding of the pain experience in this population. That knowledge then could be applied to the care of patients with similar pain experiences in cases of cancer or chronic disease. Hospice care is important in advanced disease, and more initiatives are necessary to ensure that this care also reaches the minority community.

Conclusions

The pain experience is not uniform for all patients and conditions. The findings of this secondary analysis add strength to the body of knowledge regarding pain in patients with cancer receiving hospice care. A thorough understanding of the characteristics of pain is important for managing patients’ symptoms. Patterns of pain in patients with advanced cancer are unique and multidimensional; therefore, a framework that involves all dimensions of pain is most useful in addressing pain.

Implications for Practice.

• Improve pain assessment by evaluating total pain and using standardized descriptors during routine examinations.

• Develop pain management modalities to target specific inferred mechanisms.

• Increase education about the availability of hospice care in the minority community.