INTRODUCTION
Interhospital fragmentation of care, when a patient is readmitted to a different hospital than they were initially discharged from, occurs in 20–25% of readmissions1 and is associated with poor outcomes, including 20% higher in-hospital mortality.1–3
Previous studies have examined the impact of interhospital care fragmentation on specific populations, including post-operative3 and heart failure patients.2 Older adults, especially those who migrate seasonally,4 may be at particular risk for fragmented readmissions. We describe the prevalence of fragmented readmissions among older adults that occur within and outside of their state of residence and assess if their odds of in-hospital mortality differ.
METHODS
We analyzed the 2018 National Readmissions Database (NRD).5 We included patients ≥ 65 years old with 30-day readmissions and excluded interhospital transfers and January admissions, as those could represent readmissions from December 2017.
The NRD labels admissions as “in the patient’s state of residence” or not. Readmissions were considered nonfragmented if the admission and readmission were to the same hospital. Readmissions were fragmented if the readmission hospital was different than the admission hospital; both groups were further categorized as “in the patient’s state of residence” or not. The primary outcome was in-hospital mortality during the readmission.
We used weighted chi-square tests and ANOVA to compare patient and readmission characteristics. We used weighted unadjusted and adjusted logistic regression models to compare the outcome across exposure groups of interest. We used both in-state and out-of-state non-fragmented admission-readmission dyads as reference groups. The adjusted model included demographic (age, sex, payer, quartiles of median incomes in a patient’s zip code, the rural/urban designation of the readmission hospital) and clinical variables (length-of-stay of the index admission, all patient refined diagnosis resource group risk of mortality, major diagnostic category of the readmission, alcohol use, drug use, if the patient left against medical advice, whether the readmission was elective, and whether the admission originated in the ED). All analyses were performed in SAS 9.4 (Cary, NC) using weighted survey procedures.5
This study was deemed exempt from IRB review by the Emory University Institutional Review Board.
RESULTS
78.9% (n=378,809) and 16.7% (n=79,898) of weighted readmissions were nonfragmented and fragmented, respectively, in the patient’s state of residence, and 3.9% (n=18,640) and 0.52% (n=2490) were nonfragmented and fragmented, respectively, outside of the patient’s state of residence (Table 1).
Table 1.
Demographic and Hospital Characteristics of Fragmentation Categories, National Readmissions Database
| All | Non-fragmented/same state | Fragmented/same state | Non-fragmented/different state | Fragmented/different state | p value | ||
|---|---|---|---|---|---|---|---|
| Frequency | |||||||
| 78.9% (n=378,809)* | 16.7% (n=79,898) | 3.9% (n=18,640) | 0.52% (n=2490) | ||||
| Age | 77.1 ± 7.8 | 77.2 ± 7.8 | 75.8 ± 7.3 | 76.6 ± 7.6 | 76.2 ± 7.5 | <0.001 | |
| Sex | 51.7% | 52.1% | 47.7% | 50.7% | 48.3% | <0.001 | |
| Payer | Medicare | 91.5% | 91.3% | 92.2% | 92.8% | 90.9% | <0.001 |
| Medicaid | 1.5% | 1.5% | 0.4% | 1.9% | --ǂ | ||
| Private | 5.4% | 5.4% | 5.4% | 5.6% | 5.1% | ||
| Self-pay | 0.3% | 0.3% | 0.4% | --ǂ | 0.3% | ||
| No charge | 0.04% | 0.04% | --ǂ | 0.1% | --ǂ | ||
| Other | 1.4% | 1.3% | 1.6% | 1.3% | 1.4% | ||
| Zip income quartile | 1–45,999 | 26.8% | 25.9% | 33.0% | 29.7% | 36.1% | <0.001 |
| 46,000–58,999 | 27.8% | 27.6% | 28.7% | 28.6% | 28.4% | ||
| 59,000–78,999 | 25.3% | 25.8% | 21.9% | 23.6% | 22.3% | ||
| ≥ 79,000 | 20.1% | 20.8% | 16.3% | 18.0% | 13.2% | ||
| Alcohol use (Y) | 2.8% | 2.8% | 2.5% | 3.3% | 3.6% | <0.001 | |
| Drug use | 1.7% | 1.6% | 1.0% | 2.2% | 1.6% | <0.001 | |
| Left AMA | 1.0% | 0.9% | 0.4% | 2.0% | 2.1% | <0.001 | |
| Hospital urban/rural status | Large metropolitan (>1 million residents) | 56.8% | 56.3% | 46.2% | 61.9% | 47.4% | <0.001 |
| Small metropolitan (<1 million residents) | 35.8% | 36.4% | 45.9% | 30.9% | 46.2% | ||
| Micro-politan | 5.7% | 5.8% | 7.2% | 5.0% | 5.4% | ||
| Not metropolitan or micro-politan | 1.6% | 1.5% | 0.8% | 2.2% | 1.0% | ||
| Elective admission (Y) | 8.0% | 7.6% | 12.9% | 8.9% | 11.9% | <0.001 | |
| Admission originated in ED (Y) | 82.0% | 84.1% | 66.4% | 76.0% | 68.4% | <0.001 | |
| Length-of stay of the index admission | 5.8 ± 6.2 | 5.7 ± 6.0 | 6.2 ± 7.1 | 5.9 ± 7.0 | 6.0 ± 6.6 | <0.001 | |
| APR-DRG score | 2.7 ± 0.9 | 2.7 ± 0.9 | 2.7 ± 0.9 | 2.8 ± 0.9 | 2.8 ± 0.9 | <0.001 | |
| In-hospital mortality | 6.2% | 6.0% | 5.7% | 7.0% | 7.1% | <0.001 | |
*Frequencies are weighted according to the weighting scheme provided by the Healthcare Cost and Utilization Project (HCUP)
ǂSuppressed due to weighted frequency ≤ 20 per HCUP Data Use Agreement
Compared to out-of-state non-fragmented readmissions, in-hospital mortality was higher in out-of-state fragmented readmissions (AOR 1.40, 95% CI 1.15, 1.7). Compared to in-state fragmented readmissions, in-hospital mortality was higher in in-state fragmented readmissions (AOR 1.18, 95% CI 1.14, 1.23) (Table 2).
