Trust between doctors and patients
with chronic low back pain

By David Schneider, M.D., Sandra Burge, Ph.D., and the RRNeT Investigators

Introduction

Low back pain is one of the top 10 reasons people seek care from a family physician.1,2 The prevalence in the population varies from 8 percent to 37 percent. Back pain is a leading cause of lost work time and disability, and accounts for approximately 23 percent of workers’ compensation payments.2 Most back pain is temporary and improves in the first few weeks after injury, but about 10 percent extend beyond 90 days.1 After three months of pain or disability, recovery is slow and influenced by many factors.2

For many with chronic low back pain (three months duration or longer), opioid therapy can provide pain relief and restore daily functioning. Yet many physicians are hesitant to prescribe opioids. Sinatra3 stated: “Physician concern about opioid misuse remains perhaps the most significant barrier to the optimal use of opioids in patients with chronic non-cancer pain.” An RRNeT (the Residency Research Network of Texas) survey of 209 family physicians confirmed Sinatra’s statement: Physicians’ concerns about drug dependence and addiction were the most common barriers to prescribing opioids for chronic pain. (Whitham and Burge, unpublished findings)

Physicians’ concerns about addiction can lead to mistrust of patients who request opioid medicines for pain. “Pseudo-addiction” may result.4 This is characterized by pain that is inadequately treated, creating anxiety in the patient and leading him or her to demand opioid analgesia in a way that is interpreted by physicians as excessive or “drug-seeking.” Physicians’ reluctance to provide pain relief increases patients’ anxiety and demands, which in turn reinforces physicians’ reluctance to prescribe pain medicines. A cycle of mistrust and anxiety develops resulting in serious conflict between the providers and the patients.

The purpose of this study was to examine trust from both sides of the doctor-patient relationship when managing chronic low back pain. We examined several issues that might influence levels of doctors’ and patients’ trust: characteristics of the patient; characteristics of the patient’s pain and its treatment; patient’s health, co-morbidities, physical functioning, mental health, and social circumstances; and characteristics of the visit on the day of the survey.

Methods

Participants. This cross-sectional study was conducted in RRNeT, a collaboration of nine family medicine residency programs located in Austin, Corpus Christi, Dallas, Fort Worth, Garland, Harlingen, Houston, McAllen and San Antonio. The study was approved by the IRB of each participating hospital or institution. Students working in eight outpatient clinics enrolled 254 adult patients with a diagnosis of chronic low back pain with symptom duration of three months or longer. We excluded pregnant women and new patients from the study. Students successfully gathered 254 surveys from patients and 218 visit checklists from their doctors. This study examines the data from the 218 patients for which we have both patients’ and doctors’ information.

Data Collection Procedure

Students screened ledgers and charts at the beginning of each clinic session to identify potentially eligible patients for this study. When these patients arrived, students approached them in the waiting room, provided informed consent and invited them to participate in a survey about low back pain. When patients agreed, students helped them find a quiet, private place to complete the survey. Study materials were available in both Spanish and English. No patient identifiers were required on the surveys. While the patient completed the survey and waited for their appointment, the student attached a one-page visit checklist to the patient’s chart, which the physician completed following the clinic visit.

Measurement

The patient survey addressed characteristics of low back pain and its treatment, health and functioning, mental health, stress and social support, demographic information, and characteristics of the doctor-patient relationship, including trust. Specifically, the survey included the MOS Short Form 3611 to address health and physical functioning, the PHQ-914 to address depression, the Beck Anxiety Inventory – PC15 to address anxiety, the Duke Social Support and Stress Scale13 to address social support and stress, and Adverse Childhood Experiences16 to address a history of abuse.

The patients’ trust score was an average of responses from eight likert-type items from the Primary Care Assessment Survey.12 Alpha reliability for this sample was .743. A high score represented high trust. Items included:

• I can tell my doctor anything.
• My doctor sometimes pretends to know things when he/she is not really sure. (reversed)
• I completely trust my doctor’s judgments about my medical care.
• My doctor cares more about holding costs down than about doing what is needed for my health. (reversed)
• My doctor cares as much as I do about my health and my pain.
• If a mistake was made in my treatment, my doctor would try to hide it from me. (reversed)
• Overall, how much do you trust your doctor?
• How much do you trust the other people in this clinic to take good care of you?

The doctor’s visit checklist included items about the patients’ use of and requests for medicines, and doctors’ levels of trust of the patients’ management of their back pain. Trust items were based on Sinatra’s3 recommendations for assessing addiction risk, plus two general trust questions. A factor analysis guided the construction of a Doctors’ Trust Scale which included six items with likert-type response choices. Alpha reliability for these items was .816. A high score represented high trust.

