This cross-sectional study examines the association between coded housing instability and hospitalization in the US.
Key Points
Question
What are the most common reasons for hospitalization among patients with coded housing instability?
Findings
This cross-sectional study of 87 348 604 hospitalizations found that the most common reasons for hospitalization among patients with coded housing instability were mental, behavioral, and neurodevelopmental disorders (50.3%), injury (7.3%), and circulatory system diseases (6.8%). Coded housing instability was also significantly associated with longer hospital stays (6.7 vs 4.8 days) and a cost of $9.3 billion.
Meaning
These findings suggest that coded housing instability was associated with higher rates of inpatient admissions for mental, behavioral, and neurodevelopmental disorders, longer hospital stays, and substantial health care costs.
Abstract
Importance
Housing instability and other social determinants of health are increasingly being documented by clinicians. The most common reasons for hospitalization among patients with coded housing instability, however, are not well understood.
Objective
To compare the most common reasons for hospitalization among patients with and without coded housing instability.
Design, Setting, and Participants
This cross-sectional, retrospective study identified hospitalizations of patients between age 18 and 99 years using the 2017 to 2019 National Inpatient Sample. Data were analyzed from May to September 2022.
Exposures
Housing instability was operationalized using 5 International Classification of Diseases, 10th Revision, Social Determinants of Health Z-Codes addressing problems related to housing: homelessness; inadequate housing; discord with neighbors, lodgers, and landlords; residential institution problems; and other related problems.
Main Outcomes and Measures
The primary outcome of interest was reason for inpatient admission. Bivariate comparisons of patient characteristics, primary diagnoses, length of stay, and hospitalization costs among patients with and without coded housing instability were performed.
Results
Among the 87 348 604 hospitalizations analyzed, the mean (SD) age was 58 (20) years and patients were more likely to be women (50 174 117 [57.4%]) and White (58 763 014 [67.3%]). Housing instability was coded for 945 090 hospitalizations. Hospitalized patients with housing instability, compared with those without instability, were more likely to be men (668 255 patients with coded instability [70.7%] vs 36 506 229 patients without [42.3%]; P < .001), younger (mean [SD] age 45.5 [14.0] vs 58.4 [20.2] years), Black (235 355 patients [24.9%] vs 12 929 158 patients [15.0%]), Medicaid beneficiaries (521 555 patients [55.2%] vs 15 541 175 patients [18.0%]), uninsured (117 375 patients [12.4%] vs 3 476 841 patients [4.0%]), and discharged against medical advice (28 890 patients [8.4%] vs 451 855 patients [1.6%]). The most common reason for hospitalization among patients with coded housing instability was mental, behavioral, and neurodevelopmental disorders (475 575 patients [50.3%]), which cost a total of $3.5 billion. Other common reasons included injury (69 270 patients [7.3%]) and circulatory system diseases (64 700 patients [6.8%]). Coded housing instability was also significantly associated with longer mean (SD) hospital stays (6.7 [.06] vs 4.8 [.01] days) and a cost of $9.3 billion. Hospitalized patients with housing instability had 18.6 times greater odds of having a primary diagnosis of mental, behavioral, and neurodevelopmental disorders (475 575 patients [50.3%] vs 4 470 675 patients [5.2%]; odds ratio, 18.56; 95% CI, 17.86 to 19.29).
Conclusions and Relevance
In this cross-sectional study, hospitalizations among patients with coded housing instability had higher admission rates for mental, behavioral, and neurodevelopmental disorders, longer stays, and increased costs. Findings suggest that efforts to improve housing instability, mental and behavioral health, and inpatient hospital utilization across multiple sectors may find areas for synergistic collaboration.
Introduction
Housing instability is a social determinant of health (SDOH) increasingly being addressed by hospitals, health systems, and payers.1,2,3,4,5 Variable definitions of housing instability refer to several housing-related problems, such as difficulty paying rent, spending a high proportion of income on housing, moving frequently, overcrowding, and staying with relatives.6,7 Housing instability exists on a continuum and includes the experience of homelessness.6 With increasing US rates of housing instability and attention to SDOH,8 federal and state payers (eg, Medicare and Medicaid), managed care organizations, and hospitals are investing in local efforts to create and preserve affordable housing.1,8,9,10
Despite broad interest in housing instability and use of inpatient hospital resources, the relationship remains understudied. Prior work has raised concern that patients with housing instability present with more serious infections, sequalae of preventable exposures, and health crises due to untreated chronic conditions, such as diabetes, high blood pressure, and heart disease.4,11 Housing instability has also been associated with poor primary health care utilization, as patients often delay care due to cost.12,13 Previous studies14,15,16,17,18,19,20,21,22 indicated that, as a consequence, patients experiencing housing instability have higher rates of costly hospitalization, readmission, and emergency department service use compared with patients without housing instability. Many of these studies, however, only included specific health care services, used nonrepresentative data, examined a short period of time, or focused exclusively on homelessness and not multiple levels of housing instability.23 As such, understanding of how patients with housing instability use inpatient hospital resources is incomplete.
In this context, this study aimed to identify the most common diagnoses, length of stay, and associated costs among hospitalized patients experiencing housing instability. To achieve this aim, the largest publicly available, all-payer inpatient health care database and several housing instability indicators coded by clinicians via International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10), Social Determinants of Health Z-Codes were used. This study is timely as health care systems with large inpatient practices explore ways to invest in communities and address SDOH, including housing instability.
Methods
Data Source and Participants
This cross-sectional study was exempt from approval and the need for informed consent by the University of Michigan institutional review board because a publicly available data set containing deidentified patient information was used. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines were followed.
This cross-sectional, retrospective study used the National Inpatient Sample (NIS) between January 2017 and December 2019. The NIS is part of the Healthcare Cost and Utilization Project sponsored by the Agency for Healthcare Research and Quality.24 The database contains deidentified information about each hospitalization, including patient and hospital demographics, admission status, discharge diagnoses, comorbidities, procedures, outcomes, costs, and payer information. Unweighted, the NIS data contain a 20% sample of all inpatient discharges from US community hospitals (>7 million hospitalizations annually; enabling weighted estimates of >35 million hospitalizations nationally), making it the largest publicly available all-payer inpatient care database. Approximately 85% of US hospitals are community hospitals, including all nonfederal, short-term, general, and other specialty hospitals. Rehabilitation, psychiatric, and long-term acute care hospitals, as well as hospital units of institutions, are excluded from the NIS. Detailed information about the NIS sampling methods and design are available elsewhere.24 Admissions among patients aged 18 to 99 years with complete study data were included.
