There is an urgent public health need to reduce reliance on opioids for effective long-term pain management, particularly in knee osteoarthritis (KOA). This effectiveness trial will compare commonly recommended treatments to reduce pain and functional limitations in KOA.These results will lead to improved patient selection for treatment and inform evidence based guidelines by offering well-tested, effective, non-surgical alternatives.

Knee osteoarthritis (KOA) is one of the leading causes of chronic pain and disability worldwide, affecting over 30% of older adults. It represents a major global health and economic burden to individuals and society. The rates of KOA have more than doubled in the past 70 years and continue to grow sharply, given increases in life expectancy and population body mass index (BMI). Surgery is often employed to treat KOA, but it is associated with a high rate of persistent pain, and is not a permanent solution. Numerous nonsurgical therapies have been advocated to treat pain in patients with KOA yet are not often used in clinical care. The limited pain relief and functional improvement seen in a subset of knee OA sufferers has led to a high rate of opioid use and disability in this population. The overarching goal of this study is to conduct a sequential parallel group randomized controlled trial (RCT) to evaluate the comparative effectiveness of conservative behavioral and non-opioid pharmacological treatments (Phase 1) and, among those that indicate interest in obtaining further treatment and those not eligible for conservative treatment, the benefits of procedural interventions (Phase 2). This study will also evaluate whether clinical and psychosocial phenotypes predict short- and longer-term treatment response. The results of this study will examine the effectiveness of each tested intervention and provide meaningful information regarding effectiveness across key subgroups of participants.

Inclusion Criteria:

• Knee pain score of ≥4 and ≤ 9 on the Modified 4-Item BPI Pain Scale at pre-intervention screening
• Meets at least 1 of the 3 American College of Rheumatology (ACR) Classification criteria for knee osteoarthritis.

ACR criteria are:

1. At least three of the following using history and physical examination: age >50 years old; morning stiffness <30 minutes; crepitus on knee motion; bony tenderness; bony enlargement; no palpable warmth
2. At least one of the following using history, physical examination, and radiographic findings + the presence of osteophytes: age >50 years old; morning stiffness <30 minutes; crepitus on active motion and osteophytes
3. At least 5 of the following using history, physical examination, and laboratory findings: age >50 years old; morning stiffness <30 minutes; crepitus on knee motion; bony tenderness; bony enlargement; no palpable warmth; erythrocyte sedimentation rate (ESR) <40 mm/hour; Rheumatoid Factor (RF) <1:40; synovial fluid signs of osteoarthritis

Exclusion Criteria:

• <18 years of age
• Any inability to complete study procedures, including, but not limited to inadequate resources to mitigate low English language literacy
• Refusal of randomization
• Knee pain exclusions: Pain during an average of < 4 days per week over the past 3 months; pain in the index knee from a joint disease other than OA (e.g., infectious arthritis, rheumatoid arthritis, spondyloarthropathy)
• Medication exclusions: Report changes in analgesic medication dose within 2 weeks of baseline; oral morphine equivalent dose of > 90 mg/d at baseline
• Medical condition exclusions: Severe vision or hearing impairment or any signs of cognitive impairment that would prevent comprehension of consent procedures, study measures, or procedures; unstable medical condition that presents an absolute or relative contraindication for participation in both arms (e.g., unstable angina, congestive heart failure); poorly controlled serious psychiatric condition that could prevent full participation or affect outcomes (e.g., suicidal ideation, active psychosis, poorly controlled depression, active substance abuse [excluding tobacco, caffeine or moderate alcohol use])
• Knee-specific medical condition exclusions: History of bilateral knee joint replacement arthroplasty total knee arthroplasty (TKA) or TKA in the affected knee; partial replacements may be eligible depending on physician judgment; scheduled joint replacement; history of unilateral TKA and complaints of KOA pain limited to the operated knee; Intra-articular viscosupplementation, steroid injection or arthroscopic surgery in the index knee within 12 weeks of baseline
• Pregnancy by self-report, report of intention to become pregnant (Phase 1), or as determined by urine pregnancy screening (if Standard of Care at site) (Phase 2). Due to the unknown effects of duloxetine on the developing fetus and newborn, and the potential harms of fluoroscopy in pregnancy, women who are pregnant or lactating or intend to get pregnant will not be included in this study. Those of childbearing potential will be asked to use reliable contraception during the course of their participation in the study and to notify the study team if they become pregnant during participation. Definition of reliable birth control will be defined as: Female and male sterilization (female tubal ligation or occlusion, male vasectomy); long-acting reversible contraceptives (LARC) methods (intrauterine devices, hormonal implants); short-acting hormonal methods (pill, mini pills, patch, shot, vaginal ring); barrier methods (condoms, diaphragms, sponge, cervical cap)

