Revisiting Patient-Related Barriers to Cancer Pain Management in the Context of the US Opioid Crisis

Kristine Kwekkeboom; Ronald C. Serlin; Sandra E. Ward; Thomas W. LeBlanc; Adeboye Ogunseitan; James Cleary


Pain. 2021;162(6):1840-1847. 

In This Article



Subjects were recruited from 3 sites within the Palliative Care Research Cooperative Group. The study was reviewed and approved by the institutional review board at each participating site. To be eligible, patients had to be 18 years of age or older, have active recurrent or metastatic cancer, be able to speak and read English at or above a fifth grade level, and be capable of completing the questionnaire on their own. Patients who had surgery within the past 30 days and those who did not manage their own pain medications were excluded. Three hundred and twenty-two patients were approached, 179 consented, and 157 provided data for the current analyses (Figure 1).

Figure 1.

Participant flow diagram.


Demographic and Clinical Characteristics. Participants completed a self-report questionnaire to document age, sex, ethnicity, race, and marital status. Information regarding cancer diagnosis, stage, performance status, and use of palliative and or hospice services were collected from the medical record. The performance status was documented using the Australia-modified Karnofsky Performance Status (AKPS) scale.[1] Scores range from a high of 100 (normal; no complaints; and no evidence of disease), to a midpoint of 50 (requires considerable assistance and frequent medical care), and a low of 0 (dead).

Patient-related Barriers to Analgesic use. The Barriers Questionnaire II is a 27-item self-report instrument that measures beliefs about reporting pain and using analgesic medications.[15] These beliefs include fear of addiction and worry about side effects, and they are described extensively elsewhere.[15] Response options range from 0 "do not agree at all" to 5 "agree very much". Factor analysis of the 27 items has revealed a 4-factor solution with beliefs representing: (1) harmful effects of analgesics, (2) fatalism regarding pain control, (3) physiologic side effects of analgesics, and (4) concerns about communicating pain to the care provider. In analyses, either an overall mean of the 27 item scores can be used or means of the items in each factor, with higher scores indicating greater barriers to analgesic use. In the current study, internal consistency of the total BQ-II overall mean was excellent (Cronbach's alpha = 0.92) and good or acceptable for subscales— harmful effects (Cronbach's alpha = 0.80), fatalism (Cronbach's alpha = 0.77), physiologic effects (Cronbach's alpha = 0.88), and communication (Cronbach's alpha = 0.85).

Pain Severity. A composite pain severity score was created by computing the mean of 3 items from the Brief Pain Inventory (BPI).[8,10] The items were: pain worst in the past week, pain least in the past week, and pain now, each of which is rated on a scale of 0 (no pain) to 10 (pain as bad as you can imagine). Internal consistency of the pain severity composite was acceptable (Cronbach's alpha = 0.77).

Pain Interference With Life. Seven items from the BPI were used to evaluate pain interference with daily life.[8] Participants reported the extent to which pain interfered with various aspects of their lives in the past week, with response options from 0 "does not interfere" to 10 "completely interferes." A mean score of the 7 items was used in analyses, with higher scores indicating greater interference. Internal consistency in the current study was excellent (Cronbach's alpha = 0.90).

Pharmacologic Treatment for Pain. Participants were asked to complete a single item indicating which classes of medications they were currently using to treat their cancer-related pain—strong opioids, weak opioids, and/or nonopioids. Examples of medications in each category were provided to facilitate accurate identification. Patients who reported only experiencing everyday kinds of pain (eg, minor headaches) and who had taken no pain medication in the past week did not complete questionnaire items assessing pain severity, pain interference with daily life, or pharmacologic treatment of pain.

Adequacy of Analgesic use. The Pain Management Index (PMI)[54] is based on the World Health Organization (WHO) "analgesic ladder" for treating cancer pain,[52] developed in and promoted since 1986. It compares the most potent analgesic used by a patient to the patient's level of reported pain. First, the level of analgesics used is determined based on 4 categories: (0) no analgesics, (1) nonopioid (eg, NSAID or acetaminophen), (2) weak opioid (eg, any preparations containing both opioids and nonopioids), or (3) strong opioid (eg, morphine or sustained-release Oxycontin). Then, the level of the patient's pain in the past week is determined by using the worst pain item from the BPI. The levels used are (0) pain worst rating of 0, (1) pain worst rating of 1 to 4, (2) pain worst rating of 5 to 6, and (3) pain worst rating of 7 to 10.[38] The index is computed by subtracting the value of the pain level from the level of analgesics used. The resulting scores yield a 2-category system where negative scores indicate inadequate analgesic use, and scores of 0 and greater indicate acceptable analgesic use. Validity of the PMI has been demonstrated by finding predicted relationships between the PMI and other variables such as barriers scores.[9,47]


Patients were recruited in the outpatient oncology clinics at 2 cancer centers located in the Midwestern United Sates and one cancer center in the Southeastern United States. Research coordinators screened clinic appointment lists or received patient referrals from clinical staff and met with potentially eligible patients to describe the study, review consent documents, and answer questions. Signed informed consent was obtained from all patients who agreed to participate. On consent, the research coordinator provided options for the participant to complete questionnaire in paper form or electronically. Paper forms could be completed in person, at the clinic visit, or taken home to complete and return at the next visit. Electronic data entry was completed using REDCap (Research Electronic Data Capture),[17,18] hosted by the Palliative Care Research Cooperative. Participants accessed REDCap either on computers in the clinic or at home on their personal computers by an email link to the questionnaires. Forms completed on paper were entered in REDCap by a research coordinator. Participants were asked to complete questionnaires within 2 weeks of consent.

Data Analyses

Analyses were conducted using SPSS (v.26). For all instruments, scale scores were computed if less than 20% of an individual's responses were missing. Descriptive statistics were used to summarize sample demographic and clinical characteristics, as well as scores on the BQ-II for the full sample (N = 157). Descriptive statistics were used to summarize pain severity, pain interference, and the PMI for the subset of participants (n = 97) who reported pain and/or use of pain medications in the past week.

The primary analysis compared currently observed BQ-II scores to those reported in previous studies. Comparison studies were identified through a search of CINAHL and PubMed. Criteria for inclusion were use of the BQ-II (without modifications) in samples of patients with cancer who were experiencing pain and reported scores on the harmful effects (fear of addiction) BQ-II subscale or the total BQ-II score, or reported association between the BQ-II scores and pain outcomes (pain severity, pain interference, and PMI). Where 2 or more publications reported data from a single study, we excluded publications reporting data from overlapping participants. In studies reporting barriers scores before and after an intervention, we used preintervention (baseline) scores. Change in the harmful effects subscale and total BQ-II score were examined by testing linear trend contrasts in the context of a one-factor (publication year) design, using linear trend coefficients calculated according to Gaito.[14] This is equivalent to performing a simple regression analysis predicting either harmful effects or total BQ-II score from the publication year. Changes across time in relationships between BQ-II scores (harmful effects and total) and 3 pain outcomes (pain severity, pain interference, and PMI) were evaluated. The change in correlations of BQ-II scores with pain severity and pain interference was tested using Fisher's r-to-z transform. The change in relationship between BQ-II scores and PMI was examined by comparing the differences in BQ-II means in the PMI categories across time. Finally, the change over time in the proportion of patients reporting adequate analgesic use (PMI) was examined using logistic regression.