Exposure
Screening patients for housing instability and other SDOH is not consistent or standardized.25,26 In this study, housing instability was operationalized using 5 ICD-10 Z-Codes addressing aspects of the housing instability continuum, including but not exclusive to homelessness: homelessness (Z59.0); inadequate housing (Z59.1); discord with neighbors, lodgers, and landlords (Z59.2); problems related to living in a residential institution (Z59.3); and other problems relating to housing and economic circumstances (Z59.8: housing instability, housed: with risk of homelessness; homelessness in past 12 months, unspecified; other problems related to housing and economic circumstances; and problem related to housing and economic circumstances, unspecified). Consistent with other work,27,28 this study reviewed and identified coded housing instability for hospitalizations with at least 1 of the 5 housing-related Z59-codes in the first 15 ICD-10 codes in the NIS data set. Patient-level characteristics included age (continuous variable), sex, race and ethnicity (Asian, Black, Hispanic, American Indian, White, and other, including multiple races and state-reported race categories excluded from other NIS categories), insurance type (Medicare, Medicaid, private, uninsured, other), elective or emergency admittance, major operating room procedure, and discharge type (routine transfer to short term hospital, other transfer, home health care, against medical advice). Race and ethnicity were reported to examine whether rates of coded housing instability differed by these factors.
Main Outcomes and Measures
The main outcome of interest was the primary reason for hospitalization. Diagnoses and procedures are recorded in the NIS database using International Statistical Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes for hospitalizations after October 1, 2015. Diagnosis groups were created based on the first 3 digits of the ICD-10-CM primary diagnosis codes (eTable 1 in the Supplement).
Total inpatient days were calculated by summing across admissions and diagnosis groups. Mean length of stay was calculated by dividing total inpatient days by the number of admissions. Same-day stays are coded as 0 in the NIS data.
Hospitalization costs were calculated using NIS charge data and Healthcare Cost and Utilization Project cost-to-charge ratio files. After merging the 2 data sets by year and NIS charge data by NIS hospital ID, charges were multiplied by the cost-to-charge ratio. Costs in 2017 and 2018 were adjusted for 2019 inflation.
Statistical Analysis
The goal of this analysis was to examine the association between coded housing instability and reason for inpatient hospitalization, length of stay, and costs of admission. Standard descriptive statistics were first calculated to characterize hospitalizations overall and by coded housing instability status. Procedures that incorporated survey sampling weights to account for the complex NIS sampling design and to provide representative estimates of the national US population were applied. Diagnoses, total inpatient days, mean length of stay, and annual hospitalization costs were also similarly calculated. The most common reasons for hospitalization among patients with and without coded housing instability were identified and compared via estimation of unadjusted odds ratios (ORs). Due to the large sample size, standardized differences were used to examine effect sizes (Cohen d) of the differences between groups with and without coded housing instability, independent of sample size. Standardized differences were calculated for comparisons of coded housing instability status by age, sex, race and ethnicity, insurance type, elective or emergency admittance, discharge type, primary discharge diagnosis, total inpatient days, mean length of stay, and annual costs. Resulting effect sizes were assessed according to 3 thresholds: small (0.20-0.49), medium (0.50-0.79), and large (≥0.80) effect sizes.29 Additional sensitivity analyses were performed and examined all comparisons stratified by the 5 housing instability Z59 codes. All analyses were performed using STATA statistical software version 17 (STATACorp). Data were analyzed from May to September 2022.
Results
Analysis of the 87 348 604 hospitalizations included in the study indicated that patients were mean (SD) age of 58 (20) years, more likely to be women (50 174 117 [57.4%]), and mostly of White race (58 763 014 [67.3%]). Details on selection and exclusion criteria are presented in eTable 2 in the Supplement. Table 1 summarizes these and additional characteristics of hospitalizations between 2017 to 2019. Overall, patients were also primarily Medicare beneficiaries (48.3% [42 174 988]), admitted by emergency (61.9% [54 035 986]), and routinely discharged (62.8% [18 321 765]). This study identified 945 090 admissions with at least 1 of 5 housing instability codes. Admissions with coded housing instability were more likely to be men compared with the 86 403 514 admissions without coded housing instability (70.7% [668 255] vs 42.3% [36 506 229]), younger (mean [SD] age 45.5 [14.0] years vs 58.4 [20.2] years), Black (24.9% [235 355] vs 15.0% [12 929 158]), Medicare (21.0% [198 070] vs 48.6% [41 976 918]) and Medicaid beneficiaries (55.2% [521 555] vs 18.0% [15 541 175]), uninsured (12.4% [117 375] vs 4.0% [3 476 841]), admitted by emergency (75.2% [710 635] vs 61.7% [53 325 351]), and discharged against medical advice (8.4% [28 890] vs 1.6% [451 855]). Among admissions for patients with coded housing instability, 96.8% had the Z59.0 homelessness code.
Table 1. Characteristics of Hospitalized Patients With and Without Coded Housing Instabilitya.
| Characteristic | Patient admissions, No. (%) | Standardized difference (Cohen d)c | ||
|---|---|---|---|---|
| Total admissions (N = 87 348 601) | With coded housing instability admissions (n = 945 090)b | Without coded housing instability admissions (n = 86 403 511) | ||
| Age, mean (SD), y | 58.3 (20.2) | 45.5 (14.0) | 58.4 (20.2) | 0.642 |
| Sex | ||||
| Men | 37 174 484 (42.6) | 668 255 (70.7) | 36 506 229 (42.3) | 0.577 |
| Women | 50 174 117 (57.4) | 276 835 (29.3) | 49 897 282 (57.7) | 0.577 |
| Race | ||||
| Asian or Pacific Islander | 2 441 628 (2.8) | 11 920 (1.3) | 2 429 708 (2.8) | 0.094 |
| Black | 13 164 513 (15.1) | 235 355 (24.9) | 12 929 158 (15.0) | 0.278 |
| Hispanic | 9 806 187 (11.2) | 108 320 (11.5) | 9 697 867 (11.2) | 0.008 |
| American Indian | 566 015 (0.6) | 12 895 (1.4) | 553 120 (0.6) | 0.090 |
| White | 58 763 014 (67.3) | 544 525 (57.6) | 58 218 489 (67.4) | 0.208 |
| Otherd | 2 607 244 (3.0) | 32 075 (3.4) | 2 575 169 (3.0) | 0.024 |
| Insurance type | ||||
| Private | 23 071 974 (26.4) | 72 125 (7.6) | 22 999 849 (26.6) | 0.431 |
| Medicare | 42 174 988 (48.3) | 198 070 (21.0) | 41 976 918 (48.6) | 0.554 |
| Medicaid | 16 062 729 (18.4) | 521 555 (55.2) | 15 541 175 (18.0) | 0.965 |
| Otherd | 2 444 694 (2.8) | 35 965 (3.8) | 2 408 729 (2.8) | 0.062 |
| Uninsured | 3 594 216 (4.1) | 117 375 (12.4) | 3 476 841 (4.0) | 0.423 |
| Elective procedure | 19 855 840 (22.7) | 69 160 (7.3) | 19 786 680 (22.9) | 0.372 |
| Emergency admittance | 54 035 986 (61.9) | 710 635 (75.2) | 53 325 351 (61.7) | 0.278 |
| Major operating room procedure | 9 206 992 (31.6) | 27 270 (7.9) | 9 179 722 (31.8) | 0.516 |
| Discharge description | ||||
| Routine | 18 321 765 (62.8) | 261 730 (75.7) | 18 060 035 (62.7) | 0.269 |
| Transfer to short-term hospital | 586 409 (2.0) | 5980 (1.7) | 580 429 (2.0) | 0.020 |
| Other transfer | 4 778 945 (16.4) | 42 430 (12.3) | 4 736 515 (16.4) | 0.113 |
| Home health care | 4 350 181 (14.9) | 5985 (1.7) | 4 344 196 (15.1) | 0.375 |
| Against medical advice | 480 745 (1.6) | 28 890 (8.4) | 451 855 (1.6) | 0.534 |
| Housing instability Z59-codese | ||||
| Z59.0 homelessness | NA | 915 305 (96.8) | NA | NA |
| Z59.1 inadequate housing | NA | 5030 (0.5) | NA | NA |
| Z59.2 discord with neighbors, lodgers, landlords | NA | 1220 (0.1) | NA | NA |
| Z59.3 problems related to living in a residential institution | NA | 3335 (0.4) | NA | NA |
| Z59.8 other problems | NA | 24 445 (2.6) | NA | NA |
Abbreviations: NA, not applicable; NIS, National Inpatient Sample.
