- Email: [email protected]
Prediction and management of pain in patients with advanced cancer
Prediction and Management of Pain in Patients with Advanced Cancer F. Patrick McKegney, M.D., Linda R. Bailey, R.N., B.S.N., and Jerome W. Yates, M.D. Vermon t Regional Cancer Center and the Departments Medicine, Burlington, Vermont
of Medicine and Psychiaty,
Abstract: Zn a prospectiue controlled study, patients with incurable cancer and an estimated prognosis of three months to one year were allocated to two patient groups for evaluation of supportive intewentions. One group, called “intensive,” received home visits by a nurse practitioner acting as an extension of a multidisciplinary team; the other, termed “nonintensive,” did not receivesuch visits. Both patientgroups wereperiodically evaluated in their homes by an obsewer and by self-ratings, to measure changes in quality of life as their disease progressed. Several methodologie problems were identified, which have implications for future research. Pain problems later in the course of cancer seem to be predictable in those patients with higher scores on the Cornell Medical Index M-R scales (greater emotional disturbance) and on the Rotter Lotus of Control (l-E) scale (expectation of more external con trol of self). Thestudy also found that the home visiting nurse practitioners, specially trained in pain treatment, improved pain control in the “intensive” group of patients over the last 90 days of life, when pain was an increasingly major problem. Such nurses can signifcantly improve the quality of life for patients dying outside of institutions.
Pain is one of the most common and distressing symptoms in persons dying of cancer, with 40-70% of such patients significantly afflicted (14). Pain is now acknowledged to be a psychological phenomenon with varying biological, psychological, and social etiologic factors (5). Because pain is a complex and often poorly understood symptom, adequate pain control is frequently a problem for patients with cancer (4). The
This work was supported in part by NIMH Training Grant MH0805717 and NCI grant 17868,
Gen& Hospital Psychiatry 3, 95-101,
1981
0 Elsevier North Holland, Inc., 1981 52 Vanderbilt Avenue, New York, NY 10017
University of Vermont College of
vast major@ of patients with noncurable cancer are not institutionalized and often receive less attention to their pain symptomatology because of lack of access to physicians. Patients in hospitals and nursing homes, on the other hand, can communicate their pain symptoms to readily available health professionals. Even in such settings, however, the management of pain may be more related to the person complaining of pain than to the degree of pain complaint itself (3). Even aftèr patients have communicated their pain to a health professional, that pain may be given less than optimal treatment (3). Marks and Sacher found that medical inpatients were almost universally undertreated with narcotic analgesics, according to standard pharmacologic guidelines (6). Even experienced clinicians hesitate to prescribe increasing amounts of narcotic analgesics for cancer patients, often with a nonverbalized fear of “addiction,” though the findings of Saunders and others indicate that significant tolerante develops only infrequently in patients with terminal cancer (2). Because pain is a major problem and current management is often inadequate, emphasis on this aspect of patient care was studied by the Vermont Regional Cancer Center (VRCC) at the University of Vermont. In a large prospective controlled study of patients with terminal cancer, pain was closely studied as an important reflection of quality of life in these patients. The major hypothesis of the study was that visits in the home by trained oncology nurse practitioners would result in improved quality of life for an experimental group of patients, 95 ISSN 0163.8343181/020095-07/$02.25
F. P. McKegney et al
compared with a carefully matched control group of patients who did not receive such home visits but whose care was otherwise the same as that of the experimental group. Multiple psychosocial and biological variables were also studied, as they might identify those patients with predictably greater difficulties in pain management over their last months of life.
Met,hods Vermont is a largely rural environment, unique in having a medical school and one major medical center, the Medical Center Hospita1 of Vermont (MCHV), responsible for most of the region’s cancer surgery, radiation therapy, and chemotherapy. The stability of the population, constant care resources, and geographic isolation also make the region a suitable setting for a prospective controlled trial requiring continued patient follow-up. A complete description of this study’s design can be found elsewhere (7). For the purposes of this report about pain and related variables, the following details are pertinent. Counties in Vermont were paired on the basis of population density, distance from the Medical Center, socioeconomic status, local medical facilities, referral patterns, and local social services resources. The paired counties were randomly separated into two groups, with one designated “intensive” and the other “non-intensive.” Those patients residing in the groups of counties designated “intensive” received regular home visits by nurse practitioners as described below. The patients from the counties designated non-intensive were not visited at home by nurses. Both groups of patients received the same multidisciplinary care at the MCHV andlor from their private physicians. Regular participants in the multidisciplinary team consisted of medical and radiation oncologists, psychiatists, social workers, physical therapists, nutritionists, occupational therapists, enterostomal therapists, and clergymen. The requirements for patient eligibility included: mental competente; a minimum age of 16 years; a signed informed consent; and an expected survival of greater than three months but less than one year as judged by an experienced oncologist. Al1 new patients seen by radiation therapy and medical oncology were eligible for accession to the study. Ten percent of the patients approached refused study assignment. Reviews early in the course of the study failed to demonstrate any apparent bias in 96
