Patient-reported outcomes as a tool for enhanced symptom management in gynecologic radiation oncology.

388 Background: Patient Reported Outcomes (PROs) can help guide patient-provider discussions and generate high alert values (HAVs) that prompt clinical response. We analyzed PRO use among Gynecologic Radiation Oncology (GYN RO) patients to identify predictors of HAVs and provider response. Methods: We retrospectively analyzed PRO responses among GYN RO patients seen at our institution from 2022-2024. Patients received PRO-CTCAE surveys 72 hours prior to each consult, follow up (F/U), and on-treatment visit (OTV). HAVs (prespecified high severity/frequency PRO responses) sent alerts to the care team via Epic. Demographic (age, race, marital status), clinical (cancer type, metastatic disease mets, comorbidities, ECOG performance status PS), and treatment (radiation RT, chemotherapy chemo) data were collected. Multivariable logistic regression models (MVA) were used to identify variables associated with 1) patients reporting at least 1 HAV and 2) documented provider response to HAVs within 7 days. Results: Among 369 patients completing at least 1 survey, the median age was 63 (IQR 52-70). Most common cancer types included uterine (50%) and cervix (31%). Only 34% had mets. Within this cohort, 45% generated at least 1 HAV. Common HAVs included pain (42%), vaginal bleeding (27%), anxiety (10%), and urinary symptoms (7%). On MVA, higher ECOG PS (score 2: OR 2.78, p<0.01; score 3: OR 4.12, p=0.02; score 4: OR 23.61, p<0.01; ref score 0) and diagnoses of depression (OR 2.27, p<0.01), chronic pain (OR 3.24, p<0.01), neuropathy/neuralgia (OR 1.64, p=0.04), and sleep disorders (OR 1.92, p=0.01) were associated with higher HAV risk. Compared to consult visits, OTVs were less likely to generate HAVs (OR 0.40, p<0.01). HAV risk did not differ with age, race, marital status, cancer type, mets, concurrent chemo, RT modality/fraction, or other comorbidities. Providers responded to pain, vaginal bleeding, and urinary HAVs 94%, 79% and 60% of the time, while response rates for anxiety HAVs were 33%. On MVA, provider response was more likely for patients with higher Charlson Comorbidity Index (CCI) scores and non-uterine cancers (Table). Provider response was less likely prior to F/U visits or for non-pain related HAVs. Provider response did not differ with age, race, marital status, ECOG PS, concurrent chemo, or RT modality/fraction. Conclusions: 45% of patients completing PROs generated HAVs. Providers frequently documented interventions for pain and bleeding related HAVs, but infrequently intervened for mental health-related HAVs. This may indicate room for improvement in how psychosocial care is integrated into oncology care. Significant variables Provider response (%) OR p-value Visit type Consult 79 Ref F/U 62 0.20 0.001 HAV type Pain 94 Ref Vaginal Bleeding 79 0.10 <0.001 Anxiety 33 0.02 <0.001 Urinary Symptoms 60 0.08 <0.001 Other 57 0.09 <0.001 Uterine cancer Yes ( Ref = no) 70 0.45 0.047 CCI 1.32 <0.001