National Institutes of Health Stroke Scale, recovery, stroke



  1. Jain, Ash MD, FACC
  2. Houten, Douglas Van BSN, RN, CCRN, CNRN
  3. Sheikh, Lamiya MS


Background: The National Institutes of Health Stroke Scale (NIHSS) was designed primarily as a research instrument and is used in clinical settings. Its use has not yet been examined as a predictor of patient functional outcomes and prognosis in a community hospital setting.


Objectives: The aim of this study was to determine the effectiveness of baseline NIHSS score in predicting patient functionality and disposition at discharge in a designated stroke center at a community hospital.


Methods: The study population included every transient ischemic attack and stroke encounter seen at our community hospital over the past 6 years (n = 2909). Neurological impairment at baseline was quantified using the NIHSS score on the patient's arrival. Outcomes included the patient's discharge disposition (expired or alive) and ambulatory status at discharge. Results were adjusted for age, gender, race, and stroke, myocardial infarction, diabetes, and hypertension history. Analysis was done using R-based statistical tools.


Results: Baseline NIHSS score was a strong predictor of both patient discharge disposition and ambulatory status. After adjusting for confounding factors, with every 1 point increase in the stroke scale at baseline, there was a 2.3 times increased likelihood of mortality and 3 times increased likelihood in worsening of ambulatory function.


Conclusions: In our community hospital setting, the NIHSS score was found to be a strong predictor of patient recovery after stroke. The NIHSS score at baseline may be important for clinicians to consider before patient management decisions and counseling.


Article Content


Despite recent achievements in the field of stroke care and improved awareness, stroke remains the fourth leading cause of death in the United States, with more than 140 000 mortalities due to a stroke every year.1 Recent literature has reported a mortality rate of 10%2 within 30 days of an ischemic stroke and 30% to 40% for hemorrhagic strokes. The identification of early mortality predictors is of extreme importance for clinicians so that specific treatment and management strategies can be used to provide optimum care for high-risk patients. Patient functionality after stroke is significantly associated with likelihood of a hospital readmission and is an important stroke outcome to consider, especially given recent Medicare avoidable readmission policies. Patients with poor functionality on discharge are also more likely to need long-term care and intensive rehabilitation plans, implying a greater financial impact and more serious consequences for the family of the stroke survivor.


The National Institutes of Health Stroke Scale (NIHSS) was designed primarily as a research instrument but can also be useful in clinical practice. It is commonly obtained in patients presenting with acute stroke. The scale consists of 15 items and ranges from 0 to 42 points. A score of 0 indicates no clinically relevant neurological abnormality, whereas a score of 20 or higher usually indicates a dense paralysis with impaired consciousness. Although there have been previous studies linking stroke severity with expected mortality, there is little information on using NIHSS as a predictor of patient functional outcomes and none in a district "community" hospital setting (defined as "a public entity working to meet needs of local residents through services best matching the community needs"3). This study examines the effectiveness of using the NIHSS to aid healthcare providers in predicting mortality as well as functional outcomes in patients presenting with ischemic and hemorrhagic stroke in a community district healthcare setting. Results may also be useful clinically by risk stratifying patients, to engage and educate patients and caregivers who are at high risk for poor patient outcomes on their treatment decisions.



The study was a retrospective analysis of our patient population, which included all admitted stroke and transient ischemic attack patients seen from the establishment of our stroke program in July 2006 to June 2012, regardless of age and severity of neurological deficit. Each admission was considered independently, regardless of whether it was the first stroke occurrence or a recurrent stroke. Data for all stroke patients were entered into the Get With the Guidelines program stroke registry; this included demographic data, risk factors for stroke, stroke type (ischemic or hemorrhagic), stroke outcomes, and all relevant clinical information collected during the current admission. Noncontrast computed tomography was performed on all acute stroke patients within 72 hours of admission, specifically within 25 minutes of arrival to the emergency department. The stroke severity of each patient was assessed within 12 hours of admission by a trained stroke nurse using the NIHSS. The hospital has 7 stroke nurses, certified and trained in administration of the NIHSS, and at least 1 nurse is available onsite for consultation on a 24/7 basis. Each patient was given a score from 0 to 42; patients were initially evaluated within the first 25 minutes of arrival to the emergency department.


