Housing Instability and Outcomes Among Patients With Access-Sensitive Surgical Conditions

Abstract

Introduction:

Housing instability is a significant contributor to poor health but remains understudied among surgical patients. We evaluated the association between housing instability and rates of unplanned surgical procedures, as well as resultant health and financial outcomes, for patients with access-sensitive conditions.

Methods:

Using the Healthcare Cost and Utilization Project National Inpatient Sample, we identified patients who underwent one of four selected procedures for access-sensitive surgical conditions (abdominal aortic aneurysm repair, colectomy, incisional hernia repair, and lower extremity bypass). Housing status was determined using International Classification of Disease, 10th Revision, Clinical Modification codes Z59.0 (homelessness) and Z59.1 (lack of adequate housing). Risk-adjusted multivariable logistic regression compared outcomes between patients with and without housing instability.

Results:

Of 1,761,965 individuals admitted for access-sensitive surgical conditions, 2280 were experiencing housing instability. Housing-unstable individuals had more than four times the odds of undergoing unplanned surgery than housing-stable individuals (odds ratio 4.41, P < 0.001). Across all procedures, individuals with housing instability experienced longer lengths of stay (planned: 5 d versus 4 d, P < 0.001; unplanned: 8 d versus 7 d, P < 0.001) and higher costs per admission following planned surgery ($20,379 versus $18,152, P < 0.001) than housing-stable individuals. Housing-unstable individuals had lower odds of complications and in-hospital mortality following planned surgeries. No differences in morbidity or mortality were identified following unplanned surgeries.

Conclusions:

Patients experiencing housing instability had higher odds of undergoing unplanned surgical procedures, had longer hospital stays and higher costs, and had similar or better outcomes following surgery. Efforts to expand affordable housing opportunities may improve access to timely surgical care.

Introduction

Housing stability is an influential health determinant yet remains a woefully understudied issue in the United States. The term encompasses a broad range of obstacles ranging from difficulty paying rent to frequent moves.1 The experience exists on a spectrum, with houselessness considered its most severe form.^1–3 Housing instability has been linked to poor health outcomes: previous work has demonstrated an association between houselessness and higher rates of chronic disease, emergency department use, and readmission.^4–13 Not only might housing instability co-occur with other social determinants of health, such as food insecurity and economic instability, but individuals with unstable housing have less access to medical care than individuals who are stably housed.^4,14–20 Individuals undergoing surgery are especially sensitive to delayed access to care. Timely access to care is critical in attaining good outcomes, particularly when it comes to access-sensitive surgical conditions. These are surgical conditions that are preferentially treated with elective surgery, but when access to surgical care is limited, the natural progression of the condition leads to unrelenting symptoms that may ultimately require unplanned surgery.^21–25 Thus, the proportion of unplanned surgical procedures for access-sensitive conditions can serve as an important indicator of surgical access.^21–25

Little is known regarding national trends in timely access to surgical care among individuals with housing instability despite the reality that, when these individuals require surgical care, they are at a distinct risk for poor surgical outcomes due to limited access to care. The reason for this paucity of literature is likely multifactorial. First, most studies of housing instability have primarily excluded surgical patients and focused on nonsurgical cohorts in primary, emergency, or psychiatric care.^7,12,26 Second, studies have focused exclusively on houselessness, overlooking the broader spectrum of housing instability. Third, because few nationally representative databases identify housing instability, most studies to date have relied on local registries or single-institution databases, leading to small samples and nongeneralizable results.²⁷

To better understand the unique condition of housing instability among individuals undergoing surgery, our study leveraged a nationally representative inpatient database to evaluate the association between housing instability and the likelihood of unplanned surgery for access-sensitive conditions as well as length of stay, cost, morbidity, and mortality. We hypothesized that, compared to stably housed individuals, individuals experiencing housing instability would demonstrate higher rates of unplanned surgery and worse clinical and financial outcomes.

Materials and Methods

We conducted a cross-sectional study comparing rates of unplanned procedures in individuals experiencing housing instability compared to those with stable housing using data obtained from the Agency for Healthcare Research and Quality’s Healthcare Cost and Utilization Project National Inpatient Sample (NIS) for calendar years 2016 to 2019. The NIS is used to make both regional and national estimates of inpatient hospital utilization, access, cost, quality, and outcomes and includes records from all hospital stays regardless of the expected payer.^28,29 This date range was chosen as 2016 was the first full calendar year of data with diagnosis and procedure codes reported using only the International Classification of Disease, 10th Revision, Clinical Modification (ICD-10-CM)/International Classification of Disease, 10th Revision, Procedure Coding System (ICD-10-PCS) coding system. 2019 was the last complete year of NIS data available to our group at the time of analysis.

