Optimizing Discharge Planning
Clinical Predictors of Longer Stay After Recombinant Tissue Plasminogen Activator for Acute Stroke
Background and Purpose— The length of stay (LOS) is the main cost-determining factor for inpatients with acute stroke. Although studies have identified variables associated with LOS, few have analyzed predictors of longer stay after receiving thrombolytic therapy for acute stroke.
Methods— We studied all consecutive acute stroke patients receiving intravenous recombinant tissue plasminogen activator (rtPA) admitted to the London Health Sciences Center, in London, Ontario, Canada, from 1999 to 2003. Longer stay was defined as LOS ≥7 days after admission. Demographic as well as baseline clinical, laboratory, and imaging variables were analyzed to identify predictors of LOS. Significant variables were entered into a multivariate logistic regression analysis.
Results— Among 216 acute stroke patients receiving rtPA, the median LOS was 6 days. LOS was >7 days in 102 (49%) patients. Age ≥70 (odds ratio [OR], 2.2; 95% CI, 1.2 to 4.0), lack of improvement at 24 hours (OR, 2.5; 95% CI, 1.4 to 4.4), prestroke modified Rankin Scale ≥2 (OR, 2.4; 95% CI, 1.2 to 4.9), baseline National Institutes of Health Stroke Scale score ≥15 (OR, 9.4; 95% CI, 3.2 to 27.6), cortical involvement (OR, 2.2; 95% CI, 1.2 to 3.9), and new infarction on the control computed tomography (CT; OR, 2.8; 95% CI, 1.4 to 5.9) were independent predictors of longer stay.
Conclusions— Lack of improvement at 24 hours after rtPA, cortical involvement, and new infarction on the 24-hour CT scan are relevant variables that can independently affect the LOS. These new variables may be useful for establishing policy in relation to the organization and planning of the health care system.
The length of stay (LOS) is the main cost-determining factor during hospitalization.1 The average LOS for stroke varies among different countries, which may reflect the impact of the differences in health care system organization. For example, the LOS in the United States2 for patients with acute ischemic stroke ranges from 6 to 11 days compared with much longer hospitalizations (17 to 26 days) in Canada,3 Europe,4 and Asia.5 Information that can predict longer LOS in stroke patients will be useful for clinical and systems management, for example, as a marker for discharge planning, resource utilization, and cost implications.6 Few articles identify clinical factors associated with longer hospitalization in stroke patients who receive thrombolytic therapy.7,8 Most articles concerning LOS take into account the perspective of the administrators, are focused on costs, and do not necessarily consider the clinical relevance of findings.5
The aim of the present study was to determine clinical predictors for longer hospitalization in stroke patients after receiving recombinant tissue plasminogen activator (rtPA). We hypothesized that (1) baseline clinical, imaging, or laboratory factors are associated with LOS; and (2) these factors are different from those described previously in the prethrombolytic period or in patients not receiving rtPA.
Patients and Methods
We analyzed all consecutive acute stroke patients who received rtPA. All patients were admitted to the London Health Sciences Center, University Campus (LHSC-UC) in London between January 1999 and March 2003. London is the largest city in Southwestern Ontario, with a population of 336 540 inhabitants (432 450 in the metropolitan area). It has 2 academic medical centers with 24-hour access to computed tomography (CT) and MRI. These academic hospitals are a referral center for a large part of Ontario. In addition to serving the local population, LHSC-UC receives acute stroke referrals from 33 rural hospitals from 7 counties. Hospital characteristics, population served, and catchment area were outlined in previous articles.9,10 Demographic variables, stroke severity at admission and at 24 hours, functional status at admission, comorbidity, and outcomes were prospectively collected. Time of symptom onset was defined by the time when patients were “last seen to be well.” Time of rtPA was obtained from the nurse records as onset to needle time for the rtPA infusion. Baseline National Institutes of Health Stroke Scale (NIHSS) score was performed by a certified neurologist just before treatment with rtPA. Current medications were also recorded if they were regularly consumed in the previous month.
All patients had a baseline and 24-hour CT scan. A single neuroradiologist blinded to clinical data reviewed scans to determine the presence of new infarction, cortical involvement, and extension of the ischemic lesion.
The decision to treat with rtPA was made according to the National Institute of Neurological Disorders and Stroke (NINDS) protocol. Inclusion and exclusion criteria were applied with a major difference from NINDS: stroke patients with involvement of more than one third of the middle cerebral artery territory on the baseline CT scan were excluded.
An evaluation to determine the stroke mechanism and subtype in all patients included: routine laboratory tests, ECG, transthoracic echocardiogram, and carotid ultrasound. Outcomes at 3 months were assessed according to the modified Rankin score.
Definition of Variables
We analyzed the distribution of all variables by graphic and analytic methods (frequency distribution by quartiles or quintiles). When there was no clear relationship, we used clinical criteria to analyze the variables. In our analysis, dose of rtPA, time to treatment, glucose, and white cell count were considered continuous variables. Age, baseline NIHSS, and modified Rankin Scale (mRS), were categorized a priori according to common cutoff described in the literature.
