BACKGROUND

Prescription opioid abuse in the United States is a devastating public health crisis. Many chronic opioid users were originally prescribed the medication for acute pain. Individualized risk communication, with or without a storytelling element, may improve patient outcomes such as understanding and recall of opioid risk as well as opioid use behaviors in the setting of acute pain.

OBJECTIVES

The primary objective of this study was to compare 3 strategies to inform patients of their risks associated with misuse of opioid prescriptions after treatment in the emergency department (ED) for renal colic or musculoskeletal back pain. The 3 strategies are (1) a general risk information (GRI) sheet meant to represent standardized usual care, (2) a visual probabilistic risk tool (PRT), and (3) a narrative-enhanced PRT (NE-PRT). The PRT conveys individualized risk based on the Opioid Risk Tool (ORT) of opioid misuse following prescription of opioids. We sought to determine whether the PRT, with or without video narrative enhancement, can improve functional outcomes and patient-centered outcomes.

METHODS

Patients aged 18 to 70 years presenting to the acute care facilities (ie, ED and fast track/urgent care) of 4 US hospital centers with renal colic or musculoskeletal back pain were eligible for this multicenter randomized clinical trial. Included were patients for whom discharge was planned within 24 hours. A usual-care group of patients receiving a standardized GRI sheet was compared with 2 intervention groups, 1 receiving individualized risk information via a visual PRT and another receiving the PRT plus video NE-PRT. The video narratives were a menu of 8 short stories told by real patients who had varying experiences with pain treatment and opioid use disorder. In the ED, participants in the NE-PRT group were able to watch videos narrated by different storytellers of varying race and age. We studied the effect of the PRT and NE-PRT on (1) knowledge as measured by recall of risk category; (2) patient preference for fewer opioids prescribed at discharge; (3) reduced use of opioids after discharge; (4) functional improvement; (5) patient-provider alignment as measured by concordance between patient preference and finalized prescription plan; and (6) satisfaction with pain control and the presence of shared decision-making (SDM). Participants were stratified by complaint (ie, renal colic or back/neck pain) and hospital center. We administered baseline patient surveys during acute care admission, and follow-up surveys were administered at 7, 14, and 90 days after ED discharge. We sampled a subgroup of participants to qualitatively evaluate patient experiences with pain and treatment in the ED. In-depth, semistructured telephone interviews were completed at least 4 months after enrollment to better understand the patients' experiences with pain management, risk awareness, the visual and video tools, and treatment in the acute care settings.

RESULTS

A total of 1301 participants were enrolled over 24 months. Of these, 434 participants received the GRI sheet (control group), 434 participants received the PRT intervention, and 433 participants received the NE-PRT intervention. The mean age of the participants was 40 years, 53% were female, 43% were White, 38% were Black, and 75% were in the back-pain group. Almost three-quarters of the participants (74.4%) were in the at-risk category, 15.3% were in the high-risk category, and 10.4% were in the highest-risk category of opioid misuse. PRIMARY ANALYSIS: The NE-PRT group had similar recall of their personal risk at 14 days after ED discharge as did the PRT group (43.7% vs 38.8%, respectively; absolute risk reduction, 4.9%; 95% CI, −2.98 to 12.75; = .22). Risk recall was not assessed in the control group (those who only received the GRI sheet), because participants had not received individualized risk estimates. The NE-PRT group had lower rates of preference for opioids than did the control group (25.9% vs 33.0%, respectively; difference, 7.1%; 95% CI, 0.64-0.97; = .023). Self-reported opioid use at 14 days across the NE-PRT, PRT, and GRI groups was 10.5%, 10.3%, and 13.3%, respectively ( = .44), with no difference between groups. There was no difference between groups with respect to alignment of participant preference with provider treatment decision (71.2%, 71%, and 73.6%, respectively; = .591). Participants in the NE-PRT reported higher levels of SDM when compared with participants in the GRI group (7.22/10 vs 6.83/10, respectively; difference, 0.39; 95% CI, −0.05 to 0.83; = .09). Additionally, participants in the NE-PRT group perceived higher levels of participation in decisions about pain treatment than did those in the GRI group (7.28 vs 6.52, respectively; difference, 0.76; 95% CI, 0.07-1.45; = .03). PREPLANNED SUBGROUP ANALYSIS: Outcomes by ORT risk category were examined. Among participants in the high-risk category, the NE-PRT group had greater recall of personal risk (39.5%) than did the PRT-alone group (19.2%; relative risk [RR], 2.06; 95% CI, 1.03-4.13; = .039), and preferences for opioids were decreased in the NE-PRT group (25.0%) vs the GRI group (48.7%; RR, 0.51; 95% CI, 0.32-0.83; = .001). QUALITATIVE ANALYSIS: Thirty-six participants (n = 18 each in the PRT and NE-PRT groups) were administered a structured interview. Salient themes included experiences with treatment, preferences and attitudes toward pain treatment, risk awareness, decision-making and use of tools, and video narrative experiences. Through the qualitative interviews, we found that patients had strong reactions to the interventions and that decision-making and preference regarding analgesia were often influenced by previous experiences with pain treatment.

CONCLUSIONS

We found no differences in 14-day recall of the risk of opioid misuse at the index visit in the NE-PRT group compared with the PRT or GRI groups. However, in a preplanned subgroup analysis, high-risk NE-PRT participants had greater recall of their risk category than did those in the PRT-alone group. We report a decreased preference for opioids in the NE-PRT group vs the GRI group. We found no differences in reported opioid use at 14 days in the NE-PRT group vs PRT-alone group or the NE-PRT group vs the GRI group. We also found no differences in alignment of patient preferences with provider prescribing in the NE-PRT group vs PRT-alone group or the NE-PRT group vs the GRI group. Perceived participation in the decision about pain treatment was greater in the NE-PRT group than in the GRI group. The impact of the interventions on potential mediators of opioid use (eg, recall of personal risk of opioid misuse, preference for opioids, participation in decision-making) was inconsistent and did not appear to affect the rates of opioid use at 14 days This may be due to overall low rates of self-reported opioid use, the persistence of severe pain among a small proportion of participants, or false reassurance from predominantly low-risk assessments and communication in this cohort.

LIMITATIONS

Limitations included 30% of participants being lost to follow-up by day 14, which did not differ across study arm. Social desirability biases may have influenced respondent surveys, because the participants were not blinded to the intervention. Last, the ORT, which was used to communicate risk, has not been validated for predicting opioid misuse in emergency care populations. Strengths of the Life Stories for Opioid Risk Reduction in the Emergency Department (STORRIED) trial include a rigorous randomized experimental design to address how to help patients navigate the inherent risks and benefits of different pain-relief treatment regimens after an episode of moderate to severe acute pain.