patient selection resulting from voluntary refusal. A total of 199 patients, 98 in the “intensive” and 101 in the non-intensive groups, were followed; at the close of the four-year study 139 had died. The nurse was chosen as the interface between the multidisciplinary team and patients receiving palliative therapy, since normal activities for nurses encompass many of the skills carried out by individual team members. Nurse practitioners with extensive experience in care of the patients with advance cancer were selected. Patients with an expected survival of less than three months were visited by nurses biweekly and those expected to live longer were visited monthly. The home visit by the nurses was primarily focused on attending to the needs of the patient, and interactions with family members were incidental to that task. In addition to providing physical care, much of the nurse’s time was spent in talking with the patient about their illness and its implications. The nurse frequently mobilized family and other social resources to meet the patient’s needs and also coordinated with the patient’s local physician. These nurses thus served in the well-known public health, or visiting nurse, role, with the differente that the project nurse had the benefit of a multidisciplinary health care team back-up resource. A “Protocol for Management of Pain” (8), available from the authors, was developed by the team and used by the nurses as part of their wide range of physical treatments and psychosocial interventions. This protocol was based upon sound pharmacological principles, many of which are often ignored. For example, oral methadone is an effective and longer lasting analgesic than other narcotics, and was used extensively in this program. Further, analgesics were usually prescribed on a regular time schedule, rather than on a p.r.n. basis, to ensure continuous relief rather than wide swings of degrees of discomfort. Once adequate pain relief has been achieved by a given daily regimen, the analgesic dose requirement may then be reduced, to avoid complications such as sedation or confusion. Anxiolytic agents were used when anxiety seemed to be a significant factor in pain. Tricyclic antidepressants were infrequently used because the depression syndrome was uncommon. Relaxation, hypnosis, and other pain-relieving techniques were not used because pain control was only one of many goals of the nurses’ visits and the visit rarely exceeded 30-45 minutes. Initial, on-study patient evaluation included the following instruments:
Prediction and Management of Pain in Patients
A modified Rotter Lotus of Control (1-E) scale, an eight-item self-report measure of internalexternal expectation of control (9). Higher scores indicate a person’s greater expectation of control from outside themselves. The Cornell Medical Index (CMI) M-R scales, a 51-item self-report measure of psychological symptoms (10). Higher scores indicate a greater degree of emotional disturbance. An
initial research data base (IRDB), which assessed by structured interview the patient’s medical, nutritional, physical, psychosocial, and religieus status.
The Karnofsky observer-rated status (11,13).
Performance Status (KPS), an 11-point scale of overall health
A “Pain Estimate” score described by Sternbach (12), a 0-100 self-report of pain at the present time. Follow-up data bases were gathered from both intensive and non-intensive patients by trained independent raters, using structured interviews in the patients’ homes, done at the same frequency as the nurses’ visits, which were based upon the patients’ prognosis. Follow-up interviews were carried out in the patients’ homes because earlier data had indicated greater validity of home vs. clinical evaluations for some measures (13). Initial and follow-up rater training took place during the pilot phase of the project using videotapes. At the end of the first year the raters had achieved acceptable interrater agreement on major scale items (Pearson correlation coefficient 2 0.70). The follow-up data bases covered the same areas as the initial research data base, including the self-reported “Pain Estimate.” In reporting results, pain levels over time are expressed as mean pain levels in groups of patients during blocks of time before death. Thirty-day blocks of time were used with larger numbers of patients in compared groups and 60-day blocks with smaller numbers of patients. (For each patient, the earliest pain leve1 in each 30-60-day time black was used.) Because patients came on study at varying stages of their illness, data analysis was done in terms of days before death, since the time of death is the clearest reference point for al1 patients. Mean pain scores of various groups of patients were compared for 30- or 60-day blocks of time before death, by Student’s t-test. Interactions between variables were examined by an analysis of variante.
with Advanced
Cancer
Results IntensivelNon-In
tensive Pain Comparisons
This report concerns the effect of the home nurse interventions on pain, together with some factors which seemed to influence pain in the patients studied. Results in other parameters are described in detail elsewhere (7). It should be briefly noted that the intensive and non-intensive patients did not differ in terms of length of survival, nor did these two groups differ in several other quality of life outcomes such as physical activity, nutrition, optimism, or overall health status as defined by the KPS (7). This report concerns the effect of the home nurse interventions on pain, together with some factors that seemed to influence pain in the patients studied. More than 80% of al1 the patients who died reported some degree of pain during their time on study, with an increasing frequency and severity as they approached death. Of the 139 patients who died during the study, 38 intensive and 45 non-intensive patients had a sufficient number of pain ratings (three or more) to compare the effectiveness of pain management over time in the two groups. A comparison of patient characteristics for these two groups demonstrated similarities in cancer diagnosis, sex, age, social class, and religious preferente (Table 1). The initial, on-study scores on the CMI, 1-E, and KI’S did not differ significantly between the intensive and non-intensive groups. When pain levels for these intensive and nonintensive groups were compared over time, in 30-day periods before death, the two groups had essentially the same mean pain scores until the last 90 days before death (Figure 1). In these last 90 days, the mean pain scores in the non-intensive group of patients continued to rise until death, whereas the mean pain scores in the intensive group of patients plateaued. During patients’ last 30 days of life, this differente in mean pain scores between the two groups approached customary statistical significante (P = 0.06).
Pain-Associated Variables To identify early predictors of pain problems, relationships between the on-study CM1 and scores and the pain scores over the course of disease were examined. A group of 55 patients, intensive and 33 non-intensive) were chosen on basis of having complete on-study CM1 and
the 1-E the (22 the 1-E 97
F. P. McKegney et al.
Table 1. Composition
of comparison
groupP Sex
Age (years)
Intensive (N = 40) Nonintensive (N = 43) High CM1 (N = 23) Low CM1 (N = 32) High 1-E (N = 34) Low 1-E (N = 21)
Type of cancer
Survival (days)
Mean
(SE)
M
F
Lung
Breast
GI
Other
Mean
(SE)
58.7 56.1 55.7 57.7
(1.7) (2.2) (2.6) (2.1)
21 27 10 23
19 16 13
20 14 8 19
6 6 5 1
8 9 3 4
6 14 7 8
345 309 327 299
(28) (24) (26) (40)
56.7 57.0
(1.8) (3.2)
20 13
13 14
4 2
5 2
12 3
296 334
(34) (40)
I: 8
‘NO differences significant at the P = 0.05 leve1 except: *females higher than males on the CMI, P < 0.05.
scores and at least one pain score while on study. The mean number of pain ratings per subject over time was 6.1.
differ in the 240-280-day and 180-120-day periods before death. However, the High CM1 group had higher mean pain scores than the Low CM1 group in the last 120 days before death, with this differente being statistically different in the last 60 days before death (P < 0.05). There was no significant interaction between the intensivelnon-intensive variable and the CM1 variable with regard to pain.
