The Incremental Risk of Intraoperative Fentanyl on PONV

The primary aims of this quality control study are:

• to determine whether there is an independent association between the intraoperative administration of fentanyl (both as a continuous and categorical variable) and PONV within 24 postoperative hours
• determine whether or not the intraoperative fentanyl amount can be used to successfully restratify patients for PONV beyond the simplified Apfel score for PONV

All required data is collected in regular daily practice. The following variables will be assessed by univariable models:

Preoperative variables (e.g. patient characteristics) Surgical/Anesthesia variables (e.g. type/length of surgery, antiemetics) Outcome variables (PONV, worst pain as delineated in the outcomes section), Intraoperative fentanyl measured 1) as a continuous variable and 2) by predefined limits of 0 to 0.2mg, >0.2-0.5mg, >0.5 to 0.8mg, and >8mg.

Finally for the multivariable model, the investigators will adjust for confounders, specifically for sex (m/f), history (Hx) of PONV (y/n), Hx of motion sickness (y/n), current smoker (y/n), age (years), expected duration of surgery (min; >60 min), postoperative opioids (y/n), the use of preemptive PONV Therapy (0,1,2 medications), propofol vs. inhalative agents (y/n), and surgery type (3 kinds) from the patient's electronic record. In the event that insufficient event rates are observed, we will collapse the variables into the simplified Apfel score to assess the independent value of fentanyl in predicting PONV.

In addition to a logistic model as delineated above, we will conduct the following:

• receiver operating characteristic (ROC) curves of the logistic model without fentanyl, with fentanyl (continuous), and with fentanyl (categorical).
• additionally, ROCs of a simplified model using the Apfel score, the Apfel score with fentanyl (continous), and the Apfel score with fentanyl (categorical) will be made. Comparison of area under the receiver operating characteristics(AUROCs) by DeLong
• net reclassification improvement (NRI) of both the logistic model with and without fentanyl as well as the Apfel score with and without fentanyl

To assess pain and administered fentanyl, we will examine pain scores by the numeric rating scale (NRS).

Inclusion criteria

• In-patients with an expected simplified Apfel score of 2 or higher,
• Aged >= 18 years, and
• Undergoing general anesthesia (with or without neuraxial or regional anesthesia.)
• in abdominal, gynecological or otorhinolaryngological surgery

Exclusion criteria

• Outpatients
• Patients with chronic pain, defined as recurring pain requiring intermittent hospitalization or regular intake of pain medication.
• Non-opioid naïve patients, defined as a history of abuse or having taken opioids within the last 30 days