Spinal trauma in Tanzania: current management and outcomes

Andreas Leidinger Departments of Neurological Surgery and

Search for other papers by Andreas Leidinger in
jns
Google Scholar
PubMed
Close
 MD
,
Eliana E. Kim Departments of Neurological Surgery and

Search for other papers by Eliana E. Kim in
jns
Google Scholar
PubMed
Close
 BA
,
Rodrigo Navarro-Ramirez Departments of Neurological Surgery and

Search for other papers by Rodrigo Navarro-Ramirez in
jns
Google Scholar
PubMed
Close
 MD
,
Nicephorus Rutabasibwa Muhimbili Orthopaedic Institute, Dar es Salaam, Tanzania

Search for other papers by Nicephorus Rutabasibwa in
jns
Google Scholar
PubMed
Close
 MD
,
Salim R. Msuya Departments of Neurological Surgery and
Muhimbili Orthopaedic Institute, Dar es Salaam, Tanzania

Search for other papers by Salim R. Msuya in
jns
Google Scholar
PubMed
Close
 MD
,
Gulce Askin Healthcare Policy & Research, Weill Cornell Brain and Spine Center, NewYork-Presbyterian Hospital, New York, New York; and

Search for other papers by Gulce Askin in
jns
Google Scholar
PubMed
Close
 MPH
,
Raphael Greving Muhimbili Orthopaedic Institute, Dar es Salaam, Tanzania

Search for other papers by Raphael Greving in
jns
Google Scholar
PubMed
Close
,
Hamisi K. Shabani Muhimbili Orthopaedic Institute, Dar es Salaam, Tanzania

Search for other papers by Hamisi K. Shabani in
jns
Google Scholar
PubMed
Close
 MD, PhD
, and
Roger Härtl Departments of Neurological Surgery and

Search for other papers by Roger Härtl in
jns
Google Scholar
PubMed
Close
 MD
Full access

OBJECTIVE

Spinal trauma is a major cause of disability worldwide. The burden is especially severe in low-income countries, where hospital infrastructure is poor, resources are limited, and the volume of cases is high. Currently, there are no reliable data available on incidence, management, and outcomes of spinal trauma in East Africa. The main objective of this study was to describe, for the first time, the demographics, management, costs of surgery and implants, treatment decision factors, and outcomes of patients with spine trauma in Tanzania.

METHODS

The authors retrospectively reviewed prospectively collected data on spinal trauma patients in the single surgical referral center in Tanzania (Muhimbili Orthopaedic Institute [MOI]) from October 2016 to December 2017. They collected general demographics and the following information: distance from site of trauma to the center, American Spinal Injury Association Impairment Scale (AIS), time to surgery, steroid use, and mechanism of trauma and AOSpine classification and costs. Surgical details and complications were recorded. Primary outcome was neurological status on discharge. The authors analyzed surgical outcome and determined predicting factors for positive outcome.

RESULTS

A total of 180 patients were included and analyzed in this study. The mean distance from site of trauma to MOI was 278.0 km, and the time to admission was on average 5.9 days after trauma. Young males were primarily affected (82.8% males, average age 35.7 years). On admission, 47.2% of patients presented with AIS grade A. Most common mechanisms of injury were motor vehicle accidents (28.9%) and falls from height (32.8%). Forty percent of admitted patients underwent surgery. The mean time to surgery was 33.2 days; 21.4% of patients who underwent surgery improved in AIS grade at discharge (p = 0.030). Overall, the only factor associated with improvement in neurological status was undergoing surgery (p = 0.03) and shorter time to surgery (p = 0.02).

CONCLUSIONS

This is the first study to describe the management and outcomes of spinal trauma in East Africa. Due to the lack of referral hospitals, patients are admitted late after trauma, often with severe neurological deficit. Surgery is performed but generally late in the course of hospital stay. The decision to perform surgery and timing are heavily influenced by the availability of implants and economic factors such as insurance status. Patients with incomplete deficits who may benefit most from surgery are not prioritized. The authors’ results suggest that surgery may have a positive impact on patient outcome. Further studies with a larger sample size are needed to confirm our results. These results provide strong support to implement evidence-based protocols for the management of spinal trauma.

ABBREVIATIONS

AIS = American Spinal Injury Association Impairment Scale; GDP = gross domestic product; LMICs = low- and middle-income countries; LOS = length of stay; MOI = Muhimbili Orthopaedic Institute; MVA = motor vehicle accident; SCI = spinal cord injury; ST = spinal trauma.