Comell Medical Index. Of these 55 patients, 23 had initially given 6 or more “yes” responses on the M-R section of the CMI, indicating a mild to severe emotional disturbance. This “cut score” of 6, slightly above the customary 4 or 5, was chosen in
Rotter Lotus of Control (I-E) Scale. Of the 55 patients, 34 initially gave 4 or more “external” responses on the 8-item 1-E scale, and 21 patients gave 0-3 external responses. There were no significant differences in demographic or disease characteristics between the two 1-E groups (Table 1). Mean pain scores in the 34 externally oriented High 1-E patients were compared with the mean pain scores in the more internally oriented 21 Low 1-E patients
order to increase the accuracy of positive prediction (14). There were no significant differences in demographic or disease characteristics between the two CM1 groups except that females scored higher than males (Table 1). Mean pain scores in these 23 “High CMI” patients were compared with the mean pain scores in the 32 “Low CMI” patients over time, in 60-day periods before death (Figure 2). Mean pain scores in the two CM1 groups did not
INTENSIVE
,et
(38)
NON-INTENSIVE
(451
O----
I S.E. +
I S. E.
T Figure 1. Mean pain scores (0-100) for “Intensive” and “Non-intensive” groups of patients at time on study and for each 30-day period before death. The “Intensive” group of patients has lower mean pain scores than the “Non-intensive” group over the ,last 90 days before death. The differente in the 30-O-day period approaches statistical significante (P = 0.06). *Not al1 patients had pain scores in al1 time periods.
ON
98
STUOY
-180
-ISO
-120 OAYS
-90 FROH
-60 DEATH
-30
0
Prediction
and Management
of Pain in Patients with Advanced Cancer
problems is a higher incidence than that in other studies, which may be due to our following patients until death rather than sampling a cross-section of a population with cancer. This high incidence clearly indicates the need for more attention and study than in the past. This is the first study examining the impact of standardized home intervention for pain problems in patients with cancer. The factor apparently most responsible for this lower pain in the “intensive” group over the last 90 days of life was reflected in the monitoring of nurse interventions. In the last 90 days of life approximately two-thirds of the nurse interventions involved management of pain: initiating pain medications, increasing doses of the same pain medication, and changing to or adding a more This demonstrated benefit by potent analgesic. home intervention strongly supports the current “Hospice” home care nursing movement, at least in terms of pain management. This better access to health care personnel needs to be combined with greater education in the use of pain management procedures for al1 professionals, especially appropriate pharmacologic intervention (8,16). A multidisciplinary team developed a Pain Management Protocol, available from the authors, with clear guidelines for appropriate and effective interventions in varying degrees of pain. The availability of such an instrument as an educational tool should improve professional responses to pain complaints, irrespective of setting.
over time, in 60-day periods before death (Figure 3). Mean pain scores in the High and Low 1-E groups were not different in the 240-180-day period. However, the High 1-E group had significantly higher mean pain scores than did the Low 1-E group in the 180-120-day and 120-60-day periods before death (P < 0.05). This differente persisted in the last 60 days before death but was no longer statistically significant at the 0.05 level. There was no significant interaction between the intensivelnon-intensive variable and the Rotter variable with regard to pain. The initial on-study CM1 and 1-E scores in the 55 patients showed a significant positive correlation, (Pearson r = 0.34, P = 0.003), with the more emotionally disturbed having a greater external lotus of control. However, the CM1 and 1-E scores showed no correlation with initial pain scores, with initial overall health status, as measured by the KI’S score, or with survival after entry on study.
Discussion Pain has traditionally been viewed as a cardinal “sign” of disease or inflammation (15). However, unlike swelling, redness, and heat, pain is not an objective sign but rather a subjective symptom. This time-honored misconception highlights our need to learn more about pain and its treatment, particularly in those illnesses in which pain is extremely common. Our finding that more than 80% of patients approaching death have significant pain
60
HIGH CHI123) -fl LOW
CHI
(321
50 Figure 2. Mean pain scores (0-100) for High CMI patients (M-R scales score 6-17) and Low CM1 patients (M-R scales 0-5) for each 60-day period before death. The High CM1 group (N = 23) has higher mean pain scores than the Low CMI group (N = 32) in the last two periods before death; the differente is significant (P < 0.05) in the 60-0-day period. “Not al1 patients had pain scores in al1 time periods.
?? ---+-
T
SE I SE
LO : 0 Co 30 z d 20
10
-
.
2lbo
-180
-120
DAYS FROM
-60
0
DEATH
99
F. P. McKegney
et al.
6C HIGH
1-E
134)
-ti
SE
LOW
1-E
Ik?11
?? ---+l
SE
51
Figure 3. Mean pain scores (0-100) for
High 1-E patients (4-7 external responses) and Low 1-E patients (0-3 external responses) for each 60 day period before death. The High 1-E group (N = 34) had higher mean pain scores than the Low 1-E group (N = 21) in the last three periods before death; these differences are significant (P < .05) in the 180-120-day and 120-60-day periods. *Nat al1 patients had pain scores in al1 time periods.
