Barriers to Better Pain Control in Hospitalized Patients

Transcription

1 434 Journal of Pain and Symptom Management Vol. 17 No. 6 June 1999 Original Article Barriers to Better Pain Control in Hospitalized Patients Rebecca A. Drayer, BS, Jessica Henderson, BS, and Marcus Reidenberg, MD Departments of Pharmacology and Medicine, Cornell University Medical College, New York, NY Abstract Pain is often inadequately treated. To evaluate a common method of assessing pain and to identify some barriers to improving pain control, 50 hospitalized patients in pain, their nurses, and their physicians were interviewed about the pain experienced by the patients. The patients pain behavior was assessed and literature was reviewed to identify the risk for developing iatrogenic drug addiction in patients with no prior history of drug abuse. Doctors and nurses tended to assess pain intensity as less than the patients assessments. The doctor s and nurse s assessment of the same patient correlated poorly (r 0.21). Pain behavior correlated weakly (r s 0.36) with patients pain scores. Twenty-one patients wanted more pain medicine but often staff would not give more. A common reason was fear of addiction, notwithstanding a literature that iatrogenic addiction hardly ever occurs in usual hospitalized patients. Barriers to better pain control in acutely hospitalized patients include: (1) exaggerated fear of iatrogenic addiction, (2) an attitude among staff that patients exaggerate the intensity of their pain, (3) poor correlation between pain behavior and pain intensity that can mislead staff who rely on pain behavior to assess pain intensity, and (4) a lack of agreement between doctor and nurse in estimating the intensity of a patient s pain. A new way of thinking about patients in pain, assessing pain as either present or absent, is proposed for evaluation. If pain is present, one should attempt to ameliorate it in a manner consistent with the desire of the patient and the acceptability of adverse effects. J Pain Symptom Manage 1999;17: U.S. Cancer Pain Relief Committee, Key Words Pain, opioids, analgesics Introduction Health care professionals attitudes toward analgesic therapy were strongly influenced by an effort in the 1930s by The Rockefeller Foundation, the National Research Council, the U.S. Public Health Service, the U.S. Bureau of Address reprint requests to: Marcus Reidenberg, MD, 1300 York Avenue, Box 70, New York, NY 10021, USA. Accepted for publication: September 17, Narcotics, and several universities to solve the problem of drug addiction. A paper read at the annual session of the American Medical Association in June 1940, and subsequently published in the Association s journal stated, The use of narcotics in terminal cancer is to be condemned if it can possibly be avoided. Morphine and terminal cancer are in no way synonymous. Morphine use is an unpleasant experience to the majority of human subjects because of undesirable side effects. Dominant in the list of these unfortunate effects is addiction. 1 U.S. Cancer Pain Relief Committee, /99/$ see front matter Published by Elsevier, New York, New York PII S (99)

2 Vol. 17 No. 6 June 1999 Barriers to Better Pain Control 435 One effect of this attitude is that, despite much writing that pain should be adequately controlled with adequate doses of opioids when needed, 2 10 there has been the repeated documentation of undertreatment of acute pain, cancer pain, and pain in seriously ill hospitalized patients with various diagnoses. 9,11 16 One of many hypothesized reasons for the continued undertreatment of pain is that there is poor communication between doctors, nurses, and patients so that health care professionals are not aware of the severity of pain. While this may be true, it has also been observed that physicians and nurses are often aware of a patient s perception of pain intensity but think the patient overstates its intensity. 17 The present study was done to clarify the latter barrier to improving pain control and identify others that may compromise care in hospitalized medical patients. Methods Survey The study was approved by the institutional review board. Fifty acutely hospitalized patients on the medical service, and the nurses and house staff physicians caring for them were interviewed about the pain experienced by patients. The interviewer, a medical student, visited three medical floors and learned from the charge nurse which patients admitted during the previous 24 hours had pain as one of their admitting symptoms. She then interviewed the patient, and if pain was still present at the time of the interview, invited the patient to enter the study. Ten patients refused to participate. The nurse and house officer primarily responsible for the patient were also interviewed as soon after the patient interview as possible. This was within 24 hours in 48 of the 50 cases. The patients, nurses, and physicians were asked to rate the patient s pain on a 0 to 10 scale, with a value of 0 corresponding to no pain and a value of 10 corresponding to the worst pain