Cognitive deficit identified via patient reported outcome measures may be a predictor for post-operative health care utilization (131)

• Published in: Gynecologic oncology 2022-08, Vol.166, p.S82-S82
• Main Authors: Saini, Aashna, MD, Orav, Endel, PhD, Philp, Lauren, MD, MSc, FRCSC, Mazina, Varvara, MD, Sullivan, Mackenzie, MD, Alimena, Stephanie, MD, del Carmen, Marcela, MD, MPH, Goodman, Annekathryn, MD, MPH, MS, MA, Eisenhauer, Eric, MD, Sisodia, Rachel, MD
• Format: Article
• Language: English
• Subjects: Hematology, Oncology, and Palliative Medicine; Obstetrics and Gynecology
• ISSN: 0090-8258
• DOI: 10.1016/S0090-8258(22)01356-7

Objectives: As care models become increasingly value-based, understanding healthcare utilization and predictors for high utilization are paramount. Patient Reported Outcome Measurement surveys (PROMs) offer a valuable method to evaluate a patient’s clinical course more accurately and gain insight into how they manage at home and with their disease. The purpose of this study was to determine if preoperative PROMs examining social, mental, emotional, and physical health could be used to predict increased healthcare utilization in the postoperative period. Methods: PROMs were collected monthly in all patients undergoing care for gynecologic malignancy between January 1, 2018, and September 1, 2019, as described previously. For patients who underwent surgery, charts were reviewed to ascertain the amount of healthcare utilization that occurred in the 90-day postoperative period. Healthcare utilization was grouped into the following categories: phone calls, emails/portal messages, oncology office visits, other office visits, Emergency Department encounters, and inpatient hospital admission. Descriptive statistics, non-parametric tests (Spearman’s correlation), and logistic regression were performed where appropriate. Results: A total of 236 patients completed preoperative PROMs. Of all PROMs items, “difficulty remembering things” was strongly associated with postoperative healthcare utilization (see Table 1). This relationship was present even when controlling for other variables (age, body mass index, race, length of stay, disease site, surgery/surgical complications, and ASA class). As difficulty with memory increased, healthcare utilization increased as well (regression coefficient 2.3, p=0.007). When controlling for quality of life measures against one another (i.e., trouble with strenuous activities, depression, treatment interfering with family life, difficulty remembering), memory impairment remained a statistically significant prognostic indicator compared to other variables for total hospital touches (regression coefficient 0.74, p=0.03). Conclusions: Patients who report impairment in memory are significantly more likely to have unplanned re-admissions and office visits postoperatively. Preoperative collection of PROMs can identify this cohort of vulnerable women and suggest that further studies should be done to see if additional support around surgery can mitigate this risk. As healthcare systems move towards more value-based approaches which require