S1339 Healthcare Resource Use and Associated Costs of Cyclic Vomiting Syndrome (CVS) in the US

INTRODUCTION: Cyclic vomiting syndrome (CVS) is a chronic disorder of gut-brain interaction characterized by episodic vomiting. CVS imposes a significant burden on patients and the healthcare system, but the extent of this burden remains unclear. We estimated the direct cost burden of CVS, measured by healthcare resource use (HRU) and its associated costs, using large claims databases in the US. METHODS: MarketScan Commercial and Medicare Supplemental databases were employed to define a cohort of CVS patients as those with ≥ 1 inpatient (IP) or ≥ 2 outpatient (OP) diagnosis claims for CVS (ICD-10-CM G43A0 or G43A1) between 10/1/2015 and 6/30/2019 (study period). We only included individuals with continuous insurance enrollment for ≥ 12 months prior to (baseline) and for ≥ 3 months following the first CVS diagnosis claim (index date). The follow-up extended from the index date to either insurance disenrollment or the end of study period, whichever comes first. We used propensity score of each CVS patient, reflecting baseline demographic and clinical characteristics, to match with up to 3 non-CVS controls whose index dates were randomly drawn (Table 1). HRU and its associated costs (insurer paid amounts; 2019 $) in CVS patient and non-CVS control groups were annualized to accommodate follow-up lengths. With multivariable regressions to further balance the CVS and non-CVS groups, we then measured the differences in HRU and costs between the study groups to quantify the direct cost burden of CVS. RESULTS: A total of 7,413 CVS patients were matched to 19,453 non-CVS controls (Table 1). CVS patients had significantly higher average annualized HRU, with the largest differences in hospital IP (0.9 vs. 0.1 IP stays; 11.7 vs. 7.1 IP days) and emergency room (ER) (1.9 vs. 0.4 ER visits) use (all P < 0.001, Table 2). CVS patients had significantly higher total healthcare costs ($57,140 vs. $14,912), with IP costs as the primary driver ($28,522 vs. $3,250) of the difference (all P < 0.001, Figure 1). After multivariable regression adjustments, total healthcare costs remained 4.1 times higher for CVS patients relative to non-CVS controls, with IP costs 12.3 times higher, ER costs 5.8 times higher, OP costs 2.9 times higher, and OP pharmacy costs 1.5 times higher (all P < 0.001). CONCLUSION: Newly diagnosed CVS patients have greater healthcare utilization and costs than matched non-CVS counterparts, suggesting substantial direct cost burden to the US healthcare system.Table 1.: (1) Matching covariates include, A) Demographic: age groups, sex, payer type, insurance plan type, geographic region, urbanicity; B) Clinical: Deyo-Charlson Comorbidity Index, baseline comorbid conditions (abdominal pain, anxiety, autonomic dysfunction, cannabis use/abuse, cardiac conditions, depression, fibromyalgia, gastroesophageal reflux disease [GERD], gastroparesis, irritable bowel syndrome, migraine, nausea, and seizure); and C) Economic: log transformation of baseline total costs. | (2) Only baseline comorbidities with prevalence rate ≥ 5% are presented. | *Covariates with standardized difference < ± 10 between the matched cohorts, indicating a well-balanced match. | Multivariable regressions were applied to further adjust. | § “Follow-up duration (in days)” and “Baseline total cost, observed” were NOT matching covariates, thus, they were not checked on balanceTable 2.: Note: Proportion of patients with specific services were identified during the entire follow-up period (mean length of 559 days for CVS patients and 571 days for non-CVS controls), while the quantities of healthcare services were annualized to adjust for different lengths of follow-upFigure 1.: Note: P < 0.001 for all comparisons on total healthcare costs and setting-specific costs between CVS patients and non-CVS controls.