Comparing two approaches to help patients and doctors talk about chronic pain treatments : voices in pain care / Brennan M. Spiegel.

Format:

Book

Author/Creator:

Spiegel, Brennan M., author.

Language:

English

Subjects (All):

Physician and patient.

Pain--Treatment.

Pain.

Physical Description:

1 online resource (129 pages) : illustrations

Other Title:

Comparing two approaches to help patients and doctors talk about chronic pain treatments

Place of Publication:

Washington, D.C. : Patient-Centered Outcomes Research Institute (PCORI), 2022.

Summary:

BACKGROUND: In the past decade, the use of opioids for chronic pain has increased at an alarming rate. High rates of opioid prescribing have been accompanied by a concomitant increase in opioid use disorders as well as opioid overdoses. Opioid use disorder is associated with decreased health-related quality of life and increased risk of death. Both the physician-facing Clinical Decision Support (CDS) and patient-facing Patient Education and Activation Tools (PEATs) are widely used evidence-based interventions to mitigate the risk of opioid use disorder, but they have never been tested in a head-to-head comparative effectiveness trial. OBJECTIVES: We assessed the comparative effectiveness of 2 established patient-provider communication strategies to promote appropriate use of opioids for chronic pain: (1) the physician-facing electronic health record (EHR)-based Choosing Wisely CDS alerts, and (2) the patient-facing PEATs developed by the American Chronic Pain Association and Consumer Reports and administered via an EHR patient portal. We conducted a pragmatic, cluster-randomized trial to measure outcomes that are important to patients, their families, and caregivers, with a focus on balancing 2 goals: reducing the impact of pain on patients' lives and reducing opioid overuse. METHODS: We conducted this study within the Cedars-Sinai Medical Network, a large, urban, integrated health care delivery system serving a diverse population in Southern California. All primary care physicians (PCPs) and eligible patients in the Cedars-Sinai Medical Network were enrolled in this system-wide intervention as part of a leadership-supported initiative. Members of the integrated PCP network are overseen by the same administration and medical directors, use the same EHR (Epic Systems Corporation), and are governed by the same policies and practices. We compared the effects of the CDS vs PEATs in patients with an opioid prescription of 6 weeks or longer on outcomes that are important to patients. The primary outcomes were (1) change in composite satisfaction with patient-physician communication at appointments postintervention as measured by the Clinician and Group Consumer Assessment of Healthcare Providers and Systems (CG-CAHPS) survey and (2) change in postintervention monthly pain interference with normal activities as measured by PROMIS(r). Secondary outcomes included depression as measured by the Patient Health Questionnaire-9 and rates of high-risk opioid prescribing, including prescriptions of ≥90 morphine milligram equivalents (MME) per day, a clinically important threshold identified by the Centers for Disease Control and Prevention, and coprescribing of benzodiazepines and opioids (co-Rx). The inclusion criteria allowed for generalizability of findings to individuals with many conditions, including low back pain, fibromyalgia, and neuropathic pain. The trial was cluster randomized, involving 82 PCPs and 951 of their patients. Patients of PCPs in the PEATs arm received educational materials developed to engage patients in chronic pain treatment prior to PCP office visits, while PCPs in the CDS arm received computerized reminders about appropriate opioid use during office visits for enrolled patients. Patients in both groups received questionnaires about pain interference with normal activities, health-related quality of life, and physician-patient communication through the patient portal 1 month after each visit to their PCP. Outcomes were collected in the 12 months before and after the start of the intervention. We used multilevel regression models to compare the effectiveness of these 2 communication strategies. RESULTS: We randomized 82 physicians with 951 patients to the 2 arms. There were no significant differences between the study arms in 1 of the 2 primary outcomes, pain interference, over the course of the study. Compared with the CDS arm, the PEATs arm displayed a nearly 3-times-higher change from the preintervention period to the postintervention period in the other primary outcome, the odds ratio (OR) of a CG-CAHPS response, indicating that a patient "always" received the desired care or service for a given measure (OR, 2.96; 95% CI, 1.20-7.31; P = .019). However, the baseline odds of this response were much lower in the PEATs group than in the CDS group, which complicates the comparison of the 2 interventions. Because of high proportions of missingness in CG-CAHPS survey responses (62.3% of patients did not contribute a CG-CAHPS response), we performed an analysis of imputed data that did not show differences in CG-CAHPS between the study arms. The PEATs arm experienced a higher OR of Rx ≥90 MME/day (a secondary outcome) than did the CDS arm in the postintervention period than in the corresponding preintervention period (OR, 1.63; 95% CI, 1.13-2.36; P = .010), but the study arms differed at baseline, with the CDS arm having fewer opioid prescriptions overall (805 vs 1422, respectively) and lower rates of Rx ≥90 MME/day (8.5% vs 14.8%, respectively; P < .001). There were no differences in physical function, depression, or co-Rx outcomes between arms. CONCLUSIONS: The results are equivocal because of several study limitations. Differences between the 2 groups in the preintervention period make the results challenging to interpret because the groups may not have displayed exchangeability. The results from this study do not refute previous evidence showing that the PEAT is an inexpensive, easily scaled intervention to support high-risk opioid users, but more evidence is needed to ascertain how the PEATs compare with CDS. LIMITATIONS: Sample sizes for a number of outcomes (ie, CG-CAHPS, PROMIS Pain Interference, PROMIS Physical Function) were a fraction of the proposed targets due to low response rates, threatening the study's internal validity and generalizability. The adoption of an autoenrollment procedure limited the number of active survey participants, reducing the number of available covariates available for adjustment across the whole study. Differences in baseline satisfaction and prescribing rates between the 2 arms in the preintervention period pose challenges to interpreting differences in outcomes between arms. The presence of multiple interventions coupled with differences between the 2 study arms caused by randomization failures may have meant that interventions were applied unevenly between the 2 arms. The PEATs intervention was multicomponent, and the pragmatic study design did not attempt to isolate the individual effects of those components.

Contents:

ABSTRACT 5

BACKGROUND 7

A Public Health Crisis: the Overuse of Opioids for Treating Chronic Pain 7

Burden of Chronic Pain in the United States 8

Engaging Patients in the Chronic Pain Discussion 8

Addressing Gaps in Provider Knowledge of Chronic Pain 9

Patient Education and Activation Tool 9

Clinical Decision Support Tools for Chronic Pain 10

Significance 11

Specific Aims 13

PATIENT AND STAKEHOLDER ENGAGEMENT 14

Patient Partners and Their Roles in the Study 14

PAB Members and Their Roles in the Study 18

Other Stakeholder Partners 19

METHODS 22

Study Overview 22

Study Design 22

Figure 1 Study Design and Assessments 23

Study Setting 23

Participants 23

Patient Recruitment and Inclusion/Exclusion Criteria 24

Study Interventions/Arms 25

Figure 2 Screenshot of the Coprescribing Pop-up Alert 26

Figure 3 Representative Process Flow for a CDS Alert 27

Study Outcomes 28

Figure 4 Questions Comprising the CG-CAHPS "How Well Providers Communicate With Patients" 6-Item Composite Score 30

Sample-Size Calculations and Power 31

Time Frame for the Study 32

Data Collection and Sources 32

Figure 5 Schedule of Assessments 33

Analytical and Statistical Approaches 33

Changes to the Original Study Protocol 36

RESULTS 40

DISCUSSION68

Subpopulation Considerations 73

Study Limitations 74

Future Research 78

CONCLUSIONS 80

REFERENCES 81

APPENDICES 88.

Notes:

Description based on publisher supplied metadata and other sources.