Burden of Housing Insecurity among Acute Care Surgery patients: A Patient level Social Determinants of Health Study
Joshua A. Sznol, Sarah J. Ullrich, Kevin Schuster
Surgery, Yale, New Haven, Connecticut, United States
1. Objective Social determinants of health (SDOH) are frequently studied modifiers of patient outcomes in epidemiologic datasets. Few studies have used patient level data to analyze how social determinants correlate with individual outcomes. The objective of this study is to determine how the social determinants of health impact patient outcomes in order to identify at risk populations.
2. Design Retrospective cohort study
3. Setting Large academic medical center, inpatient Acute Care Surgery service
4. Patients From the years 2019-2020, consecutive patients admitted to or operated on by the Acute Care Surgery service in a large academic hospital and who underwent social work evaluations (SWE) were identified. Patient demographics, medical history, Elixhauser score (ES), admission vitals, SWE data, NSQIP data, and operative outcomes were obtained. The SDOH identified included housing insecurity, alcohol use, financial insecurity, food vulnerability, employment status, and education level.
5. Interventions None
6. Main Outcome Measures Outcomes included: mortality, length of stay (LOS), unplanned return to the operating room (OR), and discharge disposition were studied. Secondary outcomes studied included post-operative complications of myocardial infarction (MI), deep venous thrombosis (DVT), pulmonary embolism (PE), acute kidney injury (AKI), and hospital acquired pneumonia (PNA). Mortality, return to the OR, and all secondary outcomes were analyzed utilizing bivariate logistic regression. LOS was analyzed utilizing a negative binomial regression. The SDOH were modeled individually. Additional associations of the social determinants and all outcomes were analyzed with chi-sq analysis.
7. Results 149 patients with SWE were identified. After controlling for age, gender, ES, number of ventilator days, wound class, presence of peritonitis, bowel resection, organ space SSI, education, employment, and alcohol use, none of the social risk factors were predictive of mortality or unplanned return to the OR. Housing insecurity (Table 1, p=.002) was a significant predictor of LOS. AKI varied by education level (p=.016) and was most common among those with a high school degree (69.2%). PE was more common among those with high school education (50%, p=.049). Operative complications of DVT, MI, PNA, organ space SSI, need for bowel resection, unplanned return to the OR, and disposition did not vary by any of the social determinants.
8. Conclusions Using patient level data, certain social risk factors are associated with LOS in emergency general surgery patients. Early identification of social barriers to discharge, particularly housing security, in acute care surgery patients may help facilitate disposition and reduce LOS. The mechanism by which SDOH affect LOS should be elucidated.
|Variable||Median (IQR); n=149||Parameter Estimate (95% CI)||p-value|
|Age||55 (42,69)||-0.029 (-0.081-0.022||0.269|
|Elixhauser Score||5 (3,8)||0.031 (-0.090-0.153)||0.612|
|Length of Stay||6 (3,13)||*||*|
|Ventilator Days||0 (0,0)||0.099 (-0.006-0.204)||0.064|
|Female||76 (51.01%)||-0.471 (-1.924-0.982)||0.526|
|Clean contaminated||48 (41.38%)||0.209 (-1.172-1.589)||0.767|
|Contaminated||16 (13.79%)||1.181 (-0.494-2.856)||0.167|
|Dirty||36 (31.03%)||1.822 (0.195-3.450)||0.028|
|Presence of peritonitis||50 (42.37%)||-0.743 (-1.695-0.210)||0.126|
|Bowel resected||38 (32.76%)||1.504 (0.096-2.911)||0.036|
|Organ Space SSI||21 (17.8%)||-1.301 (-2.907-0.304)||0.112|
|Housing Insecurity||13 (8.72%)||1.114 (0.397-1.83)||0.002|
|College||12 (17.91%)||-1.807 (-3.007- -0.608)||0.003|
|Graduate Degree||2 (2.99%)||*||*|
|Other||12 (17.91%)||-1.849 (-4.335-0.637)||0.145|
|Disabled||7 (6.48%)||-0.53 (-1.267-0.207)||0.159|
|Unemployed||31 (28.7%)||-0.827 (-2.554-0.899)||0.348|
|Retired||45 (41.67%)||1.625 (-0.415-3.664)||0.118|
|Alcohol use Disorder||8 (25%)||-1.935 (-5.833-1.963)||0.331|
* - not modeled / reference SSI - surgical site infection, OR - operating room IQR- Interquartile range Std Dev - standard deviation
Back to 2022 Abstracts