- Email: [email protected]
Why do ischemic stroke and transient ischemic attack patients get readmitted?
Journal of the Neurological Sciences 307 (2011) 50–54
Contents lists available at ScienceDirect
Journal of the Neurological Sciences j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j n s
Why do ischemic stroke and transient ischemic attack patients get readmitted?☆ Pratik Bhattacharya ⁎, Deependra Khanal, Ramesh Madhavan, Seemant Chaturvedi Department of Neurology, Wayne State University/Detroit Medical Center, Detroit MI, USA
a r t i c l e
i n f o
Article history: Received 25 March 2011 Accepted 17 May 2011 Available online 1 June 2011 Keywords: Readmission Acute stroke Healthcare quality
a b s t r a c t Objective: Readmission is an important indicator for the quality of healthcare services. The authors examined the reasons for 30-day readmission among urban stroke patients, and their clinical consequences. Methods: Consecutive patients admitted to a JCAHO certified primary stroke center with ischemic stroke or transient ischemic attacks (TIA) were included. Demographics, TOAST mechanism, risk factors, treatments administered and discharge destination were collected. Charts were reviewed for readmissions up to 30 days from discharge. Reasons for readmission and outcomes in terms of disability and discharge destination were determined. Results: Two hundred sixty-five patients (50.9% male; 79.6%African American; mean age 60.9 years) were included. There were 205(77.4%) strokes and 60(22.6%) TIAs. Thirteen (5%) patients died during their first admission. Of the remaining 252 patients, 25 (9.9%) were readmitted within 30 days. The reason for readmission was neurological in 8/25 patients (32%; 3 ischemic strokes, 1 hemorrhagic stroke and 4 TIAs); and non-neurological in 17/25 patients (68%). The frequent non-neurological reasons were infections (6/25), electrolyte disturbances (3/25) and trauma related to falls (2/25). Patients with coronary artery disease were more likely to be readmitted (45.5% vs. 14.7%; p = 0.001) An NIH stroke scale ≥ 10 predicted readmission (50.0% vs. 25.4% for NIHSS b 10; p value 0.02). Patients discharged home or to acute rehabilitation units were less likely to be readmitted than those discharged to subacute rehabilitation units or nursing homes (8.2% vs. 23.8%; p value = 0.01). Interpretation: Disabling strokes are more likely to be readmitted. The reason is often non-neurological, and sometimes preventable. Physicians should review cases that return within 30 days and determine best practices that prevent readmission. © 2011 Elsevier B.V. All rights reserved.
1. Introduction Readmission after hospital discharge is being accepted as an important indicator for the quality of healthcare services. Readmissions may indicate unresolved issues during the initial admission, or poor resources allocated for post-hospital care. The shorter the time between initial hospitalization and readmission, the more likely that the initial stay played a significant role in the rehospitalization [1]. It is estimated that about 90% of readmissions are unplanned, and the annual costs associated with unplanned readmission is $17.4 billion [2]. In order to limit preventable readmissions, the Center for Medicare and Medicaid services proposes to increase accountability of hospitals by making adjustments in hospital reimbursements based on their 30 day readmission rates, and making these rates public [3].
☆ Acknowledgement statement (including conflict of interest and funding sources): The authors have no disclosures relevant to the manuscript. ⁎ Corresponding author at: Department of Neurology, Wayne State University/Detroit Medical Center, 4201 St Antoine, UHC 8C, Detroit, MI-48201, USA. Tel.: +1 313 745 1540; fax: +1 313 577 4641. E-mail address: [email protected] (P. Bhattacharya). 0022-510X/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2011.05.022
Chronic debilitating medical conditions have the highest rates of 30 day readmission (e.g. congestive heart failure and chronic obstructive pulmonary disease) [2]. Stroke is a leading cause of disability and healthcare expenditure in the United States [4]. Patients with strokes, who get readmitted within 30 days are at a higher risk of mortality in the long term and incur greater healthcare costs [5]. Several population based studies from administrative databases have identified different risk factors for readmission in selected populations (e.g. Medicare beneficiaries in the US) [6]. However on account of different methodologies and variables tested in the different studies, there is no single consensus model to predict readmission risk [6]. Moreover, most databases study extended periods of time, and few report 30 day readmission rates [5, 7–10] and predictors [5, 8, 10]. Depending on the research methods, predictors of readmission at 30 days include older age, black race, discharge destination, comorbidities such as prior stroke, heart disease, peripheral vascular disease, fluid and electrolyte imbalance, anemia, dementia, PEG tube placement [5], stroke severity, need for nursing care, discharge planning, length of stay [8] and treatment by specialists during initial admission [10]. When compared to large administrative databases, hospital based record reviews help us identify reasons for readmission at an individual patient level, and determine on a case by case basis, if care during the
P. Bhattacharya et al. / Journal of the Neurological Sciences 307 (2011) 50–54
initial stay could have been altered in order to avoid the readmission. In this study, our objective is to explore the reasons why acute ischemic stroke and transient ischemic attack (TIA) patients get readmitted within 30 days after discharge. Further, we have evaluated baseline characteristics and clinical features to predict 30 day readmission among ischemic stroke patients.