imaginable. The answer do not know was accepted, and one nurse and five doctors responded in this way. Nurses and physicians were also asked to predict how the patient would rate his or her pain and to make a judgment about whether or not the patient should receive more opioid medication. An answer of unsure was also accepted for this question. If the staff member thought that the patient should not receive more analgesics, the reasons for this were elicited. The patients were also asked if they wanted more pain medicine. If they answered no, they were also asked to explain. In addition, a 5-point pain behavior scale was developed early in the study. The interviewer evaluated the patient s behavior during the interview and classified it into one of the following categories: 0: Patient is laughing or smiling; 1: Patient is not smiling, but does not appear withdrawn; 2: Patient appears withdrawn and unwilling to talk; 3: Patient has a pained facial expression; 4: Patient is writhing or screaming. Inter-rater agreement was assessed by one student interviewing the patient with both students present during this interview and independently classifying the patient s pain behavior. This was done during 5 interviews. The results were full agreement in 2 and differing by 1 scale point for the other 3. The behaviors of 45 of the 50 patients in this study were classified using this scale. The first 5 patients were entered into the study before this behavior scale was implemented. Data Analysis For analysis, the patients were classified into four groups based on the cause of their pain. There were 16 patients with cancer, 9 with sickle cell disease, 18 with miscellaneous lesions, and 7 with no identifiable lesion causing pain. The cancer patents and most in the miscellaneous group had chronic pain. Some with digestive system disease in the miscellaneous group and the patients in sickle cell crisis had acute pain. There was no procedure or treatment-related pain. All pain was due to disease. Associations between variables were tested using Spearman rank order correlation. Literature Search Because fear of addiction was often cited as a reason for limiting opioid dose, we performed a Medline search (1966 August 1996) using Knowledge Finder for papers indexed under analgesics with addiction and using Grateful Med under analgesics with pain with substance dependence. We also identified all papers in the Citation Index from 1993 to June 1996 having the paper by Porter and Jick 18 in

3 436 Drayer et al. Vol. 17 No. 6 June 1999 Fig. 1. Distribution of patients pain scores stratified by diagnoses. their reference list. Reference lists in all of the identified papers were reviewed for any title suggesting relevant papers. Articles in English, French, or German were considered. A total of 102 papers were retrieved and read. Studies were included if they presented data citing the number of patients with pain treated with opioids and the number of patients from that population who were identified as being addicted after the start of this treatment but for whom the authors had no evidence of a drug problem prior to this treatment. Since a definition of addiction was not included in most of these papers, all were included, accepting whatever the authors chose to consider addiction. Results The distribution of the patients pain scores is shown in Fig. 1. On average, the doctors and nurses believed that the pain was less intense than the patients said it was (Fig. 2). In addition, they thought that the patients would rate Fig. 3. Distribution of differences between patients pain rating and caregivers predictions of patients pain rating. The caregiver s prediction is subtracted from the patient s rating. There is a large variation in how intense the caregiver thinks the patient s pain really is and what the caregiver thinks the patient will say the pain intensity is. the pain as more severe than it really was (Fig. 3). We correlated pain behavior in the 45 patients in whom this observation was made with the patients pain scores (Fig. 4) and found a Spearman rank correlation coefficient (r s ) of (P 0.02), indicating that pain behavior correlates modestly with pain severity as expressed by the patient. There was no statistically significant correlation between the nurses and doctors assessments of the patients pain (Fig. 5). In fact, for 36% of the pairs of assessments, the ratings differed by 3 or more scale divisions. In 44 of the cases, both the doctor and nurse agreed that the patient was in pain and in 1 case, both agreed that there was no pain. Table 1 shows how many patients in each diagnostic category wanted more pain medicine and how many would be given more by the doctor or nurse. The overall numbers reveal Fig. 2. Distribution of differences between patients pain ratings and caregivers rating of patients pain for each patient. The caregiver s rating is subtracted from the patient s rating. Fig. 4. Correlation of the patient s pain rating with the observed pain behavior rating. Spearman rankorder correlation (r s ) 0.357, P 0.02 for N 45.

