Characteristics and predictors of readmission to a medical short-stay unit.
Short-stay units (SSUs) are medical units with anticipated lengths of stay (LOS) of 2472 hours. Increasing interest exists among hospitals for creating these units, as the focus is typically on treatment of mild conditions by dedicated staff incorporating a patient-centered model of care. Often SSUs can serve as an alternative to admission to a medical-surgical unit. Use of SSUs can decrease hospital costs as fewer resources are mobilized to manage less-severe conditions (Downing, Scott, & Kelly, 2008).SSUs typically differ from observational units, which are extensions of a hospital's Emergency Department (ED) used to accommodate patients needing a period of observation before determining if hospital admission is required. In contrast, SSUs function as separate medical units where patients are admitted for anticipated short stays of 3 days or less. SSUs generally are managed by a hospitalist and dedicated nursing staff. They have 14-24 beds (Abenhaim, Kahn, Raffoul, & Becker, 2000; Downing et al., 2008). Common characteristics of SSUs include level of care no higher than intermediate; narrow range of diagnoses, with limited or manageable co-morbidities; (no anticipated transfer to a traditional inpatient unit; no complex discharge needs; and no need for advanced ancillary services, such as wound care, bedside procedures, or extensive management (Abenhaim et al., 2000; Lucas et al., 2009). Most patients are admitted through the ED after initial diagnosis is made and urgent tests completed.
Significance of Research
Despite some early benefits described with integration of SSUs, little research has investigated factors associated with 30-day readmission rates after patients have been discharged from these units, and if these factors remain stable over time. Readmission rates have become a key outcome due to changes in reimbursement, particularly for diagnosis types commonly seen in SSUs (Centers for Medicare & Medicaid Services, n.d.). Information on predictors of readmission is important for medical-surgical nurses as they align discharge educational information and needed services to optimize patient outcomes and keep readmission rates low.
Purpose of Study
Purposes of this study were to compare characteristics and outcomes of patients admitted to an SSU over two periods (immediately after SSU initiation and 12 months later), and identify predictors of 30-day readmission after SSU discharge.
Review of the Literature
A literature review was conducted to frame existing knowledge regarding SSUs and identify gaps that could be addressed with nurse-led research. Databases included Medline, Pubmed, and CINAHL, and the following search terms were used: short-stay unit, observation unit, and 30-day readmission. Studies published 2000-2014 were included. The majority of articles on SSUs were published in the early 2000s, reflecting the introduction of this type of unit. More recently, a meta-analysis of outcomes from these studies was published that also highlighted areas of additional research needed to demonstrate the impact of these units on patient and organizational outcomes (Damiani et al., 2011). A summary of the early studies describing characteristics of SSU units is provided, followed by the more recent meta-analyses and areas for future research that support the need for the current study.
Characteristics of SSUs
The premise behind SSUs was to provide focused care by a specific, dedicated staff with the goal of decreasing readmissions and cost of care, easing overcrowding in EDs, reducing inappropriate admissions, and shortening LOS (Abenhaim et al., 2000; Daly, Campbell, & Cameron, 2003). Data from early observational studies indicated the typical diagnoses for patients admitted to SSUs included asthma, chronic obstructive pulmonary disease (COPD), pneumonia, heart failure (HF), acute coronary syndrome, urinary tract infection, and cellulitis. The average length of stay among patients in these studies was 2-3 days (Abenhaim et al., 2000; Downing et al., 2008; Lucas et al., 2009).
Outcomes of SSUs
Several studies compared outcomes of patients admitted to SSUs to patients admitted to traditional medicine units. Findings from observational studies using retrospective and prospective approaches reported shorter LOS, fewer hospital complications, and lower 30day readmission rates for SSUs compared to traditional medicine units, possibly due to lower patient acuity on SSUs (Abenhaim et al., 2000; Daly et al., 2003; Downing et al., 2008).
A prospective, observational cohort study by Lucas and colleagues (2009) examined predictors of LOS and eventual admission to traditional inpatient services among 738 patients admitted to SSUs. Within the study sample, 79% (n=582) were discharged successfully within 72 hours. Among patients with LOS greater than 72 hours, the need for additional diagnostic tests was the strongest predictor of LOS, while transfer to traditional inpatient services as a secondary outcome was influenced by the need for specialty consultations or diagnostic tests.
