Advertisement

Frequency and Predictors of Patient Deviation From Prescribed Opioids and Barriers to Opioid Pain Management in Patients With Advanced Cancer

Open AccessPublished:September 03, 2012DOI:https://doi.org/10.1016/j.jpainsymman.2012.02.023

      Abstract

      Context

      Approximately 80% of patients with advanced cancer report pain and receive opioids. Information is limited about deviations from prescribed opioid doses and barriers to pain control, but poor opioid adherence has been reported in 49%–70% of patients.

      Objectives

      To evaluate the frequency and severity of self-reported opioid deviation and barriers to opioid pain management in outpatients with advanced cancer.

      Methods

      We surveyed 198 patients and collected pain scores (0–10), prescribed opioid dose, confidential patient-reported opioid prescription dose and intake (as long as there was no severe opioid deviation), barriers to pain management (Barriers Questionnaire-II [BQ-II]) scores, and adherence scores. Opioid deviation was defined as <70% or >130% of the prescribed dose.

      Results

      Median patient age was 55 years; 91 (46%) were female. Median pain intensity and morphine equivalent daily dose were 4 (interquartile range=3–7) and 120 mg (interquartile range=45–270 mg), respectively. Prescribed and patient-reported prescribed doses were highly correlated for regular (r=0.90, P<0.001) and regular plus breakthrough opioid intake (r=0.94, P<0.001). Nineteen (9.6%) patients deviated. Deviation was more frequent in males (P=0.039) and nonwhites (P=0.0270). Nonwhite patients had higher scores on the BQ-II than white patients (P=0.038). Low adherence scores were significantly associated with higher BQ-II scores (1.99±0.80) for lower motivation score vs. 1.61±0.77 for higher score, P=0.007; and 2.13±0.79 for lower knowledge score vs. 1.57±0.72 for higher score, P=0.001.

      Conclusion

      Very few patients reported dose deviations, which were mostly toward lower dose. More research is necessary to better characterize the frequency and predictors of opioid deviation in this population.

      Key Words

      Introduction

      Pain is one of the most frequent and distressing symptoms among patients with advanced cancer; 75%–80% report pain.
      • van den Beuken-van Everdingen M.H.
      • de Rijke J.M.
      • Kessels A.G.
      • et al.
      Prevalence of pain in patients with cancer: a systematic review of the past 40 years.
      Because of their effectiveness, opioid analgesics are the mainstay of cancer pain management.
      • Brescia F.J.
      • Adler D.
      • Gray G.
      • et al.
      Hospitalized advanced cancer patients: a profile.
      • Portenoy R.K.
      • Lesage P.
      Management of cancer pain.
      Unfortunately, there is limited information about how cancer patients use the medications prescribed by their physicians, but poor opioid adherence has been reported in 51%–70% of patients.
      • Abbas S.Q.
      • Abbas Z.
      Is opiate compliance a problem in cancer pain? A survey of health-care professionals' views.
      • Tzeng J.I.
      • Chang C.C.
      • Chang H.J.
      • Lin C.C.
      Assessing analgesic regimen adherence with the Morisky Medication Adherence Measure for Taiwanese patients with cancer pain.
      In recent years, there has been increased awareness of the major role prescription opioids have in cases of chemical coping, emergency room visits, and death from drug overdose.
      • Bohnert A.S.
      • Valenstein M.
      • Bair M.J.
      • et al.
      Association between opioid prescribing patterns and opioid overdose-related deaths.
      • Okie S.
      A flood of opioids, a rising tide of deaths.
      • Paulozzi L.J.
      • Budnitz D.S.
      • Xi Y.
      Increasing deaths from opioid analgesics in the United States.
      • McLellan A.T.
      • Turner B.
      Prescription opioids, overdose deaths, and physician responsibility.
      One study found that patients receiving opioids for cancer pain have a higher risk of receiving an overall higher opioid dose, and among these patients, opioid dose is associated with a higher risk of overdose death.
      • Bohnert A.S.
      • Valenstein M.
      • Bair M.J.
      • et al.
      Association between opioid prescribing patterns and opioid overdose-related deaths.
      However, other authors have reported consistent underutilization of opioids by patients with cancer pain,
      • Tzeng J.I.
      • Chang C.C.
      • Chang H.J.
      • Lin C.C.
      Assessing analgesic regimen adherence with the Morisky Medication Adherence Measure for Taiwanese patients with cancer pain.
      • Valeberg B.T.
      • Miaskowski C.
      • Hanestad B.R.
      • et al.
      Prevalence rates for and predictors of self-reported adherence of oncology outpatients with analgesic medications.
      • Liang S.Y.
      • Yates P.
      • Edwards H.
      • Tsay S.L.
      Factors influencing opioid-taking self-efficacy and analgesic adherence in Taiwanese outpatients with cancer.
      • Lai Y.H.
      • Keefe F.J.
      • Sun W.Z.
      • et al.
      Relationship between pain-specific beliefs and adherence to analgesic regimens in Taiwanese cancer patients: a preliminary study.
      leading to insufficient relief of pain. It has been suggested that patients' reluctance to report pain or use analgesics prevents optimal pain management. Over the past 15 years, there has been a growing literature exploring patient-related barriers to pain management.
      • Bagcivan G.
      • Tosun N.
      • Komurcu S.
      • Akbayrak N.
      • Ozet A.
      Analysis of patient-related barriers in cancer pain management in Turkish patients.
      • Chang M.C.
      • Chang Y.C.
      • Chiou J.F.
      • Tsou T.S.
      • Lin C.C.
      Overcoming patient-related barriers to cancer pain management for home care patients. A pilot study.
      • Chung T.K.
      • French P.
      • Chan S.
      Patient-related barriers to cancer pain management in a palliative care setting in Hong Kong.
      • Gunnarsdottir S.
      • Donovan H.S.
      • Serlin R.C.
      • Voge C.
      • Ward S.
      Patient-related barriers to pain management: the Barriers Questionnaire II (BQ-II).
      • Gunnarsdottir S.
      • Donovan H.S.
      • Ward S.
      Interventions to overcome clinician- and patient-related barriers to pain management.
      • Gunnarsdottir S.
      • Serlin R.C.
      • Ward S.
      Patient-related barriers to pain management: the Icelandic Barriers Questionnaire II.
      • Jacobsen R.
      • Moldrup C.
      • Christrup L.
      • Sjogren P.
      Patient-related barriers to cancer pain management: a systematic exploratory review.
      • Lin C.C.
      • Ward S.E.
      Patient-related barriers to cancer pain management in Taiwan.
      • Ward S.E.
      • Goldberg N.
      • Miller-McCauley V.
      • et al.
      Patient-related barriers to management of cancer pain.
      • Ward S.E.
      • Hernandez L.
      Patient-related barriers to management of cancer pain in Puerto Rico.
      Outpatient cancer palliative care programs provide multidimensional and interdisciplinary assessment of multiple symptoms by specially trained nurses, physicians, counselors, and other health care professionals. Patient-reported outcomes are regularly used for the screening of symptom distress, cognition, and chemical coping.
      • Chang V.T.
      • Hwang S.S.
      • Feuerman M.
      Validation of the Edmonton Symptom Assessment Scale.
      • Buchsbaum D.G.
      • Buchanan R.G.
      • Centor R.M.
      • Schnoll S.H.
      • Lawton M.J.
      Screening for alcohol abuse using CAGE scores and likelihood ratios.
      • Lawlor P.G.
      • Nekolaichuk C.
      • Gagnon B.
      • et al.
      Clinical utility, factor analysis, and further validation of the memorial delirium assessment scale in patients with advanced cancer: assessing delirium in advanced cancer.
      • Kirsh K.L.
      • Jass C.
      • Bennett D.S.
      • Hagen J.E.
      • Passik S.D.
      Initial development of a survey tool to detect issues of chemical coping in chronic pain patients.
      The objective of this study was to evaluate the frequency and severity of self-reported opioid dose deviation and barriers to opioid pain management in patients with advanced cancer at an outpatient supportive care center operated by a palliative care team.