51
Table 1 Baseline demographics and stroke risk factor frequencies in the study population and their influence on readmissions among the strokes. Baseline characteristics
All TIAs (n = 60)
All strokes (n = 205)
Ischemic strokes discharged alive (n = 192) Readmitted (n = 22)
Not readmitted (n = 170)
P value
57.3 (15.6) 61.9 (14.3) 64.5 (11.8) 24 (40.0) 111 (54.2) 14 (63.6)
61.0 (14.2) 91 (53.5)
0.27 0.38
2. Methods The study was approved by the local Institutional Review Board and consent was waived for record review. This was a retrospective study design. Consecutive patients admitted to a single urban hospital with a primary diagnosis of acute ischemic stroke or TIA, over a period of 9 months (November 2009 through July 2010) were included in the study. Patients were selected from admission logs to the department of Neurology. Subjects who left the hospital against medical advice without completion of evaluation, and subjects determined to have non cerebral vascular etiologies such as multiple sclerosis or conversion disorder were excluded. The study location was a 243 bed hospital, equipped with a neurocritical care unit staffed by neurologists. It was a Joint Commission certified primary stroke center with a large volume of stroke patients brought in by emergency services as well as through referral from other hospitals. Electronic charts for all patients were reviewed. Demographics, insurance information (Medicare/Medicaid, private insurers or uninsured), presenting National Institute of Health Stroke scale (NIHSS), risk factors and comorbidities were abstracted. In most cases, the NIHSS was abstracted from the admission note; in a few cases, when presenting NIHSS was not documented, it was determined based on the detailed neurological examination documented in the admission note. Treatments given in the hospital and discharge medications were recorded. The investigations were reviewed to determine TOAST mechanisms of each stroke/TIA [11]. Modified Rankin scores at discharge were recorded based on assessments by physical and occupational therapists and rehabilitation physicians. Discharge destination (home, acute rehabilitation hospital, subacute rehabilitation, nursing home or transfer to another hospital) was also noted. Electronic charts of patients who survived the first admission were followed for 30 days after their discharge. The Electronic Medical records include reports from 10 hospitals in the same metropolitan region under the same Healthcare System. For subjects who were readmitted; reason for readmission (neurological vs. non neurological), treatment offered for readmission if stroke, and outcome of readmission in terms of discharge destination were noted. Finally, for subjects readmitted with a recurrent stroke or TIA, we determined whether the events could have been prevented. The main outcome of interest was readmission within 30 days of discharge. Frequencies of the different reasons for readmission were calculated. As the majority of the readmissions were among subjects with radiologically proven strokes, analysis for predictors of admission in 30 days was restricted to this group. Univariate analyses were carried out using the Pearson's Chi Square test for categorical variables and t test for continuous variables. Finally, multivariate logistic regression with backward selection of variables was conducted to determine the strongest predictors of readmission. All analyses were carried out using SAS 9.2 (Cary, NC) statistical software. 3. Results Two hundred sixty-five subjects with ischemic strokes (205; 77.4%) and TIAs (60; 22.6%) were included in the study. The demographic features, insurance information and risk factors are described in Table 1. TOAST mechanisms for stroke, treatments rendered and outcomes in terms of discharge destination are described in Table 2. Twenty five (9.4%) patients were readmitted within 30 days after discharge. Thirteen subjects (5.0% of all subjects; 6.3% of all strokes) died
Mean Age (SD) Male sex Race: African Americans White Hispanic Others
50 (83.3) 7 (11.7) 1 (1.7) 2 (3.3)
161 (78.5) 35 (13.2) 2 (1.0) 7 (3.4)
17 (77.28) 3 (13.64) 0 2 (9.1)
132 (77.65) 31 (18.2) 2 (1.2) 5 (2.9)
0.47
Insurance Medicare/Medicaid Private insurers uninsured
24 (40.7) 21 (35.6) 14 (23.7)
104 (51.0) 66 (32.3) 34 (16.7)
13 (59.1) 4 (18.2) 5 (22.7)
80 (47.3) 60 (35.5) 29 (17.2)
0.29
27 (45.0) 5 (12.5) 44 (73.3) 3 (5.0) 7 (11.7)
90 (43.9) 12 (8.3) 178 (86.8) 29 (14.2) 23 (11.2)
8 2 19 1 5
(36.4) (12.5) (86.4) (4.6) (22.7)
78 (45.9) 10 (8.3) 146 (85.9) 24 (14.1) 15 (8.8)
0.40 0.58 0.95 0.24 0.05
11 (18.3)
38 (18.5)
10 (45.5)
25 (14.7)
0.001
33 (55.0) 21 (35.0) 17 (28.3) 2 (3.3)
140 (68.3) 51 (24.9) 70 (34.2) 64 (31.4)
116 (68.2) 40 (23.5) 55 (32.4) 43 (25.4)
0.67 0.40 0.42 0.02
Risk factors: Current smokers Cocaine⁎ Hypertension Atrial fibrillation Congestive heart failure Coronary artery disease Hyperlipidemia Previous stroke Diabetes NIHSS ≥ 10
16 7 9 11
(72.7) (31.8) (40.9) (50.0)
Numbers in parenthesis represent column percentages except when specified as Standard deviations. ⁎ Cocaine usage was tested in 185 out of 265 patients.
during the first admission. When restricted to those who were discharged alive, 25/252 (9.9%) were readmitted within 30 days after discharge. Within the subgroup of subjects with ischemic strokes (excluding TIAs); 22 out of 192 (11.5%) subjects discharged alive were readmitted in 30 days. 3.1. Reasons for readmission Reasons for readmission in 30 days were primarily non-neurological (17 of 25; 68%). The most common were infections (n = 6: 3 urinary tract infections, 1 chest infection, 1 spontaneous bacterial peritonitis, 1 abdominal wall cellulitis around PEG insertion site) followed by metabolic derangements (n = 3). Other causes were fall and fracture (n = 2), hematemesis (n = 1), deep vein thrombus (n = 1), hypotension (n = 1), sickle pain crisis (n = 1), subtherapeutic INR (n = 1), delusions due to Mitochondrial encephalomyopathy, lactic acidosis and stroke-like episodes (MELAS) (n = 1). Eight subjects were readmitted for recurrent stroke (n = 4) and TIAs (n = 4). In 3 out of 25 subjects (12%), the mRS at second discharge had worsened, i.e. worse functional disability. This includes one subject who died from complications of severe sepsis during the second admission. The eventual discharge destination changed for the worse (i.e. a person readmitted from a rehabilitation institution discharged to a nursing home and so on) in 3/25 (12%) subjects. 3.2. Readmissions among TIA patients Only 3/60 (5%) patients with a TIA diagnosis were readmitted within 30 days. The causes for readmission were recurrent TIA (n = 1;
52
P. Bhattacharya et al. / Journal of the Neurological Sciences 307 (2011) 50–54
Table 2 Stroke subtypes, treatments offered in the study population and their influence on readmissions among the strokes. Variables
All TIAs (n = 60)
All strokes Ischemic strokes discharged (n = 205) alive (n = 192) P Readmitted Not (n = 22) readmitted value (n = 170)
TOAST mechanisms: Large artery atherosclerosis Cardioembolic Small vessel disease Other known Other unknown Treatments offered iv tPA Intervention Aspirin Statin Discharge destination⁎ Deaths Home/acute rehabilitation Subacute rehab/nursing home
9 (15.0)
87 (42.4)
12 (54.6)
69 (40.6)
7 (11.7) 11 (18.3)
21 (10.2) 60 (29.3)
2 (9.1) 3 (13.6)
15 (8.8) 57 (33.5)
7 (11.7) 26 (43.3)
12 (5.9) 25 (12.2)
2 (9.1) 3 (13.6)
9 (5.3) 20 (11.8)
26 29 171 161
(12.7) (14.2) (83.4) (78.5)
0 0 53 (88.3) 32 (53.3)
0 13 (6.4) 59 (98.3) 146(71.6) 1 (1.7)
42 (20.6)
4 4 17 18
(18.2) (18.2) (77.3) (81.8)
19 22 146 136
(11.2) (12.9) (85.9) (80.0)
– 12(54.6)
– 134 (80.7)
10 (45.4)
32 (19.3)
0.47
0.35 0.50 0.34 0.84
0.01
Numbers in parenthesis represent column percentages. ⁎ Three strokes were transferred to other acute care hospitals for insurance reasons and discharge status of one stroke patient was unknown.
patient 8 in Table 3), ischemic stroke (n = 1; patient 3 in Table 3) and unrelated: sickle pain crisis (n = 1). Results of univariate analysis for predictors of readmission among subjects with radiologically proven strokes in 30 days are shown in Tables 1 and 2. While not reaching statistical significance; the readmitted group had more Medicare/Medicaid subjects (59.1% vs. 47.3%) and fewer privately insured subjects (18.2% vs. 35.5%) than the non-readmitted group. There was a significantly higher proportion of congestive heart failure (22.7% vs. 8.8%) and coronary artery disease (45.5% vs. 14.7%) among the readmitted group; although the reasons for readmission were non-cardiac. If the initial stroke was severe (NIHSS ≥ 10), the subjects had higher likelihood of being readmitted (20.4% versus 8.0% for NIHSS b 10; p = 0.02). Fewer subjects with mild disability (mRS 0–2) at the first discharge were readmitted than those with moderate-severe disability (mRS 3–5), however this was not statistically significant. (10% mRS 0–2 vs. 14.7% mRS 3–5; p value 0.34). Significantly greater proportion of patients discharged to subacute rehabilitation facility or a nursing home was readmitted than those discharged home or to acute rehabilitation facility (23.8% vs. 8.2%; p value 0.01). Finally, a multivariate model was created with all aforementioned significant variables from the univariate analysis with backward selection of variables to determine the most important predictors of readmission, Coronary artery disease was the only predictor of readmission in this analysis (n = 0.001). 3.3. Readmissions with strokes Clinical scenarios of the eight subjects who were readmitted with neurological symptoms are presented in Table 3. Upon reviewing the circumstances of readmission, patients 4 and 5 could potentially have been spared readmission by altering treatments offered during the initial admission (better blood pressure management for patient 4 and carotid endarterectomy for patient 5).
