All-Cause, 30-Day Readmissions Among Persons With Intellectual and Developmental Disabilities and Mental Illness
Abstract
Objective:
Early hospital readmissions within 30 days of discharge are common and costly. This research describes predictors of all-cause, 30-day hospital readmissions among persons with intellectual and developmental disabilities (IDD), a group known to experience high rates of hospitalization.
Methods:
A cohort of 66,484 adults with IDD from Ontario, Canada, was used to create two subgroups: individuals with IDD only and those with IDD and mental illness. The rates of hospital readmission were determined and contrasted with a comparison subgroup of people without IDD who have mental illness.
Results:
Compared with those with mental illness only, individuals with IDD and mental illness were 1.7 times more likely to experience a hospital readmission within 30 days. Predictors of their readmission rates included being a young adult and having high morbidity levels.
Conclusions:
The high rate of hospital readmission suggests that individuals with IDD and mental illness need attention regarding discharge planning and outpatient follow-up.
Early readmissions to inpatient care are common and costly events that governments have been trying to reduce. In Ontario (population of 14 million), Canada, the 30-day readmission rate is reported to be 15%, accounting for $705 million in hospital costs in fiscal year 2008–2009 (1). The Ontario government considers hospital readmissions within 30 days of discharge an important indicator of health system performance, and decreasing the rate is a priority for both Canadian provincial and American state hospitals (2,3).
Persons with intellectual and developmental disabilities (IDD) experience more mental and physical health problems and use more health services compared with those without IDD (4,5). They are reported to be among the most difficult populations to serve, with very high health care costs (6), attributable in part to high hospitalization rates (7). Although it is reported that persons with IDD have poor hospital experiences (8), it is likely that persons with both IDD and mental illness are at even higher risk of poor outcomes, such as readmissions, because of their complex health needs. Currently, only one study has explored all-cause, 30-day readmission rates among those with IDD (9). However, no study has yet examined the influence of having IDD and a comorbid mental illness on hospital readmission rates.
Our goal was to study the influence of IDD and IDD plus mental illness on hospital readmission rates. We compared rates of early readmissions at the population level across three mutually exclusive subgroups: those with IDD only, those with IDD and a mental illness, and those without IDD but with a mental illness (referred henceforth as “mental illness only”). Our main objective was to determine whether IDD only and IDD and mental illness predicted all-cause, 30-day readmissions rates after adjusting for relevant demographic, health, and health-service variables. The second objective was to identify factors associated with readmissions unique to each subgroup.
Methods
To complete this retrospective cohort study, we accessed eight administrative databases containing health service, social-support, and demographic data maintained at the Institute for Clinical and Evaluative Sciences in Toronto. The population-based, longitudinal data at the Institute for Clinical and Evaluative Sciences includes all Ontarians eligible for health services, thus representing all legal residents of the province. The study included individuals 19–65 years old on April 1, 2010, and those eligible for health insurance.
This study’s main independent variable consisted of three study subgroups. Two subgroups had IDD (IDD only [N=36,496] and IDD and mental illness [N=29,988]). These were derived from the Health Care Access Research and Developmental Disabilities cohort, which has been described in a separate source (4). The subgroup of persons with IDD and mental illness consisted of individuals with IDD who received a diagnosis of mental illness (including substance use disorders) in at least one of the health databases between April 1, 2007, and March 31, 2009. The third subgroup, used for comparison, consisted of individuals with a mental illness only (N=49,130). To create this subgroup, a 20% random sample was drawn from the population of Ontario without IDD; within this sample, mental illness was determined by using the same criteria as the subgroup with IDD and mental illness (received diagnosis of mental illness, including substance use disorder, between April 1, 2007, and March 31, 2009).
Other independent variables were categorized by using Kangovi’s framework (10). The framework conceptualizes readmissions in terms of quality and access within outpatient and inpatient care and acknowledges the influence of patient health and socioeconomic factors; demographic variables were also included as possible determinants.
For objective 1, adjusted odds of being readmitted for any reason (medical, surgical, or psychiatric) within 30 days after discharge from an index hospitalization for any reason (dependent variable) were compared across the three study subgroups (main independent variable). An index hospitalization was defined as the first hospitalization for an individual occurring between April 1, 2010, and March 31, 2011.