      Materials and Methods

      The Institutional Review Board at the University of Texas M. D. Anderson Cancer Center approved this study, and all patients gave written informed consent.

      Patients

      Patients who attended the supportive care clinic for a follow-up visit between May 26, 2010 and September 23, 2010 were screened and subsequently approached if deemed eligible for this study. Inclusion criteria included a diagnosis of advanced cancer, age 18 years or older, and a prescription for regular and/or as-needed oral opioids for cancer pain written on the visit immediately before the one during which the patient was enrolled. Patients with impaired cognition or who did not speak English were excluded. Patients whose caregivers managed the administration of the opioids were excluded, as the analysis was focused on the patients', and not the caregivers', knowledge, attitudes, and beliefs about opioid use.
      Patients were enrolled in the study after providing informed consent. Patients were informed that all information would be kept confidential and not shared with the physician unless the investigators found that patients might be at serious risk for harm because of severe opioid deviation.

      Setting

      The supportive care clinic's interdisciplinary team is led by board-certified physicians. Other important members include a palliative care-trained registered nurse, pharmacist, nutritionist, chaplain, social worker, psychiatric nurse counselor, and wound care nurse. The types of clinic visits are new consultations, follow-up visits, and walk-in visits for symptom management, including pain, and counseling patients with psychosocial distress, including goals of care planning for patients with advanced cancer. All patients are initially assessed by the palliative care-trained registered nurse using the Edmonton Symptom Assessment System, Memorial Delirium Assessment Scale, and the CAGE questionnaire. The findings are discussed with the physician, who then conducts an interview with the patient and family and performs a physical examination. After the patient assessment is complete, the physician refers the patient to other interdisciplinary team members as necessary. Telephone care is provided by the nursing staff, and on nights and weekends, patients may contact the supportive care team through the on-call pager.

      Review of Medical Records

      The research assistant collected the following patient data: age, gender, ethnicity, marital status, employment status, profession, educational level, cancer diagnosis, and insurance type. When the information was not available from the medical record, patients were asked during the interview process. From the previous clinic visit, the Edmonton Symptom Assessment System pain intensity and physician opioid prescription data were collected. The estimated morphine equivalent daily dose (MEDD) was calculated by one of the investigators using an equianalgesic conversion table (Appendix).

      Interviews

      Interviews were conducted in a private room in the outpatient supportive care clinic by the research assistant, lasted approximately 20 minutes, and comprised the following questions.
      • 1.
        Patient-reported opioid prescription: Patients were asked, “What pain medications have you been told to take? What is the dose? What is the interval or frequency of administration?”
      • 2.
        Patient-reported opioid use: Patients were asked, “What medications are you taking? What is the dose? What is the interval or frequency of administration?”

      Self-Reported Questionnaire

      After the interview was completed, patients received a packet with a questionnaire and an envelope. They were left alone to answer the questions, which took approximately 20 minutes. The questionnaire was returned to the study coordinator after completion. The questionnaire contained the following assessment tools:

      Patient-Related Barriers to Pain Management (Barriers Questionnaire-II)

      The Barriers Questionnaire (BQ) was developed to measure the extent to which patients have concerns about reporting pain and using analgesics based on erroneous beliefs and misconceptions about pain and pain medications. The original questionnaire, developed in 1993, was revised in 2002. We used this revised version, the BQ-II.
      • Gunnarsdottir S.
      • Donovan H.S.
      • Serlin R.C.
      • Voge C.
      • Ward S.
      Patient-related barriers to pain management: the Barriers Questionnaire II (BQ-II).
      • Ward S.E.
      • Goldberg N.
      • Miller-McCauley V.
      • et al.
      Patient-related barriers to management of cancer pain.
      It comprises 27 items addressing eight different barriers: fear of addiction, fatalism (pain is an inevitable consequence of cancer), concerns about tolerance, the desire to be a “good” patient (good patients do not complain of pain), fear of distraction (treatment of pain distracts from treating the cancer), side effects, fear that pain medications may impair the immune system, and fear of losing pain as a marker of disease progression. Patients rate their agreement with each statement on a scale of 0–5. Higher scores are associated with higher patient-related barriers and inadequate analgesic use. The BQ-II has an internal consistency of 0.89, and in the validation study, BQ-II scores were related to measures of pain intensity and duration, mood, and quality of life, suggesting good validity.
      • Gunnarsdottir S.
      • Donovan H.S.
      • Serlin R.C.
      • Voge C.
      • Ward S.
      Patient-related barriers to pain management: the Barriers Questionnaire II (BQ-II).
      It has been validated for cancer patients.
      • Gunnarsdottir S.
      • Donovan H.S.
      • Serlin R.C.
      • Voge C.
      • Ward S.
      Patient-related barriers to pain management: the Barriers Questionnaire II (BQ-II).