More recently, Yong and colleagues (2011) investigated factors associated with successful selection of appropriate patients for SSU admission. Successful selection was determined by the percentage of patients discharged within 72 hours. Over a 4-year period, authors reported successful SSU stays under 72 hours ranging from 72% to 77% of patients (n=6453-6570). Factors contributing to prolonged stays included older patients with complex comorbidities, weekend admissions, and transfers to an intensive care unit. While 30-day readmission rates were not examined in this study, authors reported a 7-day readmission rate of 3% during the study period.
A meta-analysis of outcomes of SSUs was conducted using six studies involving 21,264 patients (Damiani et al., 2011). Studies included in this analysis compared outcomes among patients admitted to SSU to those of patients admitted to traditional inpatient units. Findings indicated insufficient data to calculate pooled estimates comparing mortality and readmission rates between the two types of units. When investigating outcomes from each study separately, authors found no statistically significant differences between units for mortality or readmission rates. Data on 30-day readmission rates among SSUs are important, but no studies have investigated factors predicting 30-day readmission or if these factors remain stable over time. The current study was conducted to address this gap in the literature.
Methodology
Design, Setting, Sample
This double cohort observational retrospective study was reviewed and approved by the hospital Institutional Review Board. The study site, a 641-bed public academic medical center, serves as the safety-net hospital for a county in northeastern Ohio. The 18-bed SSU is staffed by hospitalists, two senior residents, registered nurses, customer care partners, a case manager, and a social worker. The unit has an average of 8-10 admissions and 8-10 discharges daily.
Data were gathered on characteristics and outcomes of patients admitted to the SSU over two time periods: October-November 2011 (cohort A, soon after unit opening) and July-September 2012 (cohort B, approximately 12 months after unit opening). The study used a convenience sampling plan, which included all patients age 18 and older admitted to the SSU during the specified periods. Data were abstracted from medical records of eligible patients using a standardized data collection tool. Patient variables included gender identity, age, ethnicity, type of insurance, admitting diagnosis, comorbidities, discharge disposition, previous ED visit or hospitalization within the last 30 days, admitting location, LOS, and specialty consults. The primary outcome variable was hospital readmission within 30 days of SSU discharge.
Data Analysis
Data were analyzed using the Statistical Package for the Social Sciences (SPSS) software, version 22.0 (IBM Corp.; Armonk, NY). Descriptive statistics were calculated for clinical characteristics and discharge data. Bivariate analyses using chi-square statistics were performed to determine relationships of study variables to hospital readmission within 30 days. Multivariate logistic regression models were created separately using data from cohort A and cohort B to determine predictors of 30-day hospital readmission.
Results
Data were gathered on 276 patients in cohort A and 281 patients in cohort B (see Table 1). The mean age for both cohorts was 51. Bivariate chi-square tests were performed to determine significant differences for 30-day readmission based on patient age (>65, <65), gender identity, ED visit prior to SSU admission, and hospitalization prior to SSU admission. For cohort A, occurrences of a previous ED visit or previous hospitalization were related statistically to 30-day readmission after SSU discharge [ED visit: [chi square] = (1, N=270) = 10.61, /x0.01; previous hospitalization: [chi square] = (1, N=270) = 11.41, p<0.01)]. Patient age and gender identity were not significant for 30-day readmission in cohort A. For cohort B, chi-square analyses were performed with the same variable set. No variables were significant in these analyses; however, the relationship between previous ED visit and 30day readmission after SSU discharge approached statistical significance ([chi square] = (1, N=272) = 3.66, p=0.05).
Multivariate logistic regression analyses then were conducted for each cohort (see Table 2). Variables predictive of 30-day readmission for cohort A were private insurance and hospitalization prior to SSU admission. Patients with private insurance had a lower risk of hospital readmission (p=0.01, OR=0.27, 95% 0=0.099-0.771), and those who experienced a hospitalization prior to SSU admission were five times more likely to be readmitted 30 days after SSU discharge (pc0.001, OR=5.15, 95% 0=2.118-12.538). For cohort B, patients who had an ED visit in the previous 30 days were twice as likely to be readmitted after SSU discharge (p=0.039, OR=2.014, 95% 0=1.037-3.912).
