Forum
Distribution and Characteristics of Severe Traumatic Brain Injury at Mulago National Referral Hospital in Uganda Tu M. Tran1, Anthony T. Fuller1,2, Joel Kiryabwire3, John Mukasa3, Michael Muhumuza3, Hussein Ssenyojo3, Michael M. Haglund1,4
OBJECTIVE: Road traffic accidents are a leading cause of injury in low- and middle-income countries, where mortality rates are disproportionately higher. Patients with severe traumatic brain injury (TBI) tend to have very poor outcomes. To reduce the burden from severe TBI, we describe its distribution at Mulago National Referral Hospital (Kampala, Uganda) and identify the associations between outcomes and patient characteristics, offering insights into prevention and future research efforts to improve clinical care.
-
METHODS: This is a single-institution, retrospective chart review including patients of all ages with a Glasgow Coma Scale (GCS) score of 8 or less (measured upon admission). A database was compiled to maximize all available clinical variables. Descriptive statistics and univariable and multivariable regression models were fitted to identify significant associations with outcome (died or discharged).
CONCLUSIONS: Severe TBI was a common condition for injury-related hospital admissions at Mulago Hospital. The capacity for neurosurgery may have explained the relatively lower mortality rate than previously reported from Sub-Saharan Africa. Further investigations are needed. Targeted prevention programs focused on motorcycle users and helmet law enforcement should decrease the incidence of severe TBI.
-
-
RESULTS: One hundred twenty patients were identified between July 1, 2008, and June 30, 2009. The cumulative incidence of admissions is 89 per 100,000. Thirty-one patients died in the hospital, yielding a 25.8% mortality rate. Motorcycle road traffic accident was the leading mechanism of injury, and males ages 15e29 years comprised the predominant demographic (42.5% of patients). Initial GCS, change in GCS score during hospital stay, and the presence of hematoma were strongest predictors of outcome.
-
Key words Africa - Closed head injury - Epidemiology - Head injury - Incidence - Traumatic brain injuries -
Abbreviations and Acronyms CI: Confidence interval CT: Computed tomography GCS: Glasgow Coma Scale LMICs: Low- and middle-income countries OR: Odds ratio RTA: Road traffic accident
WORLD NEUROSURGERY 83 [3]: 269-277, MARCH 2015
INTRODUCTION
F
rom 2015 to 2030, road traffic injuries will become the fifthleading cause of death (47). Current approximations suggest 90% of deaths from injuries occur in low- and middle-income countries (LMICs) (46). Traumatic brain injury (TBI) outcomes are among the worst in LMICs and the incidence is greater; the Sub-Saharan Africa (SSA) region has an incidence of 150170 per 100,000 compared with a global average of 106 per 100,000 (18). The incidence is likely an underestimate because of the typically nascent or nonexistent trauma care infrastructure, which leaves many patients presenting late at referral hospitals. Moreover, TBI deaths are underreported because victims are likely routed first to mortuaries, as 50% of TBI mortality occurs within the first 2 hours from the moment of primary injury (14). Rapid infrastructure growth and motorization in the SSA region is likely to exacerbate this trend. In the United States, TBI has long been recognized as a “silent”
SSA: Sub-Saharan Africa TBI: Traumatic brain injury From the 1Duke University Global Health Institute, Durham, North Carolina, USA; 2Duke University Medical School, Durham, North Carolina, USA; 3Department of Neurosurgery, Mulago Hospital, Kampala, Uganda; and 4Division of Neurosurgery, Department of Surgery, Duke University, Durham, North Carolina, USA To whom correspondence should be addressed: Michael M. Haglund, M.D., Ph.D. [E-mail: [email protected]] Citation: World Neurosurg. (2015) 83, 3:269-277. http://dx.doi.org/10.1016/j.wneu.2014.12.028 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Published by Elsevier Inc.
www.WORLDNEUROSURGERY.org
269
FORUM TU M. TRAN ET AL.