4 Vol. 17 No. 6 June 1999 Barriers to Better Pain Control 437 Fig. 5. Correlation of the nurse s rating of a patient s pain with the doctor s rating of the same patient s pain. r s 0.21, P 0.1. (N 44 since a do not know answer from doctor or nurse occurred 6 times). only some of the discordances. For example, 7 cancer patients in pain wanted more pain medicine but the doctor would not prescribe more for these 7 and the nurse thought only 3 of these 7 should receive more. The one patient for whom the doctor would prescribe more was not one of the 7 who wanted more. The nurse would also give more to one patient who did not want it, but this patient was different than the one who did not want more for whom the doctor would prescribe more. Similar discordances in different staff members evaluation of the same patient occurred for the other categories as well. All those who said that they did not want more/should not receive more pain medicine were asked to explain. Many different answers were given. Some examples of patient statements are: Taking drugs is habit forming (from a cancer patient); Feeling cloudy (from a back pain patient); Trying to balance side effects with pain relief (from a patient with HIV neuropathy). Some examples of nurse statements are: What he is getting now is enough or maybe even too much (for a sickle cell crisis patient); Seems pretty well controlled (for an AIDS patient with pain but no identified lesion); I m having a hard time relating to pain (for a patient with pancreatitis). Some examples of doctors statements are: We re not concerned about the pain itself. We re concerned about what s causing it (for a patient with avascular necrosis of the hip); I think pain is well controlled (for a patient with cancer); I don t know what the dependency issues are. There s something psychological going on here (for a patient with sickle cell crisis). For 11 of the 50 patients in pain, at least one of each group of 3 interviewees (doctor, nurse, and patient) spontaneously mentioned concern with addiction as the reason for not wanting or giving more pain medicine. For the metastatic cancer patients in pain, at least one of the people in 4 of the 16 sets of interviews expressed concern about addiction. This also occurred in 4 of the 9 sets of interviews for patients in sickle cell crisis, 1 of 18 with miscellaneous diagnoses, and 2 of the 7 with no identifiable diagnoses that cause pain. The results of the literature survey are described in Table 2. It appears that iatrogenic addiction, especially in hospitalized patients, has rarely been observed. Discussion We have confirmed our previous finding 17 that medical and nursing staff think that the severity of pain is less than medical patients say it Table 1 Does Patient Want More/Should Patient Receive More Pain Medicine? Patient Nurse Doctor More Enough More Enough Unsure More Enough Unsure Cancer SS Misc None Definitions: Cancer = metastatic cancer pain; SS = sickle cell crisis pain; Misc = a variety of diagnoses known to be associated with pain; None = no diagnosis of a known painful condition; More = patient wanted more or staff would give more pain medicine; Enough = patient did not want more or staff would not prescribe/give more pain medicine; Unsure = staff member was uncertain about prescribing/giving more pain medicine.