OBJECTIVE

Spinal trauma is a major cause of disability worldwide. The burden is especially severe in low-income countries, where hospital infrastructure is poor, resources are limited, and the volume of cases is high. Currently, there are no reliable data available on incidence, management, and outcomes of spinal trauma in East Africa. The main objective of this study was to describe, for the first time, the demographics, management, costs of surgery and implants, treatment decision factors, and outcomes of patients with spine trauma in Tanzania.

METHODS

The authors retrospectively reviewed prospectively collected data on spinal trauma patients in the single surgical referral center in Tanzania (Muhimbili Orthopaedic Institute [MOI]) from October 2016 to December 2017. They collected general demographics and the following information: distance from site of trauma to the center, American Spinal Injury Association Impairment Scale (AIS), time to surgery, steroid use, and mechanism of trauma and AOSpine classification and costs. Surgical details and complications were recorded. Primary outcome was neurological status on discharge. The authors analyzed surgical outcome and determined predicting factors for positive outcome.

RESULTS

A total of 180 patients were included and analyzed in this study. The mean distance from site of trauma to MOI was 278.0 km, and the time to admission was on average 5.9 days after trauma. Young males were primarily affected (82.8% males, average age 35.7 years). On admission, 47.2% of patients presented with AIS grade A. Most common mechanisms of injury were motor vehicle accidents (28.9%) and falls from height (32.8%). Forty percent of admitted patients underwent surgery. The mean time to surgery was 33.2 days; 21.4% of patients who underwent surgery improved in AIS grade at discharge (p = 0.030). Overall, the only factor associated with improvement in neurological status was undergoing surgery (p = 0.03) and shorter time to surgery (p = 0.02).

CONCLUSIONS

This is the first study to describe the management and outcomes of spinal trauma in East Africa. Due to the lack of referral hospitals, patients are admitted late after trauma, often with severe neurological deficit. Surgery is performed but generally late in the course of hospital stay. The decision to perform surgery and timing are heavily influenced by the availability of implants and economic factors such as insurance status. Patients with incomplete deficits who may benefit most from surgery are not prioritized. The authors’ results suggest that surgery may have a positive impact on patient outcome. Further studies with a larger sample size are needed to confirm our results. These results provide strong support to implement evidence-based protocols for the management of spinal trauma.

Spinal trauma (ST) is a major cause of disability worldwide, with an annual incidence estimated at 45–80 cases per million.3,24 It mostly affects people of working age, often generating lifelong disability as well as financial burden on affected families. The impact of ST is especially severe in low-income countries, where hospital infrastructure is poor, resources are limited, and the volume of cases is high. Treatment of ST is cost-intensive and requires significant infrastructure, surgical training, and know-how. Patients in developing countries often cannot afford the high costs of treatment and rehabilitation.

Most developing countries do not offer reliable statistics about the real incidence or status in management of ST.3 In sub-Saharan Africa, fewer than 20 studies concerning 9 countries have been published since 2000. Among these studies, only 2 describe the current situation in East Africa.11,18

Tanzania is a country in East Africa that has a population of approximately 54 million people17 and a gross domestic product (GDP) per capita of $877.5 (all monetary values are reported in USD). In Tanzania, the Muhimbili Orthopaedic Institute (MOI) is the only referral center in the nation where urgent neurosurgical management for ST is available around the clock.

The main objective of this study is to describe, for the first time, the demographics, management, costs of surgery and implants, treatment decision factors, and outcomes of patients with spine trauma in Tanzania.

Methods

We prospectively collected data from patients admitted to MOI in Dar es Salaam, Tanzania, with the diagnosis of ST from October 2016 to December 2017. This study is a retrospective review of the obtained database.

Study Site and Surgical Infrastructure

MOI is the largest referral center for ST in Tanzania. Medical personnel consist of a team of 7 neurosurgeons and trainees and another team of 15 orthopedic surgeons and trainees. Both groups surgically treat ST at MOI.

MOI is equipped with more 120 beds in the general inpatient wards, 16 beds in the ICU, and 10 beds in the emergency department. Due to the high patient load in the emergency department, ST patients are generally transferred to the wards immediately on admission to the emergency department. If elected for surgery, patients are transferred to the surgical area. The ICU is managed by a team of trained anesthesiologists and surgical residents.