Data source: National Inpatient Sample, 2017 to 2019.
“With coded housing instability” data reflect hospitalizations with at least 1 of 5 housing instability Z59-codes.
“Standardized difference” displays the absolute value of the difference in proportions divided by the standard error and is an indicator of effect size (Cohen d) (0.20-0.49 indicates small; 0.50 to 0.79, medium; and ≥0.80, large effect sizes).
Other race includes multiple races and state-reported race categories excluded from other NIS categories.
“Housing instability Z59-codes” display the number of hospitalizations with each of the 5 listed Z59-codes.
There were significant differences in the most common reasons for hospitalization among patients with and without coded housing instability (Table 2). Admissions with coded housing instability had 18.6 times greater odds of having a mental, behavioral, and neurodevelopmental disorder diagnosis (50.3% [475 575] vs 5.2% [4 470 675]; OR, 18.56; 95% CI = 17.86-19.29) (Table 2 and eFigure in the Supplement). The second and third most common reasons for hospitalization among admissions with coded housing instability were injury (7.3% [69 270] vs 8.7% [7 559 439], OR = 0.82; 95% CI, 0.80-0.86), and diseases of the circulatory system (6.8% [64 700] vs 16.7% [14 452 895 ], OR, 0.37; 95% CI, 0.35-0.38). For comparison, the 3 most common reasons for admissions among patients without coded housing instability were diseases of the circulatory system (16.6% [14 452 895]), pregnancy, childbirth, and puerperium (12.8% [11 157 270]), and diseases of the digestive system (9.9% [8 581 113]).
Table 2. Hospital Admissions by Diagnosis Among Patients With and Without Coded Housing Instabilitya.
| Discharge diagnosis | Patient admissions, No. (%) | Odds ratio (95% CI) | Standardized difference (Cohen d)b | ||
|---|---|---|---|---|---|
| Total admissions | With coded housing instability admissions | Without coded housing instability admissions | |||
| Mental, behavioral, and neurodevelopmental disorders | 4 946 250 (5.7) | 475 575 (50.3) | 4 470 675 (5.2) | 18.56 (17.86-19.29) | 1.994 |
| Injury, poisoning, and certain other consequences of external causes | 7 628 709 (8.7) | 69 270 (7.3) | 7 559 439 (8.7) | 0.82 (0.80-0.86) | 0.050 |
| Diseases of the circulatory system | 14 517 595 (16.6) | 64 700 (6.8) | 14 452 895 (16.7) | 0.37 (0.35-0.38) | 0.266 |
| Certain infections and parasitic diseases | 7 184 214 (8.2) | 56 060 (5.9) | 7 128 154 (8.2) | 0.70 (0.67-0.73) | 0.084 |
| Diseases of the respiratory system | 6 970 298 (8.0) | 51 365 (5.4) | 6 918 933 (8.0) | 0.66 (0.64-0.68) | 0.095 |
| Diseases of the skin and subcutaneous tissue | 1 652 514 (1.9) | 46 665 (4.9) | 1 605 849 (1.9) | 2.74 (2.64-2.85) | 0.226 |
| Endocrine, nutritional, and metabolic diseases | 3 800 334 (4.4) | 38 710 (4.1) | 3 761 624 (4.4) | 0.94 (0.91-0.97) | 0.013 |
| Diseases of the digestive system | 8 618 458 (9.9) | 37 345 (4.0) | 8 581 113 (9.9) | 0.37 (0.36-0.39) | 0.201 |
| Diseases of the musculoskeletal system and connective tissue | 6 259 779 (7.2) | 19 310 (2.0) | 6 240 469 (7.2) | 0.27 (0.26-0.28) | 0.201 |
| Diseases of the genitourinary system | 4 103 464 (4.7) | 18 030 (1.9) | 4 085 434 (4.7) | 0.39 (0.38-0.41) | 0.133 |
| Symptoms, signs, and abnormal clinical laboratory findings, not elsewhere classified | 2 419 310 (2.8) | 17 910 (1.9) | 2 401 400 (2.8) | 0.68 (0.65-0.70) | 0.054 |
| Diseases of the nervous system | 2 243 755 (2.6) | 17 270 (1.8) | 2 226 485 (2.6) | 0.70 (0.67-0.73) | 0.047 |
| Pregnancy, childbirth, and puerperium | 11 174 720 (12.8) | 17 450 (1.8) | 11 157 270 (12.9) | 0.13 (0.12-0.13) | 0.332 |
| Neoplasms | 3 566 719 (4.1) | 7100 (0.8) | 3 559 619 (4.1) | 0.18 (0.17-0.19) | 0.170 |
| Diseases of the blood and blood-forming organs and certain disorders involving the immune mechanism | 1 079 290 (1.2) | 4420 (0.5) | 1 074 870 (1.2) | 0.37 (0.35-0.40) | 0.070 |
| Factors influencing health status and contact with health services | 911 070 (1.0) | 2415 (0.3) | 908 655 (1.1) | 0.24 (0.21-0.27) | 0.078 |
| Diseases of the eye and adnexa | 66 715 (0.1) | 895 (0.1) | 65 820 (0.1) | 1.24 (1.07-1.45) | 0.007 |
| Certain conditions originating in the perinatal periodc | 905 (<0.01) | 0 | 905 (<0.01) | NA | NA |
| Congenital malformations, deformations, and chromosomal abnormalities | 109 755 (0.1) | 225 (0.0) | 109 530 (0.1) | 0.19 (0.14-0.25) | 0.029 |
| Diseases of the ear and mastoid process | 94 350 (0.1) | 370 (<0.01) | 93 980 (0.1) | 0.36 (0.29-0.45) | 0.021 |
| External causes of morbidity | 400 (<0.01) | 5 (<0.01) | 395 (<0.01) | 1.16 (0.18-7.54) | 0.000 |
| Total No. of hospital admissions | 87 348 604 (100) | 945 090 (1.1) | 86 403 514 (98.9) | NA | NA |
Abbreviation: NA, not applicable.