SEVERE TBI AT MULAGO HOSPITAL, UGANDA
epidemic, with immense public health impact with annual costs estimated at USD 76.5 billion (9). Countries in the SSA region carries a disproportionate burden with greater mortality, worsened outcomes, and greater impact from ensuing disability. In terms of capacity, human resources are lacking. There is 1 neurosurgeon per 10,000,000 average for the African continent compared with 102 per 10,000,000 in Europe and 56 per 10,000,000 globally (45). The evidence base for TBI in SSA countries has been enriched within the previous 10 years as the result of increased descriptive studies, establishment of prospective databases, and expansion of sites for the international cohort study, Clinical Randomisation of an Antifibrinolytic in Significant Haemorrhage (CRASH) trials (1, 8, 28, 35, 36, 48). CRASH trial results of 8927 patients predict that a TBI victim in a LMIC has twice the odds of dying after severe TBI (odds ratio [OR] 2.23, 95% confidence interval [CI] 1.513.30). In Uganda, head injury is one of the top 4 common admission diagnoses, contributing to a total 45.3% mortality rate in one study of intensive care unit patients and 75% head injuryspecific mortality rate in another study of all casualty department admissions (17, 24). Head/neck injury is associated with 65% of all injury-related fatalities in urban Uganda (19). The literature from SSA countries specifically related to severe TBI is scarce. We undertook this retrospective analysis to describe distribution of severe TBI at Mulago Hospital, the national referral hospital of Uganda. Secondary objectives were to identify associations between outcomes and patient characteristics in order to potentiate platforms for more effective prevention and inform future research efforts at Mualgo Hospital. METHODS This was a single-institution study at Mulago Hospital, which has an immediate catchment area of the Kampala Capital City (estimated population 1,516,210 in 2014), but patients are referred from throughout the country (44). Injuries are triaged and treated in the Casualty Department; more severe injuries are admitted and many are referred to the Surgery Department. Mulago Hospital has 1500 beds and admits 130,000140,000 patients per year. The Neurosurgery Unit employs 4 of Uganda’s 5 formally trained neurosurgeons, uses a shared elective operating room and casualty operating rooms as necessary, and provides intensive care in a 4- to 8-bed high-dependency surgery unit (16, 24). We performed a retrospective review of medical records for severe TBI patients admitted within a 1-year period (July 1,2008 to June 30, 2009). The inclusion criterion was a confirmed diagnosis of TBI with an initial Glasgow Coma Scale (GCS) score between 3 and 8. Patients with general head injury were excluded from analysis. A database was assembled that included patient demographics, discharge status, initial and greatest GCS score, date of trauma, date of admission, date of treatments, mechanism of injury, pupillary reactivity, medications, neurosurgical procedure (if applicable), and pathologies detected by a diagnostic computed tomography (CT) scan. During the study period, Mulago’s one available CT scanner was functional without interruption, 4 neurosurgeons provided 24-hour coverage, and at least 1 of 5 operating rooms were available for neurosurgical cases; thus, the
270
www.SCIENCEDIRECT.com
management of TBI was consistently at maximum potential for capacity. Pearson c2 tests were used to compare categorical demographic and patient clinical variables effect on outcomes (dichotomized as dead or discharged). Fisher exact tests were used for smaller proportions. Logistic regression was used to determine relationship of 4 exposure variables (GCS admit, presence of hematoma, pupillary reactivity, and age) on the odds of being dead or discharged. The multivariate model was constructed based on the CRASH trials prognostic model for TBI outcomes and selection of the 4 available variables (7). The significance level was set to 0.05 for all analyses. The 1-year cumulative incidence of severe TBI admissions was calculated with the use of total severe TBI admissions and total admissions for the 1-year period (16). All statistical analyses were performed using StataCorp STATA 13 SE (College Station, Texas, USA). All graphics were produced using Microsoft Excel 2010 (Redmond, Washington, USA). Institutional review board approval was granted by the Mulago Hospital Research and Ethics Committee and Duke University Health System Institutional Review Board. RESULTS One hundred twenty patients met the inclusion criteria within the 1-year period (July 1, 2008 to June 30, 2009). Thirty-one of 120 patients died, yielding an in-hospital mortality rate of 25.8%. Demographics are depicted in Table 1 and Figure 1A. Mechanisms of Injury The most common mechanism of injury was road traffic accident (RTA), which contributed to 79% of all severe TBI cases. Thirtyfour percent were motorbike accidents, 25% were pedestrians struck by a vehicle, and 20% were vehicular accidents in which the patient was a passenger or driver in a motor vehicle. Assaults, bicycle accidents, falls, and one animal-related trauma were grouped into an “other” category because they were of low-velocity impact (Table 1, Figure 1B). Figure 1C depicts the mechanism of injury by age and sex. The incidence was greatest in male patients age 15e29 years, in whom motorcycle RTA was the leading mechanism of injury (P < 0.001). For the 0- to 14-year age group, the leading mechanism of injury was being struck as a pedestrian. Incidence of Severe TBI The cumulative incidence of admissions was 89 per 100,000. Sixtysix patients (55.0%) sustained closed-head injuries, and 54 patients (45%) sustained open (mainly depressed skull fractures). The average age was 25.0 (16.7 standard deviation, range 1e87). Of 4 age groups (014, 1529, 3044, and 45 years), the 15- to 29year age group contributed the greatest number of cases with 51 patients (42.5%). Mortality was greatest in the 45 years group at 40.0% (Table 1 and Figure 1A). The male/female ratio was 4.5:1. Median hospitalization length was 7 days (interquartile range, 5e14.5 days). Sixty-three patients (52.5%) presented directly to Mulago after injury, and 57 patients (47.5%) were referred; 71 patients (59.2%) were transferred to Mulago within the same day of primary injury (Table 1).