Radiological assessment at MOI utilizes radiography, CT scanning, and MRI. Radiography is available within the hospital. CT scanning and MRI are located at the general hospital, a few meters away from MOI. CT and MRI machines are sometimes damaged and not always available. Patients are asked to pay prior to image acquisition.

There are 5 operating rooms at MOI and 2 C-arm fluoroscopy units. Three full instrument sets for spinal decompression and fusion are available. Surgical implants (screws and rods) are supplied directly to MOI by 3 official providers: Medtronic, Pyramid (Johnson & Johnson), and Sharma. Patients are requested to pay upfront for the implants prior to surgery.

Inclusion and Exclusion Criteria

All patients presenting with history of acute ST (no longer than 14 days before admission) and radiological diagnosis of vertebral fracture were included in the study. Pregnant and pediatric patients (< 17 years old) were excluded.

Patient Data

The following variables were recorded at admission: demographics (age, sex, insurance, private vs public status), details on prehospital history (mechanism of injury, distance from site of trauma to MOI, time from trauma to first regional hospital), and time from the first hospital to MOI.

Clinical Outcomes

Neurological assessment was performed by neurosurgical residents immediately on admission and shortly prior to discharge. The American Spinal Injury Association Impairment Scale (AIS) was used. Additionally, medical complications and severity of pressure ulcers, if any, were recorded. Pressure ulcers were classified using the 2016 National Pressure Ulcer Advisor Panel Pressure Injury Staging System.5 Blood pressure monitoring and physiotherapy were documented throughout the patient’s stay until discharge from the hospital.

For patients who underwent surgery, the following variables were collected: time to surgery (from admission to surgery), surgical details (technique, decompressed levels, fused levels, instrumentation used), complications, wound infection, and the cost of surgery.

Radiological Outcomes

Diagnostic images were analyzed by local neurosurgeons or orthopedic surgeons. The AOSpine vertebral fracture classification system was used (https://aospine.aofoundation.org/). Diagnostic discrepancies were resolved by consulting a neuroradiologist from Weill Cornell Medicine.

Insurance and Financial Variables

Public patients are asked to pay upfront for all expenses and are admitted to the common wards. However, a small number of patients benefit from one of 2 modalities of insurance: 1) the National Health Insurance Fund, which insures governmental workers and covers all healthcare for a flat-rate fee on admission; and 2) coverage by private insurance companies. Finally, a few patients are considered “true” private patients because they can afford additional charges to stay in the private wards. For this study, patients benefiting from insurance of any kind were grouped in the “private” category, under the commonality of exemption from any financial delays in receiving any hospital service that set them apart from the patients in the “public” category.

Statistical Analysis

Database management was performed by a single operator, and data quality was maintained by consistent monitoring and analysis of the data. Data were statistically analyzed using IBM SPSS Statistics for Macintosh (version 25, IBM Corp.) and R Version 3.3.1 (R Foundation for Statistical Computing). A p value ≤ 0.05 was considered significant.

Ethics Approval

Our local institutional review board approved the study, and informed consent was obtained from all patients at admission for data collection and prior to surgery.

Results

Epidemiology and Demographics

We prospectively included 180 patients from October 2016 to December 2017. ST affected primarily young males (82.8% males vs 17.2% females). The mean age overall was 35.7 ± 12 years. The most common mechanisms of injury were motor vehicle accidents (MVAs, 28.9%) and falls (32.8%), followed by motorcycle accidents (16.1%), pedestrian accidents (9.4%), and blunt trauma (8.3%). Lesions affected the cervical (43.3%), thoracic (26.7%), and lumbar (30.0%) spinal regions (Table 1).

TABLE 1.

General information on patients who sustained ST in this study

VariableValue
No. of patients180
Sex
 Male149 (82.8)
 Female31 (17.2)
Mean age, yrs ± SD35.7 ± 12
Anatomical level affected
 Cervical78 (43.3)
 Thoracic48 (26.7)
 Lumbar54 (30.0)
Mechanism of injury
 MVA52 (28.9)
 Motorcycle29 (16.1)
 Pedestrian17 (9.4)
 Fall >3 m40 (22.2)
 Fall <3 m19 (10.6)
 Blunt object15 (8.3)
 Sharp object3 (1.7)
 Other5 (1.4)

Values are presented as the number of patients (%) unless stated otherwise.