Data source: National Inpatient Sample, 2017 to 2019.
Standardized difference displays the absolute value of the difference in proportions divided by the standard error and is an indicator of effect size (Cohen d) (>0.20 indicates small; >0.50, medium; and ≥.80, large effect sizes).
Certain conditions originating in the perinatal period had no coded housing instability so no odds ratio or standardized difference is reported.
Table 3 presents total inpatient days for hospitalizations with and without coded housing instability by diagnosis and overall mean length of stay. Hospitalizations with coded housing instability accounted for 1.5% (6 197 196) of the 416 199 050 total inpatient days between 2017 to 2019. When compared with hospitalizations without coded housing instability, admissions with coded housing instability had a greater mean (SD) length of stay (6.7 [.06] days vs 4.8 [.01] days). Of the 6 197 196 inpatient days among those with coded housing instability, 59.1% (3 661 816) were due to mental, behavioral, and neurodevelopmental disorders compared with 7.6% (31 111 715) among patients without coded housing instability. Diseases of the circulatory system accounted for the highest percentage of inpatient days among admissions without coded housing instability (17.4% [71 429 857]).
Table 3. Total Inpatient Days and Mean Length of Stay by Diagnosis Among Patients With and Without Coded Housing Instabilitya.
| Discharge diagnosis | Days, No. (%) | Standardized difference (Cohen d)b | ||
|---|---|---|---|---|
| Total | With coded housing instability | Without coded housing instability | ||
| Mental, behavioral, and neurodevelopmental disorders | 34 773 530 (8.4) | 3 661 816 (59.1) | 31 111 715 (7.6) | 1.920 |
| Injury, poisoning, and certain other consequences of external causes | 43 386 087 (10.4) | 408 225 (6.6) | 42 977 862 (10.5) | 0.127 |
| Certain infections and parasitic diseases | 49 113 167 (11.8) | 384 130 (6.2) | 48 729 037 (11.9) | 0.176 |
| Diseases of the circulatory system | 71 780 192 (17.2) | 350 335 (5.7) | 71 429 857 (17.4) | 0.311 |
| Diseases of the respiratory system | 34 492 570 (8.3) | 233 820 (3.8) | 34 258 750 (8.4) | 0.166 |
| Diseases of the skin and subcutaneous tissue | 7 419 371 (1.8) | 217 315 (3.5) | 7 202 057 (1.8) | 0.133 |
| Endocrine, nutritional, and metabolic diseases | 16 014 737 (3.8) | 175 525 (2.8) | 15 839 212 (3.9) | 0.054 |
| Diseases of the digestive system | 38 827 784 (9.3) | 171 090 (2.8) | 38 656 694 (9.4) | 0.229 |
| Diseases of the musculoskeletal system and connective tissue | 19 337 003 (4.6) | 133 365 (2.2) | 19 203 639 (4.7) | 0.120 |
| Diseases of the nervous system | 12 094 530 (2.9) | 118 090 (1.9) | 11 976 440 (2.9) | 0.060 |
| Diseases of the genitourinary system | 16 979 533 (4.1) | 85 495 (1.4) | 16 894 038 (4.1) | 0.138 |
| Pregnancy, childbirth, and puerperium | 29 850 763 (7.2) | 73 965 (1.2) | 29 776 798 (7.3) | 0.235 |
| Neoplasms | 21 827 586 (5.2) | 67 715 (1.1) | 21 759 871 (5.3) | 0.189 |
| Symptoms, signs, and abnormal clinical laboratory findings, not elsewhere classified | 8 467 245 (2.0) | 65 830 (1.1) | 8 401 415 (2.0) | 0.070 |
| Diseases of the blood and blood-forming organs and certain disorders involving the immune mechanism | 4 748 999 (1.1) | 21 430 (0.3) | 4 727 569 (1.2) | 0.076 |
| Factors influencing health status and contact with health services | 6 032 843 (1.4) | 18 940 (0.3) | 6 013 903 (1.5) | 0.097 |
| Diseases of the eye and adnexa | 234 540 (0.1) | 6645 (0.1) | 227 895 (0.1) | 0.022 |
| Congenital malformations, deformations, and chromosomal abnormalities | 551 060 (0.1) | 1925 (<0.01) | 549 135 (0.1) | 0.028 |
| Diseases of the ear and mastoid process | 262 290 (0.1) | 1520 (<0.01) | 260 770 (0.1) | 0.016 |
| External causes of morbidity | 1525 (<0.01) | 20 (<0.01) | 1505 (<0.01) | 0.000 |
| Certain conditions originating in the perinatal periodc | 3695 (<0.01) | 0 | 3695 (<0.01) | NA |
| Total inpatient | 416 199 050 (100) | 6 197 196 (1.5) | 410 001 857 (98.5) | NA |
| Mean length of stay (95% CI), d per admission | 4.8 (4.7-4.8) | 6.7 (6.5-6.7) | 4.8 (4.7-4.8) | 0.284 |
Abbreviation: NA, not applicable.
Data source: National Inpatient Sample, 2017 to 2019.
Standardized difference displays the absolute value of the difference in proportions divided by the standard error and is an indicator of effect size (Cohen d) (>0.20 indicates small; >0.50, medium; and ≥.80, large effect sizes).
Certain conditions originating in the perinatal period had no coded housing instability so no standardized difference is reported.
Costs of hospitalizations with and without coded housing instability by diagnoses are displayed in Table 4. Admissions with coded housing instability accounted for $9.3 billion (0.8%) of the total $1.2 trillion cost of admissions between 2017 and 2019. Of that $9.3 billion, $3.5 billion (37.9%) was attributed to the 475 575 admissions with both coded housing instability and mental, behavioral, and neurodevelopmental disorders (vs $31.7 billion [2.7%] without coded housing instability). Diseases of the circulatory system accounted for the greatest percentage of costs for admissions without coded housing instability ($254.7 billion [21.8%]). Admissions for injury, poisoning, and certain other consequences of external causes accounted for the second highest costs of admissions among patients with coded housing instability ($1.1 billion [11.7%]).