Table 3 Features of patients readmitted with neurological symptoms. No. Age/ Initial admission sex Readmission as ischemic stroke: 1 74/ History of atrial fibrillation, M prostate cancer; recent cocaine and heroin use with right frontal infarct 2 52/ History of moderate aortic M stenosis; right hemiparesis and global aphasia, received iv + ia tPA; worsened in hospital and got mechanical thrombectomy of M1 clot; discharged home on dual antiplatelet and statin 3 50/F History of breast cancer, hypercholesterolemia coronary artery disease; presented with transient left hand clumsiness; Left atrial thrombus on Echo; discharged on Lovenox anticoagulation
Readmission
Recurrent left sided weakness, non compliant with Coumadin, developed heroin withdrawal in hospital. Recurrence of symptoms within hours of discharge; got mechanical thrombectomy and stenting of M1
Presented with aphasia; left temporal lobe infarct.
Readmission as intracerebral hemorrhage: 4 88/ History of hypertension, diabetes Returned with worsening ataxia; left cerebellar hemorrhage; BP M hypercholesterolemia and CHF, 222/100. symptoms of dizziness and ataxia; left cerebellar infarct; severe basilar stenosis; discharged to subacute rehabilitation with dual antiplatelet Readmission as transient ischemic attacks: 5 65/ History of hypertension, M hypercholesterolemia; right hemiparesis and global aphasia; received iv tPA; found high grade left carotid stenosis and moderate size left frontal infarct 6 62/F History of hypertension, diabetes. Mild right hemiparesis: left putaminal lacunar stroke; discharged on aspirin and statin. 7
72/ M
8
55/F
Unable to get outpatient endarterectomy due to insurance reasons, presented with a mild Wernicke's aphasia
speech difficulty lasting for less than 12 hours, no new stroke on imaging; risk factors were controlled at the time of second admission returned with transient right hemisensory symptoms
History of; right hemiparesis, declined tPA, MRI showed multiple lacunar strokes within the left MCA territory; improved to NIHSS 0; discharged on aspirin, statin, nicotine patches Transient ataxia, blood pressure History of hypertension, and glucose were lower than hypercholesterolemia, diabetes. initial admission but not at goal. Transient left hemisensory symptoms; discharged home on aspirin and statin
4. Discussion In our hospital based series from a certified stroke center 11.5% of ischemic strokes discharged alive were readmitted within 30 days. The causes of readmission were primarily non-neurological. Severe strokes and patients discharged to nursing homes and subacute rehabilitation were more likely to be readmitted. Coronary artery disease was a strong predictor of readmission, although the reasons for readmission were non-cardiac. A similar hospital registry based study was recently reported with about 10% readmission rates at 30 days. The study reported infections to be the most frequent cause of readmission up to 1 year. The study was conducted in Taiwan where the national health insurance system provides universal and quality health care to the people of Taiwan at affordable cost [12]. In our study, we confirmed the finding that infection is the most common cause of readmission even during the
P. Bhattacharya et al. / Journal of the Neurological Sciences 307 (2011) 50–54
short term (30 days). Moreover, our sample was from an urban population with a moderate number of uninsured and underinsured patients. In our study, the most frequent cause of readmission was infection. This has been demonstrated in studies looking at administrative (Medicare fee for service) databases within the 30 day limit too [7, 10]. Experimental evidence indicates that stroke leads to suppression of immunity which leads to infections [13]. Urinary tract infection is a common complication of stroke, occurring in 17–22% of stroke patients [14,15]. Bladder dysfunction (incontinence and retention) compounded with functional and cognitive impairment; and the increased use of catheters predispose stroke patients to urine infections [16]. Reducing the use of indwelling catheters/short term use rather than long term use, use of condom catheters in patients without retention and education of healthcare staff both at hospitals and skilled facilities, about the proper use of catheters are all potential strategies to reduce urinary infections [17]. Chest infections are also common and affect up to a quarter of stroke patients [14, 18]. Not only do they involve increased costs of care but also increase the risk of mortality up to three fold [19]. Older age, more severe neurological impairment and oropharyngeal dysphagia are risk factors for chest infections [14, 18]. The first episode of pneumonia tends to occur in the first month after the stroke, and patients with oropharyngeal dysphagia may get recurrent pneumonias [20]. Dysphagia screening protocols significantly decrease pneumonia risk [21]. Other modifiable measures that potentially decrease the risk of pneumonia are early rehabilitation and mobilization, limiting sedation, influenza vaccination and pneumovax, oral hygiene, education of patient and caregiver, maintaining nutrition and avoiding gastric reflux particularly in the presence of a nasogastric tube or PEG tube. Patients discharged to nursing homes and subacute rehabilitation centers were readmitted more often than those discharged home or to an acute rehabilitation center. Most likely, this is because the patients discharged to the nursing homes had more disabling strokes. As in our hospital based study, severe strokes and increased functional disability have consistently been shown to predict readmission in administrative databases [8, 22]. Another plausible explanation is that caregivers at home and in acute care-rehabilitation settings are able to pick up problems earlier and can address issues before they become serious to require hospitalization. Some studies reflect this trend [5] whereas other studies have found contrary results, where patients discharged home were more likely to be readmitted than those in nursing homes [23]. The premise was that caregiver perception of problems may be different at home than among healthcare providers at nursing homes. We found opposite results probably due to our study design which followed patients up to 30 days instead of several months in the other studies. In one study, although readmissions were more frequent in the group discharged home at 2 years; this difference appeared at 6 months and beyond. At 30 days, readmissions after stroke were more common among the nursing home patients [23]. Coronary artery disease was a strong predictor of 30 day readmission in our dataset. Ischemic heart disease was associated with a 14% excess risk of readmission in one study [24]. Heart failure has also consistently been associated with 30 day readmission [5]. Interestingly, the reasons for readmission among coronary artery disease patients in our study were not cardiac. This coincides with a population based study on Medicare beneficiaries, where significant proportions of readmissions due to myocardial infarctions occurred after the first few months since discharge [7, 25, 26]. Finally, discharge planning, including good communication between the hospital and the community (family and primary care physicians) is an essential component to prevent readmissions [27]. Prior to discharge, a reliable link of communication between the physician and the patient/family should be set up, where they can
53