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2014.12.028
FORUM TU M. TRAN ET AL.
SEVERE TBI AT MULAGO HOSPITAL, UGANDA
Table 1. Characteristics of Patients with Severe TBI at Mulago Hospital (N ¼ 120) Characteristics Patients with severe TBI FY2008/2009 Total Mulago hospitalizations for FY2008/2009
All Patients (% of Total)
Discharged (Row %)
Died (Row %)
c2 and Fisher Statistic, P Value
120
89 (74.2%)
31 (25.8%)
N/A
135,000
Est. admission incidence Severe TBI FY2008/2009
89/100,000
Head injury Closed
66 (55.0%)
52 (73.1%)
14 (26.9%)
Open
54 (45.0%)
37 (54.1%)
17 (45.9%)
No
63 (52.5%)
47 (74.6%)
16 (25.4%)
Yes
57(47.5%)
42 (73.7%)
15 (26.3%)
0
71 (59.2%)
54 (76.1%)
17 (23.9%)
13
35 (29.2%)
28 (80.0%)
7 (20.0%)
410
14 (11.6%)
7 (50.0%)
7 (50.0%)
0.201
Referred 0.909
Days transferred 0.081
Days hospitalized Median [IQR] Range
7 [5e14.5] 0190
Age, years 0e14
33 (27.5%)
23 (69.7%)
10 (30.3%)
15e29
51 (42.5%)
43 (84.3%)
8 (15.7%)
30e44
21 (17.5%)
14 (66.7%)
7 (33.3%)
45
15 (12.5%)
9 (60.0%)
6 (40.0%)
Mean SD [range]
0.153
25.0 16.7 [1e87]
Sex Female
22 (18.3%)
16 (72.8%)
6 (27.2%)
Male
98 (81.7%)
73 (74.5%)
25 (25.5%)
24 (20.0%)
21 (87.5%)
3 (12.5%)
RTA, pedestrian
30 (25.0%)
20 (66.7%)
10 (33.3%)
RTA, motorcycle
41 (34.2%)
27 (65.9%)
14 (34.1%)
0.864
Mechanism of injury* RTA, vehicle
Other
25 (20.8%)
21 (84.0%)
4 (16.0%)
Assault
16 (13.3%)
12 (75.0%)
4 (25.0%)
Bicycle
6 (5.0%)
6 (100.0%)
0.119
0
Fall
2 (1.7%)
2 (100.0%)
0
Struck by cow
1 (0.8%)
1 (100.0%)
0
“Other” includes: falls, assaults, bicycle injuries, and one patient was struck by a cow. RTA pedestrian is a road traffic accident whereby the patient was struck by a moving vehicle while traversing as a pedestrian. RTA vehicle is a road traffic accident whereby the patient suffered from a collision while in a vehicle. RTA motorcycle is a road traffic accident whereby the patient was on or driving a motorcycle and involved in a collision with another moving or stationary object. TBI, traumatic brain injury; N/A, not available; FY, fiscal year; IQR, interquartile range; RTA, road traffic accident. *Explanation for injury mechanism.
