Stroke is becoming a serious problem in public health in developing countries, Uganda inclusive, accounting for 85% of global deaths (2, 5, 7, 8, 13, 26, 27, 28, 29). Stroke severity, its associated complications, and lack of stroke units contribute to the high mortality rate found to be as high as 62% at one year (9, 14, 20, 21, 22, 30, 31). Despite this, there is very limited data on stroke mortality in sub-Saharan Africa, and to the best of our knowledge there are no published data on stroke mortality in Uganda. In this study, we reported the 30 day stroke mortality and associated clinical and laboratory presentations among adult patients admitted with stroke at Mulago hospital in Uganda.
Setting: Mulago Hospital is Uganda’s national referral hospital and Makerere University College of Health Sciences’ teaching hospital. It is located in Kampala and has an estimated bed capacity of 1,500. It has an accident and emergency unit, intensive care unit, clinical laboratory, radiology department with a CT scan and highly trained radiologists, neurology, neurosurgery and physiotherapy units.
Study period: During a six months period from 1st July 2010 to 30th January 2011, 167 patients who presented to Mulago hospital’s accident and emergency unit with neurologic deficits suggestive of acute stroke (29) were screened. Computerised tomography scan of the brain was done to confirm stroke by radiologists with experience in this field. Classification of stroke subtypes was done using the Trial of ORG 10172 and medical disability guidelines for ischemic and hemorrhagic stroke respectively (1, 17). Patients who had a normal brain CT scan, had a repeat done on day 7 and were excluded in case it was still normal. Participants with confirmed stroke were approached for enrolment into the study. Those that consented were recruited consecutively. An interviewer based questionnaire was administered by the principal investigator and trained research assistants. Selected social demographic characteristics, medical history relevant to stroke and a comprehensive clinical examination was done. Blood samples for essential laboratory investigations (complete blood count, erythrocyte sedimentation rate, fasting lipid profile and blood sugar, HIV serology and rapid plasma reagin) were obtained. These patients were reviewed every two days from the day of admission while on the ward until discharge. Upon admission, the patients were managed on the accident and emergency unit, and general neurology unit by a neurologist, internal medicine physicians, nurses and auxiliary staffs. They received supportive treatment which included ensuring a patent airway, good oxygen saturation, haemodynamic stability, nutrition, hydration, temperature and glycaemic control, prevention of deep vein thrombosis and pressure sores. Specific treatment included use of antihypertensive drugs (labetalol and hydralazine) when blood pressure exceeded 180/105mmHg and 160/100mmHg for ischemic and haemorrhagic stroke respectively. Antiplatelet drugs including aspirin and clopidogrel for ischemic stroke and statins were administered. Rehabilitation included physiotherapy, occupational, speech and language therapy. Those that required mechanical ventilation were admitted to the intensive care unit. None of the patients with ischemic stroke benefitted from recombinant tissue plasminogen activator because they presented on average 2 days post ictal. On discharge, they were then scheduled for a neurology outpatient clinic visit that coincided with the 30th day from the date of stroke onset. Participants that did not turn up for the scheduled visit were then contacted by telephone (contacts of three immediate relatives) to ascertain whether they were still alive or dead. Participants who died within 30 days of stroke onset were recorded.
One hundred and sixty seven patients who presented to Mulago hospital’s accident and emergency unit with neurologic deficits suggestive of acute stroke (29) were screened. One hundred and fifty patients, 18 years and above, were confirmed to have stroke on brain CT scan. Seventeen patients declined to consent for participation in the study and 133 patients were enrolled consecutively into the study. Five participants were lost to follow up and therefore data from 128 participants were analysed. At the end of the follow up period of 30 days from stroke onset, 56 (43.8%) participants had died while 72 (56.2%) were still alive. Stroke mortality was not stratified at 24 hours and seven days from stroke onset as this preliminary study only established stroke mortality at 30 days among adult stroke patients admitted at Mulago hospital.
The mean age of the participants was 62.3+15.7 years with a median of 64.5 and 58% were female. Fifty six percent of the participants had never had formal education, 38% were Catholic and 66% were married. Twenty one (15.8%) were taking alcohol while only 3 (2.3%) were current smokers. Social demographic characteristics are presented in Table 1.
Majority of the participants 105 (79%) had ischemic stroke and 28 (21%) had hemorrhagic stroke. All study participants presented with focal neurological deficits, with majority 97 (76%) presenting with only motor deficits. Impaired level of consciousness was present in 78 (61%) patients. More than 50% of the participants had high blood pressure at admission with diastolic and systolic hypertension found among 106(83%) and 68 (53%) respectively. Eighty four (66%) were aware of their hypertension. The clinical characteristics of the study participants are presented in Table 2.
Most patients 103(80.47%) had haemoglobin levels less than 11.5g/dl followed by hyperglycaemia of more than 7.0 mmol/l in 48 (38%) and leucocytosis in 42 (33%) patients. The laboratory presentations of the study participants are presented in Table 3.