Because the distribution of LOS was skewed to the right, the results were summarized be median and 25th and 75th percentile values.
Stroke subtype (lacunar versus nonlacunar) was based on presenting symptoms, physical examination, and neuroimaging. The presence of cortical involvement, new infarction, or hemorrhagic transformation was established according to the neuroradiology report of the 24-hour CT scan.
Clinical Outcome Measures
Longer stay was defined as LOS ≥7 days. In previous studies, a longer LOS was defined as 6 to 8 days.11,12 This measure is based on the understanding that the provision of acute stroke care, identification of the stroke mechanism, and preventing complications is generally achieved within the first week of admission.
Outcome at 24 hours was defined as a lack of improvement (LOI) determined by a ≤3-point difference between the baseline and 24-hour NIHSS. Three-month outcomes were determined using the mRS. Poor outcome was defined as an mRS ≥3 or death.
The association between demographic characteristics, clinical and hemodynamic variables, and LOS was examined using univariate logistic regression. Stepwise multivariate logistic regression, allowing for entry at the 0.15 level of significance based on the score statistic, was used to determine a subset of these variables independently associated with LOS. Covariates were checked for collinearity and interaction effects. Discrimination of the model was assessed by the area under the receiver operating characteristic (ROC) curve, and calibration was assessed using goodness of fit test.
Statistical analysis was performed using STATA 7.0 (StataCorp LP). P values <0.05 were considered significant.
A total of 216 patients received intravenous rtPA at LHSC-UC between January 1999 and March 2003. The mean age was 71.5±12 years, and 111 (51%) were male. Baseline characteristics are described in Table 1.
Seven patients (3.2%) were excluded because of missing data. Finally, 209 patients were considered for the analysis. Median LOS in stroke patients after rtPA was 6 days (25th to 75th percentile: 4 to 12 days; Table 1).
The LOS was >7 days in 102 (49%) patients. There were no statistically significant differences in sex, city of onset (London versus other), vascular risk factors, previous medication, stroke subtype (lacunar versus non lacunar), glucose level at admission, time to treatment, and rtPA dose between both groups (LOS ≤7 days). Table 2 summarizes predictors of longer stay in the univariate analysis. The overall asymptomatic and symptomatic hemorrhage rates at 36 hours were 10.4% and 4.1%, respectively. Five patients (2.3%) with symptomatic intracranial hemorrhage died. The presence of symptomatic or any kind of bleeding was not associated with longer stay (P=0.79 and 0.20, respectively).
Forty patients (18%) were treated outside the time window (180 minutes). There was no statistically significant difference in the LOS between those patients treated within or outside 180 minutes of symptom onset (P=0.29).
A total of 118 patients (55%) had poor outcome (mRS 3 to 5 or death) at 90 days. The frequency of longer stay was significantly higher among patients with poor outcome at 3 months (74% versus 39.5%; P<0.001).
In logistic regression analysis, we identified 2 models with similar performance and predicting values. In model A, age ≥70 years (odds ratio [OR], 2.10; 95% CI, 1.12 to 3.86), baseline NIHSS ≥15 (OR, 2.22; 95% CI, 1.19 to 4.15), and the presence of a new infarction (OR, 2.52; 95% CI, 1.16 to 5.47) were independent predictors of longer stay (Table 3).
In model B, age ≥70 years (OR, 2.18; 95% CI, 1.15 to 4.12), baseline NIHSS >15 (OR, 2.48; 95% CI, 1.33 to 4.62), and LOI 24 hours after rtPA (OR, 2.70; 95% CI, 1.48 to 4.70) were independent predictors of longer stay (Table 3).
Goodness of fit test was not significant (model A P value=0.76; model B P value=0.21), indicating adequate fitness. The discrimination of models was moderate to adequate with an under the curve area (ROC curve) of 0.69 and 0.70 for models A and B, respectively.
The cost-effectiveness of rtPA in acute stroke has been analyzed previously.8,13,14 The analysis of LOS can provide valuable data for planning and policy in the health care system. After the approval of rtPA, the main focus of planning has been productivity, cost containment, and quality of care. In daily practice, hospitals and health administrators have to find strategies to manage different and often competing demands. This includes maintaining access to care in the scenario of disabling diseases that predispose patients to longer LOS balanced against the pressure to admit new patients. Clearly, shortening the LOS and using the bed and personnel resources more efficiently are ways to achieve high patient turnover, and subsequently, provide more effective acute care for stroke patients.
In the NINDS trial, use of rtPA caused a statistically significant reduction in the average LOS by 2 days. Other studies have shown that certain medical, psychological, cognitive, and physical aspects of stroke correlate with LOS. For example, NIHSS and Barthel index at admission, male gender, smoking, and stroke subtype were associated with longer stay.5,15 Our findings in relation to age, functional status, prestroke and post–rtPA, and stroke severity are in agreement with previous studies.16,17
The novel finding was that the presence of either a new infarction or cortical involvement on the 24-hour CT scan independently predicted longer stay. More interestingly, LOI at 24 hours after receiving rtPA, as measured by NIHSS, was also a predictor of LOS. These new variables were relevant explanatory factors for longer LOS according to the logistic regression analysis.