WORLD NEUROSURGERY 83 [3]: 269-277, MARCH 2015
www.WORLDNEUROSURGERY.org
271
FORUM TU M. TRAN ET AL.
SEVERE TBI AT MULAGO HOSPITAL, UGANDA
Figure 1. (A) Outcome of severe traumatic brain injury (TBI) by age group. (B) Outcome of severe TBI by mechanism of injury. (C) Mechanism of injury of patients with severe TBI by age and sex. RTA, road traffic accident.
Clinical Variables: GCS and Hematoma Statistically significant differences in proportion of dead and discharged were observed for 3 variables: 1) initial GCS score (P < 0.001); 2) DGCS, which is the difference between the greatest inpatient GCS and initial GCS (P < 0.001); and 3) presence of hematomas detected by CT (P < 0.001) (Table 2 and Figure 2A). Initial GCS measures were as follows: 41 patients had a GCS score between 3 and 5 (61% died) whereas 79 patients scored between 6 and 8, with 6 deaths (7.6% died). Figure 2B depicts number of patients and the change in GCS score during hospitalization: 74 patients who measured an increase of 6 or greater all survived. Only 1 patient with change of 8 died. Figure 2C depicts the greatest GCS score stratified by outcome: 28 patients who scored in the 3 and 8 range all died, 3 of 20 patients (15%) died in the 9 and 12 range, and none of the 72 patients scoring in 1315 range died. There was a strong association between outcome and patients who had hematoma(s): 14, 31, and 5 patients had epidural, subdural, and intracerebral hematomas, respectively. In total, 50 patients (approximately 42%) had at least one hematoma. Figure 3 depicts outcomes for each primary pathology diagnosed by CT. Weighted by frequency, the greatest proportion of deaths was observed in patients with subdural hematomas (51.6%). Intracerebral hematoma had a 60% fatality rate, but it was only observed in 5 patients. No patients with edema as primary pathology indicated died. In a multivariable logistic regression model, initial GCS score was a strong predictor of outcome with an OR of 2.58 (95% CI 1.74e3.82) per 1 point increase initial GCS (Table 3). Patients with one or more hematoma were less likely to be discharged (OR 0.22; 95% CI 0.07e0.73). All findings were statistically significant.
272
www.SCIENCEDIRECT.com
Variables not significantly associated with outcome include age, sex, days elapsed for transfer to Mulago Hospital, length of hospital stay, pupillary reactivity, administration of mannitol, and other pharmacotherapy. DISCUSSION Within the study period, Mulago Hospital’s cumulative incidence of admissions was 89 per 100,000, and compared with other tertiary hospitals in the region, the high incidence suggests a concentration of severe TBI cases. A tertiary university hospital in Durban treated 32 severe TBI patients in 6 months; Nigeria’s Lagos State University Hospital’s treated 32 severe TBI patients in 7 months; Kamuzu Central Hospital in Milawi treated 15 severe TBI cases in 4 months; and 2 major Cape Town hospitals treated 124 patients in 32 months (1, 36, 40, 50). The main surgery referral center in Benin treated 71 severe TBI cases in 8622 admissions in an 18-month period (13). The largest population-based study to date in SSA estimates an incidence in Johannesburg of 316 per 100,000 all severity TBI (33). A systematic review of European studies indicates a population incidence of 235 per 100,000 with 10% of those incident cases classified as severe (43). Compared with other injury presentations at Mulago, severe TBI alone contributes approximately 9% of all injury-related admissions, using data from Hsia et al. (17), who observed 1315 all injuryrelated admissions in a 1-year period. Severe TBI is a significant contributor to morbidity at Mulago Hospital and its catchment area. The concentration may overwhelm the resources of Mulago Hospital, considering it treats a high volume of patients with minor injures as well (17). In our study’s cohort, only half of patients were referred, reinforcing the critical role Mulago Hospital plays in severe TBI care for the capital and the country in general.
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2014.12.028
FORUM TU M. TRAN ET AL.