Table 4. Cost of Hospitalization by Diagnosis Among Patients With and Without Coded Housing Instabilitya.
| Discharge diagnosis | Cost in millions, $ (%) | Standardized difference (Cohen d)b | ||
|---|---|---|---|---|
| Total | With coded housing instability | Without coded housing instability | ||
| Mental, behavioral, and neurodevelopmental disorders | 35 276.3 (3.0) | 3531.5 (37.9) | 31 744.8 (2.7) | 2.076 |
| Injury, poisoning, and certain other consequences of external causes | 139 294.5 (11.8) | 1090.9 (11.7) | 138 203.5 (11.9) | 0.004 |
| Diseases of the circulatory system | 255 556.2 (21.7) | 878.4 (9.4) | 254 677.8 (21.8) | 0.301 |
| Certain infections and parasitic diseases | 127 900.8 (10.9) | 860.9 (9.2) | 127 040.0 (10.9) | 0.053 |
| Diseases of the respiratory system | 82 281.5 (7.0) | 522.8 (5.6) | 81 758.7 (7.0) | 0.055 |
| Diseases of the skin and subcutaneous tissue | 14 197.5 (1.2) | 432.2 (4.6) | 13 765.3 (1.2) | 0.316 |
| Diseases of the digestive system | 105 878.5 (9.0) | 416.6 (4.5) | 105 462.0 (9.0) | 0.160 |
| Endocrine, nutritional, and metabolic diseases | 42 623.5 (3.6) | 360.0 (3.9) | 42 263.5 (3.6) | 0.013 |
| Diseases of the musculoskeletal system and connective tissue | 114 480.0 (9.7) | 294.1 (3.2) | 114 185.9 (9.8) | 0.224 |
| Diseases of the nervous system | 28 464.5 (2.4) | 220.8 (2.4) | 28 243.7 (2.4) | 0.003 |
| Neoplasms | 79 219.7 (6.7) | 163.6 (1.8) | 79 056.1 (6.8) | 0.201 |
| Diseases of the genitourinary system | 38 542.7 (3.3) | 162.1 (1.7) | 38 380.6 (3.3) | 0.087 |
| Symptoms, signs, and abnormal clinical laboratory findings, not elsewhere classified | 20 327.0 (1.7) | 145.4 (1.6) | 20 181.6 (1.7) | 0.013 |
| Pregnancy, childbirth, and puerperium | 60 648.8 (5.2) | 131.0 (1.4) | 60 517.8 (5.2) | 0.171 |
| Diseases of the blood and blood-forming organs and certain disorders involving the immune mechanism | 12 014.9 (1.0) | 49.3 (0.5) | 11 965.6 (1.0) | 0.049 |
| Factors influencing health status and contact with health services | 14 123.3 (1.2) | 30.7 (0.3) | 14 092.6 (1.2) | 0.081 |
| Congenital malformations, deformations, and chromosomal abnormalities | 3351.7 (0.3) | 6.2 (0.1) | 3345.5 (0.3) | 0.041 |
| Diseases of the eye and adnexa | 636.9 (0.1) | 13.9 (0.1) | 623.0 (0.1) | 0.041 |
| Certain conditions originating in the perinatal periodc | 10.9 (<0.01) | 0.0 | 10.9 (<0.01) | NA |
| Diseases of the ear and mastoid process | 702.0 (0.1) | 4.1 (<0.01) | 697.9 (0.1) | 0.007 |
| External causes of morbidity | 4.8 (<0.01) | 0.0 | 4.7 (<0.01) | 0.000 |
| Total | 1 175 536.0 (100) | 9314.5 (0.8) | 1 166 221.5 (99.2) | NA |
Abbreviation: NA, not applicable.
Data source: National Inpatient Sample, 2017 to 2019.
“Standardized difference” displays the absolute value of the difference in proportions divided by the standard error and is an indicator of effect size (Cohen d) (>0.20 indicates small; >0.50,medium; and ≥.80, large effect sizes).
“Certain conditions originating in the perinatal period” had no coded housing instability so no standardized difference is reported.
Sensitivity analyses were performed to examine differences between hospitalizations with and without each of the 5 Z59-codes used in the study. General trends were consistent with original comparisons of patient characteristics (eTable 3 in the Supplement), admissions (eTables 4-6 in the Supplement), length of stay (eTables 7 and 8 in the Supplement), and cost (eTables 9 and 10 in the Supplement) by housing instability. When comparing hospitalizations with and without each of the 5 housing instability Z59-codes by reason for admission, large effect sizes were found for differences in admission rates (eTables 4-6 in the Supplement), length of stay (eTables 7 and 8 in the Supplement), and costs (eTables 9 and 10 in the Supplement) among patients with mental, behavioral, and neurodevelopmental disorders.
Discussion
In this cross-sectional study that compared common primary diagnoses, length of stay, and costs among hospitalized patients with and without coded housing instability between 2017 and 2019, 3 principal findings were observed. First, the most common reason for hospitalization among patients with coded housing instability was mental, behavioral, and neurodevelopmental disorders. This diagnosis accounted for 50.3% of hospitalizations and nearly 60% of inpatient days for patients with coded housing instability. Second, hospitalized patients with coded housing instability had a higher mean length of stay compared with patients without coded housing instability (6.7 vs 4.8 days). Third, hospitalization costs for patients with coded housing instability were $9.3 billion, with 37.9% of those costs from primary diagnoses of mental, behavioral, and neurodevelopmental disorders. Taken together, findings suggest that future attention to understanding the association between housing instability and inpatient admissions for mental, behavioral, and neurodevelopmental disorders is warranted.
This study adds to the existing literature on housing instability and health via analysis of 3 years of nationally representative health outcome data. Prior work23 largely relied on nonrepresentative or self-reported data. Findings from this study were consistent with prior work that documented higher prevalence of mental and behavioral health disorders, hospitalizations, and health care costs among people experiencing housing instability, and especially homelessness.4,30,31
By including multiple levels of housing instability in addition to homelessness, results suggested that trends in diagnoses, utilization, and costs may extend across the housing instability continuum. Much of the existing literature exclusively examined homelessness, which exacerbates mental health and substance use conditions due to a lack of safety, inability to adhere to treatment and manage chronic conditions, and a lack of or disrupted care that can lead to increased anxiety, depression, paranoia, and avoidable hospitalization.4 Other aspects of housing instability captured by the 5 Z59-codes and examined via sensitivity analyses highlight the potential relevance of the complete housing instability continuum when examining hospitalizations, length of stay, and costs by reason for admission. Results also supported the need for more sensitive and widely used housing instability measures. This study also adds to prior work32,33,34,35,36,37,38 quantifying health care costs associated with housing instability and homelessness by using nationally representative hospitalization data. People experiencing housing instability, homelessness, and poverty face numerous health risks and use costly hospital-based acute care at disproportionately higher rates.39 This population is among the top 5% of hospital users estimated to account for 50% of health care costs.32,39 By providing an unadjusted national estimate of inpatient hospitalization costs among patients with coded housing instability, study findings provide baseline information to build on work making a clear business case for health systems and payers to address SDOH.39
Our study findings have implications for a range of stakeholders addressing housing and social needs to improve health care delivery.1 First, the association between coded housing instability and higher rates of admissions, inpatient days, and costs has potential implications for payers and health systems. Housing and other SDOH will become increasingly important to promote health as payers transition from volume to value-based payment models32 and Medicare and Medicaid implement strategies focused on SDOH.40 Prior studies document how housing can improve health care effectiveness for patients, particularly for beneficiaries of Medicaid and similar programs.2 Specifically, health systems may find a practical business case to invest in affordable housing and supportive social services as a strategy to reduce inpatient hospital utilization associated with housing instability.1,39 Similarly, health systems and payers may consider other strategies, such as advocating for inclusionary zoning ordinances to increase the availability of affordable multifamily dwellings for renters at high risk of housing instability.31
Second, the high proportion of admissions for mental, behavioral, and neurodevelopmental disorders among patients with coded housing instability suggests that research and efforts addressing the intersection of housing instability, mental and behavioral health services, and inpatient health care utilization are likely areas of high impact. Specifically, efforts across multiple sectors may find mutually beneficial areas of interest. Findings from the present study further suggest that health care programs designed to address the needs of patients experiencing homelessness may also benefit other patients across the housing instability continuum. Third, findings support efforts to improve screening instruments and documentation methods. The increasing focus on housing instability and SDOH by clinicians and payers will require standardized, widely adopted protocols to develop interventions and evaluate effectiveness. This study generated baseline associations to inform future work in this area.