$0
-180
-120
-60
0
DAYS FROM DEATH
This prospective study also demonstrates that early screening evaluation of certain psychological parameters, such as emotional disturbance and independence or dependence upon the environment, can significantly predict pain problems later in the course of cancer. While the Rotter Lotus of Control Scale and the Cornell Medical Index M-R scores were not correlated with initial levels of pain in the population described, they did generally predict those patients who would have more pain problems later in the course of disease. Where there is limited availability of health care professionals skilled in pain management, those patients found to have more emotional disturbance and greater dependence on the environment may be those to whom more care is given. This study does confirm other, less wellestablished, observations that greater emotional disturbance and greater orientation to environmental influence are associated with more pain complaints (3,17). The finding that 1-E scores and CM1 ratings of emotional disturbance are interrelated also confirms previous observations (9). The lack of association between patient survival and the initial scores of emotional disturbances needs to be compared with the observations of others (19) who have reported such an association. Likewise, the finding that external or internal expectations of control are not associated with survival raises some question about treatment methods that emphasize independence and self-responsibility (20). For the benefit of future research, some difficul100
ties encountered in the present study must be mentioned. One major problem was that of gathering complete initial data bases: the observer-scored IRDB, and the patient-completed Cornell Medical Index, Pain Estimate and Lotus of Control scales. Two patient factors contributing to incomplete data collection were more serious physical illness and the presence of mild mental confusion (18). Observer factors included discomfort with asking upsetting questions about patients’ disabilities and unhappiness, and a reluctance to impose yet another burden on patients with progressive illness. To overcome these obstacles, which accompany the study of patients with serious illness, the data management system must be especially wel1 organized to optimize data collection. Another major problem in this study involved the trained observers who gathered follow-up data from patients in their homes. These observers were continually instructed to limit their activities to asking questions, observing behavior, and recording data. Early in the study, it became apparent that these home observers could not restrain themselves from offering help to patients and families when their needs became apparent in the process of data gathering. Although personnel changes were made later in the study, the human contact between the research observers and the patients in the home may have reduced differences in quality of life outcomes between the intensive and non-intensive groups. This hypothesis is supported by the fact that the major differente between the two groups
Prediction
was in pain management, the one problem which the nurses could address, but which the nonmedical observers could not. At any rate, the “observation effect” may wel1 be a major factor in studying the quality of life outcomes of different treatment interventions.
Conclusion This study’s design enables US to establish two comparable groups of patients for this prospective study of controlled health care interventions in patients with terminal cancer. It further found that home visits by specially trained nurses improved overall group quality of life, at least in terms of more adequate pain management. More refined studies should be carried out to confirm those risk factors that might identify early those patients who wil1 benefit most from these home interventions, if al1 patients are not able to receive such intensive home care.
Ref erences 1. Cox SS: Pain in terminal illness. Arizona Medicine
36:586-587, 1979
2. Saunders C: The treatment of intractable pain in terminal cancer. Proc R Sec Med 56:195-297, 1963 3. Pilowsky 1, Bond MR: Pain and its management in malignant disease. Psychosom Med 31:400-404,1969 4. Holden C: Pain, dying and the health care system. Science 203:484485, 1979 5. Merskey H, Spear FG: The concept of pain. J Psychosom Res 7:59-67, 1967 6. Marks RN, Sachar EJ: Undertreatment of medical inpatients with narcotic analgesics. Ann Intern Med 78:173-181, 1973 7. Yates JW, McKegney FF’, Visco G, Brown CS: Intensive rehabilitative support of the patient with advanced cancer through home health care. Proc Am Assoc Cancer Res 18:199, 1977
and Management
of Pain in Patients with Advanced Cancer
8. Bailey L, Yates JW: Protocol for Management of I’ain. Publication of Vermont Regional Cancer Center, 1976 9. Rotter JB: Generalized expectancies for internal versus external control or reinforcement. Psych Monographs 80:1-28, 1966 10. Seymour GE: The structure and predictive ability of the Cornell Medical Index for a normal sample. J. Psychosom Res 20:469478, 1976 ll. Karnofsky DA, Burchenal JH: The clinical evaluation of chemotherapeutic agents in cancer. In Macleod CM (ed). Evaluation of Chemotherapeutic Agents. New York, Columbia University Press, 1949 12. Sternbach RA: Pain Patients: Traits and Treatment. New York, Academie Press, 1974 13. Yates JW, Chalmer B, McKegney FP: Evaluation of patients with advanced cancer using the Karnofsky Performance Status. Cancer 45:228-232, 1980 14. Ransohaff DF, Feinstein AR: Problems of spectrum and bias in evaluating the efficacy of diagnostic tests. N Eng1 J Med 229:926-930, 1978 15. Florey H: Lectures on Genera1 Pathology. Philadelphia, W.B. Saunders, 1954, p 21 16. Shimm DS, Logue GL, Maltbie AA, Dugan S: Medical management of chronic cancer pain. JAMA 241:2408-2412, 1979 17. Wise TN, Hall WA, Wong 0: The relationship of cognitive styles and affective status of post-operative analgesic utilization. J Psychosom Res 22:513-518, 1978 18. Danies RK, Quinlan DM, McKegney FP, Kimball CP: Organic factors and psychological adjustment in advanced cancer patients. Psychosom Med 35:463471, 1973 19. Derogatis LR, Abeloff MD, Melisaratos N: Psychological coping mechanisms and survival times in metastatic breast cancer. JAMA 242:15041508, 1979 20. Simonton OC, Simonton S: Getting Wel1 Again. Los Angeles, JP Tarcher, 1978
Direcf reprint requesfs to: F. Patrick McKegney, M.D. Department of Psychiatry Medical Alumni Building University of Vermont College of Medicine Burlington, VT 05405
101
of Medicine and Psychiaty,
Abstract: Zn a prospectiue controlled study, patients with incurable cancer and an estimated prognosis of three months to one year were allocated to two patient groups for evaluation of supportive intewentions. One group, called “intensive,” received home visits by a nurse practitioner acting as an extension of a multidisciplinary team; the other, termed “nonintensive,” did not receivesuch visits. Both patientgroups wereperiodically evaluated in their homes by an obsewer and by self-ratings, to measure changes in quality of life as their disease progressed. Several methodologie problems were identified, which have implications for future research. Pain problems later in the course of cancer seem to be predictable in those patients with higher scores on the Cornell Medical Index M-R scales (greater emotional disturbance) and on the Rotter Lotus of Control (l-E) scale (expectation of more external con trol of self). Thestudy also found that the home visiting nurse practitioners, specially trained in pain treatment, improved pain control in the “intensive” group of patients over the last 90 days of life, when pain was an increasingly major problem. Such nurses can signifcantly improve the quality of life for patients dying outside of institutions.