      Modified Morisky Medication Adherence Scale

      The Modified Morisky Scale is a simple and validated four-item questionnaire developed to estimate medication adherence and assess adherence-related issues in studies of hypertension.
      • Morisky D.E.
      • Green L.W.
      • Levine D.M.
      Concurrent and predictive validity of a self-reported measure of medication adherence.
      • Prado Jr., J.C.
      • Kupek E.
      • Mion Jr., D.
      Validity of four indirect methods to measure adherence in primary care hypertensives.
      It also has been used in other settings, such as HIV and Parkinson's disease.
      • Elm J.J.
      • Kamp C.
      • Tilley B.C.
      • et al.
      Self-reported adherence versus pill count in Parkinson's disease: the NET-PD experience.
      • Sodergard B.
      • Halvarsson M.
      • Lindback S.
      • et al.
      Differences in adherence and motivation to HIV therapy–two independent assessments in 1998 and 2002.
      • Sodergard B.
      • Halvarsson M.
      • Tully M.P.
      • et al.
      Adherence to treatment in Swedish HIV-infected patients.
      Patients are asked to answer “yes” or “no” to the following questions: (1) Do you ever forget to take your pain medications?, (2) Are you careless at times about taking your pain medicines?, (3) When you feel better, do you sometimes stop taking your medicine?, and (4) Sometimes, if you feel worse when you take your pain medicine, do you stop taking it? The Modified Morisky four-item scale has been widely used for assessing patient adherence, and two new items were added by the Case Management Society of America to recognize patient understanding of medication benefits and refill behavior:

      Case Management Society of America. Case management adherence guidelines. 2004. Available from www.cmsa.org/portals/0/pdf/CMAG.pdf. Accessed December 2011.

      • Powell S.K.
      The case management adherence guidelines.
      (5) Do you know the long-term benefit of taking your medicine as told to you by your doctor or pharmacist? and (6) Sometimes do you forget to refill your prescription medicine on time? These modifications to the original questionnaire were made to qualify whether low adherence was related to knowledge or motivation. A knowledge domain score was computed by the sum of Questions 3, 4, and 5 and a motivation domain score by the sum of Questions 1, 2, and 6. A score of 0 or 1 for motivation and knowledge domains was considered as a low score and 3 or 4 as a high score. This scale has been widely used for assessing patients' adherence, showing acceptable validity and reliability.

      Case Management Society of America. Case management adherence guidelines. 2004. Available from www.cmsa.org/portals/0/pdf/CMAG.pdf. Accessed December 2011.

      • Powell S.K.
      The case management adherence guidelines.
      • Rudd P.
      The measurement of compliance: medication taking.

      Alcoholism Screening (CAGE Questionnaire)

      The CAGE questionnaire is a simple, four-item, validated tool to screen for a recent history of alcoholism, which is already incorporated in our palliative care routine assessment.
      • Ewing J.A.
      Detecting alcoholism. The CAGE questionnaire.
      Two positive answers have more than 85% sensitivity and 90% specificity for the diagnosis of alcohol abuse and/or dependence.
      • Mayfield D.
      • McLeod G.
      • Hall P.
      The CAGE questionnaire: validation of a new alcoholism screening instrument.
      • Liskow B.
      • Campbell J.
      • Nickel E.J.
      • Powell B.J.
      Validity of the CAGE questionnaire in screening for alcohol dependence in a walk-in (triage) clinic.
      Our group has previously reported that the frequency of CAGE-positive subjects in two separate studies of palliative care patients ranged from 13% to 28% and the mean MEDD on Day 2 was significantly higher among the CAGE-positive patients.
      • Bruera E.
      • Moyano J.
      • Seifert L.
      • et al.
      The frequency of alcoholism among patients with pain due to terminal cancer.

      Computation of MEDD

      While patients were completing the self-reported questionnaire, the research assistant contacted a physician investigator who was not involved in the care of the patient. Using the opioid prescription information documented in the medical record, the physician investigator computed the prescribed MEDD range. This was done using the equianalgesic conversion table used in our daily clinical practice at M. D. Anderson.
      • Reddy S.K.
      • Nguyen L.
      Pain management.
      The lower limit of the prescribed range was defined as the MEDD of the regular opioid alone, and the upper limit was defined as the MEDD of the regular opioid plus four breakthrough doses. The investigator also computed the patient-reported MEDD based on the opioid use (sustained-release and breakthrough) reported by the patient during the face-to-face interview with the research assistant.