Multivariable logistic regression with generalized estimating equations was used to see whether the association between the study subgroup variable and 30-day readmission was significant when controlling for the other variables. To address objective 2, we built three separate generalized estimating equations logistic regression models to identify significant predictors of all-cause, 30-day readmissions within each subgroup. [A more detailed description of the methods is available as an online supplement to this brief report.]
This study was approved by the institutional ethics review boards at Sunnybrook Health Sciences Centre (Toronto), the Centre for Addiction and Mental Health (Toronto), and the University of Ontario Institute of Technology (Oshawa, Ontario).
Results
Among the study subgroups, 55,362 experienced an index hospital admission within the analysis window. Of these, 3,979 individuals met our definition of IDD and mental illness (mean±SD age=40.3±13.5); 2,253 individuals met our definition of IDD only (mean age=41.1±13.7); and 49,130 individuals were in the subgroup with mental illness only (mean age=42.0±13.3). Among the IDD subgroups, slightly more than half were male (IDD and mental illness: N=1,997, 50.2%; IDD only: N=1,174, 52.1%); however, in the subgroup with mental illness only, most participants were female (N=32,665, 66.5%). A total of 23.2% of the subgroup with IDD and mental illness and 18.3% of the subgroup with mental illness only had a substance use disorder diagnosis.
The average length of stay for the index hospital visit was highest in the subgroup with IDD and mental illness (14.9±55.5 days; 95% confidence interval [CI]=13.2–16.6) and lowest in the subgroup with mental illness only (6.3±26.7 days, CI=6.1–6.5); for the IDD-only subgroup, the average stay was 11.3±51.1 days (CI=9.2–13.4). Overall, 8.7% of those experiencing an index visit were readmitted within 30 days. By study subgroup, the 30-day readmission rates were as follows: 14.7% (IDD and mental illness), 10.2% (IDD only), and 8.2% (mental illness only).
Model 1 in Table 1 compares the odds of readmission among the three study subgroups. After adjusting for relevant demographic, economic, health, and health service variables, the odds of experiencing a readmission remained significantly higher in the subgroup with IDD and mental illness (odds ratio [OR]=1.66) and in the IDD-only (OR=1.27) subgroup compared with the subgroup with mental illness only.
| Multivariate stratified model | ||||||||
|---|---|---|---|---|---|---|---|---|
| Model 1 (multivariatefull model) | Model 2: IDD only | Model 3: IDD and mental illness | Model 4: mental illness only | |||||
| Framework characteristic | AOR | 95% CI | AOR | 95% CI | AOR | 95% CI | AOR | 95% CI |
| Main independent variable (study subgroup) | ||||||||
| IDD and mental illness | 1.66 | 1.53–1.79* | ||||||
| IDD only | 1.27 | 1.09–1.49* | ||||||
| Mental illness only (reference) | ||||||||
| Patient demographic information | ||||||||
| Age on April 1, 2010 (years) | ||||||||
| 19–25 | 1.06 | .97–1.16 | 1.10 | .64–1.90 | 1.43 | 1.08–1.87* | 1.01 | .90–1.14 |
| 26–35 | .89 | .82–.96* | 1.27 | .81–1.98 | 1.25 | 1.00–1.55* | .85 | .78–.93* |
| 36–45 | .96 | .90–1.02 | .93 | .59–1.48 | 1.14 | .85–1.52 | .95 | .89–1.02 |
| 46–55 | .97 | .91–1.03 | 1.03 | .70–1.51 | 1.01 | .84–1.23 | .97 | .90–1.06 |
| 56–65 (reference) | ||||||||
| Gender | ||||||||
| Male | 1.22 | 1.14–1.30* | 1.32 | 1.09–1.59* | 1.01 | .81–1.27 | 1.22 | 1.15–1.32* |
| Female (reference) | ||||||||
| Region | ||||||||
| Urban | 1.00 | .89–1.12 | .99 | .68–1.46 | 1.33 | .87–2.06 | .96 | .87–1.07 |
| Rural (reference) | ||||||||