• Do you believe this patient exaggerates their pain?
• Do you believe this patient is taking more medicine than they need to control the low back pain?
• Do you have concerns about substance abuse or addiction in this patient?
• Do you have concerns about this patient’s history of legal problems or arrests?
• All things considered, how much do you trust this patient to use their pain medicines appropriately?
• How much do you think this patient trusts you?

Analysis

The analysis presented here represents the 218 patients for which we have data from both doctors and patients. The key outcome variables are Doctors’ Trust of the patient and Patients’ Trust of the doctors. We first present simple bivariable associations between the trust outcomes and demographic, health and visit-related predictors. Associations are calculated using Pearson’s r correlations for continuous predictors, and t-tests or ANOVAs for categorical predictors. We then examine independent predictors of the trust outcomes using a linear regression analysis.

Results

Demographics

Of the 218 patients with chronic low back pain, 65 percent were women, and 39 percent were Latinos. Thirty-two percent had less than a high school education. Household income was low, with 64 percent reporting less than $2,000 per month. Few patients (17 percent) had private health insurance; half (51 percent) were insured by the government.

Back Pain

Patients experienced low back pain for three months to 65 years, with a median duration of seven years (mean=11.6, std dev=11.6). Current pain severity averaged 6.0 on a 10-point scale; the median severity was 6.5. Most patients (66 percent) took opioid medications for their pain; 39 percent used them daily.

Visit Characteristics

Doctors reported that most of these patients (79 percent) were their own continuity patients. Doctors also saw walk-in patients or colleagues’ patients; new patients were excluded from the study. While half the patients reported seeing this doctor for one year or less, 54 percent had been visiting this clinic for more than three years; 17 percent had visited this clinic for more than 10 years. Trust levels were good. Doctors reported “low trust” for only 8 percent of patients and “high trust” for half (51 percent). Patients likewise trusted their doctors: only 3 percent reported “low trust,” while 61 percent reported “trust a lot.” The correlation between doctors’ trust and patients’ trust scores was positive, r =.214, p=.002.

Predictors of Trust – Bivariable Analysis

Table 1 displays the simple relationships between doctors’ and patients’ trust scores and demographic characteristics of the patients, pain and health variables, and characteristics of the visit on the day of the survey. Doctors had lower trust for younger people on disability with lower income and no health insurance (p<.05). Their trust was somewhat higher for females, Latinos and married people (p<.10). Doctors had higher trust for healthier people. Higher trust was associated with less severe pain and nonuse of opioid medications; good function; and low scores on depression, anxiety and adverse childhood experiences (p<.05). A sustained continuity relationship was associated with higher trust by doctors; however, they reported lower trust for patients who requested opioid medications for pain (p<.05).

With regard to patients’ trust of their doctors, older patients had higher trust. Male Latino patients had somewhat lower trust scores than others (p<.10). Health and pain had little to do with patients’ levels of trust, with one exception: Patients with more co-morbidities reported higher trust (p<.05). Continuity was strongly related to trust (p<.05). Also, patients who asked for opioid medications for their pain reported higher trust of their doctors (p=.049).

Predictors of Trust – Regression Analyses

To examine independent predictors of doctors’ and patients’ trust, we conducted two linear regression analyses with trust scores as outcome variables. Predictors included patient characteristics (gender, age, Latino ethnicity, education), pain characteristics (pain severity, pain duration, use of opioids for pain, disabled employment status), health variables (sum of co-morbidities, health status, role functioning, depression, anxiety), and visit characteristics (continuity patient, duration of doctor-patient relationship and request for opioids by name). Insignificant predictors were removed with backward stepwise elimination methods. The analyses included 191 subjects with complete data on all variables.

Doctors had lower trust for younger men and higher trust for Latinos and higher-educated patients (Table 2). Continuity of the

relationship was important, with higher trust for a longer relationship. Doctors were less trusting toward patients who requested opioid medications by name, and patients with higher anxiety.

Low back pain patients who were younger Latino men were less trusting of their doctors (Table 3). As with doctors, patient trust was higher when the doctor-patient relationship was longer. Patients who requested opioid medications by name had higher trust scores than others.

Discussion

This study demonstrated that doctors and patients with chronic low back pain had high levels of trust for each other. Bivariable analyses showed that many variables influenced doctors’ trust, including health indicators; doctors had higher trust for well-functioning patients with less severe pain. However the multivariable analysis, controlling for other influences, showed that doctors’ trust toward chronic low back pain patients was influenced primarily by patients’ demographics and visit characteristics. Doctors were less trusting toward younger non-Latino men, and patients with lower education and higher anxiety. A sustained doctor-patient relationship increased doctors’ levels of trust, but requests for opioid medications by name decreased it. The associations between doctors’ trust, patients’ anxiety and specific requests for opioids are consistent with descriptions of pseudo-addiction, where patients who are worried about pain control become more overt in their bids for pain relief, and doctors in turn become more suspicious about their drug-seeking behavior.4 Our findings revealed a paradox: Patients with higher levels of trust felt more comfortable asking for opioids by name; however this behavior appeared to make their doctors less trusting.