Pain is one of the most common and distressing symptoms in persons dying of cancer, with 40-70% of such patients significantly afflicted (14). Pain is now acknowledged to be a psychological phenomenon with varying biological, psychological, and social etiologic factors (5). Because pain is a complex and often poorly understood symptom, adequate pain control is frequently a problem for patients with cancer (4). The
This work was supported in part by NIMH Training Grant MH0805717 and NCI grant 17868,
Gen& Hospital Psychiatry 3, 95-101,
1981
0 Elsevier North Holland, Inc., 1981 52 Vanderbilt Avenue, New York, NY 10017
University of Vermont College of
vast major@ of patients with noncurable cancer are not institutionalized and often receive less attention to their pain symptomatology because of lack of access to physicians. Patients in hospitals and nursing homes, on the other hand, can communicate their pain symptoms to readily available health professionals. Even in such settings, however, the management of pain may be more related to the person complaining of pain than to the degree of pain complaint itself (3). Even aftèr patients have communicated their pain to a health professional, that pain may be given less than optimal treatment (3). Marks and Sacher found that medical inpatients were almost universally undertreated with narcotic analgesics, according to standard pharmacologic guidelines (6). Even experienced clinicians hesitate to prescribe increasing amounts of narcotic analgesics for cancer patients, often with a nonverbalized fear of “addiction,” though the findings of Saunders and others indicate that significant tolerante develops only infrequently in patients with terminal cancer (2). Because pain is a major problem and current management is often inadequate, emphasis on this aspect of patient care was studied by the Vermont Regional Cancer Center (VRCC) at the University of Vermont. In a large prospective controlled study of patients with terminal cancer, pain was closely studied as an important reflection of quality of life in these patients. The major hypothesis of the study was that visits in the home by trained oncology nurse practitioners would result in improved quality of life for an experimental group of patients, 95 ISSN 0163.8343181/020095-07/$02.25
F. P. McKegney et al
compared with a carefully matched control group of patients who did not receive such home visits but whose care was otherwise the same as that of the experimental group. Multiple psychosocial and biological variables were also studied, as they might identify those patients with predictably greater difficulties in pain management over their last months of life.
Met,hods Vermont is a largely rural environment, unique in having a medical school and one major medical center, the Medical Center Hospita1 of Vermont (MCHV), responsible for most of the region’s cancer surgery, radiation therapy, and chemotherapy. The stability of the population, constant care resources, and geographic isolation also make the region a suitable setting for a prospective controlled trial requiring continued patient follow-up. A complete description of this study’s design can be found elsewhere (7). For the purposes of this report about pain and related variables, the following details are pertinent. Counties in Vermont were paired on the basis of population density, distance from the Medical Center, socioeconomic status, local medical facilities, referral patterns, and local social services resources. The paired counties were randomly separated into two groups, with one designated “intensive” and the other “non-intensive.” Those patients residing in the groups of counties designated “intensive” received regular home visits by nurse practitioners as described below. The patients from the counties designated non-intensive were not visited at home by nurses. Both groups of patients received the same multidisciplinary care at the MCHV andlor from their private physicians. Regular participants in the multidisciplinary team consisted of medical and radiation oncologists, psychiatists, social workers, physical therapists, nutritionists, occupational therapists, enterostomal therapists, and clergymen. The requirements for patient eligibility included: mental competente; a minimum age of 16 years; a signed informed consent; and an expected survival of greater than three months but less than one year as judged by an experienced oncologist. Al1 new patients seen by radiation therapy and medical oncology were eligible for accession to the study. Ten percent of the patients approached refused study assignment. Reviews early in the course of the study failed to demonstrate any apparent bias in 96
patient selection resulting from voluntary refusal. A total of 199 patients, 98 in the “intensive” and 101 in the non-intensive groups, were followed; at the close of the four-year study 139 had died. The nurse was chosen as the interface between the multidisciplinary team and patients receiving palliative therapy, since normal activities for nurses encompass many of the skills carried out by individual team members. Nurse practitioners with extensive experience in care of the patients with advance cancer were selected. Patients with an expected survival of less than three months were visited by nurses biweekly and those expected to live longer were visited monthly. The home visit by the nurses was primarily focused on attending to the needs of the patient, and interactions with family members were incidental to that task. In addition to providing physical care, much of the nurse’s time was spent in talking with the patient about their illness and its implications. The nurse frequently mobilized family and other social resources to meet the patient’s needs and also coordinated with the patient’s local physician. These nurses thus served in the well-known public health, or visiting nurse, role, with the differente that the project nurse had the benefit of a multidisciplinary health care team back-up resource. A “Protocol for Management of Pain” (8), available from the authors, was developed by the team and used by the nurses as part of their wide range of physical treatments and psychosocial interventions. This protocol was based upon sound pharmacological principles, many of which are often ignored. For example, oral methadone is an effective and longer lasting analgesic than other narcotics, and was used extensively in this program. Further, analgesics were usually prescribed on a regular time schedule, rather than on a p.r.n. basis, to ensure continuous relief rather than wide swings of degrees of discomfort. Once adequate pain relief has been achieved by a given daily regimen, the analgesic dose requirement may then be reduced, to avoid complications such as sedation or confusion. Anxiolytic agents were used when anxiety seemed to be a significant factor in pain. Tricyclic antidepressants were infrequently used because the depression syndrome was uncommon. Relaxation, hypnosis, and other pain-relieving techniques were not used because pain control was only one of many goals of the nurses’ visits and the visit rarely exceeded 30-45 minutes. Initial, on-study patient evaluation included the following instruments:
Prediction and Management of Pain in Patients
A modified Rotter Lotus of Control (1-E) scale, an eight-item self-report measure of internalexternal expectation of control (9). Higher scores indicate a person’s greater expectation of control from outside themselves. The Cornell Medical Index (CMI) M-R scales, a 51-item self-report measure of psychological symptoms (10). Higher scores indicate a greater degree of emotional disturbance. An
initial research data base (IRDB), which assessed by structured interview the patient’s medical, nutritional, physical, psychosocial, and religieus status.