In this study, we used access-sensitive surgical conditions as a marker allowing insight into disparities in access to care. We identified individuals aged 18 years and more who underwent one of four selected procedures for access-sensitive surgical conditions: abdominal aortic aneurysm (AAA) repair, colectomy, incisional hernia repair, and lower extremity bypass. These procedures are frequently performed in the United States and the conditions they treat have clear guidelines indicating elective repair before symptoms worsen to the point of necessitating an unplanned or emergent operation.^23,30–38 For example, colectomy is a procedure that can be performed to treat a number of conditions that would ideally be treated with planned operations given that access to appropriate screening and surgeon referral are optimal (colon cancer, diverticulitis, etc.). Thus, the ratio of planned to unplanned procedures can be used as a metric for surgical access.²² Prior work has demonstrated that access-sensitive conditions are a reliable metric in understanding access to surgical care.^21–25 Procedures for our selected conditions were identified using codes from the ICD-10-PCS.

Individuals’ housing status was defined by the presence of ICD-10-CM codes Z59.0 (“homelessness”) and Z59.1 (“lack of adequate housing”) in their patient records. Herein, we will refer to patients either with or without housing instability, in reference to the presence or absence of coded housing instability. The acuity of a patient’s index procedure was defined using the admission type variable in the NIS database which indicates the type of admission (elective, urgent, or emergent). We classified unplanned surgeries as those occurring either urgently or emergently. Planned surgeries were those that were performed electively. The primary outcome of interest in this study was the rate of unplanned surgery for each access-sensitive condition among individuals experiencing housing instability. Secondary outcomes included length of stay, total cost of admission, complications, and in-hospital mortality. Cost per admission was calculated using the Healthcare Cost and Utilization Project cost-to-charge ratio files and NIS charge data and adjusted for 2019 inflation. Complications were identified using ICD-10-PCS codes in line with previously validated methods, and included pulmonary failure, pneumonia, myocardial infarction, deep venous thrombosis, acute renal failure, postoperative hemorrhage, surgical site infection, and gastrointestinal bleeding.^39–41 Serious complications were defined as having at least one complication and a length of stay above the 75th percentile for the specific index procedure.^42–44 In-hospital mortality was defined as death during hospitalization for the index procedure.

This study was deemed exempt from review by the University of Michigan Institutional Review Board (IRB# HUM00195316).

Statistical analysis

The primary purpose of this analysis was to evaluate the association between housing instability and unplanned operations for access-sensitive surgical conditions. Descriptive statistics characterizing patient demographics and admissions were compared between patients with and without housing instability. Continuous data were visually and statistically evaluated for normality. Data that met criteria for normality were analyzed with appropriate parametric tests and are presented as mean (standard deviation). Data that were not normally distributed were analyzed with nonparametric tests as appropriate and are presented as median (Q1, Q3). All analyses incorporated NIS-provided survey sampling weights to account for sampling design to provide estimates representative of the national US population.

We began by calculating unadjusted rates of planned and unplanned surgeries, inpatient mortality, and complications. Next, we calculated risk-adjusted odds ratios (aORs) and 95% confidence intervals (95% CIs) for each outcome using multivariable logistic regression models. We adjusted for various factors, including patient age, sex, race (Asian, American Indian, Black, White, and other) and ethnicity (Hispanic), insurance type (Medicare, Medicaid, private, and uninsured), and number of comorbidities. We used procedure year as an additional categorical variable to account for potential secular trends in our multiyear data. Similar multivariable generalized linear regression models were used to evaluate risk-adjusted differences in length of stay and total cost between patients with and without housing instability.

For all analyses, tests of statistical significance were two-sided and a P value of < 0.05 was considered statistically significant. All analyses were performed using STATA MP version 16.1 (College Station, TX). Data were analyzed from February to June 2023.