Patients’ trust was not related to their pain or functioning, but associated with personal characteristics; younger Latino men were less trusting. Like doctors, patients’ trust was higher when they had a sustained doctor-patient relationship.

This study was limited by its cross-sectional design; findings reported here can demonstrate statistical associations but not causal relationships. Self-report is a second limitation that could bias findings in either direction. Some patients may exaggerate pain and illness to gain more intense treatments, or may exaggerate good functioning to please their physicians. To maximize the reliability and validity of patients’ responses, we used standardized self-report questions which enjoy wide use in clinical research and have good psychometric properties.

To conclude, we encourage family physicians to re-examine their relationships with patients with chronic low back pain to determine if mistrust is influencing pain management. Personal characteristics of the patient (youth and male gender) and patient behaviors (anxiety and treatment requests) appear to raise family physicians’ mistrust and suspicion. However, Quinn4 suggests that these very behaviors may be signs of pain that is poorly controlled, rather than risk factors for drug abuse. Our study patients demonstrate that the act of asking for specific medicines is an act of faith. Many patients trust their doctors to provide good care for their pain, and they are comfortable requesting effective treatments. Finally, we observe that continuity of care, a cornerstone of family practice, is associated with trust from both the doctors’ and patients’ perspectives. Multiple visits with chronic pain patients over time will determine the effectiveness of pain treatment, reveal potential problems with addiction, and build the trust relationship between family doctors and patients.

This study was funded in part by a research grant from the Texas Academy of Family Physicians Foundation.

---

Literature Cited

1. Bratton R. Assessment and management of acute low back pain. American Family Physician 1999;60.
2. Rives PA, Douglass AB. Evaluation and treatment of low back pain in family practice. Journal of the American Board of Family Medicine 2004;17:S24-S31.
3. Sinatra R. Opioid analgesics in primary care: challenges and new advances in the management of noncancer pain. Journal of the American Board of Family Medicine 2006;19:165-77.
4. Quinn TE. What is pseudoaddiction? Pain Relief Connection 2004;2.
5. Burge SK, Albright T, RRNeST Investigators. Complementary and alternative medicine use among family practice patients in south Texas: an RRNeST project. American Journal of Public Health 2002;92:1614-16.
6. McCutchon TA, Parchman M, Burge SK, RRNeST Investigators. Predictors of self care behavior in adults with type 2 diabetes. Family Medicine 2001;33:354-60.
7. Parchman ML, Burge SK, RRNeST Investigators. Continuity and quality of care in type 2 diabetes. A RRNeST Study. Journal of Family Practice 2002;51:619-24.
8. Burge SK, White D, Bajorek E, Bazaldua O, RRNeST Investigators. Correlates of Medication Knowledge and Compliance: Findings from RRNeST. Family Medicine 2005;37:548-54.
9. Albright T, Burge SK. Improving firearm storage habits: the impact of brief office counseling by family physicians. Journal of the American Board of Family Medicine 2003;16:40-46.
10. Burge SK and Oscos-Sanchez M. SAFE TEENS: A Study From the Residency Research Network of South Texas (RRNeST). Paper presented at the Annual Spring Conference of the Society of Teachers of Family Medicine. Toronto, Ontario, Canada. 2004. 5-12-2004.
11. Stewart AL, et al. The MOS Short-Form General Health Survey: reliability and validity in a patient population. Medical Care 1988;26:724-35.
12. Safran DG, Kosinski M, Tarlov AR, Rogers WH, Taira DA, Lieberman N, et al. The Primary Care Assessment Survey: tests of data quality and measurement performance. Medical Care 1998;36:728-39.
13. Parkerson GR, et al. Associations among family support, family stress, and personal functional health status. Journal of Clinical Epidemiology 1989;42:217-29.
14. Kroenke K, et al. The PHQ-9: validity of a brief depression severity measure. Journal of General Internal Medicine 2001;16:606-13.
15. Mori DL, et al. The BAI-PC as a screen for anxiety, depression, and PTSD in primary care. Journal of Clinical Psychology in Medical Settings 2003;10:187-92.
16. Felitti VJ, Anda RF, Nordenberg D, Williamson DF, Spitz AM, Edwards V, et al. Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults: the Adverse Childhood Experiences (ACE) study. American Journal of Preventive Medicine 1998;14:245-58.