SEVERE TBI AT MULAGO HOSPITAL, UGANDA
Table 2. Significant Clinical Variables of Patients with Severe TBI at Mulago Hospital (N ¼ 120) All Patients (% of Total)
Discharged (Row %)
Died (Row %)
c2 and Fisher Statistic, P Value
3e5
41 (34.2%)
16 (39.0%)
25 (61.0%)
Distribution and Characteristics of Severe Traumatic Brain Injury at Mulago National Referral Hospital in Uganda Tu M. Tran1, Anthony T. Fuller1,2, Joel Kiryabwire3, John Mukasa3, Michael Muhumuza3, Hussein Ssenyojo3, Michael M. Haglund1,4
OBJECTIVE: Road traffic accidents are a leading cause of injury in low- and middle-income countries, where mortality rates are disproportionately higher. Patients with severe traumatic brain injury (TBI) tend to have very poor outcomes. To reduce the burden from severe TBI, we describe its distribution at Mulago National Referral Hospital (Kampala, Uganda) and identify the associations between outcomes and patient characteristics, offering insights into prevention and future research efforts to improve clinical care.
-
METHODS: This is a single-institution, retrospective chart review including patients of all ages with a Glasgow Coma Scale (GCS) score of 8 or less (measured upon admission). A database was compiled to maximize all available clinical variables. Descriptive statistics and univariable and multivariable regression models were fitted to identify significant associations with outcome (died or discharged).
CONCLUSIONS: Severe TBI was a common condition for injury-related hospital admissions at Mulago Hospital. The capacity for neurosurgery may have explained the relatively lower mortality rate than previously reported from Sub-Saharan Africa. Further investigations are needed. Targeted prevention programs focused on motorcycle users and helmet law enforcement should decrease the incidence of severe TBI.
-
-
RESULTS: One hundred twenty patients were identified between July 1, 2008, and June 30, 2009. The cumulative incidence of admissions is 89 per 100,000. Thirty-one patients died in the hospital, yielding a 25.8% mortality rate. Motorcycle road traffic accident was the leading mechanism of injury, and males ages 15e29 years comprised the predominant demographic (42.5% of patients). Initial GCS, change in GCS score during hospital stay, and the presence of hematoma were strongest predictors of outcome.
-
Key words Africa - Closed head injury - Epidemiology - Head injury - Incidence - Traumatic brain injuries -
Abbreviations and Acronyms CI: Confidence interval CT: Computed tomography GCS: Glasgow Coma Scale LMICs: Low- and middle-income countries OR: Odds ratio RTA: Road traffic accident
WORLD NEUROSURGERY 83 [3]: 269-277, MARCH 2015
INTRODUCTION
F
rom 2015 to 2030, road traffic injuries will become the fifthleading cause of death (47). Current approximations suggest 90% of deaths from injuries occur in low- and middle-income countries (LMICs) (46). Traumatic brain injury (TBI) outcomes are among the worst in LMICs and the incidence is greater; the Sub-Saharan Africa (SSA) region has an incidence of 150170 per 100,000 compared with a global average of 106 per 100,000 (18). The incidence is likely an underestimate because of the typically nascent or nonexistent trauma care infrastructure, which leaves many patients presenting late at referral hospitals. Moreover, TBI deaths are underreported because victims are likely routed first to mortuaries, as 50% of TBI mortality occurs within the first 2 hours from the moment of primary injury (14). Rapid infrastructure growth and motorization in the SSA region is likely to exacerbate this trend. In the United States, TBI has long been recognized as a “silent”
SSA: Sub-Saharan Africa TBI: Traumatic brain injury From the 1Duke University Global Health Institute, Durham, North Carolina, USA; 2Duke University Medical School, Durham, North Carolina, USA; 3Department of Neurosurgery, Mulago Hospital, Kampala, Uganda; and 4Division of Neurosurgery, Department of Surgery, Duke University, Durham, North Carolina, USA To whom correspondence should be addressed: Michael M. Haglund, M.D., Ph.D. [E-mail: [email protected]] Citation: World Neurosurg. (2015) 83, 3:269-277. http://dx.doi.org/10.1016/j.wneu.2014.12.028 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Published by Elsevier Inc.
www.WORLDNEUROSURGERY.org
269
FORUM TU M. TRAN ET AL.
