Letter to the editor about preliminary evaluation of the clinical efficacy of salmon calcitonin nasal spray in patients with rotator cuff tears after arthroscopic repair: a multi-center retrospective cohort study

Dear Editor, We read with interest the multicenter retrospective cohort study by Lyu et al, evaluating salmon calcitonin nasal spray (SCNS) after arthroscopic rotator cuff repair, which reported improved early function and lower re-tear rates over 2 years[1]. Given the clinical appeal of a low-burden pharmacologic adjunct during postoperative immobilization, several clarifications could strengthen causal interpretation and facilitate translation into practice. First, the postoperative strategy in the SCNS group was not a single-variable intervention: patients received SCNS alongside calcium/vitamin D supplementation, whereas controls received “conventional medication” only[1]. This creates an attribution problem – any observed benefit could reflect SCNS, calcium/vitamin D, their interaction, or differential adherence. Contemporary causal-inference guidance emphasizes specifying the estimand and the well-defined intervention(s) being compared in observational analyses[2]. At a minimum, reporting adherence to each component and performing sensitivity analyses that treat supplementation as a co-intervention (rather than background care) would reduce ambiguity; meta-epidemiological evidence also suggests imbalanced co-interventions can bias effect estimates[3]. Second, the conclusion regarding re-tear reduction appears to rely primarily on between-group comparisons without comprehensive confounding control. Re-tear risk is strongly related to tendon and muscle quality, tear size/chronicity, bone quality, rehabilitation intensity, and center/surgeon effects – many of which may influence postoperative prescribing. Propensity-score approaches (matching/weighting) and transparent balance diagnostics are widely recommended for nonrandomized treatment comparisons[4]. We encourage re-analysis using a prespecified confounder set (ideally DAG-informed), incorporating center as a stratification variable or random effect, and reporting robustness checks (e.g., quantitative bias analysis/E-values, where appropriate). Third, the imaging endpoint requires more methodological detail. Re-tear ascertainment depends on modality, timing, definitions (e.g., structural integrity grades), and reader reliability. Blinded, independent assessment and inter-/intra-observer agreement (kappa/ICC) are particularly important because subjectivity can inflate apparent treatment effects when outcome assessment is not optimized[3]. This is salient in shoulder imaging, where reliability varies across commonly used classifications and features[5]. A brief description of reader training, blinding to treatment allocation, and agreement metrics (or a plan for central adjudication in future studies) would materially increase confidence in the reported structural outcomes. Overall, Lyu et al provide hypothesis-generating data supporting further evaluation of perioperative antiresorptive/analgesic strategies after rotator cuff repair[1]. Addressing the co-intervention structure, confounding control, and imaging adjudication would clarify whether SCNS offers an independent, scalable benefit and would inform the design of a pragmatic randomized trial (e.g., standardized calcium/vitamin D in both arms with blinded imaging endpoints). Finally, the manuscript adheres to the TITAN guidelines and was assessed using the TITAN checklist[6]. Ethical approval Not applicable. Consent Not applicable. Sources of funding Not applicable. Author contributions X.Q.: and C.L.: conceived the manuscript, B.C.: revised the manuscript.