Results

Of 1,863,120 patients undergoing procedures for the four selected access-sensitive conditions initially identified, 101,155 (5.3%) were excluded due to missing housing or demographic data. Of the 1,761,965 individuals with complete data, 2280 (0.13%) were experiencing housing instability at the time of surgery. Across the entire sample, 275,285 procedures (15.6%) were AAA repairs, 867,955 (49.3%) were colectomies, 319,505 (18.1%) were incisional hernia repairs, and 299,220 (17.0%) were lower extremity bypass procedures. Compared to stably housed patients, patients experiencing housing instability were younger (54 years versus 65 years; P < 0.001), more likely to be male (79% versus 50%; P < 0.001) and more likely to be Black (23% versus 11%; P < 0.001) or Native American (1.8% versus 0.5%; P = 0.04). They were more likely to rely on Medicaid (54% versus 9%; P < 0.001) or be uninsured (13% versus 2%; P < 0.001) and had more comorbidities (2.5 versus 2.3; P = 0.004). Individuals with housing instability also had significantly higher rates of admission through the emergency department (69% versus 30%; P < 0.001) and discharge against medical advice (4% versus 0.3%; P < 0.001). Table 1 summarizes these as well as other characteristics of patients and hospitalizations included in this analysis.

Table 1.

Characteristics of the study sample overall and by housing status.

Characteristic	Total (n = 1,761,965)	No housing instability (n = 1,759,685)	Housing instability (n = 2280)	P value
Age, mean (SD), y	65 (15)	65 (15)	54 (12)	<0.001
Sex, n (%)
Female	875,480 (49.7)	875,000 (49.7)	480 (21.1)	<0.001
Male	886,485 (50.3)	884,685 (50.3)	1800 (79.0)	<0.001
Race, n (%)
Asian or Pacific Islander	29,665 (1.7)	29,635 (1.7)	30 (1.3)	0.49
Black	190,245 (10.8)	189,715 (10.8)	530 (23.3)	<0.001
Hispanic	149,185 (8.5)	148,970 (8.5)	215 (9.4)	0.49
Native American	8680 (0.5)	8640 (0.5)	40 (1.8)	0.04
White	1,342,740 (76.2)	1,341,330 (76.2)	1410 (61.8)	<0.001
Other	41,450 (2.4)	41,395 (2.4)	55 (2.4)	0.93
Insurance type, n (%)
Private	519,890 (29.5)	519,765 (29.5)	125 (5.5)	<0.001
Medicaid	164,210 (9.3)	162,975 (9.3)	1235 (54.2)	<0.001
Medicare	996,590 (56.6)	996,040 (56.6)	550 (24.1)	<0.001
Uninsured	42,155 (2.4)	41,855 (2.4)	300 (13.2)	<0.001
Elixhauser comorbidities, mean (SD)	2.3 (1.8)	2.3 (1.8)	2.5 (1.6)	0.004
Elixhauser comorbidity groups, n (%)
0	303,315 (17.2)	303,135 (17.2)	180 (7.9)	<0.001
1–2	752,375 (42.7)	751,255 (42.7)	1120 (49.1)	0.01
≥3	706,275 (40.1)	705,295 (40.1)	980 (43.0)	0.21
Admission through the emergency department, n (%)	528,135 (30.0)	526,555 (29.9)	1580 (69.3)	<0.001
Type of operation, n (%)
AAA repair	275,285 (15.6)	275,160 (15.6)	125 (5.5)	<0.001
Colectomy	867,955 (49.3)	867,130 (49.3)	825 (36.2)	<0.001
Incisional hernia repair	319,505 (18.1)	318,915 (18.1)	590 (25.9)	<0.001
Lower extremity bypass	299,220 (17.0)	298,480 (17.0)	740 (32.5)	<0.001
Discharge location, n (%)
Routine	1,138,870 (64.6)	1,137,510 (64.6)	1360 (59.7)	0.03
Transfer to short-term hospital	11,265 (0.6)	11,255 (0.6)	10 (0.4)	0.52
Other transfer	247,950 (14.1)	247,320 (14.1)	630 (27.6)	<0.001
Home healthcare	323,375 (18.4)	323,210 (18.4)	165 (7.2)	<0.001
Against medical advice	4470 (0.3)	4375 (0.3)	95 (4.2)	<0.001

Cohort: NIS 2016–2019.

AAA = abdominal aortic aneurysm; SD = standard deviation.