The average distance to MOI was 278 km (1–1378 km). Figure 1 shows the geographical distribution of sites of trauma; 26.1% of patients came from farther than 500 km away. The mean time from injury to MOI was 5.9 ± 11 days. Among patients who were injured within the Dar es Salaam region (< 100-km radius), the mean time to arrival at MOI was 2.8 ± 9 days, and for those patients transferred from regional hospitals, the time from injury to arrival at MOI was 6 ± 12 days. The latter group of patients would spend on average 4 ± 4 days at regional hospitals before being transferred to MOI.

FIG. 1.
FIG. 1.

Geographical and chronological information from trauma to treatment. A: Geographical distribution of trauma sites of the patients in our sample. Circles are proportional to the number of patients. Radial proximity to the hospital is denoted by color gradients. B: Mean delay in number of days at different time points posttrauma. C: Tabulation of values in panel B. Range is reported in parentheses. Map data from Google Maps, accessed May 31, 2018. Figure is available in color online only.

Length of Hospital Stay

The mean hospital length of stay (LOS) was 43.7 days for patients who underwent surgery and 26.7 days for those who did not undergo surgery (p < 0.001). Shorter hospital LOS was significantly and independently associated with MRI acquisition (p = 0.001). Patients with incomplete injuries, classified by AIS grades B–E, also tended to have a shorter mean hospital LOS (p = 0.118).

The injury level was found to be associated with LOS in the multivariable model. Specifically, thoracic-level injury was associated with a 40% longer LOS compared with cervical injuries (p = 0.023). Patients with AOSpine class B lesions had a 60% longer LOS than those with AOSpine class A lesions (p = 0.035).

During the hospital stay, blood pressure was monitored daily in 15.6% of all patients. Steroids were used in only 1 patient (0.6%) (Table 2).

TABLE 2.

Hospital stay

Value
Mean hospital LOS, days
 All pts (n = 180)33.2
 Surgical pts (n = 72)43.7
 Nonsurgical pts (n = 108)26.7
Associations
 Shorter stay
  Incomplete injuryp = 0.118*
  MRI acquisitionp = 0.001
 Longer stay
  Thoracic injuryp = 0.023
  AOSpine class Bp = 0.035
BP monitoring
 All pts (n = 180)28 (15.6%)
 ICU pts (n = 34)6 (17.6%)
 Ward pts (n = 146)22 (15.1%)
Use of steroids1 (0.6%)

BP = blood pressure; pts = patients.

Bivariate analysis.

Multivariate analysis.

Clinical Data

On admission, 47.2% of patients presented with AIS grade A injuries, 15.0% with grade B, 9.4% with grade C, 11.7% with grade D (incomplete lesions), and 16.7% with grade E (intact patients). The thoracic segment had the lowest proportion (35.4%) of incomplete lesions, whereas the lumbar segment had the highest (66.7%) (Table 3).

TABLE 3.

Neurological severity of injury and AOSpine fracture classification on admission by level of injury

No. of Patients
CervicalThoracicLumbarTotalUnderwent Surgery (%)
AOSpine class
 A3220267815 (19.2%)
 B04101412 (85.7%)
 C4225138041 (51.3%)
 Total744949172*68 (39.5%)*
AIS grade
 A3631188538 (44.7%)
 B12782713 (48.1%)
 C1214176 (35.3%)
 D10382110 (47.6%)
 E8616305 (16.7%)
 Total78485418072 (40.0%)

The quality of imaging data was not sufficient to make an accurate AOSpine classification for 8 patients.

Type of Fracture

Table 3 summarizes all fractures according to the AOSpine fracture classification.

Neurological Assessment

Of the patients who underwent surgery, 21.4% had improvement in their AIS grade at discharge; 8.2% of patients who did not undergo surgery had improvement in AIS grade (p = 0.030). Mortality in patients who underwent surgery was 1.6% versus 16.1% in patients who did not undergo surgery (p = 0.07); 93.8% of patients who died had cervical injury and 87.5% had a complete injury (p = 0.018). Factors associated with higher mortality by bivariate associations were shorter time to surgery (p = 0.031), shorter time from trauma to MOI (p = 0.016), cervical injury (p = 0.001), and ICU stay (p < 0.001).

Factors significantly associated with positive change in AIS grade were time from admission to surgery (p = 0.02) and undergoing surgery (p = 0.03). The independent effect of surgery on AIS grade improvement could not be isolated by multivariate analysis due to the small sample size of patients who had an improvement in AIS grade at discharge. However, among patients with incomplete injuries who underwent surgery (34 patients, 47.2%), 10 patients (29.4%) showed improvement after surgery. Overall, 82.4% of patients experienced no favorable change after surgical management.

