Skip to main content

Association between preoperative medication lists and postoperative hospital length of stay after endoscopic transsphenoidal pituitary surgery

Cite this dataset

Saad, Mary et al. (2022). Association between preoperative medication lists and postoperative hospital length of stay after endoscopic transsphenoidal pituitary surgery [Dataset]. Dryad.


Background: Endoscopic transsphenoidal surgery is the most common technique for resection of pituitary adenoma. Data on factors associated with extended hospital stay after this surgery are limited. We aimed to characterize the relationship between preoperative medications and the risk of prolonged postoperative length of stay after this procedure.

Methods: This single-center, retrospective cohort study included all adult patients scheduled for transsphenoidal pituitary surgery from July 1st, 2016 to December 31st, 2019. Anatomical Therapeutic Chemical codes were used to identify patients’ preoperative medications. The primary outcome was prolonged postoperative hospital length of stay. Secondary outcomes included unplanned admission to the Intensive Care Unit, in-hospital and one-year mortality. We developed a descriptive logistic model that included preoperative medications, obesity, and age.

Results: Median postoperative length of stay was 3 days for the 704 analyzed patients. A prolonged length of stay was defined as > 4 days. Patients taking ATC-H drugs were at increased risk of prolonged length of stay (OR 1.56, 95% CI 1.26-1.95, p<0.001). No association was found between preoperative ATC-H medication and unplanned ICU admission or in-hospital mortality. Patients with multiple preoperative ATC-H medications had significantly higher mean LOS (5.4 ± 7.6 days) and one-year mortality (p<0.02).

Conclusions: Clinicians should be aware of the possible vulnerability of patients taking systemic hormones preoperatively. Future studies should test this medication-based approach on endoscopic transsphenoidal pituitary surgery populations from different hospitals and countries.


Patients’ characteristics and preoperative medications were collected from Cesare™, computerized software for preoperative anesthetic evaluation (Bow Médical, 80440 Boves, France). Preoperative medications were defined using the Anatomical Therapeutic Chemical (ATC) classification system. In the ATC classification system, the active substances are divided into different groups according to the organ or system on which they act and their therapeutic, pharmacological and chemical properties. Drugs are classified in groups at five different levels. Cesare™ associates each medication trade name first with its corresponding International Non-proprietary Name (INN), then with its ATC code according to the ATC Classification System using the first four digits. In our analysis, we used the first level of the code, which indicates the anatomical main group and consists of one letter. There are 14 main groups. Each patient's medical prescription was transformed into numerical variables that indicated the number of drugs in each level-1 ATC class that was present in the preoperative treatment. Surgical procedures were identified using the French classification of homogeneous patient groups.14 This classification, as well as most of the other medico-economic classifications used in the rest of the world, are derived from the Diagnosis Related Groups (DRGs) classification system developed in the 1970s in the United States15 and based on the classification of hospital stays into a deliberately limited number of groups characterized by a double medical and economic homogeneity. Post-operative length of stay and hospital readmission were obtained using the hospital electronic health record system.

Mortality data on the system was synchronized with France’s National Institute of Statistics and Economic Studies (INSEE) database, ensuring a near-complete follow-up after hospital discharge. The INSEE register of death is regularly updated as municipalities send weekly reports of death certificates of their citizens. We looked at in-hospital mortality, defined as any death occurring during a hospital stay, and one-year mortality, defined as any death reported on the INSEE register within one year of surgery.

Data were anonymized before entry onto a secure internet-based electronic case record form designed specifically for our study.

Usage notes

Missing values are identified as NA.


Hôpital Foch