The potential benefits of the TIP framework to characterize health services interventions for network meta-analysis

The recent growth of networkmeta-analysis (NMA) has revolutionized outcomes research and our abilities to compare treatments to improve patient outcomes.1 Despite this innovation, most evidence synthesis of health service interventions (HSI) have focused on exploring the effect of interventions rather than how to deliver such interventions.2 To fill this gap, a framework classifying delivery by Theme (T), Intensity (I), and Provider/Platform (P) (TIP framework) was developed.3 In brief, the TIP framework was designed to be analogous to three elements of a medication intervention that contribute to its pharmaceutical effects: (1) active ingredient, (2) dose, and (3) dosage form or route of administration.3 These elements also the constitute of HSI. The theme corresponds to the active ingredient and specifies the main element of the intervention. Intensity corresponds to the dose regimen, while the provider or platform corresponds to the dosage form or route of administration and indicates how the main element is delivered. Three NMAs have been published utilizing this framework to provide granular data on effective delivery techniques. Compared to T-only NMA, which only reports data on type of intervention, this new framework provides further practical data on real-world clinical delivery. In this paper, we compared the findings of three NMAs of HSI with and without the use of the TIP framework. The first NMA to incorporate TIP evaluated patient selfmanagement of asthma, where “T” was the strategy to support asthma self-management (i.e., behavior, educational and psychosocial interventions), “I” was frequency of providing these interventions, and “P” was the provider/platform (i.e., health care personal and e-Health) that delivered information to the patient.4 The secondNMAwith a TIP framework evaluated warfarin self-care strategies (T = Patient selftesting, patient self-management; I = frequency which is more (high) or less (low) often or flexible; and P =Healthcare practitioner, patient, and e-Health) and the third one evaluated music interventions to reduce depression in older adults (T = Active music therapy, receptive music therapy, music medicine; I = >60 min per week (high), ≤60 min per week (low); and P=with or without music therapist). As the use of the framework grows, a gap remains in if the knowledge presented by TIP NMA is pertinent to clinical practice compared to T-only NMA. To investigate the potential benefits of the TIP framework, T-only NMAs were performed using the same data as previously published TIP NMAs and the surface under the cumulative ranking curve