Table 2.
Association Between Fragmented Care and Readmission In-Hospital Mortality for Older Adults Readmitted in the State They Reside in and a Different State than They Reside in, 30-Day Readmissions
| Fragmented/same state | Nonfragmented/different state | Fragmented/different state | |
|---|---|---|---|
| Nonfragmented/same state (REF) | 1.18 (1.14, 1.23) | 0.92 (0.86, 0.99) | 1.29 (1.08, 1.55) |
| Fragmented/same state (REF) | -- | 0.78 (0.72, 0.84) | 1.09 (0.91, 1.31) |
| Nonfragmented/different state (REF) | 1.29 (1.19, 1.39) | -- | 1.40 (1.15, 1.71) |
All estimates correspond to a weighted adjusted odds ratio for in-hospital mortality (95% confidence intervals) for an exposure (column) relative to the reference group (row). Statistically significant results are in bold
The model adjusts for the major diagnostic category (MDC) of the readmission, all patients refined diagnosis related groups (APR-DRG) mortality score, alcohol use, drug use, left against medical advice, age, sex, payer, zip income quartile, hospital urban/rural status, if the admission was elective or not, if the admission originated in the emergency department or not
DISCUSSION
Older adults are at risk for poor outcomes following hospital readmissions. When these readmissions are fragmented, mortality risk increases and may be compounded by cross-state fragmentation. We found that the odds of dying during a fragmented readmission, whether out-of-state or in-state, were higher than the odds of dying during a nonfragmented readmission. While inter-state care fragmentation has not been previously examined, our findings are consistent with higher rates of in-hospital mortality in fragmented readmissions observed in other studies.1–3
There are reasons beyond seasonal migration why older adults would receive care in a different state which are not ascertainable in a secondary data analysis, including traveling for specialized care6 or living close to a state border. It is likely that in some of these instances, i.e., specialized care, the patient brings their medical records, so information discontinuity may not exist at the same level across all situations. Similarly, providers may be more apt to obtain outside records if they know their patient receives care elsewhere. Additionally, because the NRD does not include state names, we cannot measure if certain states are disproportionately impacted by out-of-state fragmentation.
These findings highlight possible “information discontinuity” between states, and further the case for the need to extend information sharing beyond the borders of a single state. These findings should be considered when counseling older adults and their care partners about where to seek care should the need for hospitalization arise. If a patient knows they will be traveling and at risk for fragmented, out-of-state inpatient care, they can be advised on the risks and to prepare, perhaps, by bringing their medical records with them.
Declarations
Disclosure Statement
Research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health under Award Number K23AG065505. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Conflict of Interest
The authors declare that they do not have a conflict of interest.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Turbow S, Sudharsanan N, Rask KJ, Ali MK. Association between interhospital care fragmentation, readmission diagnosis, and outcomes. Am J Manag Care. 2021;27(5):e164–e70. doi: 10.37765/ajmc.2021.88639. [DOI] [PubMed] [Google Scholar]
- 2.McAlister FA, Youngson E, Kaul P. Patients with heart failure readmitted to the original hospital have better outcomes than those readmitted elsewhere. Journal of the American Heart Association. 2017;6(5):e004892–7. doi: 10.1161/JAHA.116.004892. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Brooke BS, Goodney PP, Kraiss LW, Gottlieb DJ, Samore MH, Finlayson SR. Readmission destination and risk of mortality after major surgery: an observational cohort study. Lancet. 2015;386(9996):884–95. doi: 10.1016/S0140-6736(15)60087-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Shariff SZ, Paterson JM, Dixon SN, Garg AX, Clemens KK. Prevalence of winter migration to warmer destinations among ontarians (“snowbirds”) and patterns of their use of health care services: a population-based analysis. CMAJ Open. 2021;9(2):E491–E9. doi: 10.9778/cmajo.20200270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.National Readmission Database. Healthcare Cost and Utilization Project (HCUP). 2013-2014. Agency for Healthcare Research and Quality, Rockville, MD. www.hcup-us.ahrq.gov/databases.jsp. [Internet]. [PubMed]
- 6.Fujiwara T, Ogura K, Healey J. Greater travel distance to specialized facilities is associated with higher survival for patients with soft-tissue sarcoma: US Nationwide Patterns. PLoS One. 2021;16(6):e0252381. doi: 10.1371/journal.pone.0252381. [DOI] [PMC free article] [PubMed] [Google Scholar]