Surgery

Forty percent of all patients underwent surgery. Forty-five percent of all patients with complete injury had surgery at any point in time, compared with 36% of those with incomplete injury (p = 0.198). Among patients who did not undergo surgery, 43.5% were AIS grade A (p < 0.001).

The mean time to surgery was 33.2 days. Time to surgery for patients with incomplete injuries was 36 days versus 30 days for those with complete injuries, and this difference was not statistically significant (p = 0.344). Ninety percent of patients with incomplete injuries who arrived within 24 hours posttrauma did not undergo surgery within 1 week.

Cervical Injuries

Twenty-three patients underwent surgery for cervical injuries. The anterior approach was performed in 12 patients, among whom 3 had anterior cervical discectomy and fusion, 6 had single-level corpectomy, and 3 had multiple-level corpectomy (Table 4). Surgeons at MOI prefer autograft over allograft, mainly due to limited availability of the latter. Iliac crest grafting was used in 75% of these patients (n = 9). For the remaining patients, cages were used. Anterior plating was used only in multiple-level corpectomy. In the remaining 11 cases, a posterior approach with lateral mass screw fixation and laminectomy was performed.

TABLE 4.

Type of surgery performed by approach and segment

No. of Patients
CervicalThoracicLumbarTotal
Posterior approach
 Laminectomy only0011
 Laminectomy & fusion w/ instrumentation11242358
Anterior approach
 Discectomy & fusion3003
 Corpectomy (1 level)6006
 Corpectomy (multiple levels)3003
Total23242471

Thoracic and Lumbar Injuries

Forty-seven patients with injury at the thoracic or lumbar level were treated with posterior fusion and stabilization using pedicle screws and rods. In 72.3% of these cases, fixation included one level above and below the fractured vertebra. Only 1 patient, with an L1 AOSpine class A1 fracture of the anterior vertebral aspect and spinal cord contusion and presenting with AIS grade A, underwent simple laminectomy with no fusion. Table 4 gives a detailed account of the performed procedures.

Imaging Acquisition

Radiography was performed in 88.8% of patients on admission for initial assessment of fracture. Preoperative radiography and MRI were performed in 61.1% of all patients who underwent surgery. Figure 2 shows a Venn diagram depicting the pattern of radiograph acquisition in these patients.

FIG. 2.
FIG. 2.

Venn diagram showing number of patients by type of imaging done. Two patients did not undergo imaging.

Costs

Overall, 11.7% of patients were admitted under “private” status. Private patients undergoing surgery (17.7% of all surgical patients) paid on average $2322.20 compared with public patients, who paid on average $873. Private patients tended to have a shorter time from admission to surgery than public patients (18.9 days vs 34.2 days for public patients, p = 0.057). A significantly higher proportion of private patients (25%) underwent surgery within 1 week of arrival, compared with 6.7% of public patients (p = 0.05). In general, private patients were more likely to undergo surgery (57%) than public patients (37.7%, p = 0.088).

Postoperative Care, Complications, and Rehabilitation

Thirty-four (18.9%) patients presented with pressure ulcers, and we recorded 2 cases of surgical wound infections. The presence of pressure ulcers was associated with the absence of postsurgical improvement (p = 0.073). Surgical wound infection was present in 2 of 78 surgical cases. In both cases, empirical treatment was used.

Physiotherapy was done for 87.5% of patients who underwent surgery as opposed to 48.1% of patients who did not undergo surgery. Bedridden patients were visited 3 times per week, and passive mobilization and positional changes were performed. Patients with some mobility (even those in wheelchairs) were transported daily to the rehabilitation unit where they performed active exercises, depending on tolerance.

Discussion

Epidemiology and Demographics

In this study, distribution by sex and age corresponds to previous global and African reports.3,4 Unlike other African countries, such as South Africa and Zimbabwe,4,6,8,10,22 violence is not a representative cause of ST in Tanzania. Although MVA is the most common mechanism of injury described in the literature, a high prevalence of occupational “falls” in all Tanzanian studies was noted.11,18 These were mostly documented as falls from coconut trees.

