Osteoarthritis (OA) affects 528 million individuals globally, predominantly in knee and hip joints, with a notable impact on females aged over 55, resulting in a substantial economic burden. However, the efficacy of modalities used in physiotherapy to manage OA pain for reducing the need for joint replacement remains an open question, and guidelines differ. Our systematic narrative review, drawing from reputable databases (e.g., PubMed, Cochrane, and CINAHL) with specific Mesh terms investigated evidence from 23 Randomized Controlled Trials (that included a control or a sham group in 30 different protocols) using therapeutic modalities like ultrasound, diathermy, and electrical stimulation for knee and hip OA pain, involving a total of 1055 subjects. We investigated the attainment of minimal clinically important differences in pain reduction, operationalized through a 20% decrement in the Western Ontario and McMaster University Arthritis Index or Visual Analog Scale (VAS) score. Our results indicated that 15 protocols out of 30 reach that level, but there were no statistical differences among modalities. Half of the protocol presented in the literature reached clinical efficiency but studies on hip remains scarce. We recommend a comprehensive, sequential, and multimodal intervention plan for individuals with joint OA with initial transcutaneous electrical nerve stimulation and progressing to a 2-week protocol of continuous ultrasound, potentially combined with deep microwave diathermy. Long-term intervention involves the use of pulsed electrical stimulation. For hip OA, a cautious approach and discussions with healthcare providers about potential benefits of spinal cord nerve stimulation.