CHARACTERIZING REGIONAL VARIABILITY IN HEALTH RESOURCE UTILIZATION DURING THE DIAGNOSTIC PHASE OF LUNG CANCER ACROSS ONTARIO: A POPULATION-BASED ANALYSIS

Abstract

TOPIC: Lung Cancer TYPE: Original Investigations PURPOSE: Lung cancer (LC) is the leading cause of cancer-related mortality among Canadians. Significant differences in LC relative survival exist within Ontario (ON), though the reasons for this require further elucidation. We aimed to characterize regional variability in LC care through assessment of health resource utilization (HRU) across ON health regions by Local Health Integration Networks (LHIN) to assess variability in access and equity of care and evaluate the impact of HRU on LC outcomes. METHODS: A retrospective population-based cohort study was conducted with ICES, an administrative database in ON, from January 2007 - December 2017. Patients with newly diagnosed LC were identified from the ON Cancer Registry, including if LC was the cause of death. Data were collected from 90 days prior to LC diagnosis until first LC treatment or up to 120 days after diagnosis if no treatment was received, and included: number of staging tests, biopsies, emergency department (ED) visits and admissions. Through OHIP billing codes, specialty appointments with internal medicine, pulmonology, thoracic surgery (TS), medical oncology, radiation oncology, and palliative care were identified. Descriptive data and unpaired t-tests are reported; a multivariate analysis is in progress. RESULTS: LC patients (n=70,749) in ON underwent a mean of 3.20 staging investigations, which varied by LHIN (range 2.52 to 3.61) and by stage (Stage I mean 2.97+/- 1.75 vs Stage IV mean 3.42+/- 1.88, p<0.001). Patients with stage I disease underwent more biopsies compared to stage IV disease (mean all patients = 1.47, stage I = 1.86+/-1.17, stage IV = 1.34+/-1.08, p<0.001), with variability across LHINs (range mean of all patients 1.25 to 1.61). ED visits were more common for patients with stage IV disease compared to stage I (Stage I = 0.92+/-1.56 visits vs Stage IV 1.68+/-1.67 visits, p<0.001), but hospital admissions occurred more frequently in stage I disease (stage I = 0.94+/-0.79 admissions vs stage IV = 0.91+/-0.87 admissions, p=0.001). The number of specialist visits prior to first LC treatment varied between LHIN from an average of 2.9 to 4.9 visits per patient, with patients in some remote areas seeing fewer specialists than in urban regions. There was significant variability in types of specialist visits between LHINs, with mean pulmonology visits ranging from 0.08 to 1.00 per patient and mean TS visits ranging from 0.21 to 1.62. Only 19.9% of LC patients received smoking cessation counselling ranging from 15.1% - 22.4% across LHINs. CONCLUSIONS: HRU varies by disease stage and by LHIN across ON. Further evaluation of regional differences between may explain the variation in LC survival outcomes. CLINICAL IMPLICATIONS: The knowledge gained from this project has the potential to identify disparities in care that may contribute to the variability in LC outcomes across ON and identify target areas for improvement in LC care. DISCLOSURES: no disclosure on file for Micheal Brundage; No relevant relationships by Genevieve Digby, source=Web Response No relevant relationships by Elizabeth Eisenhauer, source=Web Response Advisory Committee Member relationship with AstraZeneca Precision Program Please note: 2017 - present Added 05/14/2021 by M. Diane Lougheed, source=Web Response, value=Honoraria Research relationship with Astra Zeneca Inc Please note: 2018-present Added 05/14/2021 by M. Diane Lougheed, source=Web Response, value=Grant/Research Support Research relationship with GlaxoSmithKlein Please note: 2018-present Added 05/14/2021 by M. Diane Lougheed, source=Web Response, value=Grant/Research Support Speaker/Speaker's Bureau relationship with Canadian Thoracic Society Please note: 2019 Added 05/14/2021 by M. Diane Lougheed, source=Web Response, value=Honoraria Research relationship with Government of Ontario Please note: 2019-present Added 05/14/2021 by M. Diane Lougheed, source=Web Response, value=Grant/Research Support No relevant relationships by Monica Mullin, source=Web Response No relevant relationships by Chris Parker, source=Web Response No relevant relationships by Marlo Whitehead, source=Web Response