SEVERE TBI AT MULAGO HOSPITAL, UGANDA
epidemic, with immense public health impact with annual costs estimated at USD 76.5 billion (9). Countries in the SSA region carries a disproportionate burden with greater mortality, worsened outcomes, and greater impact from ensuing disability. In terms of capacity, human resources are lacking. There is 1 neurosurgeon per 10,000,000 average for the African continent compared with 102 per 10,000,000 in Europe and 56 per 10,000,000 globally (45). The evidence base for TBI in SSA countries has been enriched within the previous 10 years as the result of increased descriptive studies, establishment of prospective databases, and expansion of sites for the international cohort study, Clinical Randomisation of an Antifibrinolytic in Significant Haemorrhage (CRASH) trials (1, 8, 28, 35, 36, 48). CRASH trial results of 8927 patients predict that a TBI victim in a LMIC has twice the odds of dying after severe TBI (odds ratio [OR] 2.23, 95% confidence interval [CI] 1.513.30). In Uganda, head injury is one of the top 4 common admission diagnoses, contributing to a total 45.3% mortality rate in one study of intensive care unit patients and 75% head injuryspecific mortality rate in another study of all casualty department admissions (17, 24). Head/neck injury is associated with 65% of all injury-related fatalities in urban Uganda (19). The literature from SSA countries specifically related to severe TBI is scarce. We undertook this retrospective analysis to describe distribution of severe TBI at Mulago Hospital, the national referral hospital of Uganda. Secondary objectives were to identify associations between outcomes and patient characteristics in order to potentiate platforms for more effective prevention and inform future research efforts at Mualgo Hospital. METHODS This was a single-institution study at Mulago Hospital, which has an immediate catchment area of the Kampala Capital City (estimated population 1,516,210 in 2014), but patients are referred from throughout the country (44). Injuries are triaged and treated in the Casualty Department; more severe injuries are admitted and many are referred to the Surgery Department. Mulago Hospital has 1500 beds and admits 130,000140,000 patients per year. The Neurosurgery Unit employs 4 of Uganda’s 5 formally trained neurosurgeons, uses a shared elective operating room and casualty operating rooms as necessary, and provides intensive care in a 4- to 8-bed high-dependency surgery unit (16, 24). We performed a retrospective review of medical records for severe TBI patients admitted within a 1-year period (July 1,2008 to June 30, 2009). The inclusion criterion was a confirmed diagnosis of TBI with an initial Glasgow Coma Scale (GCS) score between 3 and 8. Patients with general head injury were excluded from analysis. A database was assembled that included patient demographics, discharge status, initial and greatest GCS score, date of trauma, date of admission, date of treatments, mechanism of injury, pupillary reactivity, medications, neurosurgical procedure (if applicable), and pathologies detected by a diagnostic computed tomography (CT) scan. During the study period, Mulago’s one available CT scanner was functional without interruption, 4 neurosurgeons provided 24-hour coverage, and at least 1 of 5 operating rooms were available for neurosurgical cases; thus, the
270
www.SCIENCEDIRECT.com
management of TBI was consistently at maximum potential for capacity. Pearson c2 tests were used to compare categorical demographic and patient clinical variables effect on outcomes (dichotomized as dead or discharged). Fisher exact tests were used for smaller proportions. Logistic regression was used to determine relationship of 4 exposure variables (GCS admit, presence of hematoma, pupillary reactivity, and age) on the odds of being dead or discharged. The multivariate model was constructed based on the CRASH trials prognostic model for TBI outcomes and selection of the 4 available variables (7). The significance level was set to 0.05 for all analyses. The 1-year cumulative incidence of severe TBI admissions was calculated with the use of total severe TBI admissions and total admissions for the 1-year period (16). All statistical analyses were performed using StataCorp STATA 13 SE (College Station, Texas, USA). All graphics were produced using Microsoft Excel 2010 (Redmond, Washington, USA). Institutional review board approval was granted by the Mulago Hospital Research and Ethics Committee and Duke University Health System Institutional Review Board. RESULTS One hundred twenty patients met the inclusion criteria within the 1-year period (July 1, 2008 to June 30, 2009). Thirty-one of 120 patients died, yielding an in-hospital mortality rate of 25.8%. Demographics are depicted in Table 1 and Figure 1A. Mechanisms of Injury The most common mechanism of injury was road traffic accident (RTA), which contributed to 79% of all severe TBI cases. Thirtyfour percent were motorbike accidents, 25% were pedestrians struck by a vehicle, and 20% were vehicular accidents in which the patient was a passenger or driver in a motor vehicle. Assaults, bicycle accidents, falls, and one animal-related trauma were grouped into an “other” category because they were of low-velocity impact (Table 1, Figure 1B). Figure 1C depicts the mechanism of injury by age and sex. The incidence was greatest in male patients age 15e29 years, in whom motorcycle RTA was the leading mechanism of injury (P < 0.001). For the 0- to 14-year age group, the leading mechanism of injury was being struck as a pedestrian. Incidence of Severe TBI The cumulative incidence of admissions was 89 per 100,000. Sixtysix patients (55.0%) sustained closed-head injuries, and 54 patients (45%) sustained open (mainly depressed skull fractures). The average age was 25.0 (16.7 standard deviation, range 1e87). Of 4 age groups (014, 1529, 3044, and 45 years), the 15- to 29year age group contributed the greatest number of cases with 51 patients (42.5%). Mortality was greatest in the 45 years group at 40.0% (Table 1 and Figure 1A). The male/female ratio was 4.5:1. Median hospitalization length was 7 days (interquartile range, 5e14.5 days). Sixty-three patients (52.5%) presented directly to Mulago after injury, and 57 patients (47.5%) were referred; 71 patients (59.2%) were transferred to Mulago within the same day of primary injury (Table 1).