The Karnofsky observer-rated status (11,13).
Performance Status (KPS), an 11-point scale of overall health
A “Pain Estimate” score described by Sternbach (12), a 0-100 self-report of pain at the present time. Follow-up data bases were gathered from both intensive and non-intensive patients by trained independent raters, using structured interviews in the patients’ homes, done at the same frequency as the nurses’ visits, which were based upon the patients’ prognosis. Follow-up interviews were carried out in the patients’ homes because earlier data had indicated greater validity of home vs. clinical evaluations for some measures (13). Initial and follow-up rater training took place during the pilot phase of the project using videotapes. At the end of the first year the raters had achieved acceptable interrater agreement on major scale items (Pearson correlation coefficient 2 0.70). The follow-up data bases covered the same areas as the initial research data base, including the self-reported “Pain Estimate.” In reporting results, pain levels over time are expressed as mean pain levels in groups of patients during blocks of time before death. Thirty-day blocks of time were used with larger numbers of patients in compared groups and 60-day blocks with smaller numbers of patients. (For each patient, the earliest pain leve1 in each 30-60-day time black was used.) Because patients came on study at varying stages of their illness, data analysis was done in terms of days before death, since the time of death is the clearest reference point for al1 patients. Mean pain scores of various groups of patients were compared for 30- or 60-day blocks of time before death, by Student’s t-test. Interactions between variables were examined by an analysis of variante.
with Advanced
Cancer
Results IntensivelNon-In
tensive Pain Comparisons
This report concerns the effect of the home nurse interventions on pain, together with some factors which seemed to influence pain in the patients studied. Results in other parameters are described in detail elsewhere (7). It should be briefly noted that the intensive and non-intensive patients did not differ in terms of length of survival, nor did these two groups differ in several other quality of life outcomes such as physical activity, nutrition, optimism, or overall health status as defined by the KPS (7). This report concerns the effect of the home nurse interventions on pain, together with some factors that seemed to influence pain in the patients studied. More than 80% of al1 the patients who died reported some degree of pain during their time on study, with an increasing frequency and severity as they approached death. Of the 139 patients who died during the study, 38 intensive and 45 non-intensive patients had a sufficient number of pain ratings (three or more) to compare the effectiveness of pain management over time in the two groups. A comparison of patient characteristics for these two groups demonstrated similarities in cancer diagnosis, sex, age, social class, and religious preferente (Table 1). The initial, on-study scores on the CMI, 1-E, and KI’S did not differ significantly between the intensive and non-intensive groups. When pain levels for these intensive and nonintensive groups were compared over time, in 30-day periods before death, the two groups had essentially the same mean pain scores until the last 90 days before death (Figure 1). In these last 90 days, the mean pain scores in the non-intensive group of patients continued to rise until death, whereas the mean pain scores in the intensive group of patients plateaued. During patients’ last 30 days of life, this differente in mean pain scores between the two groups approached customary statistical significante (P = 0.06).
Pain-Associated Variables To identify early predictors of pain problems, relationships between the on-study CM1 and scores and the pain scores over the course of disease were examined. A group of 55 patients, intensive and 33 non-intensive) were chosen on basis of having complete on-study CM1 and
the 1-E the (22 the 1-E 97
F. P. McKegney et al.
Table 1. Composition
of comparison
groupP Sex
Age (years)
Intensive (N = 40) Nonintensive (N = 43) High CM1 (N = 23) Low CM1 (N = 32) High 1-E (N = 34) Low 1-E (N = 21)
Type of cancer
Survival (days)
Mean
(SE)
M
F
Lung
Breast
GI
Other
Mean
(SE)
58.7 56.1 55.7 57.7
(1.7) (2.2) (2.6) (2.1)
21 27 10 23
19 16 13
20 14 8 19
6 6 5 1
8 9 3 4
6 14 7 8
345 309 327 299
(28) (24) (26) (40)
56.7 57.0
(1.8) (3.2)
20 13
13 14
4 2
5 2
12 3
296 334
(34) (40)
I: 8
‘NO differences significant at the P = 0.05 leve1 except: *females higher than males on the CMI, P < 0.05.
scores and at least one pain score while on study. The mean number of pain ratings per subject over time was 6.1.
differ in the 240-280-day and 180-120-day periods before death. However, the High CM1 group had higher mean pain scores than the Low CM1 group in the last 120 days before death, with this differente being statistically different in the last 60 days before death (P < 0.05). There was no significant interaction between the intensivelnon-intensive variable and the CM1 variable with regard to pain.