      Statistical Analyses

      Our primary objective was to determine the frequency with which patients with advanced cancer confidentially self-reported the over- or underuse of opioid analgesics. We defined deviation as a patient-reported MEDD that falls 30% above or below the prescribed MEDD range. The range of 30% was chosen because this is what is recommended in the clinical setting when dose changes are needed.
      • Reddy S.K.
      • Nguyen L.
      Pain management.
      Severe deviation from the prescribed dose was defined as a patient-reported use of double or more or one-half or less of the prescribed dose range. If a patient was found to have a severe deviation, the responsible palliative care physician was notified to assist in the patient's care. We expected that approximately 67% of patients would self-report within the prescribed MEDD range and the remaining 33% would report deviations equally distributed between underuse and overuse (16.5% for each group). If this assumption was correct, by enrolling 200 patients in the study, we would be able to estimate each of the proportions of persons who under- or overuse opioids to within a 95% CI of 16.5±5.5 or 11–22. We also would be able to estimate the proportion of nondeviations (expected to be 67%), with a 95% CI of 60–74.
      Our secondary objective was to correlate the actual percentage of misuse of opioid analgesics (deviation from the prescribed dose) with the variables of knowledge and motivation (Modified Morisky Scale), alcohol abuse/chemical coping (CAGE), and patient-related barriers (BQ-II). We used Pearson correlation coefficients to estimate these associations. Given a sample size of 200, we were able to declare as significant correlation coefficients that were 0.2 or greater (or −0.2 or less), assuming a two-sided significance level of 0.05 and 80% power. Spearman correlation coefficients were used if the data were not approximately normally distributed. We used logistic regression analyses to determine if secondary variables were associated with the categorized use/misuse of opioid analgesics. In addition, we summarized the results for all variables by the category of opioid use (deviation toward underuse, deviation toward overuse, or nondeviation).

      Results

      A convenience sample of 198 patients with advanced cancer took part in this study. Participation was high as demonstrated by completion of accrual within approximately four months (May 26, 2010 to September 23, 2010) in a small clinic with only two physicians. Unfortunately, because of a methodological error by the data coordinator, data on the number of patients who did not want to participate were not collected. Patient demographics are reported in Table 1. Of the 198 evaluable patients, most did not deviate from the prescribed opioid dose (Fig. 1). Nineteen (9.6%) patients deviated: 11 (6%) used lower-than-prescribed and eight (4%) used higher doses. Only five (2.5%) cases required notification of the palliative care physician for severe opioid deviation, and all reported less opioid use than the prescribed dose.
      Table 1Patient Demographic and Clinical Characteristics (n=198)
      Patient Characteristicsn (%)
      Female107 (54)
      Ethnicity
       Caucasian148 (74)
       African American32 (16)
       Other18 (9)
      Median age (range) (years)55 (22–82)
      Type of cancer
       Gastrointestinal41 (20)
       Breast35 (17)
       Lung30 (12)
       Gynecologic23 (11)
       Genitourinary21 (10)
       Head and neck14 (7)
       Sarcoma14 (7)
       Other20 (9)
      CAGE positive20 (10)
      Median ESAS pain intensity (IQR)4 (3–7)
      ESAS=Edmonton Symptom Assessment System; IQR=interquartile range.
      Figure thumbnail gr1
      Fig. 1Frequency of confidential patient-reported use of opioids (n=198).
      Opioid deviation was more frequent in males (P=0.039) and nonwhites (P=0.027). There were no other associations between opioid deviation and demographic, socioeconomic (age, gender, ethnicity, marital status, employment status, profession, educational level, cancer diagnosis, and insurance type), or clinical characteristics (Table 2). There was no variation between the prescribed dose and the patient-reported prescription in 168 (85%, 95% CI 80–90) patients. There was no variation between the prescribed dose and the patient-reported intake in 157 (80%, 95% CI 74–86) patients. For 133 (67%, 95% CI 60.5–73.5) patients, the patient-reported intake for the regular and the breakthrough medication was within ±30% of the prescribed doses for the regular and the breakthrough medication. The MEDD of prescribed doses correlated highly with both the MEDD of the patient-reported prescription and the MEDD of the opioid intake (Table 3).