Rates of unplanned surgery differed significantly between patients with and without housing instability (Table 2). Overall, patients experiencing housing instability had increased odds of undergoing unplanned surgeries when compared to their stably housed counterparts (aOR 4.41, P < 0.001). This difference was the largest among unstably housed patients undergoing AAA repair (aOR 23.78, P < 0.001), followed by lower extremity bypass (aOR 6.18, P < 0.001), incisional hernia repair (aOR 3.51, P < 0.001), and colectomy (aOR 3.33, P < 0.001).

Table 2.

Rates of unplanned procedures for access-sensitive surgical conditions.

			Risk-adjusted odds ratio (95% CI)	P value
	No housing instability	Housing instability
Total	39.35 (38.96 to 39.73)	82.02 (78.47 to 85.57)	4.41 (3.42 to 5.67)	<0.001
AAA repair	21.24 (20.43 to 22.04)	92.00 (80.83 to 100.00)	23.78 (5.41 to 104.57)	<0.001
Colectomy	38.31 (37.91 to 38.72)	76.97 (70.57 to 83.37)	3.33 (2.30 to 4.83)	<0.001
Incisional hernia repair	47.62 (46.96 to 48.28)	80.51 (73.41 to 87.61)	3.51 (2.20 to 5.61)	<0.001
Lower extremity arterial bypass	50.21 (49.54 to 50.88)	87.16 (81.66 to 92.67)	6.18 (3.73 to 10.24)	<0.001

Cohort: NIS (2016–2019), AAA repair, colectomy, incisional hernia repair, and lower extremity arterial bypass.

Risk-adjustment model: Sex, age, race, insurance type, comorbidity groups, and surgery year.

AAA = abdominal aortic aneurysm; CI = confidence interval.

Table 3 shows median length of stay and total cost per admission for planned and unplanned procedures. Compared to stably housed patients, those with housing instability spent longer in the hospital following both planned (5 days versus 4 days, P < 0.001) and unplanned surgeries (8 days versus 7 days, P < 0.001). Housing instability was associated with higher cost per admission for patients undergoing planned procedures ($20,379 versus $18,152, P < 0.001) and this trend persisted when analyzed by procedure type (P < 0.001 for all). Across all unplanned procedures, housing instability was not associated with a statistically significant difference in cost ($25,538 versus $24,147, P = 0.68). Comparisons of median length of stay and cost per admission by procedure type are included in Supplemental Tables 1 and 2.

Table 3.

Comparison of median length of stay and cost following procedures for access-sensitive surgical conditions.

			Risk-adjusted P value
	No housing instability	Housing instability
Planned procedures
Length of stay (d)	3 (2, 5)	5 (4, 8)	<0.001
Cost (2019 USD)	18,152 (12,284, 31,051)	20,379 (14,420, 30,937)	<0.001
Unplanned procedures
Length of stay (d)	7 (4, 12)	8 (5, 13)	<0.001
Cost (2019 USD)	24,147 (15,326, 39,849)	25,538 (16,152, 37,950)	0.68

Cohort: NIS (2016–2019), AAA repair, colectomy, incisional hernia repair, and lower extremity arterial bypass. Risk-adjustment model: sex, age, race, insurance type, comorbidity groups, and surgery year.

Mortality and complication rates for planned and unplanned cases by housing cohort are shown in Table 4. For planned procedures, housing instability was associated with lower odds of any complications (aOR 0.48, P < 0.001), any serious complications (aOR 0.51, P < 0.001), and in-hospital mortality (aOR 0.28, P = 0.03). For unplanned procedures, no statistically significant differences in any complications, serious complications, or in-hospital mortality were observed between housing cohorts (P > 0.05 for all). Mortality and complication rates were not analyzed by procedure type due to an inadequate number of patients in some subgroups. Results of a sensitivity analysis comparing outcomes following planned versus unplanned procedures for each housing cohort are included in Supplemental Table 3.

Table 4.

Comparison of clinical outcomes following procedures for access-sensitive surgical conditions.

