Following the 2011 guidelines of the American Society of Anesthesiologists, capnography, a noninvasive measurement of exhaled carbon dioxide, has become the standard of care for ventilation monitoring of sedated patients. This invention quantitatively identifies a patient's sedation level and predicts adverse events based on information from the capnogram and outputs from pharmacokinetic, pharmacodynamics or ventilatory models. This system is comprised of a breath receiver, a sensor and a processor. The sensor measures the carbon dioxide concentration of air exhaled by the patient into the breath receiver. The processor processes the sensor data to generate capnograms, and computes the outputs of pharmacokinetic, pharmacodynamic and ventilatory models. The processor also extracts the resulting features from these models and a multi-parameter metric is computed based on the extracted features to determine the current or predicted sedation level of the patient. Assessing an individual's ventilatory state during sedation provides a unique perspective on patient's clinical status. In particular, parameter extraction and predictive methods derived from capnography and other physiological data can improve the safety and operational effectiveness of the sedation.