Limitations
Results of this cross-sectional study must be interpreted within the context of its limitations. First, because this study focused on adults aged 18 years and older, correlational findings may not extend to the younger US population. The NIS, however, is the largest publicly available all-payer data set in the US and therefore provides the most generalizable outcome estimates for adults. Second, large claims data sets including the NIS do not provide the clinical granularity of health records. To minimize effects of this limitation, only reliably identifiable outcomes in the NIS were used (primary diagnoses, inpatient days, and cost).
Third, cross-sectional study results cannot be interpreted causally. The unadjusted comparisons did not address potential confounders, such as differential use of Z-codes by reason for admission. More representative data collected via consistent and standardized screening are required to fully understand the association between housing instability and reason for hospitalization. Nevertheless, the descriptive comparisons produced by this study using a nationally representative data set provide critical baseline information. This information is necessary to improve social risk documentation and develop interventions focused on the association between housing instability and inpatient health care utilization, especially among patients admitted for mental, behavioral, and neurodevelopmental disorders.
Fourth, housing instability definitions widely vary and no standardized measure exists. This study relied on 5 imperfect and underutilized ICD-10 Z59-codes that, according to the literature, are likely more specific than sensitive indicators of housing instability.41,42,43,44,45 The Z59-codes likely capture the most extreme experience of housing instability: homelessness. Studies26,41,43,46,47,48,49 have also shown that these and other SDOH-related Z-codes (Z55-65) inconsistently map to SDOH screening instruments. Moreover, study results were consistent with prior work25,42,43,44 that demonstrated that Z-codes Z55-65 are only found in 1% to 2% of inpatient medical records. An additional study16 found that SDOH, including housing instability, were underrepresented in administrative data compared with physician notes. This prior work further supports that results of the present study reveal the “tip of the iceberg.”
Despite these limitations, ICD-10 Z-codes remain an important metric for measuring health disparities. They are already being used ubiquitously in clinical practice across US hospitals and internationally.50 ICD-10 codes are used by all World Health Organization (WHO) member nations, translated into 43 languages, and serve as the WHO’s basis for reporting health status, mortality, and medical reimbursements.
Conclusions
This cross-sectional analysis of a nationally representative data set found that coded housing instability among hospitalized patients was associated with higher rates of inpatient admissions for mental, behavioral, and neurodevelopmental disorders, longer hospital stays, and substantial health care costs. This association, however, is likely the tip of the iceberg and requires attention and cooperation from multiple sectors to improve screening, deepen understanding of the housing and health care utilization linkage by diagnosis, and develop effective interventions. These findings suggest that multisector efforts aimed at improving housing instability, mental and behavioral health services, and efficient utilization of inpatient hospital services may find opportunities for synergistic collaboration.
eTable 1. ICD-10 Codes Used to Create Primary Discharge Diagnoses Categories
eTable 2. NIS Sample Details
eTable 3. Characteristics of Hospitalized Patients Without and With Each of 5 Housing Instability Z59-Codes
eTable 4. Number of Hospital Admissions by Diagnosis Among Patients Without and With Housing Instability Codes Z59.0 (Homelessness) and Z59.1 (Inadequate Housing)
eTable 5. Number of Hospital Admissions by Diagnosis Among Patients Without and With Housing Instability Codes Z59.2 (Discord With Neighbors, Lodgers, and Landlords) and Z59.3 (Problems Related to Living in a Residential Institution)
eTable 6. Number of Hospital Admissions by Diagnosis Among Patients Without and With Housing Instability Code Z59.8 (Other Housing and Economic Problems)
eTable 7. Total Inpatient Days and Mean Length of Stay by Diagnosis Among Patients With and Without Housing Instability Codes Z59.0 (Homelessness), Z59.1 (Inadequate Housing), and Z59.2 (Discord With Neighbors, Lodgers, and Landlords)
eTable 8. Total Inpatient Days and Mean Length of Stay by Diagnosis Among Patients With and Without Housing Instability Codes Z59.3 (Problems Related to Living in a Residential Institution) and Z59.8 (Other Housing and Economic Problems)
eTable 9. Cost (Millions, USD) of Hospitalization by Diagnosis Among Patients With and Without Housing Instability Codes Z59.0 (Homelessness), Z59.1 (Inadequate Housing), and Z59.2 (Discord With Neighbors, Lodgers, and Landlords)
eTable 10. Cost (Millions, USD) of Hospitalization by Diagnosis Among Patients With and Without Housing Instability Codes Z59.3 (Problems Related to Living in a Residential Institution) and Z59.8 (Other Housing and Economic Problems)
eFigure. Most Common Reasons for Hospitalization Among Patients With and Without Coded Housing Instability
References
- 1.Koeman J, Mehdipanah R. Prescribing housing: a scoping review of health system efforts to address housing as a social determinant of health. Popul Health Manag. 2021;24(3):316-321. doi: 10.1089/pop.2020.0154 [DOI] [PubMed] [Google Scholar]
- 2.Kottke T, Abariotes A, Spoonheim JB. Access to affordable housing promotes health and well-being and reduces hospital visits. Perm J. 2018;22:17-079. doi: 10.7812/tpp/17-079 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Long CL, Franklin SM, Hagan AS, et al. Health-related social needs among older adults enrolled in Medicare Advantage. 2022;41(4). doi: 10.1377/hlthaff.2021.01547 [DOI] [PubMed] [Google Scholar]
- 4.Bailey P. Housing and health partners can work together to close the housing affordability gap. January 17, 2020. Center on Budget and Policy Priorities . Accessed May 16, 2022. https://www.cbpp.org/sites/default/files/atoms/files/1-17-20hous.pdf
