Pragmatic Comparative Effectiveness Trial of Evidence-based, On-demand, Digital Behavioral Treatments for Chronic Pain

Transcending COVID-19 Barriers to Pain Care in Rural America: Pragmatic Comparative Effectiveness Trial of Evidence-based, On-demand, Digital Behavioral Treatments for Chronic Pain

This study compared two available, evidence-based, at-home digital pain treatment programs. The goal is to see if one approach is better than the other, and whether certain patients respond to one more than the other. Study participants will be randomly assigned to receive one of two treatment programs: Skills-Based VR or painTRAINER. VR devices and painTRAINER manuals will be delivered to the participant's home with instructions for use via FedEx; participants will receive remote technical support. They will be followed for 12 weeks and complete Patient Reported Outcome (PRO) questionnaires to assess functional status, pain levels, and use of pain medications (including opioids). Participants will also be asked to provide consent/authorization to access medical records from their treating facility.

We performed a two-arm, multi-center, virtual randomized controlled trial (vRCT) in a geographically diverse group of patients with mixed-etiology chronic pain. Using a random number generator, patients were allocated in a 1:1 ratio between two self-administered, remotely deployed CBT delivery platforms, stratified by site: (1) a 2D mHealth app called PainTRAINER; and (2) 3D virtual reality (VR) app called EaseVRx+. The primary analysis compared changes in pain intensity over two months upon completion of the standardized 8-week digital CBT protocols. Secondary outcomes include pain catastrophizing, anxiety, pain interference, self-efficacy, and opioid use. Patient blinding is not possible in VR versus non-VR studies; however, researchers exhibited equipoise when describing the competing interventions, Investigators and data analysts were blinded to patient allocation; coordinators administering the intervention were unblinded in order onboard participants to their assigned program and troubleshoot any difficulties throughout the study.

The population eligible for the study include individuals with any of more than 130 ICD codes for chronic pain conditions, including somatic, musculoskeletal, neuropathic, and visceral pain who also recorded an average pain intensity of 4 or greater on a standard 0-10 numeric rating scale (NRS) over the past week. Patients were recruited from three major sites: (1) Cedars-Sinai Health System (CSHS); (2) Ochsner Health; (3) UAB. All sites used search tools to identify participants within their respective electronic health record (EHR) with at least one of the inclusionary chronic pain ICD codes and who did not meet several coded ineligibility criteria, included but not limited to ongoing end-of-life care and an age less than 13 years old. Patients meeting preliminary eligibility were chart reviewed for confirmation, then contacted through methods approved by each respective site's IRB, which included combinations of email, physical mail, and patient-portal notifications. Patients were offered an opportunity to opt-in or opt-out, and those who do not respond were contacted by phone call to assess interest and eligibility.

This trial was conducted remotely using patient identification, screening, and monitoring processes developed by the study team and applied in many trials involving virtual reality, including other NIH-sponsored trials using VR therapies for pain (NCT04409353, NCT04907643). Participants who provided eConsent and completed a 7-day "screener week" on REDCap were randomized and sent study materials by mail. Study onboarding was performed over the phone and did not require in-person interactions with the study team.

We employed a biopsychosocial conceptual framework for monitoring outcomes in the trial, focusing on pain intensity as the primary outcome, and included a range of other relevant PROs selected in partnership with our patient partners. PROs were collected via REDCap. Pain intensity was measured using a standard 11-point numeric rating scale (NRS) with a 24-hour recall. Consistent with NIH Helping to End Addiction Long-Term (HEAL) guidance, we measured daily pain NRS for 7-days during baseline (week 0), and again during the final week of the digital program study (week 8). Intention-to-Treat (ITT) analysis of the primary endpoint involved comparing the proportion of individuals who recorded a minimally clinically importance difference (MCID) of two on the average pain NRS recorded between Week 8 and Week 0. Secondary outcomes include the 4-item short form pain catastrophizing scale, 4-item NIH PROMIS Anxiety scale, 4-item NIH PROMIS Pain Interference scale, 2-item Pain Self-Efficacy Questionnaire, and average opioid usage derived from total morphine milligram equivalents (MMEs) in self-reported medication use. In addition, VR presence and cybersickness at the start of intervention use using the Presence Scale and Simulator Sickness Questionnaire (SSQ), respectively.

We will recruit patients who:

(1) have chronic pain, from any underlying condition, using the administrative definition of ICD-10 code series G89.X or one or more of 134 chronic overlapping pain condition codes, as previously standardized and validated by an expert panel, (2) have experienced average pain intensity of >3 out of 10 within the previous week, (3) are ≥13 years of age, (4) are able to read/write English, (5) have either a personal computer or a smartphone, and (6) live in a designated rural zip code as defined by the Federal Office of Rural Health Policy (FORHP) data (RUCA Codes 4-10).

We will exclude patients who:

(1) have a condition that interferes with use of the intervention (e.g., significant visual or auditory impairment, any history of seizure), (2) are hospitalized, (3) are receiving active cancer treatment, (4) are receiving end-of-life care, (5) have cognitive impairment that affects participation, or (6) have previously used the EaseVRx+ or painTRAINER programs.

We will stratify analyses across key patient characteristics, including type of pain condition, sex, age, race, ethnicity, State and zip code, Rural-Urban Commuting Area (RUCA) codes (a composite measure of population density, urbanization, and daily commuting), pain severity, opioid use, comorbidities, social support, and COVID-19 impacts (unemployment, financial hardship, family or personal COVID-19 diagnosis).