Comell Medical Index. Of these 55 patients, 23 had initially given 6 or more “yes” responses on the M-R section of the CMI, indicating a mild to severe emotional disturbance. This “cut score” of 6, slightly above the customary 4 or 5, was chosen in
Rotter Lotus of Control (I-E) Scale. Of the 55 patients, 34 initially gave 4 or more “external” responses on the 8-item 1-E scale, and 21 patients gave 0-3 external responses. There were no significant differences in demographic or disease characteristics between the two 1-E groups (Table 1). Mean pain scores in the 34 externally oriented High 1-E patients were compared with the mean pain scores in the more internally oriented 21 Low 1-E patients
order to increase the accuracy of positive prediction (14). There were no significant differences in demographic or disease characteristics between the two CM1 groups except that females scored higher than males (Table 1). Mean pain scores in these 23 “High CMI” patients were compared with the mean pain scores in the 32 “Low CMI” patients over time, in 60-day periods before death (Figure 2). Mean pain scores in the two CM1 groups did not
INTENSIVE
,et
(38)
NON-INTENSIVE
(451
O----
I S.E. +
I S. E.
T Figure 1. Mean pain scores (0-100) for “Intensive” and “Non-intensive” groups of patients at time on study and for each 30-day period before death. The “Intensive” group of patients has lower mean pain scores than the “Non-intensive” group over the ,last 90 days before death. The differente in the 30-O-day period approaches statistical significante (P = 0.06). *Not al1 patients had pain scores in al1 time periods.
ON
98
STUOY
-180
-ISO
-120 OAYS
-90 FROH
-60 DEATH
-30
0
Prediction
and Management
of Pain in Patients with Advanced Cancer
problems is a higher incidence than that in other studies, which may be due to our following patients until death rather than sampling a cross-section of a population with cancer. This high incidence clearly indicates the need for more attention and study than in the past. This is the first study examining the impact of standardized home intervention for pain problems in patients with cancer. The factor apparently most responsible for this lower pain in the “intensive” group over the last 90 days of life was reflected in the monitoring of nurse interventions. In the last 90 days of life approximately two-thirds of the nurse interventions involved management of pain: initiating pain medications, increasing doses of the same pain medication, and changing to or adding a more This demonstrated benefit by potent analgesic. home intervention strongly supports the current “Hospice” home care nursing movement, at least in terms of pain management. This better access to health care personnel needs to be combined with greater education in the use of pain management procedures for al1 professionals, especially appropriate pharmacologic intervention (8,16). A multidisciplinary team developed a Pain Management Protocol, available from the authors, with clear guidelines for appropriate and effective interventions in varying degrees of pain. The availability of such an instrument as an educational tool should improve professional responses to pain complaints, irrespective of setting.
over time, in 60-day periods before death (Figure 3). Mean pain scores in the High and Low 1-E groups were not different in the 240-180-day period. However, the High 1-E group had significantly higher mean pain scores than did the Low 1-E group in the 180-120-day and 120-60-day periods before death (P < 0.05). This differente persisted in the last 60 days before death but was no longer statistically significant at the 0.05 level. There was no significant interaction between the intensivelnon-intensive variable and the Rotter variable with regard to pain. The initial on-study CM1 and 1-E scores in the 55 patients showed a significant positive correlation, (Pearson r = 0.34, P = 0.003), with the more emotionally disturbed having a greater external lotus of control. However, the CM1 and 1-E scores showed no correlation with initial pain scores, with initial overall health status, as measured by the KI’S score, or with survival after entry on study.
Discussion Pain has traditionally been viewed as a cardinal “sign” of disease or inflammation (15). However, unlike swelling, redness, and heat, pain is not an objective sign but rather a subjective symptom. This time-honored misconception highlights our need to learn more about pain and its treatment, particularly in those illnesses in which pain is extremely common. Our finding that more than 80% of patients approaching death have significant pain
60
HIGH CHI123) -fl LOW
CHI
(321
50 Figure 2. Mean pain scores (0-100) for High CMI patients (M-R scales score 6-17) and Low CM1 patients (M-R scales 0-5) for each 60-day period before death. The High CM1 group (N = 23) has higher mean pain scores than the Low CMI group (N = 32) in the last two periods before death; the differente is significant (P < 0.05) in the 60-0-day period. “Not al1 patients had pain scores in al1 time periods.
?? ---+-
T
SE I SE
LO : 0 Co 30 z d 20
10
-
.
2lbo
-180
-120
DAYS FROM
-60
0
DEATH
99
F. P. McKegney
et al.
6C HIGH
1-E
134)
-ti
SE
LOW
1-E
Ik?11
?? ---+l
SE
51
Figure 3. Mean pain scores (0-100) for
High 1-E patients (4-7 external responses) and Low 1-E patients (0-3 external responses) for each 60 day period before death. The High 1-E group (N = 34) had higher mean pain scores than the Low 1-E group (N = 21) in the last three periods before death; these differences are significant (P < .05) in the 180-120-day and 120-60-day periods. *Nat al1 patients had pain scores in al1 time periods.