contact the physician with issues with the post discharge plan, medication use, home care et cetera. Home follow up visits and home physical therapy visits have shown to significantly reduce readmissions [28]. The use of telemedicine technology to facilitate post-discharge care is another possibility. The present study has certain limitations. First, our data represents a single center experience. Admissions to other region hospitals are unknown, so we may underestimate the true frequency of readmissions. However, our electronic records system reflects admissions to 10 different hospitals within the same healthcare system serving our metropolitan area, so, if patients chose to come to any of those hospitals in the system for readmission, they would have been included in our data. Second, the data only captures hospital admissions. Emergency room visits for problems that did not warrant admission were not captured in this data. Third, majority of the patients in our study were admitted by neurologists. It is possible that patients treated primarily by other services may have different rates of readmission, although this has not been shown in prior studies [10]. We did not collect data about post-discharge care, medication compliance and coordination of care between outpatient follow-up: all variables that influence readmission rates. We also do not have data of previous hospitalizations; a known predictor of readmissions. Finally, the data was collected in a retrospective manner where we relied on the physician provider definitions rather than standardized definitions for various complications. The main advantage of the hospital based design of our study is that it gave us the opportunity to review the charts of patients returning with neurological conditions. As a quality assessment exercise, we could evaluate whether there were any deviations from standards of care which could otherwise have spared a readmission. Such evaluations are not possible with administrative databases. Administrative databases collect data on large numbers of patients and are not limited to a single hospital. However, they have inherent problems such as reliance on ICD codes which are not always accurate. Also many studies rely on the Medicare database which includes subjects 65 years and older, whereas a significant proportion of patients admitted with stroke is younger than 65 (mean age in our data was 60.9 years). 5. Conclusion We found that 9.9% of patients with ischemic strokes and TIAs discharged alive were readmitted within 30 days. Readmission within 30 days incurs excess healthcare costs, which may potentially be preventable. The reasons for early readmission in urban stroke patients are largely non-neurological, and can conceivably be minimized by the measures outlined above. At a local level, physicians caring for stroke patients should review cases that return within 30 days to determine best practices that prevent readmission, as these may vary from hospital to hospital. Preventing readmissions within 30 days is likely to be an important indicator of quality of care in the future. References [1] Ashton CM, Wray NP. A conceptual framework for the study of early readmission as an indicator of quality of care. Soc Sci Med 1996;43:1533–41. [2] Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med 2009;360:1418–28. [3] Center for Medicare and Medicaid Services: http://www.cms.gov/ AcuteInpatientPPS/downloads/CMS-1390-F.pdf. (Accessed December 21, 2010. [4] Lloyd-Jones D, Adams RJ, Brown TM, et al. Heart disease and stroke statistics—2010 update: a report from the American Heart Association. Circulation 2010;121: e46–e215. [5] Kind AJ, Smith MA, Liou JI, Pandhi N, Frytak JR, Finch MD. The price of bouncing back: one-year mortality and payments for acute stroke patients with 30-day bounce-backs. J Am Geriatr Soc 2008;56:999–1005. [6] Lichtman JH, Leifheit-Limson EC, Jones SB, et al. Predictors of hospital readmission after stroke: a systematic review. Stroke 2010;41:2525–33. [7] Bravata DM, Ho SY, Meehan TP, Brass LM, Concato J. Readmission and death after hospitalization for acute ischemic stroke: 5-year follow-up in the medicare population. Stroke 2007;38:1899–904.
54
P. Bhattacharya et al. / Journal of the Neurological Sciences 307 (2011) 50–54
[8] Chuang KY, Wu SC, Ma AH, Chen YH, Wu CL. Identifying factors associated with hospital readmissions among stroke patients in Taipei. J Nurs Res 2005;13:117–28. [9] Johansen HL, Wielgosz AT, Nguyen K, Fry RN. Incidence, comorbidity, case fatality and readmission of hospitalized stroke patients in Canada. Can J Cardiol 2006;22: 65–71. [10] Smith MA, Liou JI, Frytak JR, Finch MD. 30-day survival and rehospitalization for stroke patients according to physician specialty. Cerebrovasc Dis 2006;22:21–6. [11] Adams Jr HP, Bendixen BH, Kappelle LJ, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in acute stroke treatment. Stroke 1993;24:35–41. [12] Lin HJ, Chang WL, Tseng MC. Readmission after stroke in a hospital-based registry: risk, etiologies, and risk factors. Neurology 2011;76:438–43. [13] Klehmet J, Harms H, Richter M, et al. Stroke-induced immunodepression and poststroke infections: lessons from the preventive antibacterial therapy in stroke trial. Neuroscience 2009;158:1184–93. [14] Aslanyan S, Weir CJ, Diener HC, Kaste M, Lees KR. Pneumonia and urinary tract infection after acute ischaemic stroke: a tertiary analysis of the GAIN International trial. Eur J Neurol 2004;11:49–53. [15] Langhorne P, Stott DJ, Robertson L, et al. Medical complications after stroke: a multicenter study. Stroke 2000;31:1223–9. [16] Kong KH, Young S. Incidence and outcome of poststroke urinary retention: a prospective study. Arch Phys Med Rehabil 2000;81:1464–7. [17] Poisson SN, Johnston SC, Josephson SA. Urinary tract infections complicating stroke: mechanisms, consequences, and possible solutions. Stroke 2010;41: e180–4. [18] Mann G, Hankey GJ, Cameron D. Swallowing function after stroke: prognosis and prognostic factors at 6 months. Stroke 1999;30:744–8.
[19] Katzan IL, Cebul RD, Husak SH, Dawson NV, Baker DW. The effect of pneumonia on mortality among patients hospitalized for acute stroke. Neurology 2003;60: 620–5. [20] Masiero S, Pierobon R, Previato C, Gomiero E. Pneumonia in stroke patients with oropharyngeal dysphagia: a six-month follow-up study. Neurol Sci 2008;29: 139–45. [21] Hinchey JA, Shephard T, Furie K, Smith D, Wang D, Tonn S. Formal dysphagia screening protocols prevent pneumonia. Stroke 2005;36:1972–6. [22] Bohannon RW, Lee N. Association of physical functioning with same-hospital readmission after stroke. Am J Phys Med Rehabil 2004;83:434–8. [23] Camberg LC, Smith NE, Beaudet M, Daley J, Cagan M, Thibault G. Discharge destination and repeat hospitalizations. Med Care 1997;35:756–67. [24] Kennedy BS. Does race predict stroke readmission? An analysis using the truncated negative binomial model. J Natl Med Assoc 2005;97:699–713. [25] Christensen MC, Munro V. Ischemic stroke and intracerebral hemorrhage: the latest evidence on mortality, readmissions and hospital costs from Scotland. Neuroepidemiology 2008;30:239–46. [26] Dhamoon MS, Sciacca RR, Rundek T, Sacco RL, Elkind MS. Recurrent stroke and cardiac risks after first ischemic stroke: the Northern Manhattan Study. Neurology 2006;66:641–6. [27] Naylor MD, Brooten D, Campbell R, et al. Comprehensive discharge planning and home follow-up of hospitalized elders: a randomized clinical trial. JAMA 1999;281:613–20. [28] Andersen HE, Schultz-Larsen K, Kreiner S, Forchhammer BH, Eriksen K, Brown A. Can readmission after stroke be prevented? Results of a randomized clinical study: a postdischarge follow-up service for stroke survivors. Stroke 2000;31:1038–45.