| Patient health status(morbidity)b | ||||||||
| Very high | 2.66 | 2.26–3.11* | 1.87 | 1.03–3.42* | 3.00 | 1.26–7.15* | 2.77 | 2.26–3.40* |
| High | 1.48 | 1.24–1.77* | 1.20 | .74–1.94 | 1.87 | .77–4.56 | 1.40 | 1.11–1.75* |
| Moderate | 1.39 | 1.17–1.65* | 1.51 | .19–2.47 | 2.12 | .92–4.89 | 1.49 | 1.20–1.85* |
| Low morbidity (reference) | ||||||||
| Patient socioeconomic resources (income quintile)b | ||||||||
| 1st (low) | 1.07 | .96–1.18 | .74 | .60–.91* | .91 | .69–1.21 | 1.11 | 1.00–1.23* |
| 2nd | 1.14 | 1.00–1.30* | .94 | .73–1.21 | 1.00 | .71–1.40 | 1.17 | 1.03–1.33* |
| 3rd | .99 | .90–1.10 | .97 | .70–1.35 | .87 | .57–1.31 | 1.01 | .89–1.14 |
| 4th | .97 | .89–1.06 | .73 | .49–1.10 | .76 | .60–.96* | 1.01 | .91–1.12 |
| 5th (high) (reference) | ||||||||
| Services: quality and accessb | ||||||||
| Inpatient health services | ||||||||
| Length of stay (quality) | 1.001 | 1.000–1.001* | 1.001 | .999–1.003 | 1.000 | .999–1.002 | 1.001 | 1.000–1.002* |
| LHIN beds per 1,000 population (access)c | 1.10 | 1.02–1.17* | 1.45 | .99–2.12 | 1.12 | .97–1.28 | 1.08 | 1.01–1.15* |
| Outpatient health services | ||||||||
| Continuity of Care Index (quality)d | ||||||||
| <3 visits | 1.23 | 1.05–1.45* | .99 | .62–1.59 | .88 | .55–1.39 | 1.34 | 1.14–1.58* |
| UPC <.75 | 1.24 | 1.14–1.35* | 1.08 | .88–1.33 | 1.13 | .97–1.31 | 1.26 | 1.15–1.38* |
| UPC ≥.75 (reference) | ||||||||
| Number of visits to primary physician or psychiatrist (access) | 1.01 | 1.00–1.01* | 1.01 | .99–1.03 | 1.01 | 1.00–1.01* | 1.01 | 1.00–1.01* |
| Primary care physician supply (access)e | 1.07 | 1.00–1.15* | .94 | .77–1.15 | 1.07 | .92–1.23 | 1.08 | 1.01–1.15* |
| Psychiatrist supply (access)e | .97 | .95–.99* | .84 | .76–.93* | .97 | .91–1.02 | .98 | .96–1.00 |
a
AOR, adjusted odds ratio
b
Framework categories are based on work by Kangovi and Grande (10)
c
LHIN, local health integration network
d
At least three visits are required to calculate the usual provider care (UPC) value. Low continuity of care was defined as a UPC value less than .75, whereas high continuity was defined as .75 or greater.
e
Number of full-time equivalent specialists per 10,000 population in the LHIN where community is located.
*
p<.05
Table 1 also shows the factors associated with 30-day readmissions for each subgroup (see stratified models 2–4). Few consistent patterns are discernible across the subgroups. In terms of demographic characteristics, the subgroup with IDD and mental illness had higher odds of being readmitted if they were younger (OR=1.43 for those ages 19–25 and OR=1.25 for those ages 26–35); in contrast, among members of the subgroup with mental illness only, those who were between ages 26 and 35 showed lower odds (OR=.85) of being readmitted compared with the oldest age group. Being male significantly increased the odds of being readmitted in the IDD-only subgroup (OR=1.32) and the subgroup with mental illness only (OR=1.22). Across all three study subgroups, those with the highest morbidity levels were more likely to be readmitted compared with the lowest morbidity level (IDD only, OR=1.87; IDD and mental illness, OR=3.00; and mental illness only, OR=2.77).
Discussion
This study found that persons with IDD (with or without a mental illness) were much more likely to experience a 30-day readmission compared with persons with mental illness but no IDD. After adjusting for factors known to be related to readmissions, the OR was greatest when comparing the subgroup with IDD and mental illness with the subgroup with mental illness only. When separate models were created for the three study subgroups, the only consistent predictor of readmissions was morbidity.