Phase 1 specific Exclusion Criteria- An individual who meets any of the following criteria will be excluded from participation in Phase 1 of this study and will be enrolled and randomized directly into Phase 2:

• Known allergic reaction or medical condition that renders an individual unsuitable for Phase 1 study interventions, including closed-angle glaucoma, kidney disease (creatinine clearance < 30 mL/ min), severe liver disease, known adverse reaction to duloxetine or another selective serotonin-norepinephrine reuptake inhibitor (SNRI), bipolar disorder or mania, high likelihood of drug interactions that could lead to side effects (e.g., serotonin syndrome in people on multiple drugs that inhibit serotonin reuptake including monoamine oxidase (MAO) inhibitors).
• Report failed trial of an adequate dose of duloxetine to relieve KOA symptoms over a 1-month period
• Have tried and failed two of the following: NSAIDS, physical therapy (there are many physical therapies so clinicians should exercise their judgment as to what constitutes 'failed' therapy), or weight loss (need determined by clinician) and refuses participation in Phase 1
• End-stage renal disease
• Unreliable access to the internet on a daily basis, i.e., sufficient access to participate in the study and may include public library access, cafe/coffee shop spaces, access to a friend or neighbor's wifi or hotspot, etc. (reliability determined on a site-by-site basis)

Phase 2 specific exclusion criteria- An individual who meets any of the following is excluded from sequential participation in Phase 1 and then Phase 2, and will instead be randomized as a "solo" recruit into Phase 1:

• Inability to pay for interventions (insurance or otherwise)
• Medical condition exclusions: Untreated coagulopathy that could interfere with Phase 2 interventions; for automated implantable cardioverter-defibrillator that cannot be disabled before radiofrequency ablation (RFA), the investigator can consult cardiology, bioengineering or the device manufacturer before enrolling in phase 2 (i.e. not a definite exclusion criterion)
• Knee specific medical condition exclusions: Current local infection in the knee; ulcers or an open wound in the region of the index knee; underwent an adequate trial of any Phase 2 procedural study intervention in the study knee; severe needle phobia that cannot be addressed pharmacologically

Upon completing data collection, specifically after the last patient's main outcome visit (excluding long-term follow-up), the HEAL Pain Management Effectiveness Research Network (ERN) Data Coordinating Center (DCC) at the University of Utah will compile the final study dataset. This dataset will undergo statistical analyses and form the basis for publishing both primary and secondary trial results. The SKOAP dataset, recognized as valuable for pain management and opioid use research communities, will be formatted for accessibility to non-ERN investigators. Additionally, a data dictionary will accompany the dataset, offering concise definitions for each included data element. Any peculiarities impacting interpretation or analysis will be outlined within the dictionary. Unreliable data elements will be removed, with documentation provided accordingly.

These policies are expected to focus primarily on the timing of data release. The investigators preliminary plan is to release the database (defined below) and supporting information at the time of publication of the primary manuscript, or within 12 months of last patient's final follow-up, whichever comes first. Implementation of the plan will follow the HEAL Public Access and Data Sharing Policy as outlined at https://heal.nih.gov/about/public-access-data.

Access to the releasable database housed in the NIH-assigned repository will be in accordance with procedures and regulations of the NIH or specific institute. The data coordinating center will not provide any support for investigators using the releasable database.

Individuals randomized to a nerve procedure will be blinded to whether they have a long acting block or nerve ablation.