      Table 2Comparison of Opioid Deviators and Nondeviators
      Patient CharacteristicsDeviators n=19 (n [%])Nondeviators n=179 (n [%])P
      Age (mean [SD])52 (13)55 (12)0.427
      Wilcoxon two-sample test.
      Male gender13 (68)78 (44)0.039
      Chi-squared test.
      Ethnicity
       Caucasian10 (53)138 (977)
       African American6 (32)26 (15)
       Other3 (16)15 (8)0.027
      Chi-squared test.
      (Caucasians vs. non-Caucasians)
      CAGE positive3 (16)16 (10)0.416
      Fisher's exact test.
      Level of education
       High school education or less5 (26)42 (25)0.714
      Fisher's exact test.
       Some college or vocational school8 (42)57 (34)
       Completed bachelor's or higher4 (21)58 (35)
       Did not respond2 (11)22 (13)
      Profession
       Professional13 (68)92 (52)0.150
      Fisher's exact test.
       Administrative and management033 (18)
       Laborers and workers3 (16)20 (11)
       Other1 (5)20 (11)
       Did not respond2 (11)14 (8)
      Employment
       Working4 (21)40 (22)0.289
      Fisher's exact test.
       Retired7 (37)48 (27)
       Homemaker1 (5)13 (7)
       Unemployed030 (17)
       Other6 (32)43 (24)
       Did not respond1 (5)5 (3)
      Marital status
       Married13 (68)116 (65)0.627
      Fisher's exact test.
       Divorced/separated3 (16)32 (18)
       Single3 (16)18 (10)
       Widowed013 (7)
      Insurance
       Private insurance/self-pay14 (74)110 (61)0.295
      Chi-squared test.
       Government5 (26)69 (39)
      a Wilcoxon two-sample test.
      b Chi-squared test.
      c Fisher's exact test.
      Table 3Morphine Equivalent Daily Dose
      Prescribed and Reported Opioidmg/day (IQR)Correlation With Prescribed Dose
      Regular
       Prescribed dose120 (45–270)
       Patient-reported prescribed dose100 (40–270)r=0.90, P<0.0001
       Patient opioid intake100 (30–250)r=0.94, P<0.0001
      Regular+breakthrough
       Prescribed dose194 (90–360)
       Patient opioid intake160 (65–330)r=0.94, P<0.0001
      The overall score on the BQ-II was 1.67 (±0.76). There was a high level of motivation and knowledge as demonstrated by both scores from the Modified Morisky Scale (Table 4). Nonwhite patients had higher scores on the BQ-II than white patients: mean scores 2.16 (±0.93) for other races, 1.76 (±0.81) for African Americans, and 1.60 (±0.72) for whites (P=0.038). In subgroup analyses, we found that low Modified Morisky Scale scores were significantly associated with higher BQ-II scores: mean score for the BQ-II was 1.99 (±0.80) for lower score for motivation domain vs. 1.61 (±0.77) for higher score (P=0.007) and mean score for the BQ-II was 2.13 (±0.79) for lower score for knowledge domain vs. 1.57 (±0.72) for higher score (P=0.001). There was no association between high BQ-II scores and opioid deviation toward lower than the prescribed dose. Opioid deviation toward higher than the prescribed dose was significantly associated with high BQ-II scores, with a mean BQ-II score of 2.17 (±0.63) vs. 1.65 (±0.77) (P=0.047).
      Table 4Barriers Questionnaire-II and Modified Morisky Medication Adherence Scale
      Barriers Questionnaire II
      SubscaleMedian (IQR)
       Physiological effects1.75 (0.94–2.5)
       Fatalism1.66 (1.00–2.00)
       Communication0.83 (0–1.38)
       Harmful effects2.50 (1.50–3.00)
       Total1.63 (1.05–2.19)
      Modified Morisky Medication Adherence Scalen (%)
      Do you ever forget to take your pain medications? (Question 1)115 (58)
      Are you careless at times about taking your pain medicines? (Question 2)168 (85)
      When you feel better, do you sometimes stop taking your medicine? (Question 3)147 (74)
      Sometimes, if you feel worse when you take your pain medicine, do you stop taking it? (Question 4)153 (77)
      Do you know the long-term benefit of taking your medicine as told to you by your doctor or pharmacist? (Question 5)152 (77)
      Sometimes do you forget to refill your prescription medicine on time? (Question 6)152 (77)
      Motivation domain score
      Sum of Questions 1, 2, and 6; lowest score describes lowest motivation.
       011 (5)
       120 (10)
       269 (35)
       393 (48)
      Knowledge domain score
      Sum of Questions 3, 4, and 5; lowest score describes lowest knowledge.
       03 (1)
       129 (14)
       258 (29)
       3103 (53)
      a Sum of Questions 1, 2, and 6; lowest score describes lowest motivation.
      b Sum of Questions 3, 4, and 5; lowest score describes lowest knowledge.
      Of the 19 cases of deviation, there were three (16%) CAGE-positive patients who had severe deviation toward less than the prescribed doses, and the palliative care physicians responsible were notified (Table 2). Table 2 summarizes the comparison of opioid deviators and nondeviators.