5 438 Drayer et al. Vol. 17 No. 6 June 1999 Table 2 Addiction Due to Medically Administered Opioids Author Ref. Frequency Percent Porter and Jick 18 4 out of 11,882 hospitalized patients 0.03% Pegelow 21 1 of 510 sickle cell patients 0.2% Brozovic et al of 610 sickle cell patients 0.0% Vichinsky et al of 101 sickle cell patients 3.0% Schug et al of 550 cancer patients 0.2% Kanner and Foley 25 0 of 86 cancer pain patients 0.0% Evans 26 7 of 130 cancer and nonmalignant pain patient 5.4% Taub 27 5 of 313 chronic nonmalignant pain patients 1.6% Zenz et al of 100 chronic nonmalignant pain patients 0.0% Moulin et al of 46 chronic nonmalignant pain patients 0.0% Portenoy and Foley 30 0 of 36 chronic nonmalignant pain patients 0.0% Kanner and Foley 25 0 of 15 chronic nonmalignant pain patients 0.0% France et al of 16 chronic nonmalignant pain patients 0.0% Chapman and Hill 32 0 of 26 bone marrow recipients 0.0% Perry and Heidrich 33 0 of 10,000 burn patients 0.0% Sun et al of 538 patients given opioids 0.0% Medina and Diamond 35 3 of 23 chronic headache patients 13.0% is. This is in agreement with the finding of Grossman et al. 19 In addition, staff also believe that patients exaggerate the severity of their pain. This may account for some of the discordance observed between physicians reported knowledge and attitudes, and their behavior. 36 We have found a modest correlation between the behavior of the patient and the patient s statement of pain severity. This is in agreement with the observations of Teske et al. 20 and should not be surprising because some people appear stoical whereas others are demonstrative. Furthermore, there is a degree of subjectivity in rating a patient s pain behavior. Our data suggested that using pain behavior to assess pain severity can be helpful but may be misleading. * We agree with the conclusion of Von Roenn et al. that poor pain assessment is a barrier to good pain management. 37 However, poor pain assessment may be due to failure of the common assessment method to work in an acute care hospital setting under present conditions of practice. Our finding of no correlation between the doctor s and nurse s assessment of pain in the same patient is striking. It indicates *The 5 patients with pain scores of 9 or 10 and pain behavior scores of 0 included 2 cancer patients, one with very strong denial and the other with a history of serious psychiatric illness; a sickle cell patient worried that the pain interviewer would lower his opioid dose; and two miscellaneous patients. that this method of pain assessment, while effective in a research or special care setting, does not function adequately for staff in an acute care general hospital environment today. A simpler method may be more useful in this environment with both the time pressures and the staff turnover which occurs. Assessing the intensity of pain experienced by patients in acute care general hospitals is a complex task. The patients are receiving analgesics and pain intensity may vary with the time since last dose. The patients often are being taken from their rooms for imaging and other studies, and procedures in the room or elsewhere are being done. Different hospital staff members interact with the patients in various ways, and information about the illness may be acquired over time. Any of these activities may change the sensory or the emotional component of pain and lead to a change in pain intensity. For these reasons, a meaningful assessment of an acutely hospitalized patient s pain intensity is a difficult task. Given the marked discordance among doctors, nurses, and patients in assessing the intensity of a patient s pain, we question whether a meaningful and useful routine assessment of pain intensity by house staff or nursing staff is feasible at the present time for patients admitted to a medical service in a general hospital. Assessing pain as either present or absent appears more feasible. Whether this will improve current patient care in this setting is not clear. Improvement will only occur if staff accept the