Demographic variables included in this study were patient's age, race, gender, and body mass index (BMI). History of smoking, hypertension, diabetes, stroke, and myocardial infarction was examined as binary variables. Ambulatory function was assessed at both admission and discharge as ability to ambulate independently with or without a device, ability to ambulate with assistance from a person, and inability to ambulate. For the purposes of this article, ambulatory function was analyzed as a categorical variable: improvement in ambulatory function, worsening of ambulatory function, and no changes in ambulatory function from admission to discharge. Patient discharge disposition was collected as in-hospital mortality, discharged to a skilled nursing facility, discharged home (with or without home health services), discharged to a hospice facility/home hospice care, discharged to another acute care facility (including acute rehabilitation facilities), and left against medical advice. For the purposes of this article, the main discharge disposition we focused on was in-hospital mortality, which was analyzed as a binary variable. Bivariate data were analyzed using [chi]2 test for categorical variables and t test for continuous variables. Nonparametric variables were analyzed using the Wilcoxon rank-sum test. Multivariate analysis was done using logistic regression. A P value of <.05 (2 sided) was considered to be statistically significant and all data were analyzed using R-based statistical tools (R version 2.14.1). Institutional review board approval was received from our institution.



A total of 2909 patients diagnosed with ischemic or hemorrhagic stroke were included in our study during the 5-year enrollment period from July 2006 to June 2012. Among these patients, only 2664 had NIHSS assessed on the day of admission. As seen in Table 1, the mean (SD) age of patients was 69 (15.7) years. About half of the patients were men (48.3%) and half were women (51.7%). About a tenth of patients (9.8%) were smokers. A little more than half of patients were white (58.7%), followed by Asian (26.5%), African American (5.7%), Hawaiian/Pacific Islander (3.9%), and other races (5.1%). About three-fourths of the population had hypertension (72.5%). Almost a third of the population had diabetes (29.2%). Most patients had same or better ambulatory function on discharge (71.2%), whereas 28.8% had worse ambulatory function or died during the study period (Table 2).

Table 1 - Click to enlarge in new windowTABLE 1 Demographic and Clinical Characteristics of the Stroke Patients (n = 2909) From July 2006 to June 2012
Table 2 - Click to enlarge in new windowTABLE 2 Outcomes of the Stroke Patients (n = 2909) From July 2006 to June 2012

As seen in Table 3, after adjusting for confounding factors, with every 1-point increase in the stroke scale at baseline, patients were twice as likely to die (relative risk [RR], 2.34; 95% confidence interval [CI], 1.33-4.11), and there was a 3.28 times increase in likelihood of worse ambulatory function after stroke (RR,3.28; 95% CI, 2.16-4.98). Age was a significant confounding factor (older patients were less likely to survive, had worsening ambulatory function, and had higher NIHSS score; P < .0001, P = .01, and P = .04, respectively) and was therefore adjusted for in the logistic regression analyses. Male stroke patients were significantly less likely than women to survive, irrespective of the patient's stroke severity and medical history.

Table 3 - Click to enlarge in new windowTABLE 3 Relative Risks of Stroke Patient Outcomes Given Patient Characteristics (n = 2909)

Body mass index was not significantly associated with NIHSS in our study. However, with every 1-point increase in BMI, patients were 92% less likely to die (95% CI, 0.01-0.46). When this protective effect of BMI was examined more closely, it was determined that all older patients (>70 years old) were overweight, that is, had a BMI of 25 kg/m2 or higher, whereas younger patients had a mostly normal to low BMI. Looking at the data more closely, it was determined that underweight patients (BMI <18 kg/m2) were at highest risk of dying, followed by normal-weight patients (BMI 18-25 kg/m2) and overweight patients had the lowest risk of mortality. Gender was not found to have a significant effect on the relationship between BMI and risk of mortality.