      Discussion

      In this study, we found that most patients (90%) reported no opioid deviation, which is a remarkable finding. Patients showed excellent understanding of their opioid prescriptions as demonstrated by the very high agreement between the prescribed dose and the patient-reported prescription, which may explain the very small number of deviators. Men and minority patients were more likely to deviate from the prescribed dose. However, other patient characteristics were not predictive of deviation, including cancer type, age, level of education, marital status, occupation, employment status, religious affiliation, or insurance type. Minority patients have been reported, in other settings and oncology, to have a lower adherence,
      • Bickell N.A.
      • Wang J.J.
      • Oluwole S.
      • et al.
      Missed opportunities: racial disparities in adjuvant breast cancer treatment.
      • Blackman D.J.
      • Masi C.M.
      Racial and ethnic disparities in breast cancer mortality: are we doing enough to address the root causes?.
      • Hershman D.
      • McBride R.
      • Jacobson J.S.
      • et al.
      Racial disparities in treatment and survival among women with early-stage breast cancer.
      as have men.
      • Kirigaya D.
      • Nakayama T.
      • Ishizaki T.
      • Ikeda S.
      • Satoh T.
      Management and treatment of osteoporosis in patients receiving long-term glucocorticoid treatment: current status of adherence to clinical guidelines and related factors.
      • Turner B.J.
      • Hollenbeak C.S.
      • Weiner M.
      • Tang S.S.
      A retrospective cohort study of the potency of lipid-lowering therapy and race-gender differences in LDL cholesterol control.
      For minority patients, this might be partially explained by the presence of comorbidities and lack of insurance.
      • Bickell N.A.
      • Wang J.J.
      • Oluwole S.
      • et al.
      Missed opportunities: racial disparities in adjuvant breast cancer treatment.
      To our knowledge, this is the first study that assessed opioid deviation in the U.S. palliative care setting. Other researchers did not report these results about ethnicity and gender.
      • Tzeng J.I.
      • Chang C.C.
      • Chang H.J.
      • Lin C.C.
      Assessing analgesic regimen adherence with the Morisky Medication Adherence Measure for Taiwanese patients with cancer pain.
      • Zeppetella G.
      How do terminally ill patients at home take their medication?.
      More research is necessary to better characterize why deviation is higher in these groups.
      One previous study suggested that, in the palliative care and hospice population, the frequency of death related to opioid overdose was not common and may be related to better monitoring, which mitigates the risk of overdose.
      • Bohnert A.S.
      • Valenstein M.
      • Bair M.J.
      • et al.
      Association between opioid prescribing patterns and opioid overdose-related deaths.
      Our study confirms the finding that opioid deviation from prescribed doses is not problematic in palliative care and hospice settings. All groups showed very high adherence, with a high level of motivation and knowledge as demonstrated by both scores on the Modified Morisky Scale. It is likely that this high level of motivation and knowledge results from the intensive interaction with palliative care and hospice interdisciplinary teams. These interdisciplinary teams use systematic assessment tools and counseling and appropriately refer patients who may require more psychosocial counseling and interventions to other interdisciplinary team members.
      Previous studies have observed that chemical coping or a history of drug use was a predictor for opioid deviation toward higher intake.
      • Gonzales G.R.
      • Coyle N.
      Treatment of cancer pain in a former opioid abuser: fears of the patient and staff and their influence on care.
      • McCorquodale S.
      • De Faye B.
      • Bruera E.
      Pain control in an alcoholic cancer patient.
      • Kaplan R.
      • Slywka J.
      • Slagle S.
      • Ries K.
      A titrated morphine analgesic regimen comparing substance users and non-users with AIDS-related pain.
      • Fainsinger R.L.
      • Nekolaichuk C.L.
      • Lawlor P.G.
      • et al.
      A multicenter study of the revised Edmonton Staging System for classifying cancer pain in advanced cancer patients.
      The frequency of patients with CAGE-positive results was similar to previous findings by our group and other groups.
      • Parsons H.A.
      • Delgado-Guay M.O.
      • El Osta B.
      • et al.
      Alcoholism screening in patients with advanced cancer: impact on symptom burden and opioid use.
      However, in our study, we found no association between CAGE positivity and opioid deviation to higher intake. This finding suggests that counseling might be effective in preventing patients with alcoholism and/or patients at risk for chemical coping from deviating from treatment.
      Opioid deviation to underuse was infrequent compared with other studies.
      • Tzeng J.I.
      • Chang C.C.
      • Chang H.J.
      • Lin C.C.
      Assessing analgesic regimen adherence with the Morisky Medication Adherence Measure for Taiwanese patients with cancer pain.
      • Lai Y.H.
      • Keefe F.J.
      • Sun W.Z.
      • et al.
      Relationship between pain-specific beliefs and adherence to analgesic regimens in Taiwanese cancer patients: a preliminary study.
      This suggests that the support provided by the palliative care professionals in reinforcing adherence to the opioid treatment plan was effective. Specifically, the teams may have better educated the patients regarding the treatable side effects of opioids, such as nausea and constipation. Only five patients had severe deviation of opioid dose toward underuse. It is possible that these patients were opioidphobic, as previously described
      • Bashayreh A.
      Opioidphobia and cancer pain management.
      and would have benefited from education about the low potential for psychological addiction in the treatment of cancer pain.
      Eight patients deviated from the prescribed dose to higher intake, and in all cases the deviation was mild and did not require palliative care physician notification. This overuse of opioid medication may be explained by the patients having developed a tolerance to the analgesics. Patients with a history of chemical coping have been reported to have a poor prognostic factor for pain control
      • Gonzales G.R.
      • Coyle N.
      Treatment of cancer pain in a former opioid abuser: fears of the patient and staff and their influence on care.
      • McCorquodale S.
      • De Faye B.
      • Bruera E.
      Pain control in an alcoholic cancer patient.
      and higher pain expression.
      • Compton P.
      • Charuvastra V.C.
      • Kintaudi K.
      • Ling W.
      Pain responses in methadone-maintained opioid abusers.
      Subsequently, this can lead to higher opioid use by the patient.
      • Kaplan R.
      • Slywka J.
      • Slagle S.
      • Ries K.
      A titrated morphine analgesic regimen comparing substance users and non-users with AIDS-related pain.
      In our study, we did not find an association between a previous history of chemical coping and opioid deviation.
      A previous study observed that higher barrier scores were associated with inadequate analgesic use.
      • Chung T.K.
      • French P.
      • Chan S.
      Patient-related barriers to cancer pain management in a palliative care setting in Hong Kong.
      • Gunnarsdottir S.
      • Donovan H.S.
      • Serlin R.C.
      • Voge C.
      • Ward S.
      Patient-related barriers to pain management: the Barriers Questionnaire II (BQ-II).
      • Gunnarsdottir S.
      • Serlin R.C.
      • Ward S.
      Patient-related barriers to pain management: the Icelandic Barriers Questionnaire II.
      • Lin C.C.
      • Ward S.E.
      Patient-related barriers to cancer pain management in Taiwan.
      • Ward S.E.
      • Hernandez L.
      Patient-related barriers to management of cancer pain in Puerto Rico.
      • Breitbart W.
      • Passik S.
      • McDonald M.V.
      • et al.
      Patient-related barriers to pain management in ambulatory AIDS patients.
      However, despite observing higher barriers to pain management in our patient population as compared with the previous study, we found no association between BQ-II scores and opioid deviation to underuse. This could be explained by the opioid education provided to patients by the interdisciplinary team in the supportive care clinic. Our results confirm our expectation that patients with high levels of knowledge and motivation have lower barrier scores. More research is necessary to determine if the findings of the BQ-II can be predictive of opioid deviation in the advanced cancer population.
      Because patients were informed during the consent process of the safety plan to notify the palliative care physician if they reported opioid use that severely deviated from the prescription, some patients may have not been candid about their opioid use. It would have been unethical, however, to withhold this information from the palliative care physician. Nevertheless, our findings are quite reassuring that when patients are educated, counseled, and carefully monitored by a supportive care clinic, opioid deviation is uncommon and most commonly trends toward lower opioid use. Because so few patients reported opioid deviation in our study, more research and a larger sample size is needed to confirm the results. Furthermore, these findings in a convenience sample of cancer patients from a single institution may not be generalizable to other patient populations with less frequent follow-up and non-cancer pain management.
      One limitation of our study was that our exclusion criteria did not allow us to assess patients who did not speak English or who were not in charge of their own medication. Another possible limitation is that the patients might have perceived the research assistant as a staff member, leading them to not report deviation from the prescribed regimen. Finally, although accrual to this study was excellent, as demonstrated by completion in four months in a two-physician outpatient clinic, it is possible that some patients engaging in severe opioid deviation chose not to participate, limiting the generalizability of our findings. More research is necessary to better characterize the frequency and predictors of opioid deviation in this important population.

      Disclosures and Acknowledgments

      This study had no specific funding. Eduardo Bruera is supported in part by National Institutes of Health grants R01NR010162-01A1, R01CA1222292.01, and R01CA124481-01. The authors have no conflicts of interest.