Prehospital Management and Delays in Transportation

We received patients from all Tanzanian regions (Fig. 1); 79.4% of all patients were first admitted to regional hospitals. However, the average referral time from these hospitals to MOI was 5 days (± 10). Regional hospitals are generally under-resourced and ill-equipped to deal with patients with ST, and it seems plausible that lack of education and expertise on ST in these local health centers hampers prompt referral to the major treatment center (MOI). We do not know how many patients died at these local hospitals or during transportation; hence, we cannot estimate with certainty the national prevalence and incidence of ST.

In our sample, even though time from trauma to admission at MOI was 6 ± 11 days, it only accounts for 17.6% of the total time from trauma to surgery, as there were further delays after admission.

Initial Hospital Management and Neurological Assessment

On arrival at the emergency department, basic initial workup included hemogram, ionogram, and blood grouping. The patients were assigned to the different surgical teams (neurosurgery or orthopedics) based on who initially admitted them at the emergency department.

In the wards, patients were assessed daily by residents and visited weekly by the assigned surgical team. No protocol exists to aid surgeons with the scheduling of patients undergoing surgery.

The AIS grade distribution of our series is similar to that in the only previous description for Tanzania.18 On admission, 47.2% of patients presented with complete (AIS grade A) injuries. In other studies from sub-Saharan Africa, as much as 77% of the patients with ST had a complete lesion (AIS grade A).4,7

Admission to the ICU was reserved for patients presenting with hemodynamic instability or associated abdominal, thoracic, or cranial injuries. No invasive blood pressure monitoring was available, and the use of central lines is reserved for very specific cases. Thirty-four patients (18.9%) required ICU admission.

Preoperative Imaging

In our study, the majority of patients underwent radiography on admission (88.8%). Vertebral fractures were seen on nearly all (96.4%) of the acquired radiographs. Radiography was complemented mainly by MRI (n = 96) rather than CT (n = 18), and rarely both (n = 7), even though CT could provide a better image for AOSpine classification of vertebral fractures.16,20

In MOI, there is no clear decision-making algorithm in electing imaging studies in ST patients. In most cases, the surgeon’s preference and the patient’s ability to pay in advance will determine the imaging modality. The latter also determines the waiting time for imaging acquisition. At MOI, MRI alone seems to be preferred over CT alone even at a higher cost, because it offers more information in a single study, providing information about soft-tissue structures (e.g., disc, blood, ligaments, cord compression) while also providing some information on spine alignment and structure. Surgeons at MOI choose MRI over CT most of the time and argue that the information regarding bone anatomy and fracture characteristics can be obtained from a much cheaper plain radiograph.

Time to Surgery

In our study, the only factor related to shorter time to surgery was insurance status (p = 0.057). Twenty-five percent of all private patients underwent surgery within 1 week, compared with 6.7% of public patients (p = 0.05). Being a private patient was also related to shorter times from admission to surgery (p = 0.057). It seems that the economic resources available to the private patients reduce some of the delays before surgery.

At MOI, there is no established consensus or protocol about surgical priority based on the AIS grade. Some surgeons operate first on patients with the most severe neurological deficit, while others prioritize those with incomplete and progressive injuries.

Surgery

The surgical indication for spinal cord injury (SCI) was determined by the treating neurosurgeons and orthopedic surgeons independently. Table 3 shows the total number of patients undergoing surgery depending on AIS score on admission and AOSpine fracture classification.

Cervical spine trauma was treated using posterior and anterior approaches with similar frequency. The anterior approach is performed by a few trained neurosurgeons at MOI. The use of interbody cages is very limited, and usually iliac crest bone is grafted. Cages were only used in 3 patients. Plating is never done for single-level fusion. To our understanding, shortening the length of the procedure and reducing the cost of the instrumentation also makes the intervention financially more feasible for the patient and reduces delays.

Surgical Outcome and Mortality

Agreeing with other authors,9,13,21 AIS grade A patients consistently experience little to no improvement. In contrast, and despite the long delays to surgery, among patients presenting with incomplete injuries who underwent surgery at any point (n = 34), 29.4% showed improvement prior to discharge.

In our study, 21.4% of patients who underwent surgery had improvement in AIS grade at discharge (p = 0.030), which brings the results from MOI close to those of the only other published study analyzing surgical outcomes in sub-Saharan Africa. Ojo et al. found a 25.7% improvement rate in SCI patients 6 months after spinal surgery in a study conducted in Nigeria in 2017.15 Figure 3 shows the improvement in AIS grade in both surgical and nonoperative groups.