Only one other study has examined readmissions among persons with IDD. Kelly et al. (9) found no difference in the rate of 30-day readmissions between persons with and without IDD in a single hospital in England. They reported that 13% of their sample with IDD and 11% of those without IDD experienced a readmission. In contrast, our population-based study found that IDD status was a significant contributor to readmissions after controlling for multiple other variables; furthermore, the subgroup with IDD and mental illness had almost two times higher odds of readmission (OR=1.66) compared with the subgroup with mental illness only.
The increased risk of readmission among those with IDD and mental illness may be related to the amount and quality of discharge planning for these individuals but also to the availability and quality of clinical services after discharge to the community. A literature review concluded that the needs of people with IDD are not being adequately met when they are hospitalized (7). It may be of benefit to flag persons with IDD, and specifically those with IDD and mental illness, during index admissions to target their unique health and social care needs, thereby potentially preventing readmission. In the United States, increased spending on occupational therapy services was associated with lower readmissions rates, an outcome attributed to the profession’s focus on patient functional and social needs when preparing for discharge (11). In order for persons with IDD and mental illness to be successfully discharged into the community, housing models are needed that provide access to specialized clinical support that address the complex mental health and frequent behavioral challenges exhibited by individuals in this subgroup (12).
Among members of the subgroup with IDD and mental illness, age played a role in readmissions. Persons between 19 and 25 years old in this subgroup experienced significantly higher odds (OR=1.43) of being readmitted compared with persons ages 56 to 65. The higher risk of readmission in the younger group may be due to the lack of health and social supports for young people with IDD and mental illness and their families after they are discharged home. Older individuals, in contrast, are more likely to reside in supported living environments (for example, group homes) because of aging parents. Better health service outcomes among persons living in group homes has been attributed, in part, to the presence of trained caregivers with quicker access to community services and more structured health care planning (13).
Among persons with only mental illness, the odds of readmission to a hospital were highest among those living in the poorest neighborhoods. Interestingly, we failed to detect this pattern among the IDD subgroups. It is possible that the risk of readmission is (mostly) equally high among persons with IDD regardless of where they live (that is, no socioeconomic gradient). This may be due to fewer barriers to health care because of income.
When examining factors representing quality and access of inpatient and outpatient health services, no compelling pattern of significant variables could be discerned. Our findings suggest that issues other than factors such as continuity of care affect the readmission rates of those in the subgroup with IDD and mental illness and the IDD-only subgroup.
There were limitations to our study. For instance, it was not possible to distinguish between avoidable and nonavoidable readmissions. This is possibly important to consider in future research because some authors have found that only around one in four readmissions are avoidable (14). Our data did not permit us to identify potentially relevant social factors such as marital status or type of residence after index hospitalization. As suggested by some researchers, such information may be especially relevant among vulnerable populations (10). It can be difficult to correctly diagnosis mental illness among people with IDD because of problems they experience communicating symptoms and behaviors that can mimic characteristics of mental illness (15). Failure to properly diagnose mental illness in this group would lead to misclassification, thus influencing effect size estimates.
Conclusions
This exploratory study has shown that adults with IDD, especially those with comorbid mental illness, are at significant risk of being readmitted to a hospital within 30 days of discharge. Future research should investigate the impact of tailored discharge planning and well-supported community housing. Such initiatives have the potential to reduce costs and improve quality of life.
Footnote
This study is part of the Health Care Access Research and Developmental Disabilities program (www.hcardd.ca). Parts of this material are based on data and information compiled and provided by the Canadian Institute for Health Information (CIHI). The opinions, results and conclusions herein are those of the authors and are independent from the funding sources or the data providers. No endorsement by the Province of Ontario, MOHLTC, CIHR, ICES, or CIHI is intended or should be inferred.
Supplementary Material
File (appi.ps.201600534.ds001.pdf)
- View/Download
- 81.82 KB
References
1.
Enhancing the Continuum of Care: Report of the Avoidable Hospitalization Advisory Panel. Toronto, Ontario Ministry of Health and Long-Term Care, 2011
2.
Excellent Care for All. Toronto, Ontario Ministry of Health and Long-Term Care, 2014
3.