Contents lists available at ScienceDirect
Journal of the Neurological Sciences j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j n s
Why do ischemic stroke and transient ischemic attack patients get readmitted?☆ Pratik Bhattacharya ⁎, Deependra Khanal, Ramesh Madhavan, Seemant Chaturvedi Department of Neurology, Wayne State University/Detroit Medical Center, Detroit MI, USA
a r t i c l e
i n f o
Article history: Received 25 March 2011 Accepted 17 May 2011 Available online 1 June 2011 Keywords: Readmission Acute stroke Healthcare quality
a b s t r a c t Objective: Readmission is an important indicator for the quality of healthcare services. The authors examined the reasons for 30-day readmission among urban stroke patients, and their clinical consequences. Methods: Consecutive patients admitted to a JCAHO certified primary stroke center with ischemic stroke or transient ischemic attacks (TIA) were included. Demographics, TOAST mechanism, risk factors, treatments administered and discharge destination were collected. Charts were reviewed for readmissions up to 30 days from discharge. Reasons for readmission and outcomes in terms of disability and discharge destination were determined. Results: Two hundred sixty-five patients (50.9% male; 79.6%African American; mean age 60.9 years) were included. There were 205(77.4%) strokes and 60(22.6%) TIAs. Thirteen (5%) patients died during their first admission. Of the remaining 252 patients, 25 (9.9%) were readmitted within 30 days. The reason for readmission was neurological in 8/25 patients (32%; 3 ischemic strokes, 1 hemorrhagic stroke and 4 TIAs); and non-neurological in 17/25 patients (68%). The frequent non-neurological reasons were infections (6/25), electrolyte disturbances (3/25) and trauma related to falls (2/25). Patients with coronary artery disease were more likely to be readmitted (45.5% vs. 14.7%; p = 0.001) An NIH stroke scale ≥ 10 predicted readmission (50.0% vs. 25.4% for NIHSS b 10; p value 0.02). Patients discharged home or to acute rehabilitation units were less likely to be readmitted than those discharged to subacute rehabilitation units or nursing homes (8.2% vs. 23.8%; p value = 0.01). Interpretation: Disabling strokes are more likely to be readmitted. The reason is often non-neurological, and sometimes preventable. Physicians should review cases that return within 30 days and determine best practices that prevent readmission. © 2011 Elsevier B.V. All rights reserved.
1. Introduction Readmission after hospital discharge is being accepted as an important indicator for the quality of healthcare services. Readmissions may indicate unresolved issues during the initial admission, or poor resources allocated for post-hospital care. The shorter the time between initial hospitalization and readmission, the more likely that the initial stay played a significant role in the rehospitalization [1]. It is estimated that about 90% of readmissions are unplanned, and the annual costs associated with unplanned readmission is $17.4 billion [2]. In order to limit preventable readmissions, the Center for Medicare and Medicaid services proposes to increase accountability of hospitals by making adjustments in hospital reimbursements based on their 30 day readmission rates, and making these rates public [3].
☆ Acknowledgement statement (including conflict of interest and funding sources): The authors have no disclosures relevant to the manuscript. ⁎ Corresponding author at: Department of Neurology, Wayne State University/Detroit Medical Center, 4201 St Antoine, UHC 8C, Detroit, MI-48201, USA. Tel.: +1 313 745 1540; fax: +1 313 577 4641. E-mail address: [email protected] (P. Bhattacharya). 0022-510X/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2011.05.022
Chronic debilitating medical conditions have the highest rates of 30 day readmission (e.g. congestive heart failure and chronic obstructive pulmonary disease) [2]. Stroke is a leading cause of disability and healthcare expenditure in the United States [4]. Patients with strokes, who get readmitted within 30 days are at a higher risk of mortality in the long term and incur greater healthcare costs [5]. Several population based studies from administrative databases have identified different risk factors for readmission in selected populations (e.g. Medicare beneficiaries in the US) [6]. However on account of different methodologies and variables tested in the different studies, there is no single consensus model to predict readmission risk [6]. Moreover, most databases study extended periods of time, and few report 30 day readmission rates [5, 7–10] and predictors [5, 8, 10]. Depending on the research methods, predictors of readmission at 30 days include older age, black race, discharge destination, comorbidities such as prior stroke, heart disease, peripheral vascular disease, fluid and electrolyte imbalance, anemia, dementia, PEG tube placement [5], stroke severity, need for nursing care, discharge planning, length of stay [8] and treatment by specialists during initial admission [10]. When compared to large administrative databases, hospital based record reviews help us identify reasons for readmission at an individual patient level, and determine on a case by case basis, if care during the
P. Bhattacharya et al. / Journal of the Neurological Sciences 307 (2011) 50–54
initial stay could have been altered in order to avoid the readmission. In this study, our objective is to explore the reasons why acute ischemic stroke and transient ischemic attack (TIA) patients get readmitted within 30 days after discharge. Further, we have evaluated baseline characteristics and clinical features to predict 30 day readmission among ischemic stroke patients.
51
Table 1 Baseline demographics and stroke risk factor frequencies in the study population and their influence on readmissions among the strokes. Baseline characteristics
All TIAs (n = 60)
All strokes (n = 205)
Ischemic strokes discharged alive (n = 192) Readmitted (n = 22)
Not readmitted (n = 170)
P value
57.3 (15.6) 61.9 (14.3) 64.5 (11.8) 24 (40.0) 111 (54.2) 14 (63.6)
61.0 (14.2) 91 (53.5)
0.27 0.38
2. Methods The study was approved by the local Institutional Review Board and consent was waived for record review. This was a retrospective study design. Consecutive patients admitted to a single urban hospital with a primary diagnosis of acute ischemic stroke or TIA, over a period of 9 months (November 2009 through July 2010) were included in the study. Patients were selected from admission logs to the department of Neurology. Subjects who left the hospital against medical advice without completion of evaluation, and subjects determined to have non cerebral vascular etiologies such as multiple sclerosis or conversion disorder were excluded. The study location was a 243 bed hospital, equipped with a neurocritical care unit staffed by neurologists. It was a Joint Commission certified primary stroke center with a large volume of stroke patients brought in by emergency services as well as through referral from other hospitals. Electronic charts for all patients were reviewed. Demographics, insurance information (Medicare/Medicaid, private insurers or uninsured), presenting National Institute of Health Stroke scale (NIHSS), risk factors and comorbidities were abstracted. In most cases, the NIHSS was abstracted from the admission note; in a few cases, when presenting NIHSS was not documented, it was determined based on the detailed neurological examination documented in the admission note. Treatments given in the hospital and discharge medications were recorded. The investigations were reviewed to determine TOAST mechanisms of each stroke/TIA [11]. Modified Rankin scores at discharge were recorded based on assessments by physical and occupational therapists and