			Risk-adjusted odds ratio (95% CI)	P value
	No housing instability	Housing instability
Planned procedures
Any complication	35.45 (35.15 to 35.76)	20.32 (16.21 to 24.43)	0.48 (0.37 to 0.63)	<0.001
Serious complication	26.13 (25.85 to 26.41)	14.97 (11.31 to 18.64)	0.51 (0.38 to 0.70)	<0.001
Mortality, in-hospital	4.11 (4.00 to 4.22)	0.80 (0.00 to 1.71)	0.28 (0.09 to 0.87)	0.03
Unplanned procedures
Any complication	11.74 (11.57 to 11.91)	10.98 (4.14 to 17.81)	0.86 (0.40 to 1.86)	0.71
Serious complication	5.73 (5.61 to 5.84)	4.88 (0.17 to 9.59)	0.72 (0.25 to 2.04)	0.53
Mortality, in-hospital	0.70 (0.67 to 0.74)	1.22 (−1.18 to 3.62)	2.84 (0.37 to 21.60)	0.31

Cohort: NIS (2016–2019), AAA repair, colectomy, incisional hernia repair, and lower extremity arterial bypass.

Risk-adjustment model: Sex, age, race, insurance type, comorbidity groups, and surgery year.

CI = confidence interval.

Discussion

This is the first national analysis of the association between housing instability and access-sensitive surgical care. We found that individuals with housing instability had four times greater odds of undergoing unplanned surgery for access-sensitive surgical conditions compared to housing-stable individuals, indicating significant disparities in timely access to surgery. These individuals also had a longer length of stay and higher total cost of admission following planned surgeries. Length of stay was similarly increased following unplanned surgery; however, for patients undergoing an unplanned procedure, there was no difference in total cost between housing-unstable and housing-stable individuals. While housing instability was associated with lower morbidity and mortality following planned surgery, there were no observed differences following unplanned surgery. Taken together, these findings demonstrate profound disparities in timely access to surgical care for individuals with housing instability and highlight the need for focused interventions to improve access to care for this population.

Our study found that individuals with housing instability had significantly higher odds of undergoing unplanned surgery (more than four times the odds overall and almost 24 times the odds for AAA repairs) indicating drastic differences in timely access to care between cohorts. These findings are consistent with prior work that has used the ratio of emergent to elective surgeries as a metric to evaluate disparities in access to surgical care.22 For example, in the United States, residing in a community with a higher Social Vulnerability Index score is associated with higher rates of unplanned surgeries, postoperative complications, reoperations, readmissions, and mortality.²¹ Differences in outcomes were attributable to the differential rates of unplanned surgery between those in the most and least vulnerable groups. Another study found that for patients undergoing cholecystectomy, those with high social vulnerability were more likely to present emergently.⁴⁵ Our study builds on this work and identifies inadequate housing as a significant vulnerability and a critical barrier to receiving timely surgical care.

This study found that housing instability is associated with increased length of stay, regardless of procedure acuity, and variable differences in cost across procedures, consistent with the extant literature exploring healthcare utilization by individuals experiencing housing instability. For example, in one study, patients experiencing homelessness admitted to medical or surgical services were found to have longer lengths of stay and higher costs per admission, driven by more alternate care days (days spent in the hospital despite no longer needing the level of care of an acute care facility).¹² Alternate care days may have been one factor driving our finding that housing instability was associated with longer hospital stays and increased cost following planned surgery. Due to uncertainty surrounding discharge planning and accessibility of follow-up care, it is likely that housing-unstable individuals required more inpatient support (including higher utilization of inpatient physical therapy, occupational therapy, and social work) compared to their stably housed counterparts who could receive these services as an outpatient.⁴⁶ Prior work has also demonstrated a disproportionate rate of behavioral health disorders in individuals experiencing homelessness as compared to housed individuals and that behavioral health disorders are associated with increased length of stay following surgery.^12,47–49 Although we didn’t explicitly identify psychiatric diagnoses within our cohorts, it is plausible that such diagnoses were more prevalent among patients experiencing housing instability. This may have further compounded the observed increase in length of stay. While we saw a similar increase in length of stay following unplanned surgery, we found similar costs per admission. While our findings may not have shown an increase in cost at the individual level, a disproportionate number of housing-unstable individuals undergo unplanned surgery, which is associated with a substantial financial burden on the healthcare system. In one study exploring the incremental cost of emergent versus elective surgeries, authors estimated that reducing emergency surgeries by just 10% across three key procedures could save nearly one billion dollars over a 10-year period.^{50 A} predicted increase in rates of unplanned surgery, growing associated costs (predicted to reach $41 billion by 2060), and the growing number of individuals facing housing instability will have a large impact on an already overburdened healthcare system.^51,52 To our knowledge, this is the first study that has explored these issues in tandem and provides insight into the needs of a previously overlooked subset of surgical patients.