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2014.12.028
FORUM TU M. TRAN ET AL.
SEVERE TBI AT MULAGO HOSPITAL, UGANDA
Table 1. Characteristics of Patients with Severe TBI at Mulago Hospital (N ¼ 120) Characteristics Patients with severe TBI FY2008/2009 Total Mulago hospitalizations for FY2008/2009
All Patients (% of Total)
Discharged (Row %)
Died (Row %)
c2 and Fisher Statistic, P Value
120
89 (74.2%)
31 (25.8%)
N/A
135,000
Est. admission incidence Severe TBI FY2008/2009
89/100,000
Head injury Closed
66 (55.0%)
52 (73.1%)
14 (26.9%)
Open
54 (45.0%)
37 (54.1%)
17 (45.9%)
No
63 (52.5%)
47 (74.6%)
16 (25.4%)
Yes
57(47.5%)
42 (73.7%)
15 (26.3%)
0
71 (59.2%)
54 (76.1%)
17 (23.9%)
13
35 (29.2%)
28 (80.0%)
7 (20.0%)
410
14 (11.6%)
7 (50.0%)
7 (50.0%)
0.201
Referred 0.909
Days transferred 0.081
Days hospitalized Median [IQR] Range
7 [5e14.5] 0190
Age, years 0e14
33 (27.5%)
23 (69.7%)
10 (30.3%)
15e29
51 (42.5%)
43 (84.3%)
8 (15.7%)
30e44
21 (17.5%)
14 (66.7%)
7 (33.3%)
45
15 (12.5%)
9 (60.0%)
6 (40.0%)
Mean SD [range]
0.153
25.0 16.7 [1e87]
Sex Female
22 (18.3%)
16 (72.8%)
6 (27.2%)
Male
98 (81.7%)
73 (74.5%)
25 (25.5%)
24 (20.0%)
21 (87.5%)
3 (12.5%)
RTA, pedestrian
30 (25.0%)
20 (66.7%)
10 (33.3%)
RTA, motorcycle
41 (34.2%)
27 (65.9%)
14 (34.1%)
0.864
Mechanism of injury* RTA, vehicle
Other
25 (20.8%)
21 (84.0%)
4 (16.0%)
Assault
16 (13.3%)
12 (75.0%)
4 (25.0%)
Bicycle
6 (5.0%)
6 (100.0%)
0.119
0
Fall
2 (1.7%)
2 (100.0%)
0
Struck by cow
1 (0.8%)
1 (100.0%)
0
“Other” includes: falls, assaults, bicycle injuries, and one patient was struck by a cow. RTA pedestrian is a road traffic accident whereby the patient was struck by a moving vehicle while traversing as a pedestrian. RTA vehicle is a road traffic accident whereby the patient suffered from a collision while in a vehicle. RTA motorcycle is a road traffic accident whereby the patient was on or driving a motorcycle and involved in a collision with another moving or stationary object. TBI, traumatic brain injury; N/A, not available; FY, fiscal year; IQR, interquartile range; RTA, road traffic accident. *Explanation for injury mechanism.