rehabilitation physicians. Discharge destination (home, acute rehabilitation hospital, subacute rehabilitation, nursing home or transfer to another hospital) was also noted. Electronic charts of patients who survived the first admission were followed for 30 days after their discharge. The Electronic Medical records include reports from 10 hospitals in the same metropolitan region under the same Healthcare System. For subjects who were readmitted; reason for readmission (neurological vs. non neurological), treatment offered for readmission if stroke, and outcome of readmission in terms of discharge destination were noted. Finally, for subjects readmitted with a recurrent stroke or TIA, we determined whether the events could have been prevented. The main outcome of interest was readmission within 30 days of discharge. Frequencies of the different reasons for readmission were calculated. As the majority of the readmissions were among subjects with radiologically proven strokes, analysis for predictors of admission in 30 days was restricted to this group. Univariate analyses were carried out using the Pearson's Chi Square test for categorical variables and t test for continuous variables. Finally, multivariate logistic regression with backward selection of variables was conducted to determine the strongest predictors of readmission. All analyses were carried out using SAS 9.2 (Cary, NC) statistical software. 3. Results Two hundred sixty-five subjects with ischemic strokes (205; 77.4%) and TIAs (60; 22.6%) were included in the study. The demographic features, insurance information and risk factors are described in Table 1. TOAST mechanisms for stroke, treatments rendered and outcomes in terms of discharge destination are described in Table 2. Twenty five (9.4%) patients were readmitted within 30 days after discharge. Thirteen subjects (5.0% of all subjects; 6.3% of all strokes) died
Mean Age (SD) Male sex Race: African Americans White Hispanic Others
50 (83.3) 7 (11.7) 1 (1.7) 2 (3.3)
161 (78.5) 35 (13.2) 2 (1.0) 7 (3.4)
17 (77.28) 3 (13.64) 0 2 (9.1)
132 (77.65) 31 (18.2) 2 (1.2) 5 (2.9)
0.47
Insurance Medicare/Medicaid Private insurers uninsured
24 (40.7) 21 (35.6) 14 (23.7)
104 (51.0) 66 (32.3) 34 (16.7)
13 (59.1) 4 (18.2) 5 (22.7)
80 (47.3) 60 (35.5) 29 (17.2)
0.29
27 (45.0) 5 (12.5) 44 (73.3) 3 (5.0) 7 (11.7)
90 (43.9) 12 (8.3) 178 (86.8) 29 (14.2) 23 (11.2)
8 2 19 1 5
(36.4) (12.5) (86.4) (4.6) (22.7)
78 (45.9) 10 (8.3) 146 (85.9) 24 (14.1) 15 (8.8)
0.40 0.58 0.95 0.24 0.05
11 (18.3)
38 (18.5)
10 (45.5)
25 (14.7)
0.001
33 (55.0) 21 (35.0) 17 (28.3) 2 (3.3)
140 (68.3) 51 (24.9) 70 (34.2) 64 (31.4)
116 (68.2) 40 (23.5) 55 (32.4) 43 (25.4)
0.67 0.40 0.42 0.02
Risk factors: Current smokers Cocaine⁎ Hypertension Atrial fibrillation Congestive heart failure Coronary artery disease Hyperlipidemia Previous stroke Diabetes NIHSS ≥ 10
16 7 9 11
(72.7) (31.8) (40.9) (50.0)
Numbers in parenthesis represent column percentages except when specified as Standard deviations. ⁎ Cocaine usage was tested in 185 out of 265 patients.
during the first admission. When restricted to those who were discharged alive, 25/252 (9.9%) were readmitted within 30 days after discharge. Within the subgroup of subjects with ischemic strokes (excluding TIAs); 22 out of 192 (11.5%) subjects discharged alive were readmitted in 30 days. 3.1. Reasons for readmission Reasons for readmission in 30 days were primarily non-neurological (17 of 25; 68%). The most common were infections (n = 6: 3 urinary tract infections, 1 chest infection, 1 spontaneous bacterial peritonitis, 1 abdominal wall cellulitis around PEG insertion site) followed by metabolic derangements (n = 3). Other causes were fall and fracture (n = 2), hematemesis (n = 1), deep vein thrombus (n = 1), hypotension (n = 1), sickle pain crisis (n = 1), subtherapeutic INR (n = 1), delusions due to Mitochondrial encephalomyopathy, lactic acidosis and stroke-like episodes (MELAS) (n = 1). Eight subjects were readmitted for recurrent stroke (n = 4) and TIAs (n = 4). In 3 out of 25 subjects (12%), the mRS at second discharge had worsened, i.e. worse functional disability. This includes one subject who died from complications of severe sepsis during the second admission. The eventual discharge destination changed for the worse (i.e. a person readmitted from a rehabilitation institution discharged to a nursing home and so on) in 3/25 (12%) subjects. 3.2. Readmissions among TIA patients Only 3/60 (5%) patients with a TIA diagnosis were readmitted within 30 days. The causes for readmission were recurrent TIA (n = 1;
52
P. Bhattacharya et al. / Journal of the Neurological Sciences 307 (2011) 50–54
Table 2 Stroke subtypes, treatments offered in the study population and their influence on readmissions among the strokes. Variables
All TIAs (n = 60)
All strokes Ischemic strokes discharged (n = 205) alive (n = 192) P Readmitted Not (n = 22) readmitted value (n = 170)
TOAST mechanisms: Large artery atherosclerosis Cardioembolic Small vessel disease Other known Other unknown Treatments offered iv tPA Intervention Aspirin Statin Discharge destination⁎ Deaths Home/acute rehabilitation Subacute rehab/nursing home
9 (15.0)
87 (42.4)
12 (54.6)
69 (40.6)
7 (11.7) 11 (18.3)
21 (10.2) 60 (29.3)
2 (9.1) 3 (13.6)
15 (8.8) 57 (33.5)
7 (11.7) 26 (43.3)
12 (5.9) 25 (12.2)
2 (9.1) 3 (13.6)
9 (5.3) 20 (11.8)
26 29 171 161
(12.7) (14.2) (83.4) (78.5)
0 0 53 (88.3) 32 (53.3)
0 13 (6.4) 59 (98.3) 146(71.6) 1 (1.7)
42 (20.6)
4 4 17 18
(18.2) (18.2) (77.3) (81.8)
19 22 146 136
(11.2) (12.9) (85.9) (80.0)
– 12(54.6)
– 134 (80.7)
10 (45.4)
32 (19.3)
0.47
0.35 0.50 0.34 0.84
0.01
Numbers in parenthesis represent column percentages. ⁎ Three strokes were transferred to other acute care hospitals for insurance reasons and discharge status of one stroke patient was unknown.
patient 8 in Table 3), ischemic stroke (n = 1; patient 3 in Table 3) and unrelated: sickle pain crisis (n = 1). Results of univariate analysis for predictors of readmission among subjects with radiologically proven strokes in 30 days are shown in Tables 1 and 2. While not reaching statistical significance; the readmitted group had more Medicare/Medicaid subjects (59.1% vs. 47.3%) and fewer privately insured subjects (18.2% vs. 35.5%) than the non-readmitted group. There was a significantly higher proportion of congestive heart failure (22.7% vs. 8.8%) and coronary artery disease (45.5% vs. 14.7%) among the readmitted group; although the reasons for readmission were non-cardiac. If the initial stroke was severe (NIHSS ≥ 10), the subjects had higher likelihood of being readmitted (20.4% versus 8.0% for NIHSS b 10; p = 0.02). Fewer subjects with mild disability (mRS 0–2) at the first discharge were readmitted than those with moderate-severe disability (mRS 3–5), however this was not statistically significant. (10% mRS 0–2 vs. 14.7% mRS 3–5; p value 0.34). Significantly greater proportion of patients discharged to subacute rehabilitation facility or a nursing home was readmitted than those discharged home or to acute rehabilitation facility (23.8% vs. 8.2%; p value 0.01). Finally, a multivariate model was created with all aforementioned significant variables from the univariate analysis with backward selection of variables to determine the most important predictors of readmission, Coronary artery disease was the only predictor of readmission in this analysis (n = 0.001). 3.3. Readmissions with strokes Clinical scenarios of the eight subjects who were readmitted with neurological symptoms are presented in Table 3. Upon reviewing the circumstances of readmission, patients 4 and 5 could potentially have been spared readmission by altering treatments offered during the initial admission (better blood pressure management for patient 4 and carotid endarterectomy for patient 5).