Discussion of Findings
Findings contribute information on patients most suitable for SSU admission; specifically, additional consideration should be given to admission of patients with hospitalization or ED visit within the last 30 days because this could influence 30day rehospitalization after SSU discharge. In this study, LOS remained comparable between the two cohorts, and most patients still were discharged home. LOS for the SSU was similar to other reports at 2-3 days (Abenhaim et al., 2000; Juan et al., 2006; Martin-Sanchez et al., 2013; Sempere-Montes, Morales-Suarez-Varela, Garijo-Gomez, Illa-Gomez, & Palau-Munoz, 2010; Yong et al., 2011).
Data from both cohorts suggested type of insurance and previous ED visit or hospitalization were predictive of 30-day readmission. In both cohorts, patients who experienced previous hospitalization or ED visit were more likely to experience 30-day readmission after SSU discharge. A decrease in 30-day readmission rates was seen in cohort B, after the SSU had been open for 12 months. Readmission rates in other studies evaluating SSUs included 7-day readmission rates of 4.9%-7%, 30-day rates of 6.1%-9.6%, and 8-week readmission rates of 9% (Abenhaim et al., 2000; Arendts, MacKenzie, & Lee, 2006; Downing et al., 2008; Juan et al., 2006; Martin-Sanchez et al., 2013; Yong et al., 2011). The study institution is the county safety net hospital, and readmission rates in this study were similar to those reported by other safety net hospitals (20%-25%) (Ross et al., 2012).
The number of patients readmitted to the hospital within 30 days of SSU discharge with COPD, HF, diabetes, and pneumonia were significantly lower than readmission rates for patients with the same diagnoses discharged from the traditional inpatient medical unit for both study cohorts. Patients with asthma or cellulitis in cohort A had a higher number of readmissions than those discharged from the traditional medical unit. Cohort A had 10 readmissions for asthma and 14 for cellulitis, compared to four and five respectively for the traditional medicine unit. However, for cohort B, the number of readmissions for all diagnoses (asthma, cellulitis, COPD, HF, diabetes, and pneumonia) were lower for patients treated in the SSU compared to those receiving care on the traditional medical unit. SSUs typically have lower readmission rates than traditional medical units, often due to decreased acuity and comorbidities of patients selected for SSU admission (Abenhaim et al., 2000). The fact that readmission rates for all diagnosis types were lower in cohort B (after the SSU had been open for a year) suggested better selection of suitable patients for SSU admission as well as improved discharge education and coordination by nursing staff.
Limitations
Limitations included the retrospective design and data collection at one study site, hindering generalizability of findings. The study also was based on research published after SSUs were introduced in health care, much of it outside the 3-5 year period recommended for current literature reviews. However, findings from an earlier meta-analysis on outcomes associated with SSU highlighted the need for additional research to determine the effectiveness of these units as well as factors that could impact 30-day readmission rates (Daly et al., 2003). The question of 30-day readmission rates is a relatively recent concern in care delivery due to reimbursement changes. Data from the current study can be used as a guide when determining patients who will benefit from care in an SSU and when identifying discharge needs, with the overall goal of keeping readmission rates low.
Implications for Nursing
Information on factors that influence readmission rates is important for medical-surgical nurses, particularly for those working on SSUs as these units focus on streamlining care and discharging patients quickly. Discharge education provided by nurses must be clear and concise, and include information to help prevent 30-day readmissions. When preparing discharge information for patients in an SSU, nurses should note if patients have experienced recent ED visits or hospitalization as this may predispose them to additional subsequent hospitalization and adversely affect 30-day readmission rates for the unit and organization. Previous ED visits and hospitalization may have resulted from patients' knowledge deficits regarding medications or treatment adherence. Nurses are in a key position to address these points directly and reduce the likelihood of readmissions.