      Appendix. Equianalgesic Table Used for Computing the MEDD

      Tabled 1Equianalgesic Opioid Dose and Conversion Table
      OpioidOral DoseParenteral (IV/SC) DoseConversion Factor for Changing Parenteral Opioid to Oral OpioidConversion Factor for Changing Oral Opioid to Oral Morphine
      Morphine15 mg6 mg2.51
      Codeine100 mgN/AN/A0.1
      Oxycodone10 mgN/AN/A1.5
      Oxymorphone5 mg0.5 mg103
      Hydromorphone3 mg1.5 mg25
      Fentanyl
      1 mg IV morphine or 2.5 mg oral morphine=10 mcg of IV fentanyl
      Morphine 50 mg oral in 24 hours=fentanyl 25 mcg patch q72 hours
      Tabled 1Morphine-Methadone Conversion
      24-Hour Total Dose of Oral MorphineConversion Ratio (Oral Morphine to Oral Methadone)
      <30 mg2:1 (2 mg morphine to 1 mg methadone)
      31–99 mg4:1
      100–299 mg8:1
      300–499 mg12:1
      500–999 mg15:1
      >1000 mg20:1

      References

        • van den Beuken-van Everdingen M.H.
        • de Rijke J.M.
        • Kessels A.G.
        • et al.
        Prevalence of pain in patients with cancer: a systematic review of the past 40 years.
        Ann Oncol. 2007; 18: 1437-1449
        • Brescia F.J.
        • Adler D.
        • Gray G.
        • et al.
        Hospitalized advanced cancer patients: a profile.
        J Pain Symptom Manage. 1990; 5: 221-227
        • Portenoy R.K.
        • Lesage P.
        Management of cancer pain.
        Lancet. 1999; 353: 1695-1700
        • Abbas S.Q.
        • Abbas Z.
        Is opiate compliance a problem in cancer pain? A survey of health-care professionals' views.
        Int J Palliat Nurs. 2003; 9: 56-63
        • Tzeng J.I.
        • Chang C.C.
        • Chang H.J.
        • Lin C.C.
        Assessing analgesic regimen adherence with the Morisky Medication Adherence Measure for Taiwanese patients with cancer pain.
        J Pain Symptom Manage. 2008; 36: 157-166
        • Bohnert A.S.
        • Valenstein M.
        • Bair M.J.
        • et al.
        Association between opioid prescribing patterns and opioid overdose-related deaths.
        JAMA. 2011; 305: 1315-1321
        • Okie S.
        A flood of opioids, a rising tide of deaths.
        N Engl J Med. 2010; 363: 1981-1985
        • Paulozzi L.J.
        • Budnitz D.S.
        • Xi Y.
        Increasing deaths from opioid analgesics in the United States.
        Pharmacoepidemiol Drug Saf. 2006; 15: 618-627
        • McLellan A.T.
        • Turner B.
        Prescription opioids, overdose deaths, and physician responsibility.
        JAMA. 2008; 300: 2672-2673
        • Valeberg B.T.
        • Miaskowski C.
        • Hanestad B.R.
        • et al.
        Prevalence rates for and predictors of self-reported adherence of oncology outpatients with analgesic medications.
        Clin J Pain. 2008; 24: 627-636
        • Liang S.Y.
        • Yates P.
        • Edwards H.
        • Tsay S.L.
        Factors influencing opioid-taking self-efficacy and analgesic adherence in Taiwanese outpatients with cancer.
        Psychooncology. 2008; 17: 1100-1107
        • Lai Y.H.
        • Keefe F.J.
        • Sun W.Z.
        • et al.
        Relationship between pain-specific beliefs and adherence to analgesic regimens in Taiwanese cancer patients: a preliminary study.
        J Pain Symptom Manage. 2002; 24: 415-423
        • Bagcivan G.
        • Tosun N.
        • Komurcu S.
        • Akbayrak N.
        • Ozet A.
        Analysis of patient-related barriers in cancer pain management in Turkish patients.
        J Pain Symptom Manage. 2009; 38: 727-737
        • Chang M.C.
        • Chang Y.C.
        • Chiou J.F.
        • Tsou T.S.
        • Lin C.C.
        Overcoming patient-related barriers to cancer pain management for home care patients. A pilot study.
        Cancer Nurs. 2002; 25: 470-476
        • Chung T.K.
        • French P.
        • Chan S.
        Patient-related barriers to cancer pain management in a palliative care setting in Hong Kong.
        Cancer Nurs. 1999; 22: 196-203
        • Gunnarsdottir S.
        • Donovan H.S.
        • Serlin R.C.
        • Voge C.
        • Ward S.
        Patient-related barriers to pain management: the Barriers Questionnaire II (BQ-II).
        Pain. 2002; 99: 385-396
        • Gunnarsdottir S.
        • Donovan H.S.
        • Ward S.
        Interventions to overcome clinician- and patient-related barriers to pain management.
        Nurs Clin North Am. 2003; 38: 419-434
        • Gunnarsdottir S.
        • Serlin R.C.
        • Ward S.
        Patient-related barriers to pain management: the Icelandic Barriers Questionnaire II.
        J Pain Symptom Manage. 2005; 29: 273-285
        • Jacobsen R.
        • Moldrup C.
        • Christrup L.
        • Sjogren P.
        Patient-related barriers to cancer pain management: a systematic exploratory review.
        Scand J Caring Sci. 2009; 23: 190-208
        • Lin C.C.
        • Ward S.E.
        Patient-related barriers to cancer pain management in Taiwan.
        Cancer Nurs. 1995; 18: 16-22
        • Ward S.E.
        • Goldberg N.
        • Miller-McCauley V.
        • et al.
        Patient-related barriers to management of cancer pain.
        Pain. 1993; 52: 319-324
        • Ward S.E.
        • Hernandez L.
        Patient-related barriers to management of cancer pain in Puerto Rico.
        Pain. 1994; 58: 233-238
        • Chang V.T.
        • Hwang S.S.
        • Feuerman M.
        Validation of the Edmonton Symptom Assessment Scale.
        Cancer. 2000; 88: 2164-2171
        • Buchsbaum D.G.
        • Buchanan R.G.
        • Centor R.M.
        • Schnoll S.H.
        • Lawton M.J.
        Screening for alcohol abuse using CAGE scores and likelihood ratios.
        Ann Intern Med. 1991; 115: 774-777
        • Lawlor P.G.
        • Nekolaichuk C.
        • Gagnon B.
        • et al.
        Clinical utility, factor analysis, and further validation of the memorial delirium assessment scale in patients with advanced cancer: assessing delirium in advanced cancer.
        Cancer. 2000; 88: 2859-2867
        • Kirsh K.L.
        • Jass C.
        • Bennett D.S.
        • Hagen J.E.
        • Passik S.D.
        Initial development of a survey tool to detect issues of chemical coping in chronic pain patients.
        Palliat Support Care. 2007; 5: 219-226
        • Morisky D.E.
        • Green L.W.
        • Levine D.M.
        Concurrent and predictive validity of a self-reported measure of medication adherence.
        Med Care. 1986; 24: 67-74
        • Prado Jr., J.C.
        • Kupek E.
        • Mion Jr., D.
        Validity of four indirect methods to measure adherence in primary care hypertensives.
        J Hum Hypertens. 2007; 21: 579-584
        • Elm J.J.
        • Kamp C.
        • Tilley B.C.
        • et al.
        Self-reported adherence versus pill count in Parkinson's disease: the NET-PD experience.
        Mov Disord. 2007; 22: 822-827
        • Sodergard B.
        • Halvarsson M.
        • Lindback S.
        • et al.
        Differences in adherence and motivation to HIV therapy–two independent assessments in 1998 and 2002.
        Pharm World Sci. 2006; 28: 248-256
        • Sodergard B.
        • Halvarsson M.
        • Tully M.P.
        • et al.
        Adherence to treatment in Swedish HIV-infected patients.
        J Clin Pharm Ther. 2006; 31: 605-616
      1. Case Management Society of America. Case management adherence guidelines. 2004. Available from www.cmsa.org/portals/0/pdf/CMAG.pdf. Accessed December 2011.