WORLD NEUROSURGERY 83 [3]: 269-277, MARCH 2015
www.WORLDNEUROSURGERY.org
271
FORUM TU M. TRAN ET AL.
SEVERE TBI AT MULAGO HOSPITAL, UGANDA
Figure 1. (A) Outcome of severe traumatic brain injury (TBI) by age group. (B) Outcome of severe TBI by mechanism of injury. (C) Mechanism of injury of patients with severe TBI by age and sex. RTA, road traffic accident.
Clinical Variables: GCS and Hematoma Statistically significant differences in proportion of dead and discharged were observed for 3 variables: 1) initial GCS score (P < 0.001); 2) DGCS, which is the difference between the greatest inpatient GCS and initial GCS (P < 0.001); and 3) presence of hematomas detected by CT (P < 0.001) (Table 2 and Figure 2A). Initial GCS measures were as follows: 41 patients had a GCS score between 3 and 5 (61% died) whereas 79 patients scored between 6 and 8, with 6 deaths (7.6% died). Figure 2B depicts number of patients and the change in GCS score during hospitalization: 74 patients who measured an increase of 6 or greater all survived. Only 1 patient with change of 8 died. Figure 2C depicts the greatest GCS score stratified by outcome: 28 patients who scored in the 3 and 8 range all died, 3 of 20 patients (15%) died in the 9 and 12 range, and none of the 72 patients scoring in 1315 range died. There was a strong association between outcome and patients who had hematoma(s): 14, 31, and 5 patients had epidural, subdural, and intracerebral hematomas, respectively. In total, 50 patients (approximately 42%) had at least one hematoma. Figure 3 depicts outcomes for each primary pathology diagnosed by CT. Weighted by frequency, the greatest proportion of deaths was observed in patients with subdural hematomas (51.6%). Intracerebral hematoma had a 60% fatality rate, but it was only observed in 5 patients. No patients with edema as primary pathology indicated died. In a multivariable logistic regression model, initial GCS score was a strong predictor of outcome with an OR of 2.58 (95% CI 1.74e3.82) per 1 point increase initial GCS (Table 3). Patients with one or more hematoma were less likely to be discharged (OR 0.22; 95% CI 0.07e0.73). All findings were statistically significant.
272
www.SCIENCEDIRECT.com
Variables not significantly associated with outcome include age, sex, days elapsed for transfer to Mulago Hospital, length of hospital stay, pupillary reactivity, administration of mannitol, and other pharmacotherapy. DISCUSSION Within the study period, Mulago Hospital’s cumulative incidence of admissions was 89 per 100,000, and compared with other tertiary hospitals in the region, the high incidence suggests a concentration of severe TBI cases. A tertiary university hospital in Durban treated 32 severe TBI patients in 6 months; Nigeria’s Lagos State University Hospital’s treated 32 severe TBI patients in 7 months; Kamuzu Central Hospital in Milawi treated 15 severe TBI cases in 4 months; and 2 major Cape Town hospitals treated 124 patients in 32 months (1, 36, 40, 50). The main surgery referral center in Benin treated 71 severe TBI cases in 8622 admissions in an 18-month period (13). The largest population-based study to date in SSA estimates an incidence in Johannesburg of 316 per 100,000 all severity TBI (33). A systematic review of European studies indicates a population incidence of 235 per 100,000 with 10% of those incident cases classified as severe (43). Compared with other injury presentations at Mulago, severe TBI alone contributes approximately 9% of all injury-related admissions, using data from Hsia et al. (17), who observed 1315 all injuryrelated admissions in a 1-year period. Severe TBI is a significant contributor to morbidity at Mulago Hospital and its catchment area. The concentration may overwhelm the resources of Mulago Hospital, considering it treats a high volume of patients with minor injures as well (17). In our study’s cohort, only half of patients were referred, reinforcing the critical role Mulago Hospital plays in severe TBI care for the capital and the country in general.
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2014.12.028
FORUM TU M. TRAN ET AL.
SEVERE TBI AT MULAGO HOSPITAL, UGANDA
Table 2. Significant Clinical Variables of Patients with Severe TBI at Mulago Hospital (N ¼ 120) All Patients (% of Total)
Discharged (Row %)
Died (Row %)
c2 and Fisher Statistic, P Value
3e5
41 (34.2%)
16 (39.0%)
25 (61.0%)