Table 3 Features of patients readmitted with neurological symptoms. No. Age/ Initial admission sex Readmission as ischemic stroke: 1 74/ History of atrial fibrillation, M prostate cancer; recent cocaine and heroin use with right frontal infarct 2 52/ History of moderate aortic M stenosis; right hemiparesis and global aphasia, received iv + ia tPA; worsened in hospital and got mechanical thrombectomy of M1 clot; discharged home on dual antiplatelet and statin 3 50/F History of breast cancer, hypercholesterolemia coronary artery disease; presented with transient left hand clumsiness; Left atrial thrombus on Echo; discharged on Lovenox anticoagulation
Readmission
Recurrent left sided weakness, non compliant with Coumadin, developed heroin withdrawal in hospital. Recurrence of symptoms within hours of discharge; got mechanical thrombectomy and stenting of M1
Presented with aphasia; left temporal lobe infarct.
Readmission as intracerebral hemorrhage: 4 88/ History of hypertension, diabetes Returned with worsening ataxia; left cerebellar hemorrhage; BP M hypercholesterolemia and CHF, 222/100. symptoms of dizziness and ataxia; left cerebellar infarct; severe basilar stenosis; discharged to subacute rehabilitation with dual antiplatelet Readmission as transient ischemic attacks: 5 65/ History of hypertension, M hypercholesterolemia; right hemiparesis and global aphasia; received iv tPA; found high grade left carotid stenosis and moderate size left frontal infarct 6 62/F History of hypertension, diabetes. Mild right hemiparesis: left putaminal lacunar stroke; discharged on aspirin and statin. 7
72/ M
8
55/F
Unable to get outpatient endarterectomy due to insurance reasons, presented with a mild Wernicke's aphasia
speech difficulty lasting for less than 12 hours, no new stroke on imaging; risk factors were controlled at the time of second admission returned with transient right hemisensory symptoms
History of; right hemiparesis, declined tPA, MRI showed multiple lacunar strokes within the left MCA territory; improved to NIHSS 0; discharged on aspirin, statin, nicotine patches Transient ataxia, blood pressure History of hypertension, and glucose were lower than hypercholesterolemia, diabetes. initial admission but not at goal. Transient left hemisensory symptoms; discharged home on aspirin and statin
4. Discussion In our hospital based series from a certified stroke center 11.5% of ischemic strokes discharged alive were readmitted within 30 days. The causes of readmission were primarily non-neurological. Severe strokes and patients discharged to nursing homes and subacute rehabilitation were more likely to be readmitted. Coronary artery disease was a strong predictor of readmission, although the reasons for readmission were non-cardiac. A similar hospital registry based study was recently reported with about 10% readmission rates at 30 days. The study reported infections to be the most frequent cause of readmission up to 1 year. The study was conducted in Taiwan where the national health insurance system provides universal and quality health care to the people of Taiwan at affordable cost [12]. In our study, we confirmed the finding that infection is the most common cause of readmission even during the
P. Bhattacharya et al. / Journal of the Neurological Sciences 307 (2011) 50–54
short term (30 days). Moreover, our sample was from an urban population with a moderate number of uninsured and underinsured patients. In our study, the most frequent cause of readmission was infection. This has been demonstrated in studies looking at administrative (Medicare fee for service) databases within the 30 day limit too [7, 10]. Experimental evidence indicates that stroke leads to suppression of immunity which leads to infections [13]. Urinary tract infection is a common complication of stroke, occurring in 17–22% of stroke patients [14,15]. Bladder dysfunction (incontinence and retention) compounded with functional and cognitive impairment; and the increased use of catheters predispose stroke patients to urine infections [16]. Reducing the use of indwelling catheters/short term use rather than long term use, use of condom catheters in patients without retention and education of healthcare staff both at hospitals and skilled facilities, about the proper use of catheters are all potential strategies to reduce urinary infections [17]. Chest infections are also common and affect up to a quarter of stroke patients [14, 18]. Not only do they involve increased costs of care but also increase the risk of mortality up to three fold [19]. Older age, more severe neurological impairment and oropharyngeal dysphagia are risk factors for chest infections [14, 18]. The first episode of pneumonia tends to occur in the first month after the stroke, and patients with oropharyngeal dysphagia may get recurrent pneumonias [20]. Dysphagia screening protocols significantly decrease pneumonia risk [21]. Other modifiable measures that potentially decrease the risk of pneumonia are early rehabilitation and mobilization, limiting sedation, influenza vaccination and pneumovax, oral hygiene, education of patient and caregiver, maintaining nutrition and avoiding gastric reflux particularly in the presence of a nasogastric tube or PEG tube. Patients discharged to nursing homes and subacute rehabilitation centers were readmitted more often than those discharged home or to an acute rehabilitation center. Most likely, this is because the patients discharged to the nursing homes had more disabling strokes. As in our hospital based study, severe strokes and increased functional disability have consistently been shown to predict readmission in administrative databases [8, 22]. Another plausible explanation is that caregivers at home and in acute care-rehabilitation settings are able to pick up problems earlier and can address issues before they become serious to require hospitalization. Some studies reflect this trend [5] whereas other studies have found contrary results, where patients discharged home were more likely to be readmitted than those in nursing homes [23]. The premise was that caregiver perception of problems may be different at home than among healthcare providers at nursing homes. We found opposite results probably due to our study design which followed patients up to 30 days instead of several months in the other studies. In one study, although readmissions were more frequent in the group discharged home at 2 years; this difference appeared at 6 months and beyond. At 30 days, readmissions after stroke were more common among the nursing home patients [23]. Coronary artery disease was a strong predictor of 30 day readmission in our dataset. Ischemic heart disease was associated with a 14% excess risk of readmission in one study [24]. Heart failure has also consistently been associated with 30 day readmission [5]. Interestingly, the reasons for readmission among coronary artery disease patients in our study were not cardiac. This coincides with a population based study on Medicare beneficiaries, where significant proportions of readmissions due to myocardial infarctions occurred after the first few months since discharge [7, 25, 26]. Finally, discharge planning, including good communication between the hospital and the community (family and primary care physicians) is an essential component to prevent readmissions [27]. Prior to discharge, a reliable link of communication between the physician and the patient/family should be set up, where they can
53
