USC Roybal Center for Behavioral Interventions in Aging Pilot Projects
Applications for funding for the 2021-2022 pilots are closed. Application information can be found here.
Year 2 Roybal Center Pilot Awardees
A clinician-focused nudging intervention to optimize post-surgical prescribing
PI: Daniel B. Larach, MD, MSTR, MA
There are considerable data that postoperative opioids are commonly prescribed in excessive amounts but few evidence-based techniques to optimize such prescribing. The specific objectives of this study are to (1) evaluate the hypothesis that a novel nudge intervention will reduce excess postoperative opioid prescribing; and (2) determine whether this technique affects opioid consumption, refill requests, medical visits for pain, analgesia satisfaction, and opioid misuse. This pilot randomized controlled trial will randomize surgeons (1:1) to intervention (direct feedback to surgeons about patients’ opioid prescription-to-consumption ratios) or control (no feedback) arms. Patients undergoing elective surgery during days 1-30 will be contacted by telephone 30 days postoperatively (study days 31-60) for an opioid pill count; they will also be asked about a variety of secondary pain and opioid-related measures. After study day 60, surgeons in the intervention arm will be provided procedure-specific direct feedback on opioid prescribing and consumption for their patients who had surgery during days 1-30. Where available, extant surgery-specific evidence-based prescribing recommendations will also be communicated to surgeons in the intervention group. Pre-post change in opioid prescription size (measured in oral morphine equivalents) from baseline between the two groups for surgeries performed during days 61-90 (the primary outcome) will be assessed. The specific primary outcome will be mean percentage change in procedure-specific prescription size. Mean per-surgeon percentage change in prescription size will then be compared between the direct feedback and no-direct-feedback groups. Secondary outcomes will be assessed by contacting patients having post-intervention surgeries (days 61-90) 30 days postoperatively (days 91-120).
Helping hypertension patients to interpret blood pressure readings
PI: Wandi Bruine de Bruin, PhD
Only about 50% of hypertension patients have good blood pressure control. Patients with low health literacy have worse blood pressure control, perhaps because they find it harder to interpret whether blood pressure readings are high or low, reflect good or bad blood pressure control, and indicate a need for behavior change or medication us The American Heart Association has called for broad efforts to help empower hypertension patients from different backgrounds to control their blood pressure. Therefore, we will evaluate different ways for helping hypertension patients to interpret their blood pressure readings and motivate blood pressure control, in hypertension patients varying in health literacy, age, and socio-economic status. Our strategy is based on insights from behavioral science studies, which suggest that people find it easier to interpret numbers when they can see the range of possible numbers. Aim 1: Based on insights from behavioral science and our team’s communication expertise, we will create 3 blood pressure communications: (A) a basic table showing only the normal blood pressure range, which is often used in clinical practice and online communications about blood pressure, but may make it hard to interpret numbers outside of the normal range, potentially undermining behavior change intentions; (B) an enhanced table showing how combinations of diastolic and systolic blood pressure reflect normal, elevated and hypertension ranges, from the American Heart Association; (C) an enhanced graph to be adapted from Blood Pressure UK to show the same color-coded ranges as the enhanced table, with diastolic blood pressure on the x-axis and systolic blood pressure on the y-axis. Aim 2: : In a sample of 650 diagnosed hypertension patients recruited through the Pitt+Me Patient Registry at the University of Pittsburgh Medical Center (UPMC), the investigators will evaluate whether being presented with the enhanced table or graph (vs. basic table) affects patients’ self-reported blood pressure measurement (as averaged across two measurements taken at the time of the survey at least 1 minute apart, as per directions of the American Heart Association), and improves interpretations of these two blood pressure readings and of hypothetical blood pressure readings, as well as behavior change intentions Aim 3: We will examine whether Aim 2 findings vary by health literacy, age, and SES.
Year 1 Roybal Center Pilot Awardees
Comparative Effectiveness of Two Letters to Encourage Judicious Prescribing of Opioids: A County-wide Project in Los Angeles
PI: Jason Doctor, PhD
Prescribers invited to join efforts to exercise more careful use of opioids may be more likely to do so after they have been made aware of an opioid death in their practice. Many physicians underestimate the risk of opioids in their own practice, and currently, there is no mechanism for medical examiners to identify prescribing physicians, nor for physicians to be alerted about deaths, despite substantial interest from the CDC to solve this problem. In collaboration with the Los Angeles County Medical Examiner’s Office and the State of California’s controlled Substance Utilization Review and Evaluation System (CURES), we reviewed opioid poisonings over a 12 month timeframe and sent letters to prescribers in Los Angeles County when at least one of the provider’s prescriptions was filled by a patient who died of an opioid poisoning. Prescribers were randomized to one of two letter versions (Letter A and B). The letters were factual and nonjudgmental, signed by the County Medical Examiner, and stated that a patient they had treated with controlled substances died of an opioid poisoning. The letters encourage judicious prescribing and provide information on how to identify and taper unsafe regimens (high dose, polypharmacy, or use of multiple prescribers); how to identify addiction and compassionately refer patients for medication-assisted treatment; and recommendations to avoid bad outcomes (e.g. do not fire your patient for signs of addiction.) The letters also encourage use of the CURES system before prescribing, as well as co-prescribing of naloxone. Letter B also included additional text involving an “if/when/then statement” along with an injunction to providers to share safety information with patients so that they identify as a “safe prescriber.” Our analysis will assess changes over time before and after the intervention in natural log-transformed milligram morphine equivalents (MME) between the two treatment conditions.
Feasibility of Nudging Providers to Initiate Buprenorphine
PI: Rebecca Trotzky-Sirr, PhD
Buprenorphine is an opioid medication that treats opioid use disorder by tightly binding brain opioid receptors and blocking the euphoric effects of other opioids. Despite widespread provider training and institutional support, only a fraction of eligible patients has initiated buprenorphine to date. This implies that behavioral interventions that impact physician practice may be required to overcome barriers like stigma and peer support. This project aims to (1) Assess barriers and attitudes associated with physician-initiated Medication-Assisted Treatment (MAT) in uninsured, underserved, and homeless populations with opioid use disorder; and (2) Develop nudge-based intervention strategies targeting providers’ decisions. We will purposefully select 10 physicians at the Urgent Care Center (UCC) at Los Angeles County & University of Southern California and the Star Clinic that represent different levels of prescribing and different levels of factors predictive of prescribing patterns determined in a retrospective analysis. Structured interviews will be designed to inform how to overcome clinical inertia with regard to buprenorphine initiation. Interview topics will orient towards (a) key variables identified as barriers to initiation and (b) aspects of provider behavior including dimensions of commitment and professional and social competitiveness. Nudge development will depend on the results of the qualitative analysis but will consider key social motivations and workflow barriers that may be addressed with nudges; a preliminary evaluation of nudge feasibility will also be conducted.