$0
-180
-120
-60
0
DAYS FROM DEATH
This prospective study also demonstrates that early screening evaluation of certain psychological parameters, such as emotional disturbance and independence or dependence upon the environment, can significantly predict pain problems later in the course of cancer. While the Rotter Lotus of Control Scale and the Cornell Medical Index M-R scores were not correlated with initial levels of pain in the population described, they did generally predict those patients who would have more pain problems later in the course of disease. Where there is limited availability of health care professionals skilled in pain management, those patients found to have more emotional disturbance and greater dependence on the environment may be those to whom more care is given. This study does confirm other, less wellestablished, observations that greater emotional disturbance and greater orientation to environmental influence are associated with more pain complaints (3,17). The finding that 1-E scores and CM1 ratings of emotional disturbance are interrelated also confirms previous observations (9). The lack of association between patient survival and the initial scores of emotional disturbances needs to be compared with the observations of others (19) who have reported such an association. Likewise, the finding that external or internal expectations of control are not associated with survival raises some question about treatment methods that emphasize independence and self-responsibility (20). For the benefit of future research, some difficul100
ties encountered in the present study must be mentioned. One major problem was that of gathering complete initial data bases: the observer-scored IRDB, and the patient-completed Cornell Medical Index, Pain Estimate and Lotus of Control scales. Two patient factors contributing to incomplete data collection were more serious physical illness and the presence of mild mental confusion (18). Observer factors included discomfort with asking upsetting questions about patients’ disabilities and unhappiness, and a reluctance to impose yet another burden on patients with progressive illness. To overcome these obstacles, which accompany the study of patients with serious illness, the data management system must be especially wel1 organized to optimize data collection. Another major problem in this study involved the trained observers who gathered follow-up data from patients in their homes. These observers were continually instructed to limit their activities to asking questions, observing behavior, and recording data. Early in the study, it became apparent that these home observers could not restrain themselves from offering help to patients and families when their needs became apparent in the process of data gathering. Although personnel changes were made later in the study, the human contact between the research observers and the patients in the home may have reduced differences in quality of life outcomes between the intensive and non-intensive groups. This hypothesis is supported by the fact that the major differente between the two groups
Prediction
was in pain management, the one problem which the nurses could address, but which the nonmedical observers could not. At any rate, the “observation effect” may wel1 be a major factor in studying the quality of life outcomes of different treatment interventions.
Conclusion This study’s design enables US to establish two comparable groups of patients for this prospective study of controlled health care interventions in patients with terminal cancer. It further found that home visits by specially trained nurses improved overall group quality of life, at least in terms of more adequate pain management. More refined studies should be carried out to confirm those risk factors that might identify early those patients who wil1 benefit most from these home interventions, if al1 patients are not able to receive such intensive home care.
Ref erences 1. Cox SS: Pain in terminal illness. Arizona Medicine
36:586-587, 1979
2. Saunders C: The treatment of intractable pain in terminal cancer. Proc R Sec Med 56:195-297, 1963 3. Pilowsky 1, Bond MR: Pain and its management in malignant disease. Psychosom Med 31:400-404,1969 4. Holden C: Pain, dying and the health care system. Science 203:484485, 1979 5. Merskey H, Spear FG: The concept of pain. J Psychosom Res 7:59-67, 1967 6. Marks RN, Sachar EJ: Undertreatment of medical inpatients with narcotic analgesics. Ann Intern Med 78:173-181, 1973 7. Yates JW, McKegney FF’, Visco G, Brown CS: Intensive rehabilitative support of the patient with advanced cancer through home health care. Proc Am Assoc Cancer Res 18:199, 1977
and Management
of Pain in Patients with Advanced Cancer
8. Bailey L, Yates JW: Protocol for Management of I’ain. Publication of Vermont Regional Cancer Center, 1976 9. Rotter JB: Generalized expectancies for internal versus external control or reinforcement. Psych Monographs 80:1-28, 1966 10. Seymour GE: The structure and predictive ability of the Cornell Medical Index for a normal sample. J. Psychosom Res 20:469478, 1976 ll. Karnofsky DA, Burchenal JH: The clinical evaluation of chemotherapeutic agents in cancer. In Macleod CM (ed). Evaluation of Chemotherapeutic Agents. New York, Columbia University Press, 1949 12. Sternbach RA: Pain Patients: Traits and Treatment. New York, Academie Press, 1974 13. Yates JW, Chalmer B, McKegney FP: Evaluation of patients with advanced cancer using the Karnofsky Performance Status. Cancer 45:228-232, 1980 14. Ransohaff DF, Feinstein AR: Problems of spectrum and bias in evaluating the efficacy of diagnostic tests. N Eng1 J Med 229:926-930, 1978 15. Florey H: Lectures on Genera1 Pathology. Philadelphia, W.B. Saunders, 1954, p 21 16. Shimm DS, Logue GL, Maltbie AA, Dugan S: Medical management of chronic cancer pain. JAMA 241:2408-2412, 1979 17. Wise TN, Hall WA, Wong 0: The relationship of cognitive styles and affective status of post-operative analgesic utilization. J Psychosom Res 22:513-518, 1978 18. Danies RK, Quinlan DM, McKegney FP, Kimball CP: Organic factors and psychological adjustment in advanced cancer patients. Psychosom Med 35:463471, 1973 19. Derogatis LR, Abeloff MD, Melisaratos N: Psychological coping mechanisms and survival times in metastatic breast cancer. JAMA 242:15041508, 1979 20. Simonton OC, Simonton S: Getting Wel1 Again. Los Angeles, JP Tarcher, 1978
Direcf reprint requesfs to: F. Patrick McKegney, M.D. Department of Psychiatry Medical Alumni Building University of Vermont College of Medicine Burlington, VT 05405
101