contact the physician with issues with the post discharge plan, medication use, home care et cetera. Home follow up visits and home physical therapy visits have shown to significantly reduce readmissions [28]. The use of telemedicine technology to facilitate post-discharge care is another possibility. The present study has certain limitations. First, our data represents a single center experience. Admissions to other region hospitals are unknown, so we may underestimate the true frequency of readmissions. However, our electronic records system reflects admissions to 10 different hospitals within the same healthcare system serving our metropolitan area, so, if patients chose to come to any of those hospitals in the system for readmission, they would have been included in our data. Second, the data only captures hospital admissions. Emergency room visits for problems that did not warrant admission were not captured in this data. Third, majority of the patients in our study were admitted by neurologists. It is possible that patients treated primarily by other services may have different rates of readmission, although this has not been shown in prior studies [10]. We did not collect data about post-discharge care, medication compliance and coordination of care between outpatient follow-up: all variables that influence readmission rates. We also do not have data of previous hospitalizations; a known predictor of readmissions. Finally, the data was collected in a retrospective manner where we relied on the physician provider definitions rather than standardized definitions for various complications. The main advantage of the hospital based design of our study is that it gave us the opportunity to review the charts of patients returning with neurological conditions. As a quality assessment exercise, we could evaluate whether there were any deviations from standards of care which could otherwise have spared a readmission. Such evaluations are not possible with administrative databases. Administrative databases collect data on large numbers of patients and are not limited to a single hospital. However, they have inherent problems such as reliance on ICD codes which are not always accurate. Also many studies rely on the Medicare database which includes subjects 65 years and older, whereas a significant proportion of patients admitted with stroke is younger than 65 (mean age in our data was 60.9 years). 5. Conclusion We found that 9.9% of patients with ischemic strokes and TIAs discharged alive were readmitted within 30 days. Readmission within 30 days incurs excess healthcare costs, which may potentially be preventable. The reasons for early readmission in urban stroke patients are largely non-neurological, and can conceivably be minimized by the measures outlined above. At a local level, physicians caring for stroke patients should review cases that return within 30 days to determine best practices that prevent readmission, as these may vary from hospital to hospital. Preventing readmissions within 30 days is likely to be an important indicator of quality of care in the future. References [1] Ashton CM, Wray NP. A conceptual framework for the study of early readmission as an indicator of quality of care. Soc Sci Med 1996;43:1533–41. [2] Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med 2009;360:1418–28. [3] Center for Medicare and Medicaid Services: http://www.cms.gov/ AcuteInpatientPPS/downloads/CMS-1390-F.pdf. (Accessed December 21, 2010. [4] Lloyd-Jones D, Adams RJ, Brown TM, et al. Heart disease and stroke statistics—2010 update: a report from the American Heart Association. Circulation 2010;121: e46–e215. [5] Kind AJ, Smith MA, Liou JI, Pandhi N, Frytak JR, Finch MD. The price of bouncing back: one-year mortality and payments for acute stroke patients with 30-day bounce-backs. J Am Geriatr Soc 2008;56:999–1005. [6] Lichtman JH, Leifheit-Limson EC, Jones SB, et al. Predictors of hospital readmission after stroke: a systematic review. Stroke 2010;41:2525–33. [7] Bravata DM, Ho SY, Meehan TP, Brass LM, Concato J. Readmission and death after hospitalization for acute ischemic stroke: 5-year follow-up in the medicare population. Stroke 2007;38:1899–904.
54
P. Bhattacharya et al. / Journal of the Neurological Sciences 307 (2011) 50–54
[8] Chuang KY, Wu SC, Ma AH, Chen YH, Wu CL. Identifying factors associated with hospital readmissions among stroke patients in Taipei. J Nurs Res 2005;13:117–28. [9] Johansen HL, Wielgosz AT, Nguyen K, Fry RN. Incidence, comorbidity, case fatality and readmission of hospitalized stroke patients in Canada. Can J Cardiol 2006;22: 65–71. [10] Smith MA, Liou JI, Frytak JR, Finch MD. 30-day survival and rehospitalization for stroke patients according to physician specialty. Cerebrovasc Dis 2006;22:21–6. [11] Adams Jr HP, Bendixen BH, Kappelle LJ, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in acute stroke treatment. Stroke 1993;24:35–41. [12] Lin HJ, Chang WL, Tseng MC. Readmission after stroke in a hospital-based registry: risk, etiologies, and risk factors. Neurology 2011;76:438–43. [13] Klehmet J, Harms H, Richter M, et al. Stroke-induced immunodepression and poststroke infections: lessons from the preventive antibacterial therapy in stroke trial. Neuroscience 2009;158:1184–93. [14] Aslanyan S, Weir CJ, Diener HC, Kaste M, Lees KR. Pneumonia and urinary tract infection after acute ischaemic stroke: a tertiary analysis of the GAIN International trial. Eur J Neurol 2004;11:49–53. [15] Langhorne P, Stott DJ, Robertson L, et al. Medical complications after stroke: a multicenter study. Stroke 2000;31:1223–9. [16] Kong KH, Young S. Incidence and outcome of poststroke urinary retention: a prospective study. Arch Phys Med Rehabil 2000;81:1464–7. [17] Poisson SN, Johnston SC, Josephson SA. Urinary tract infections complicating stroke: mechanisms, consequences, and possible solutions. Stroke 2010;41: e180–4. [18] Mann G, Hankey GJ, Cameron D. Swallowing function after stroke: prognosis and prognostic factors at 6 months. Stroke 1999;30:744–8.
[19] Katzan IL, Cebul RD, Husak SH, Dawson NV, Baker DW. The effect of pneumonia on mortality among patients hospitalized for acute stroke. Neurology 2003;60: 620–5. [20] Masiero S, Pierobon R, Previato C, Gomiero E. Pneumonia in stroke patients with oropharyngeal dysphagia: a six-month follow-up study. Neurol Sci 2008;29: 139–45. [21] Hinchey JA, Shephard T, Furie K, Smith D, Wang D, Tonn S. Formal dysphagia screening protocols prevent pneumonia. Stroke 2005;36:1972–6. [22] Bohannon RW, Lee N. Association of physical functioning with same-hospital readmission after stroke. Am J Phys Med Rehabil 2004;83:434–8. [23] Camberg LC, Smith NE, Beaudet M, Daley J, Cagan M, Thibault G. Discharge destination and repeat hospitalizations. Med Care 1997;35:756–67. [24] Kennedy BS. Does race predict stroke readmission? An analysis using the truncated negative binomial model. J Natl Med Assoc 2005;97:699–713. [25] Christensen MC, Munro V. Ischemic stroke and intracerebral hemorrhage: the latest evidence on mortality, readmissions and hospital costs from Scotland. Neuroepidemiology 2008;30:239–46. [26] Dhamoon MS, Sciacca RR, Rundek T, Sacco RL, Elkind MS. Recurrent stroke and cardiac risks after first ischemic stroke: the Northern Manhattan Study. Neurology 2006;66:641–6. [27] Naylor MD, Brooten D, Campbell R, et al. Comprehensive discharge planning and home follow-up of hospitalized elders: a randomized clinical trial. JAMA 1999;281:613–20. [28] Andersen HE, Schultz-Larsen K, Kreiner S, Forchhammer BH, Eriksen K, Brown A. Can readmission after stroke be prevented? Results of a randomized clinical study: a postdischarge follow-up service for stroke survivors. Stroke 2000;31:1038–45.