Clinical and laboratory parameters that were significantly associated with mortality at bivariate analysis with P- value set at < 0.05 were entered into the multivariate linear logistic regression model. Factors that were significantly associated with 30 day stroke mortality were; age 51-60 years with OR 0.18 (95% CI; 0.04- 0.82) p value 0.044 and Glasgow coma scale < 9 OR 0.13 (95%CI; 0.05-0.25) p value <0.001. Temperature, though not statistically significant at multivariate analysis, it was associated with increased mortality at bivariate analysis OR 2.81 (95%CI; 1.2-6.60) P<0.005. On the other hand, though mortality was higher among patients with hemorrhagic stroke (54%) compared to ischemic stroke (41%), there was no statistically significant difference at bivariate analysis OR 1.71 (95% CI 0.68-4.27). Factors significantly associated with 30 day stroke mortality are presented in Table 4.
Stroke has been demonstrated in this study as a challenge among patients admitted to Mulago hospital. The majority of patients in this study were female 74 (58%). They were also relatively young with mean age of 62.3years similar to other studies in sub-Saharan Africa (21) unlike in the West where the mean age is above 75years (30).
The 30-day mortality among stroke patients was 43.8% which is similar to other hospital based stroke case fatality studies in Africa: 54% in Gambia, 33.6% in Nigeria and 49.6% in Maputo Mozambique (6, 10, 20, 21, 22, 25).
However in developed countries where patients present early following stroke and care takes place in specialized stroke units, the 30-day stroke mortality is much lower. A study done in Toronto, Canada showed 30 day mortality among stroke patients of 20% (30).
Majority of the patients presented with history of loss of consciousness 78 (61 %), Glasgow coma scale less than 9, 38(30 %) and diastolic hypertension 106 (83%) which is comparable with other studies done in resource limited countries (3). A Glasgow coma scale score of less than 9 was associated with increased mortality which is also comparable with other studies done on stroke mortality (23, 30).
Regarding stroke subtypes, majority of the patients had ischemic stroke (79%), mortality was higher among those with hemorrhagic stroke, however there was no statistically significant difference in mortality between hemorrhagic and ischemic stroke. This is in contrast to other studies, which demonstrated a higher mortality among patients with hemorrhagic stroke (20, 23). In this study, probably we needed a bigger sample size to conclusively come up with a more realistic association.
In this study, patients who were between the ages of 51 and 60 years (6th decade) had a higher 30 day stroke mortality rate compared to those who were below 51 and above 60 years. However, those above 70 years of age had the highest 30 day mortality rate which is similar to other studies that showed mortality worsens with advancing age (30). Smoking and alcohol use were not associated with increased 30 day stroke mortality among the study participants which is similar to other studies (4). Studies on 30 day stroke mortality and associated factors demonstrated a trend that hyperpyrexia worsens stroke mortality (11, 12). This trend was not sustained at multivariate analysis.
Prevalence of HIV among stroke patients was 12%, which is higher compared to 5.8% in an earlier study done in Mulago among stroke patients in 2007 (19). In the general population the average HIV seroprevalence in Uganda is 7.3% as reported by Ministry of Health, Uganda (24). At Mulago hospital, HIV seroprevalence on the medical wards currently stands at 50 % (16). The HIV seroprevalence among our study participants was much lower than reported from studies done in Muhimbili, Tanzania and Blantyre, Malawi 20.9% and 48% respectively (15, 18). In these two studies, the average mean age of the patients was 47years, which is lower than the mean age in our study which was 63.3 years.
Thirty day mortality among adult stroke patients admitted at Mulago hospital was 43.8%. Delay in hospital presentation and lack of organised inpatient stroke care (stroke units) is likely responsible for this high mortality. Poor prognostic factors for 30 day stroke mortality in our study population were GCS of <9 and age 51-60 years. Future studies should be directed towards increasing stroke prevention and public recognition of stroke to enhance early presentation to hospital as well as setting up organised inpatient stroke care facilities to decrease morbidity and mortality from stroke.
Table 1: Social demographic characteristics of the study participants
|Level of education|
Table 2: Clinical presentation of the study participants
|Variable||Frequency N = 128||Percentage|
|Sudden trouble in walking||108||84|
|Sudden weakness of the face||109||85|
|Sudden numbness of the face||80||63|
|Sudden onset of vomiting||37||29|
|Loss of consciousness||78||61|
|Past medical history|
|Glasgow coma scale <9||38||30|
|Glasgow coma scale >9||90||70|
|Motor & sensory||29||22|
Table 3: Laboratory presentation of study participants
|Complete blood cell count|
|Fasting lipid profile|
Table 4: Clinical and laboratory findings significantly associated with 30 day stroke mortality at multivariate analysis
|Age group (years)|