Recommendations for Future Research
Future research should focus on identifying optimal discharge interventions to keep 30-day readmission rates low among patients in SSUs. Use of nurse-led transitional care teams or structured discharge information may be worthy of future investigation. Evaluation of the effect of these programs on readmission rates will be integral to continued efforts to generate cost-effective quality care.
Conclusion
SSUs continue to be an alternative to traditional inpatient services. Findings from this study highlight possible contributors to increased 30-day readmission rates among patients in these settings. Nurses can use this information when educating patients and aligning follow up after discharge to improve patient and organizational outcomes. CEU
REFERENCES
Abenhaim, H.A., Kahn, S.R., Raffoul, J., & Becker, M.R. (2000). Program description: A hospitalist-run, medical short-stay unit in a teaching hospital. Canadian Medical Association Journal, 763(11), 1477-1480.
Arendts, G., Mackenzie, J., & Lee, J.K. (2006). Discharge planning and patient satisfaction in an emergency short-stay unit. Emergency Medicine Australasia, 18(1), 7-14.
Centers for Medicare & Medicaid Services, (n.d.). Hospital readmissions reduction program. Retrieved from https://www.medicare.gov/hospitalcompare/readmission- reduction-program.html
Daly, S., Campbell, D.A., & Cameron, P.A. (2003). Short-stay units and observation medicine: A systematic review. Medical Journal of Australia, 178(11), 559-563.
Damiani, G., Pinnarelli, L, Sommella, L, Vena, V., Magrini, R, & Ricciardi, W. (2011). The short stay unit as a new option for hospitals: A review of the scientific literature. Med Science Monitor, 77(6), S15-19.
Downing, H., Scott, C., & Kelly, C. (2008). Evaluation of a dedicated short-stay unit for acute medical admissions. Clinical Medicine, 8(1), 18-20.
Juan, A., Salazar, A., Alvarez, A., Perez, J.R., Garcia, L, & Corbella, X. (2006). Effectiveness and safety of an emergency department short-stay unit as an alternative to standard inpatient hospitalization. Emergency Medicine Journal, 23(11), 833-837. doi:10.1136/emj.2005. 033647
Lucas, B.P., Kumapley, R., Mba, B., Nisar, I., Lee K., Ofori-Ntow S..... Bienias, J.L. (2009). A hospitalist-run short stay unit: Features that predict length of stay and eventual admission to traditional Inpatient services. Journal of Hospital Medicine, 4(5), 276-284. doi:10.1002/jhm.386
Martin-Sanchez, F.J., Carbajosa, V., Llorens, R, Herrero, R, Jacob, J., Perez-Dura, M.J., ... Miro O. (2013). Prolonged hospitalization in patients admitted for acute heart failure in the short stay unit: Study of associated factors. Medicina Clinica (Bare), 743(6), 245-251. doi:10.1016/j.meddi.2013.06.028.
Ross, J.S., Bernheim, S.M., Zhenqiu, L, Drye, E.E., Chen, J., Normand, S.L., & Krumholz, H. (2012). Mortality and readmission at safety net and non-safety net hospitals for three common medical conditions. Health Affairs, 37(8), 1739-1748. doi: 10.1377/hlthaff.2011.1028
Sempere-Montes, G., Morales-Suarez-Varela, M., Garijo-Gomez, E., Illa-Gomez, M.D., & Palau-Munoz, P. (2010). Impact of a short stay unit in a tertiary hospital. Revista Clinica Espanola, 210(6), 279283. doi:10.1016/j.rae.2009.11.016
Yong, T.Y., LI, J.Y., Roberts, S., Hakendorf, R, Ben-Tovim, D.I., & Thompson, C.H. (2011). The selection of acute medical admissions for a short-stay unit. Internal and Emergency Medicine, 6(4), 321327. doi: 10.1007/s11739-010-0490-6
Cheryl Bradas, MSN, RN, is Geriatric Clinical Nurse Specialist, Specialty Care, The MetroHealth System, Cleveland, OH.
Wendy Sarver, PhD(c), MSN, RN, is Senior Nurse Researcher, The MetroHealth System, Cleveland, OH.
Katie Carney, MPA, BSN, RN, NEA-BC, is Associate Chief Nursing Officer, Ambulatory Nursing Operations, Integration and Specialty Care, The MetroHealth System, Cleveland, OH.