Our small sample size may limit the generalizability of the results. However, this is an exploratory analysis aimed to identify clinical predictors of longer hospitalization after thrombolytic therapy rather than a predictive model.
Potential Implications and Future Directions
LOS is the major determinant of acute care costs. Most institutions focus their resources according to stroke severity, as measured by mRS, Barthel index, or NIHSS. However, other variables can add relevant clinical information that may impact discharge planning. Our results suggest that age, functional status prestroke, LOI at 24 hours after rtPA, stroke severity, and CT findings can be independent predictors of longer stay.
This information can be used to optimize discharge planning by (1) beginning the discharge planning at the time of admission or 24 hours after the patient received rtPA; (2) optimizing the level of care; (3) matching care provided to care required; and (4) earlier rehabilitation or discharge to long-term care. Further, resources can be reorganized and guidelines developed that would ultimately provide more efficient management for acute stroke patients.
We thank previous stroke fellows for their contribution in collecting data. We also appreciate the daily help provided by the urgent TIA clinical nurse, assistants, and research nurses. This research was supported in part by a competitive grant of the Heart Stroke Foundation of Canada (HSFC) and Canadian Institute for Health Research (CIHR) given to G.S.
- Received September 1, 2004.
- Revision received October 13, 2004.
- Accepted October 19, 2004.
Caro JJ, Huybrechts KF, Duchesne I. Management patterns and costs of acute ischemic stroke: an international study. For the Stroke Economic Analysis Group. Stroke. 2000; 31: 582–590.
Samsa GP, Bian J, Lipscomb J, Matchar DB. Epidemiology of recurrent cerebral infarction: a medicare claims-based comparison of first and recurrent strokes on 2-year survival and cost. Stroke. 1999; 30: 338–349.
van Straten A, van der Meulen JH, van den Bos GA, Limburg M. Length of hospital stay and discharge delays in stroke patients. Stroke. 1997; 28: 137–140.
Chang KC, Tseng MC, Weng HH, Lin YH, Liou CW, Tan TY. Prediction of length of stay of first-ever ischemic stroke. Stroke. 2002; 33: 2670–2674.
Caro JJ, Huybrechts KF, Kelley HE. Predicting treatment costs after acute ischemic stroke on the basis of patient characteristics at presentation and early dysfunction. Stroke. 2001; 32: 100–106.
Hacke W, Kaste M, Fieschi C, Toni D, Lesaffre E, von Kummer R, Boysen G, Bluhmki E, Hoxter G, Mahagne MH, et al. Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke. The European Cooperative Acute Stroke Study (ECASS). J Am Med Assoc. 1995; 274: 1017–1025.
Fagan SC, Morgenstern LB, Petitta A, Ward RE, Tilley BC, Marler JR, Levine SR, Broderick JP, Kwiatkowski TG, Frankel M, Brott TG, Walker MD. Cost-effectiveness of tissue plasminogen activator for acute ischemic stroke. NINDS rt-PA Stroke Study Group. Neurology. 1998; 50: 883–890.
Merino JG, Silver B, Wong E, Foell B, Demaerschalk B, Tamayo A, Poncha F, Hachinski V; Southwestern Ontario Stroke Program. Extending tissue plasminogen activator use to community and rural stroke patients. Stroke. 2002; 33: 141–146.
Williams LS, Rotich J, Qi R, Fineberg N, Espay A, Bruno A, Fineberg SE, Tierney WR. Effects of admission hyperglycemia on mortality and costs in acute ischemic stroke. Neurology. 2002; 59: 67–71.
Wade DT, Langton Hewer R. Hospital admission for acute stroke: who, for how long, and to what effect? J Epidemiol Community Health. 1985; 39: 347–352.
Patel A, Knapp M, Perez I, Evans A, Kalra L. Alternative strategies for stroke care: cost-effectiveness and cost-utility analyses from a prospective randomized controlled trial. Stroke. 2004; 35: 196–203.
Galski T, Bruno RL, Zorowitz R, Walker J. Predicting length of stay, functional outcome, and aftercare in the rehabilitation of stroke patients. The dominant role of higher-order cognition. Stroke. 1993; 24: 1794–1800.
Effect of intravenous recombinant tissue plasminogen activator on ischemic stroke lesion size measured by computed tomography. NINDS; The National Institute of Neurological Disorders and Stroke (NINDS) rt-PA Stroke Study Group. Stroke. 2000; 31: 2912–2919.
Chapman KM, Woolfenden AR, Graeb D, Johnston DC, Beckman J, Schulzer M, Teal PA. Intravenous tissue plasminogen activator for acute ischemic stroke: a Canadian hospital’s experience. Stroke. 2000; 31: 2920–2924.