There were no significant association between NIHSS and recurrent stroke occurrence in our population; however, patients who were admitted with a recurrent stroke were almost 3 times as likely to die compared with patients with no medical history of stroke (RR, 2.71; 95% CI, 1.89-3.89). Patients with a history of stroke were also 55% more likely to have worsening in ambulatory function (RR, 1.55; 95% CI, 1.19-2.01).



Understanding the mortality and functional potential based on initial assessments can have benefits to both providers and patients in any hospital setting. It is well known that stroke is the number 1 cause of serious, long-term disability. Our study reinforces the importance of using the NIHSS score as a risk modifier in prognostic models for clinical care. Specifically, the 2-fold increased risk of mortality and approximately 3-fold increased risk of worsened ambulatory function with every 1-point increase in NIHSS score (even after adjusting for patient demographics and confounding factors) provide a quantitative picture to identify which patients are at highest risk of negative outcomes. A simple, initial test that can assist with planning for a stroke survivor's long-term medical care may allow for the opportunity to organize a long-term medical plan right from the beginning of a patient's stay in community district hospitals.


Teaching and discharge planning may help the families of stroke survivors make long-term arrangements specific to the expected functional outcome. Families may need to make alterations in the home environment such as ramps, shower grab bars, and easily accessible beds. The earlier these plans are put into action, the sooner an impaired stroke patient may be able to be discharged home, reducing patient's length of stay in the hospital.


Our analysis showed that sex was found to have a significant effect on in-hospital mortality after stroke, with men tending toward poorer outcomes. Adjustment for other demographic characteristics, NIHSS and clinical factors did not eliminate the disparity. One possible explanation for this is the generally higher all-cause mortality rate among men compared with women in the US population.4,54,5 Our results are also consistent with previous studies that have found that mortality due to cerebrovascular diseases is higher among male compared with female stroke patients.4-64-64-6 Although there have been some studies that have found an increased risk of stroke mortality among women, other research has shown this higher risk of mortality exists only among subarachnoid hemorrhage female patients.6-86-86-8 In future studies, we plan on delineating the specific causes of mortality as an outcome.


One interesting finding was that stroke patients who were overweight had better survival rates after a stroke, irrespective of NIHSS. It is important to note that although overweight patients had better prognosis, nearly 87% of patients who presented with stroke were overweight, implying that high BMI is still a significant risk factor for occurrence of stroke and possibly also for other poor cardiovascular9-129-129-129-12 outcomes. In addition, we were unable to determine using our data whether underweight patients had lost weight rapidly, a factor that may contribute to poor cardiovascular outcomes. Lastly, although this study examined the effect of BMI on in-hospital mortality, it may be important to investigate its effect on long-term prognosis. Our regression analysis also revealed that recurrent stroke was an independent predictor of poor long-term outcomes. This is in agreement with previous studies performed in acute stroke care settings.13 It is interesting to note that BMI and stroke recurrence were not associated with NIHSS score in our population.


Our study has several limitations; our population was a hospital-based cohort in a specific community district hospital setting. A larger sample and multisite study is needed to prove causality and predictive capability and for generalizability. Our study population is also reflective of a highly ethnically diverse community in our region; although the present article looks at ethnicity from a more traditional categorization, further research will delineate the effect of specific ethnicities of stroke outcomes. Our study was also limited to a single measure of NIHSS, at admission. Further studies evaluating within-subject changes of NIHSS during a patient's stay as a predictor of outcomes may be useful to accurately represent the entire spectrum of neurological deficit. In addition, the retrospective nature of this analysis introduces potential biases (selection bias due to loss to follow up and misclassification bias) that should be addressed in prospective studies.