- 5.Peng Y, Hahn RA, Finnie RKC, et al. ; Community Preventive Services Task Force . Permanent supportive housing with housing first to reduce homelessness and promote health among homeless populations with disability: a community guide systematic review. J Public Health Manag Pract. 2020;26(5):404-411. doi: 10.1097/phh.0000000000001219 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Frederick TJ, Chwalek M, Hughes J, Karabanow J, Kidd S. How stable is stable? Defining and measuring housing stability. J Community Psychol. 2014;42(8):964-979. doi: 10.1002/jcop.21665 [DOI] [Google Scholar]
- 7.Office of Disease Prevention and Health Promotion . Housing instability. Accessed June 6, 2022. https://www.healthypeople.gov/2020/topics-objectives/topic/social-determinants-health/interventions-resources/housing-instability#7#7
- 8.Hansen H, Metzl J. Structural competency in the US healthcare crisis: putting social and policy interventions into clinical practice. J Bioeth Inq. 2016;13(2):179-183. doi: 10.1007/s11673-016-9719-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Scally CP, Waxman E, Gourevitch R, Adeeyo S. Emerging strategies for integrating health and housing: innovations to sustain, expand, and replicate. July 2017. Urban Institute. Accessed October 5, 2022. https://www.urban.org/sites/default/files/publication/91941/emerging_strategies_in_integrating_health_and_housing_final_6.pdf
- 10.Kuehn BM. Hospitals turn to housing to help homeless patients. JAMA. 2019;321(9):822-824. doi: 10.1001/jama.2018.21476 [DOI] [PubMed] [Google Scholar]
- 11.Sandel M, Desmond M. Investing in housing for health improves both mission and margin. JAMA. 2017;318(23):2291-2292. doi: 10.1001/jama.2017.15771 [DOI] [PubMed] [Google Scholar]
- 12.Kushel MB, Vittinghoff E, Haas JS. Factors associated with the health care utilization of homeless persons. JAMA. 2001;285(2):200-206. doi: 10.1001/jama.285.2.200 [DOI] [PubMed] [Google Scholar]
- 13.Laurent AA, Matheson A, Escudero K, Lazaga A. Linking health and housing data to create a sustainable cross-sector partnership. Am J Public Health. 2020;110(S2):S222-S224. doi: 10.2105/ajph.2020.305693 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Saab D, Nisenbaum R, Dhalla I, Hwang SW. Hospital readmissions in a community-based sample of homeless adults: a matched-cohort study. J Gen Intern Med. 2016;31(9):1011-1018. doi: 10.1007/s11606-016-3680-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Titan A, Graham L, Rosen A, et al. Homeless status, postdischarge health care utilization, and readmission after surgery. Med Care. 2018;56(6):460-469. doi: 10.1097/mlr.0000000000000915 [DOI] [PubMed] [Google Scholar]
- 16.Navathe AS, Zhong F, Lei VJ, et al. Hospital readmission and social risk factors identified from physician notes. Health Serv Res. 2018;53(2):1110-1136. doi: 10.1111/1475-6773.12670 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Clemenzi-Allen A, Neuhaus J, Geng E, et al. Housing instability results in increased acute care utilization in an urban HIV clinic cohort. Open Forum Infect Dis. 2019;6(5):ofz148. doi: 10.1093/ofid/ofz148 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Fischer PJ, Breakey WR. The epidemiology of alcohol, drug, and mental disorders among homeless persons. Am Psychol. 1991;46(11):1115-1128. doi: 10.1037//0003-066x.46.11.1115 [DOI] [PubMed] [Google Scholar]
- 19.Susser E, Moore R, Link B. Risk factors for homelessness. Epidemiol Rev. 1993;15(2):546-556. doi: 10.1093/oxfordjournals.epirev.a036133 [DOI] [PubMed] [Google Scholar]
- 20.National Academies of Sciences, Engineering, and Medicine . Integrating Social Care Into the Delivery of Health Care: Moving Upstream to Improve the Nation's Health. The National Academies Press. 2019. [PubMed] [Google Scholar]
- 21.National Academies of Sciences, Engineering, and Medicine . Permanent Supportive Housing: Evaluating the Evidence For Improving Health Outcomes Among People Experiencing Chronic Homelessness. The National Academies Press. 2018. [PubMed] [Google Scholar]
- 22.Sadowski LS, Kee RA, VanderWeele TJ, Buchanan D. Effect of a housing and case management program on emergency department visits and hospitalizations among chronically ill homeless adults: a randomized trial. JAMA. 2009;301(17):1771-1778. doi: 10.1001/jama.2009.561 [DOI] [PubMed] [Google Scholar]
- 23.Lachaud J, Mejia-Lancheros C, Durbin A, et al. The effect of a housing first intervention on acute health care utilization among homeless adults with mental illness: long-term outcomes of the At Home/Chez-Soi Randomized Pragmatic Trial. J Urban Health. 2021;98(4):505-515. doi: 10.1007/s11524-021-00550-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Agency for Healthcare Research and Quality . Healthcare Cost and Utilization Project National Inpatient Sample (NIS). Accessed October 5, 2022. https://www.hcup-us.ahrq.gov/nisoverview.jsp
- 25.Hatef E, Rouhizadeh M, Tia I, et al. Assessing the availability of data on social and behavioral determinants in structured and unstructured electronic health records: a retrospective analysis of a multilevel health care system. JMIR Med Inform. 2019;7(3):e13802. doi: 10.2196/13802 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.De Marchis EH, Ettinger de Cuba SA, Chang L, et al. Screening discordance and characteristics of patients with housing-related social risks. Am J Prev Med. 2021;61(1):e1-e12. doi: 10.1016/j.amepre.2021.01.027 [DOI] [PubMed] [Google Scholar]
- 27.Montgomery AE, Sorrentino AE, Cusack MC, et al. Recent intimate partner violence and housing instability among women veterans. Am J Prev Med. 2018;54(4):584-590. doi: 10.1016/j.amepre.2018.01.020 [DOI] [PubMed] [Google Scholar]
- 28.Weleff J, Dalton J, Yuhas J, et al. 298 Characteristics and utilization patterns of homeless and housing-insecure patients presenting to emergency departments across a large health system. Ann Emerg Med. 2019;74(4):S117. doi: 10.1016/j.annemergmed.2019.08.256 [DOI] [Google Scholar]
- 29.Yang D, Dalton JE. A unified approach to measuring the effect size between two groups using SAS. 2012. Accessed October 5, 2022. http://support.sas.com/resources/papers/proceedings12/335-2012.pdf
- 30.Kaplan LM, Vella L, Cabral E, et al. Unmet mental health and substance use treatment needs among older homeless adults: results from the HOPE HOME Study. J Community Psychol. 2019;47(8):1893-1908. doi: 10.1002/jcop.22233 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Padgett DK. Homelessness, housing instability and mental health: making the connections. BJPsych Bull. 2020;44(5):197-201. doi: 10.1192/bjb.2020.49 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Blumenthal D, Abrams MK. Tailoring complex care management for high-need, high-cost patients. JAMA. 2016;316(16):1657-1658. doi: 10.1001/jama.2016.12388 [DOI] [PubMed] [Google Scholar]