Of note, our study did not find an association between housing instability and complications or in-hospital mortality following unplanned surgery. Following planned surgery, we observed that housing instability was associated with decreased odds of experiencing complications and in-hospital mortality. There are a number of potential reasons for these findings. Given the complexity of accessing care while unstably housed, it is likely that many individuals experiencing housing instability who have conditions for which planned operations would be preferable are never offered them. In our study, only 20% of all operations among patients with housing instability were planned. Prior work has demonstrated clear disparities in patterns of care for patients experiencing houselessness, with one study finding that among patients admitted with acute myocardial infarction, those experiencing houselessness were less likely to undergo coronary angiography, revascularization, or receive drug-eluting stents despite their proven benefits.⁵³ If individuals experiencing housing stability are less likely to be offered any surgery, there may be significant selection bias such that those housing-unstable individuals who do undergo surgical procedures are disproportionately healthy compared to the underlying housing-unstable population or stably housed individuals undergoing similar procedures. Additionally, these findings could be due to the inherent risk and clinical condition of patients require urgent or emergent procedures, as these factors would not be expected to differ based on housing status.^54–56 It is also possible that this portion of our analysis was underpowered, as mortality and serious morbidity were rare within the smaller unstable housing cohort. Analysis of postoperative morbidity and mortality in patients with housing instability would benefit from improved documentation and data collection for these patients and these results warrant further examination to elucidate contributing factors.

This study’s findings should be considered in the context of the complex nature of identifying individuals experiencing housing instability and the dataset used. Our study was limited by lack of standardized definitions of housing insecurity and homelessness, and inconsistency nationally in screening for social determinants of health.⁵⁷ These challenges limit the extent to which the true relationship between housing instability and access to surgical care can be studied, as we are likely underestimating the number of patients receiving surgical care while experiencing any form of housing instability. Not only is lack of screening for housing status an issue but the underutilization of ICD-10-Z codes in administrative databases may also stem from their nonbillable status.^58,59 Additionally, due to the nature of screening and the variety of ways in which housing instability can present, coding is likely to better capture patients experiencing houselessness versus other forms of housing instability.

Identifying patients via ICD-10-Z codes is an imperfect method and cannot ascertain the nuances of an individual’s housing status or reflect changes in housing status with time; however, it provides an objective way to identify patients for inclusion. While it is possible that a longer duration of housing instability contributes to an increased likelihood of eventually requiring unplanned surgery, we were unable to determine this given the limitations of our dataset. While the number of patients identified via this method is likely an underestimate, use of housing-related ICD codes has consistently produced larger sample sizes than other methods, and the number of patients we identified is consistent with prior work.²⁷ Another limitation is the lack of specific clinical information and susceptibility to coding inaccuracies of the NIS.⁶⁰ With only admission data, we cannot consider in our analyses outpatient care history, postdischarge outcomes, or readmissions that could offer more insight into a patient’s interactions with the healthcare system. We also were unable to fully account for variables that could impact length of stay, such as hospital safety net status, as the NIS no longer includes all admissions for individual hospitals, thus making it challenging to accurately determine safety-net status. Despite these limitations, using the largest publicly available and representative inpatient database in the United States improves generalizability of the results and aids in understanding the true extent of the relationship between housing instability and unplanned surgery.

Notwithstanding these limitations, this study provides the first nationally representative in-depth analysis of the association among housing instability, access to surgical care, and clinical and financial outcomes. Our findings are particularly timely as healthcare systems continue to explore strategies to optimize allocation of resources to achieve more equitable access to care.^{4,15,18,19,61–63} Interventions such as medical respite centers, robust case management programs and wraparound services, and investments in affordable housing are cost-effective, reduce acute care utilization rates, and reduce readmissions; however, their implementation has been limited to a small number of communities.^19,64–71 For example, a randomized controlled trial at two sites in Chicago found that an intervention offering housing and access to a case management program for chronically ill adults experiencing homelessness resulted in fewer hospitalizations, hospital days, and emergency department visits.⁶⁶ If effective, interventions aimed at reducing housing instability and tailoring healthcare to better serve this unique population may have the potential to decrease rates of unplanned surgery for access-sensitive surgical conditions, leading to potential cost-savings and shorter hospital stays.

Conclusions

We found that housing instability was associated with four times higher odds of undergoing an unplanned surgery, a longer length of stay, and increased costs. These findings reinforce housing status as a critical social determinant of health.