Within-case train-of-four response to neuromuscular blockade redosing: an interval-level analysis, and why case-level re-blockade rates are a monitoring artefact

This is a single-centre, retrospective, observational pharmacodynamic (exposure–response) study using a public de-identified operating-room database (the SIS subset of MOVER; 19,114 surgeries). The unit of analysis is the interval between two consecutive intraoperative train-of-four (TOF) readings, not the case. For every adjacent pair of readings (t_i, v_i) -> (t_j, v_j), the outcome is the change dv = v_j - v_i (integer, -3 to +3; negative = deepening block), and the exposure is whether, and how much, non-depolarising neuromuscular blocking agent (NMBA) was given within that interval. We estimate the adjusted association between interval-level dosing and the change in TOF, and separately quantify the extent to which case-level "re-blockade" endpoints are explained by monitoring intensity. No blinding is involved (observational study; no treatment was assigned by the investigators); the analysis uses the full available sample with no interim stopping rule. We make no causal claim and do not examine postoperative outcomes (the source database contains no outcome fields).