        • Powell S.K.
        The case management adherence guidelines.
        Lippincotts Case Manag. 2005; 10: 1-2
        • Rudd P.
        The measurement of compliance: medication taking.
        in: Krasnegor N.A. Epstein L. Johnson S.B. Yaffe S.J. Developmental aspects of health compliance behavior. Lawrence Erlbaum Associates, Hillsdale, NJ1993: 185-213
        • Ewing J.A.
        Detecting alcoholism. The CAGE questionnaire.
        JAMA. 1984; 252: 1905-1907
        • Mayfield D.
        • McLeod G.
        • Hall P.
        The CAGE questionnaire: validation of a new alcoholism screening instrument.
        Am J Psychiatry. 1974; 131: 1121-1123
        • Liskow B.
        • Campbell J.
        • Nickel E.J.
        • Powell B.J.
        Validity of the CAGE questionnaire in screening for alcohol dependence in a walk-in (triage) clinic.
        J Stud Alcohol. 1995; 56: 277-281
        • Bruera E.
        • Moyano J.
        • Seifert L.
        • et al.
        The frequency of alcoholism among patients with pain due to terminal cancer.
        J Pain Symptom Manage. 1995; 10: 599-603
        • Reddy S.K.
        • Nguyen L.
        Pain management.
        in: Bruera E. Elsayem A. The M. D. Anderson supportive and palliative care handbook. 4th ed. UT M. D. Anderson Cancer Center, Houston, TX2011: 29-55
        • Bickell N.A.
        • Wang J.J.
        • Oluwole S.
        • et al.
        Missed opportunities: racial disparities in adjuvant breast cancer treatment.
        J Clin Oncol. 2006; 24: 1357-1362
        • Blackman D.J.
        • Masi C.M.
        Racial and ethnic disparities in breast cancer mortality: are we doing enough to address the root causes?.
        J Clin Oncol. 2006; 24: 2170-2178
        • Hershman D.
        • McBride R.
        • Jacobson J.S.
        • et al.
        Racial disparities in treatment and survival among women with early-stage breast cancer.
        J Clin Oncol. 2005; 23: 6639-6646
        • Kirigaya D.
        • Nakayama T.
        • Ishizaki T.
        • Ikeda S.
        • Satoh T.
        Management and treatment of osteoporosis in patients receiving long-term glucocorticoid treatment: current status of adherence to clinical guidelines and related factors.
        Intern Med. 2011; 50: 2793-2800
        • Turner B.J.
        • Hollenbeak C.S.
        • Weiner M.
        • Tang S.S.
        A retrospective cohort study of the potency of lipid-lowering therapy and race-gender differences in LDL cholesterol control.
        BMC Cardiovasc Disord. 2011; 11: 58
        • Zeppetella G.
        How do terminally ill patients at home take their medication?.
        Palliat Med. 1999; 13: 469-475
        • Gonzales G.R.
        • Coyle N.
        Treatment of cancer pain in a former opioid abuser: fears of the patient and staff and their influence on care.
        J Pain Symptom Manage. 1992; 7: 246-249
        • McCorquodale S.
        • De Faye B.
        • Bruera E.
        Pain control in an alcoholic cancer patient.
        J Pain Symptom Manage. 1993; 8: 177-180
        • Kaplan R.
        • Slywka J.
        • Slagle S.
        • Ries K.
        A titrated morphine analgesic regimen comparing substance users and non-users with AIDS-related pain.
        J Pain Symptom Manage. 2000; 19: 265-273
        • Fainsinger R.L.
        • Nekolaichuk C.L.
        • Lawlor P.G.
        • et al.
        A multicenter study of the revised Edmonton Staging System for classifying cancer pain in advanced cancer patients.
        J Pain Symptom Manage. 2005; 29: 224-237
        • Parsons H.A.
        • Delgado-Guay M.O.
        • El Osta B.
        • et al.
        Alcoholism screening in patients with advanced cancer: impact on symptom burden and opioid use.
        J Palliat Med. 2008; 11: 964-968
        • Bashayreh A.
        Opioidphobia and cancer pain management.
        J Pediatr Hematol Oncol. 2011; 33: S60-S61
        • Compton P.
        • Charuvastra V.C.
        • Kintaudi K.
        • Ling W.
        Pain responses in methadone-maintained opioid abusers.
        J Pain Symptom Manage. 2000; 20: 237-245
        • Breitbart W.
        • Passik S.
        • McDonald M.V.
        • et al.
        Patient-related barriers to pain management in ambulatory AIDS patients.
        Pain. 1998; 76: 9-16