Johnbuck Creamer, MD, is Medical Director, Short Stay Unit, The MetroHealth System, Cleveland, OH; and Assistant Professor of Medicine, Case Western Reserve University, Cleveland, OH.
Alex Velotta, BSN, RN, is Clinical Nurse, Medical Short Stay Unit, The MetroHealth System, Cleveland, OH.
Sheila Byrnes, RN, is Clinical Nurse, Medical Short Stay Unit, The MetroHealth System, Cleveland, OH.
Molly McNett, PhD, RN, CNRN, is Director, Nursing Research, The MetroHealth System, Cleveland, OH.
TABLE 1.
Demographic Characteristics of Sample
Cohort A Cohort B
Variable (n = 275) (n = 281)
Gender Identity
Male 50.9% (140) 42% (118)
Female 49.1% (135) 58% (163)
Ethnicity
Caucasian 58.3% (161) 50.4% (142)
African American 32.6% (90) 33% (93)
Flispanic 6.2% (17) 8.2% (23)
Asian 1.1% (3) 0.7% (2)
Primary Diagnosis
HF 4% (11) 2.1% (6)
Pneumonia 4.7% (13) 4.3% (12)
DM 1.1% (3) 1.4% (4)
Asthma 12.3% (34) 5.3% (15)
Gl 14.9% (41) 10.6% (30)
Cellulitis 16.3% (45) 11.3% (32)
COPD 15.6% (43) 10.6% (30)
Angina 12% (33) 2.8% (8)
Other 18.9% (52) 51.2% (144)
Previous Hospitalization: 30 Days
Yes 25.8% (70) 9.7% (27)
No 74.2% (201) 90.3% (252)
Previous ED Visit: 30 Days
Yes 24.7% (68) 26.2% (72)
No 75.3% (207) 73.8% (203)
Specialty Consultations
Yes 66.7% (182) 68.2% (189)
No 33.3% (91) 31.8% (88)
Specialty Types
Dietary 9.8% (27) 7.8% (22)
PT 12% (33) 13.8% (39)
Cardiology 7.6% (21) 6% (17)
Neurology 2.9% (8) 6.7% (19)
Surgery 8.3% (23) 9.9% (28)
SW 31.9% (88) 16.3% (46)
CM 13% (36) 12.8% (36)
Gl 8.3% (23) 7.1% (20)
Other 3.3% (9) 4.3% (15)
30-Day Hospital Readmission
Yes 25.8% (70) 17.3% (48)
No 74.2% (201) 82.7% (229)
CM = case management; COPD = chronic obstructive pulmonary disease;
DM = diabetes mellitus; GI = gastrointestinal; HF = heart failure;
PT = physical therapy; SW = social work.
TABLE 2.
Results of Multiple Regression Analyses
Test
Parameter Estimate SE Statistic
Cohort A
Private insurance -1.2890 0.5251 6.0253
Previous 30-day hospitalization 1.6395 0.4537 13.0604
Cohort B
Previous 30-day ED visit 0.7003 0.3387 4.2747
Parameter p-Value Estimate
Cohort A
Private insurance 0.0141 0.276
Previous 30-day hospitalization 0.0003 5.153
Cohort B
Previous 30-day ED visit 0.039 2.014
95%
Parameter Confidence Limits
Cohort A
Private insurance 0.098 0.771
Previous 30-day hospitalization 2.118 12.438
Cohort B
Previous 30-day ED visit 1.037 3.912
 Printer friendly  Cite/link  Email  Feedback |
|
| Title Annotation: | Research for Practice |
|---|---|
| Author: | Bradas, Cheryl; Sarver, Wendy; Carney, Katie; Creamer, Johnbuck; Velotta, Alex; Byrnes, Sheila; McNe |
| Publication: | MedSurg Nursing |
| Article Type: | Report |
| Date: | Nov 1, 2016 |
| Words: | 3215 |
| Previous Article: | Hyperglycemia in the hospital setting. |
| Next Article: | Reducing 30-day readmissions for chronic obstructive pulmonary disease. |
| Topics: | |