6 Vol. 17 No. 6 June 1999 Barriers to Better Pain Control 439 notion that if pain is present, one should attempt to ameliorate it unless there is a good reason not to. This concept would reduce the need to determine the intensity of the pain. One would intensify treatment for pain consistent with the desire of the patient and the acceptability of adverse effects. One result of present attitudes toward pain and the difficulty in assessing pain intensity is that pain is often inadequately treated Although the barriers to adequate pain management with opioid analgesics have been classified into 3 groups knowledge deficits, cultural and attitudinal barriers, and influence of drug regulatory authorities 38 we are not sure of the relative importance of these barriers in the hospitalized patients in our survey. The most frequent reason that more medication was not wanted or prescribed when a patient was still in pain was fear of addiction. This was so despite a literature survey that showed that the actual risk of causing iatrogenic addiction in hospitalized patients is very low. This survey could not apply a statistical metaanalysis because of the heterogeneity of the studies and variation in the definition of addiction. We have tried to group studies of patients with similar diagnoses together (Table 2). The Diagnostic and Statistical Manual of the American Psychiatric Association does not have an entry of drug addiction. Substance abuse is defined as a maladaptive pattern of substance use manifested by recurrent and significant adverse consequences related to the repeated use of substances. 39 Substance dependence is a cluster of cognitive, behavioral, and physiological symptoms indicating that the individual continues use of the substance despite significant substance-related problems. 39 The experience cited in Table 2 clearly indicates that patients with pain receiving proper medical care including adequate doses of opioid analgesic drugs rarely develop the drug-related problems that would allow these diagnoses. Current experience suggests that pain patients use opioid drugs to interact with their families, to get out of the hospital, and to go back to work. 40 The risk of iatrogenic addiction when the use of opioids is clearly indicated appears to be greatly exaggerated by some health care professionals and by some patients and their families. The impediment to optimal care produced by this misconception has been repeatedly emphasized, even by those in the regulatory community. 41 Additional research is needed to learn how this barrier of exaggerated fear of addiction can be overcome. The reduction of all the barriers we identified is needed for pain to be controlled to the maximum extent possible with the present technologies. Another method for assessing pain in acutely hospitalized patients may also be valuable. Simply denoting pain as present or absent may be sufficient for many patient care decisions when used in combination with assessment of adverse effects and patient desire. Given the present state of pain control in many acute care general hospitals, such a concept is worthy of evaluation. This type of research may be as important as studies designed to develop better technologies to treat pain. Acknowledgment This study was partially supported by grants from The Emily Davie and Joseph S. Kornfeld Foundation, Hoechst Marion Roussel, Inc., and Sandoz Pharmaceuticals Corporation. The authors appreciate the advice and encouragement given them by the members of the Pain Management Subcommittee of The New York Hospital Medical Board. References 1. Lee LE Jr. Medication in the control of pain in terminal cancer. JAMA 1941;116: Angell M. The quality of mercy. N Engl J Med 1982;306: Health and Public Policy Committee of the American College of Physicians. Drug therapy for severe, chronic pain in terminal illness. Ann Intern Med 1983;99: McGivney WT, Crooks GM. The care of patients with severe chronic pain in terminal illness. JAMA 1984;251: Foley KM. The treatment of cancer pain. N Engl J Med 1985;313: Jacox A, Carr DB, Payne R. New clinical-practice guidelines for the management of pain in patients with cancer. N Engl J Med 1994;330: Reidenberg MM, Portenoy RK. The need for an open mind about the treatment of chronic nonmalignant pain. Clin Pharmacol Ther 1994;55: Levy MH. Pharmacologic treatment of cancer pain. N Engl J Med 1995;335:

7 440 Drayer et al. Vol. 17 No. 6 June Marks RM, Sachar EJ. Undertreatment of medical inpatients with narcotic analgesics. Ann Intern Med 1973;78: Max M. Improving outcomes of analgesic treatment: Is education enough. Ann Intern Med 1990; 113: The National Institutes of Health. The integrated approach to the management of pain. National Institutes of Health Consensus Development Conference Statement 1986;6: No Zenz M, Willweber-Strumpf A. Opiophobia and cancer pain in Europe. Lancet 1993;341: Grossman SA. Undertreatment of cancer pain: barriers and remedies. Support Care Cancer 1993;1: Cleeland CS, Gonin R, Hatfield AK, Edmonson JH, Blum RH, Stewart JA, Pandya KJ. Pain and its treatment in outpatients with metastatic cancer. N Engl J Med 1994;330: Cleeland CS, Gonin R, Baez L, Loehrer P, Pandya KJ. Pain and treatment of pain in minority patients with cancer. Ann Intern Med 1997;127: Lynn J, Teno JM, Phillips RS, et al. Perceptions by family members of the dying experience of older and seriously ill patients. Ann Intern Med 1997;126: Krivo S, Reidenberg MM. Assessment of patients pain. N Engl J Med 1996;334: Porter J, Jick H. Addiction rare in patients treated with narcotics. N Engl J Med 1980;302: Grossman SA, Sheidler VR, Swedeen K, Mucenski J, Piantadosi S. Correlation of patient and caregiver ratings of cancer pain. J Pain Symptom Manage 1991;6: Teske K, Daut RL, Cleeland CS. Relationship between nurses observations and patients self-reports of pain. Pain 1983;16: Pegelow CH. Survey of pain management therapy provided for children with sickle cell disease. Clin Ped 1992;31: Brozovic M, Davies SC, Yardumian A, Bellingham A, Marsh G, Stephens AD. Pain relief in sickle cell crisis. Lancet 1986;2: Vichinsky EP, Johnson R, Lubin BH. Multidisciplinary approach to pain management in sickle cell disease. Am J Ped Hematol Oncol 1982;4: Shug SA, Zech D, Grond S, Jung H, Meuser T, Stobbe B. A long-term survey of morphine in cancer pain patients. J Pain Symptom Manage 1992;7: Kanner RM, Foley KM. Patterns of narcotic drug use in a cancer pain clinic. Ann NY Acad Sci 1981; 362: Evans PJD. Narcotic addiction in patients with chronic pain. Anaesth 1981;36: Taub A. Opioid analgesics in the treatment of chronic intractable pain of non-neoplastic origin. In: Kitahata LM, Collins JG, eds. Narcotic analgesics in anesthesiology. Baltimore: Williams and Wilkins, Zenz M, Strumpf M, Tryba M. Long-term oral opioid therapy in patients with chronic nonmalignant pain. J Pain Symptom Manage 1992;7: Moulin DE, Iezzi A, Amireh R, Sharpe WKJ, Boyd D, Merskey H. Randomized trial of oral morphine for chronic non-cancer pain. Lancet 1996; 347: Portenoy RK, Foley KM. Chronic use of opioid analgesics in non-malignant pain: report of 38 cases. Pain 1986;25: France RD, Urban BJ, Keefe FJ. Long-term use of narcotic analgesics in chronic pain. Soc Sci Med 1984;19: Chapman CR, Hill HF. Prolonged morphine self-administration and addiction liability. Cancer 1989;63: Perry S, Heidrich G. Management of pain during debridement: a survey of U.S. burn units. Pain 1982;13: Sun WZ, Chen TL, Fan SZ, Peng WL, Wang MS, Haung FY. Can cancer pain attenuate the physical dependence on chronic long-term morphine treatment? J Formosan Med Assoc 1992;91: Medina JL, Diamond S. Drug dependency in patients with chronic headaches. Headache 1977;17: Levin ML, Berry JI, Leiter J. Management of pain in terminally ill patients: physician reports of knowledge, attitudes, and behavior. J Pain Symptom Manage 1998;15: Von Roenn JH, Cleeland CS, Gonin R, Hatfield AK, Pandya KJ. Physician attitudes and practice in cancer pain management: a survey from the Eastern Cooperative Oncology Group. Ann Intern Med 1993;119: Hill CS Jr. The barriers to adequate pain management with opioid analgesics. Semin Oncol 1993; 20(suppl 1): Task Force on DSM-IV. Diagnostic and statistical manual of mental disorders: DSM-IV. Washington, DC: American Psychiatric Association, 1994: Friedman DP. Perspectives on the medical use of drugs of abuse. J Pain Symptom Manage 1990; 5(suppl): International Narcotics Control Board. Availability of opiates for medical needs. United Nations Publication E/INCR/1995/1/Supp.1, New York, 1996.