- 33.Byrne T, Nelson RE, Montgomery AE, Brignone E, Gundlapalli AV, Fargo JD. Comparing the utilization and cost of health services between veterans experiencing brief and ongoing episodes of housing instability. J Urban Health. 2017;94(1):54-63. doi: 10.1007/s11524-016-0110-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Pollack CE, Blackford AL, Du S, Deluca S, Thornton RLJ, Herring B. Association of receipt of a housing voucher with subsequent hospital utilization and spending. JAMA. 2019;322(21):2115-2124. doi: 10.1001/jama.2019.17432 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Larimer ME, Malone DK, Garner MD, et al. Health care and public service use and costs before and after provision of housing for chronically homeless persons with severe alcohol problems. JAMA. 2009;301(13):1349-1357. doi: 10.1001/jama.2009.414 [DOI] [PubMed] [Google Scholar]
- 36.Wright BJ, Vartanian KB, Li H-F, Royal N, Matson JK. Formerly homeless people had lower overall health care expenditures after moving into supportive housing. Health Aff (Millwood). 2016;35(1):20-27. doi: 10.1377/hlthaff.2015.0393 [DOI] [PubMed] [Google Scholar]
- 37.Fenwick E, Macdonald C, Thomson H. Economic analysis of the health impacts of housing improvement studies: a systematic review. J Epidemiol Community Health. 2013;67(10):835-845. doi: 10.1136/jech-2012-202124 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Ly A, Latimer E. Housing First impact on costs and associated cost offsets: a review of the literature. Can J Psychiatry. 2015;60(11):475-487. doi: 10.1177/070674371506001103 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Doran KM, Misa EJ, Shah NR. Housing as health care—New York’s boundary-crossing experiment. N Engl J Med. 2013;369(25):2374-2377. doi: 10.1056/NEJMp1310121 [DOI] [PubMed] [Google Scholar]
- 40.Centers for Medicare & Medicaid Services . Letter from Anne Marie Costello to state health officials regarding “Opportunities in Medicaid and CHIP to address social determinants of health (SDOH).” January 7, 2021. Accessed October 7, 2022. https://www.medicaid.gov/federal-policy-guidance/downloads/sho21001
- 41.Vest JR, Wu W, Mendonca EA. Sensitivity and specificity of real-world social factor screening approaches. J Med Syst. 2021;45(12):111. doi: 10.1007/s10916-021-01788-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Guo Y, Chen Z, Xu K, et al. International Classification of Diseases, Tenth Revision, Clinical Modification social determinants of health codes are poorly used in electronic health records. Medicine (Baltimore). 2020;99(52):e23818. doi: 10.1097/MD.0000000000023818 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Truong HP, Luke AA, Hammond G, Wadhera RK, Reidhead M, Joynt Maddox KE. Utilization of social determinants of health ICD-10 Z-codes among hospitalized patients in the United States, 2016-2017. Med Care. 2020;58(12):1037-1043. doi: 10.1097/MLR.0000000000001418 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Weeks WB, Cao SY, Lester CM, Weinstein JN, Morden NE. Use of Z-codes to record social determinants of health among fee-for-service Medicare beneficiaries in 2017. J Gen Intern Med. 2020;35(3):952-955. doi: 10.1007/s11606-019-05199-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Dapkins I, Blecker SB. Homelessness and Medicaid churn. Ethn Dis. 2021;31(1):89-96. doi: 10.18865/ed.31.1.89 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Jacobs ZG. Codifying social determinants of health: a gap in the ICD-10-CM. J Gen Intern Med. 2021;36(10):3205-3207. doi: 10.1007/s11606-021-06742-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Gottlieb L, Tobey R, Cantor J, Hessler D, Adler NE. Integrating social and medical data to improve population health: opportunities and barriers. Health Aff (Millwood). 2016;35(11):2116-2123. doi: 10.1377/hlthaff.2016.0723 [DOI] [PubMed] [Google Scholar]
- 48.Olson DP, Oldfield BJ, Morales Navarro S. Standardizing social determinants of health assessments. Health Affairs Blog . March 18, 2019. Accessed October 5, 2022. https://www.healthaffairs.org/do/10.1377/hblog20190311
- 49.Arons A, DeSilvey S, Fichtenberg C, Gottlieb L. Documenting social determinants of health-related clinical activities using standardized medical vocabularies. JAMIA Open. 2018;2(1):81-88. doi: 10.1093/jamiaopen/ooy051 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Billioux A, Verlander K, Anthony S, Alley D. Standardized screening for health-related social needs in clinical settings: The Accountable Health Communities Screening Tool. National Academy of Medicine ; May 30, 2017. Accessed October 5, 2022. https://nam.edu/standardized-screening-for-health-related-social-needs-in-clinical-settings-the-accountable-health-communities-screening-tool/
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. ICD-10 Codes Used to Create Primary Discharge Diagnoses Categories
eTable 2. NIS Sample Details
eTable 3. Characteristics of Hospitalized Patients Without and With Each of 5 Housing Instability Z59-Codes
eTable 4. Number of Hospital Admissions by Diagnosis Among Patients Without and With Housing Instability Codes Z59.0 (Homelessness) and Z59.1 (Inadequate Housing)
eTable 5. Number of Hospital Admissions by Diagnosis Among Patients Without and With Housing Instability Codes Z59.2 (Discord With Neighbors, Lodgers, and Landlords) and Z59.3 (Problems Related to Living in a Residential Institution)
eTable 6. Number of Hospital Admissions by Diagnosis Among Patients Without and With Housing Instability Code Z59.8 (Other Housing and Economic Problems)
eTable 7. Total Inpatient Days and Mean Length of Stay by Diagnosis Among Patients With and Without Housing Instability Codes Z59.0 (Homelessness), Z59.1 (Inadequate Housing), and Z59.2 (Discord With Neighbors, Lodgers, and Landlords)
eTable 8. Total Inpatient Days and Mean Length of Stay by Diagnosis Among Patients With and Without Housing Instability Codes Z59.3 (Problems Related to Living in a Residential Institution) and Z59.8 (Other Housing and Economic Problems)
eTable 9. Cost (Millions, USD) of Hospitalization by Diagnosis Among Patients With and Without Housing Instability Codes Z59.0 (Homelessness), Z59.1 (Inadequate Housing), and Z59.2 (Discord With Neighbors, Lodgers, and Landlords)
eTable 10. Cost (Millions, USD) of Hospitalization by Diagnosis Among Patients With and Without Housing Instability Codes Z59.3 (Problems Related to Living in a Residential Institution) and Z59.8 (Other Housing and Economic Problems)
eFigure. Most Common Reasons for Hospitalization Among Patients With and Without Coded Housing Instability