Joynt KE, Jha AK: Characteristics of hospitals receiving penalties under the Hospital Readmissions Reduction Program. JAMA 309:342–343, 2013
4.
Lunsky Y, Klein-Geltink J, Yates E: Atlas on the Primary Care of Adults With Developmental Disabilities in Ontario. Toronto, Institute for Clinical Evaluative Sciences, 2013
5.
Cooper S-A, McLean G, Guthrie B, et al: Multiple physical and mental health comorbidity in adults with intellectual disabilities: population-based cross-sectional analysis. BMC Family Practice 16:110, 2015
6.
Polder JJ, Meerding WJ, Bonneux L, et al: Healthcare costs of intellectual disability in the Netherlands: a cost-of-illness perspective. Journal of Intellectual Disability Research 46:168–178, 2002
7.
Backer C, Chapman M, Mitchell D: Access to secondary healthcare for people with intellectual disabilities: a review of the literature. Journal of Applied Research in Intellectual Disabilities 22:514–525, 2009
8.
Iacono T, Bigby C, Unsworth C, et al: A systematic review of hospital experiences of people with intellectual disability. BMC Health Services Research 14:505, 2014
9.
Kelly CL, Thomson K, Wagner AP, et al: Investigating the widely held belief that men and women with learning disabilities receive poor quality healthcare when admitted to hospital: a single-site study of 30-day readmission rates. Journal of Intellectual Disability Research 59:835–844, 2015
10.
Kangovi S, Grande D: Hospital readmissions—not just a measure of quality. JAMA 306:1796–1797, 2011
11.
Rogers AT, Bai G, Lavin RA, et al: Higher hospital spending on occupational therapy is associated with lower readmission rates. Medical Care Research and Review (Epub ahead of print, Sept 2, 2016)
12.
Chartier K, Feldman M: Outcome evaluation of a specialized treatment home for adults with dual diagnosis and challenging behaviour. Journal on Developmental Disabilities 21:69–82, 2015
13.
Lunsky Y, Balogh R, Cairney J: Predictors of emergency department visits by persons with intellectual disability experiencing a psychiatric crisis. Psychiatric Services 63:287–290, 2012
14.
van Walraven C, Jennings A, Forster AJ: A meta-analysis of hospital 30-day avoidable readmission rates. Journal of Evaluation in Clinical Practice 18:1211–1218, 2012
15.
Mason J, Scior K: ‘Diagnostic overshadowing’ amongst clinicians working with people with intellectual disabilities in the UK. Journal of Applied Research in Intellectual Disabilities 17:85–90, 2004
Information & Authors
Information
Published In
Cover: Flying Geese, by Felix Bracquemond, 19th century. Black chalk, brush, and watercolor, highlighted with white gouache. Museum purchase, Davis Museum, Welleseley College. Photo credit: Davis Museum/Art Resource, New York City.
History
Received: 22 November 2016
Revision received: 19 April 2017
Revision received: 11 July 2017
Accepted: 25 August 2017
Published online: 15 November 2017
Published in print: March 01, 2018
Keywords
Authors
Competing Interests
The authors report no financial relationships with commercial interests.
Funding Information
Ministry of Health and Long-Term Care (MOHLTC) Health System Research Fund (HSRF)
This study was supported by the Province of Ontario through its research grants program, the Canadian Institutes of Health Research (CIHR) (PHE 103973), and the Institute for Clinical Evaluative Sciences (ICES), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC). The Johns Hopkins ACG system, version 7, was used for this research.Metrics & Citations
Metrics
Citations
Export Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
For more information or tips please see 'Downloading to a citation manager' in the Help menu.
View Options
View options
PDF/EPUB
View PDF/EPUBLogin options
Already a subscriber? Access your subscription through your login credentials or your institution for full access to this article.
Personal login Institutional Login Open Athens loginNot a subscriber?
PsychiatryOnline subscription options offer access to the DSM-5-TR® library, books, journals, CME, and patient resources. This all-in-one virtual library provides psychiatrists and mental health professionals with key resources for diagnosis, treatment, research, and professional development.
Need more help? PsychiatryOnline Customer Service may be reached by emailing [email protected] or by calling 800-368-5777 (in the U.S.) or 703-907-7322 (outside the U.S.).