As mentioned earlier, severe acute stroke is the fourth leading cause of death. Often, medical personnel may be in the position of conferring with families of elderly acute stroke patients where the expected outcome is very poor. With a more definite perception of a patient's outcome, medical personnel may be able to help families faced with difficult decisions to focus on palliative measures rather than aggressive treatments in the "hope" of a cure. Early prediction of a patient's prognosis may also help in adherence to a patient's wishes as expressed in an advanced healthcare directive; sometimes a more natural death is preferred if the expectation is that death or a greatly impaired existence would be the most likely outcome.


Understanding significance of the implications of initial NIHSS is a valuable instrument for the clinical practitioner in the care of acute stroke patients, and further research is needed on the use of our model as a predictive instrument for counseling and education. Patients with higher stroke severity, leaner patients, men, and patients with recurrent stroke should be carefully monitored after stroke for cardiovascular risk and other comorbidities that may confer increased risk of death. Results from our study may also be useful for pay-for-performance reimbursement, especially in light of the Center for Medicare and Medicaid's regulations on value-based purchasing and public reporting of stroke mortality rate. By correctly identifying patients who are at highest risk for poor outcomes, providers are able to create a system of stroke care that is customizable per patient, in alignment with a patient-centered care approach.


What's New and Important


* Our study reinforces the importance of using the NIHSS score as a risk modifier in prognostic models to aid providers in identifying which patients are at highest risk of negative outcomes.


* Adherence to rapid and more consistent use of IV TPA among severe NIHSS score patients may lead to better neurological outcomes.


* Leaner patients, men, and patients with recurrent stroke should be carefully monitored after stroke for cardiovascular risk and other comorbidities. By correctly identifying high-risk patients, providers are able to create a patient-specific treatment plan.




1. Minino A, Xu J, Kochanek K. Deaths: preliminary data for 2008. Natl Vital Stat Rep. 2010; 59 (2): 1-72. [Context Link]


2. de Jong G, van Raak L, Kessels F, Lodder J. Stroke subtype and mortality, a follow-up study in 998 patients with a first cerebral infarct. J Clin Epidemiol. 2003; 56: 262-568. [Context Link]


3. California Department of Health Care Services. Providers & Partners. Accessed July 18, 2014. [Context Link]


4. Holroyd-Leduc JM, Kapral MK, Austin PC, Tu JV. Sex differences and similarities in the management and outcome of stroke patients. Stroke. 2000; 31: 1833-1837. [Context Link]


5. Bravata DM, Ho SY, Brass LM, Concato J, Scinto J, Meehan TP. Long-term mortality in cerebrovascular disease. Stroke. 2003; 34: 699-704. [Context Link]


6. Sheikh K, Bullock CM. Effect of measurement on sex difference in stroke mortality. Stroke. 2007; 38 (3): 1085-1087. [Context Link]


7. Johnston SC, Selvin S, Gress DR. The burden, trends, and demographics of mortality from subarachnoid hemorrhage. Neurology. 1998; 50: 1413-1418. [Context Link]


8. Ayala C, Croft JB, Greenlund KJ, et al. Sex differences in US mortality rates for stroke and stroke subtypes by race/ethnicity and age, 1995-1998. Stroke. 2002; 33: 1197-1201. [Context Link]


9. Jee SH, Sull JW, Park J, et al. Body-mass index and mortality in Korean men and women. N Engl J Med. 2006; 355: 779-787. [Context Link]


10. Pischon T, Boeing H, Hoffmann K, et al. General and abdominal adiposity and risk of death in Europe. N Engl J Med. 2008; 359: 2105-2120. [Context Link]


11. Yusuf S, Hawken S, Ounpuu S, et al. INTERHEART Study Investigators. Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case-control study. Lancet. 2005; 366: 1640-1649. [Context Link]


12. Kenchaiah S, Evans JC, Levy D, et al. Obesity and the risk of heart failure. N Engl J Med. 2002; 347 (5): 305-313. [Context Link]


13. Kozdag G, Yaymac[latin dotless i] M, Iseri P, et al. Ischemic stroke history predicts increased cardiovascular mortality in chronic heart failure. Anadolu Kardiyol Derg. 2011; 11 (5): 421-427. [Context Link]