Social vulnerability and fragmentation of postoperative surgical care among patients undergoing hepatopancreatic surgery

Abstract

Presenter: Djhenne M Dalmacy MS | The Ohio State University Background: Regionalization of hepatopancreatic (HP) surgery to high-volume hospitals has been associated with fragmentation of post-discharge surgical care and, in turn, inferior outcomes. The objective of the current study was to assess whether patient social vulnerability was associated with the chance of fragmented postoperative care following HP surgery. Methods: Medicare beneficiaries who underwent HP surgery and had at least one readmission within 90 days were identified using Medicare 100% Standard Analytical Files from 2013 to 2017. Patients were categorized into four groups based on their social vulnerability level (i.e. low, low-moderate, moderate-high, and high). Fragmented postoperative care was defined as readmission at a hospital other than the index institution where the initial surgery was performed. The association of social vulnerability index (SVI) and the individual components comprising SVI [Socioeconomic, Household Composition & Disability (HC & Disability), Minority Status & Language (MS & Language), Housing Type & Transportation (HT & Transportation)] with fragmentation of postoperative care was examined using multivariable logistic regression analysis. Results: Among 11,142 patients who met the inclusion criteria, 8,053 (72.3%) had undergone pancreatectomy, while 3,089 (27.7%) had undergone hepatectomy. The overall incidence of fragmented care was 32.9% (n = 3,667). Patients who had fragmented care were older (fragmented: 73 years [IQR: 69-77] vs. 72 years [IQR: 68-77]) and had a higher comorbidity score (fragmented: 4 [IQR: 2-8] vs. non-fragmented: 3 [IQR: 2-8]) (both p<0.001). Median overall SVI was higher among patients who received fragmented care (fragmented: 52.5 [IQR: 29.3 – 70.4] vs. non-fragmented: 51.3 [IQR: 27.9 – 69.4], p = 0.03). When examining each of the components comprising SVI, patients who received fragmented care had higher socioeconomic index (fragmented: 52.5[IQR: 29.3-70.4] vs. non-fragmented: 51.3 [IQR: 27.9-69.4]), higher HC & disability index (fragmented: 41 [IQR: 22.1-59.8] vs. non-fragmented: 39.3 [IQR: 19.4-57.4]) and lower MS & language index (fragmented: 76.3 [IQR: 54.5-90.3] vs. non-fragmented: 78.4 [IQR: 58.6-91.4]) (all p<0.05). On multivariable analysis, odds of receiving fragmented care were higher among patients with high socioeconomic vulnerability (referent low; OR: 1.24; 95% CI: 1.11 – 1.39), high HC & Disability (OR: 1.30; 95% CI: 1.16 – 1.46; p<0.001), whereas high MS & Language vulnerability was inversely associated with receipt of fragmented postoperative care (OR: 0.84; 95% CI: 0.75 – 0.94; p = 0.002) (Figure 1a). The possibility of fragmented care increased with increasing overall SVI (high vs low SVI; OR: 1.12; 95% CI: 1.01 – 1.26; p = 0.04) (Figure 1b). Conclusion: Approximately one-third of Medicare beneficiaries had fragmented postoperative surgical care following HP surgery. Social vulnerability was strongly associated with the chance of receiving fragmented postoperative care. Social determinants of health including social vulnerability should be considered when assessing patient risk of fragmented care after HP surgery.