FIG. 3.
FIG. 3.

AIS outcome grades of surgical patients and nonsurgical patients. Figure is available in color online only.

As shown in Table 5, the only factors associated with improvement in AIS score from admission to discharge were surgery (p < 0.003) and time to surgery (p < 0.020), which shows the importance of reducing time to surgery and improving the surgical indication by homogenizing selection criteria and surgical techniques at MOI. No significant difference was observed in outcomes when comparing patients from the Dar es Salaam area (< 100 km) against longer-distance referrals.

TABLE 5.

Associations with 4 different outcome measures:surgery, time to surgery, improvement in AIS grade, and mortality

p Value
Op (yes/no; bivariate)Time to Op (multivariate)Improvement in AIS Grade (bivariate)Mortality (bivariate)
Age0.108NS0.0920.326
Sex0.987NS0.3820.168
Mechanism of injury0.593NS0.6430.678
Referral time0.229NC0.2430.016*
Time to opNANA0.020*0.031*
Complete injury (yes/no)0.281Not significant0.1060.018*
Surgery (yes/no)NANA0.003*0.070*
Private status0.155Significant*0.179>0.99
AOSpine class<0.001*NS0.2680.176
ICU stay>0.99NS0.372<0.001*

NA = not applicable; NC = not correlated; NS = not significant.

Statistically significant.

As described by other authors, mortality in our study was strongly associated with cervical injuries (p = 0.001) and complete lesions (p = 0.018).1,2,11,18 However, in our study the difference in mortality between patients undergoing surgery and patients managed conservatively could be partially explained by a selection bias, since critically ill patients were admitted to the ICU and surgery was delayed until hemodynamic stability was achieved.

Hospital Stay

In addition to the acquisition of presurgical imaging and the prolonged time to surgery, we observed that 41.4% of the patients who underwent surgery and had no medical complications stayed in the hospital for longer than 10 days after surgery. Such delays in discharge were often associated with financial incapacity of paraplegic and tetraplegic patients to purchase wheelchairs. The social welfare department is constantly involved in the acquisition of such rehabilitative equipment for SCI patients.

Costs

The average public patient undergoing surgery paid $873 throughout the stay at MOI. Considering the mean total income per earner (both formal and informal sectors) in urban Dar es Salaam ($25) and in rural Tanzania ($10),12 cost of care certainly imposes a large burden on the patients and families and is a major cause of delay since payment in advance is required for diagnostic imaging, medication, and surgery. As revealed by Nyberger et al., 65.8% of the Tanzanian households that have a member who requires surgery are at risk of catastrophic expenditures, and 85.5% could experience impoverishing expenditures.14,25

A few studies have explored the cost of hospitalization for injuries in low- and middle-income countries (LMICs). Only 14 studies have been conducted among the 144 LMICs defined by the World Bank.19,23 In some LMICs, the cost of injury was 15% of the GDP per capita, and those studies, which included surgical and recovery costs, estimated a 14-fold increase in this number, reaching 97% of GDP per capita.23 This devastating financial impact explains why in East Africa costs are often considered the tie breaker when decisions are made regarding how and where to prioritize health interventions on individuals.

Limitations

The scope of our study is limited to the pool of patients who survived long enough to eventually be referred to MOI and thus does not allow us to calculate the total number of ST patients in Tanzania or its absolute mortality and morbidity.

We did not collect information concerning the type of transportation to MOI, which limits our understanding of the characteristics and area of coverage of medicalized transportation in Tanzania.

Conclusions

This is the first study describing the management and outcomes of ST in East Africa. There is no hospital-based protocol for the management of ST in Tanzania. Patients are admitted late after trauma, often with severe neurological deficits. Surgery is performed but generally late in the course. The decision to perform surgery and its timing are heavily influenced by surgeon preference, economic factors such as availability of implants, and patient insurance status. Patients with incomplete deficits are left untreated or treated late.

Our results, however, suggest that surgery has a positive impact on patient outcome. These results provide strong support to work with the local surgeons and the hospitals in order to implement treatment protocols that prioritize ICU admission, hemodynamic monitoring, and early surgical management, especially of patients with incomplete deficits. Education and more collaboration with surgical societies and academic groups dedicated to global surgery are also necessary. These efforts are currently underway but will also require healthcare-related changes that are beyond our direct influence, such as insurance coverage and availability and pricing of surgical implants. Ongoing studies with a larger sample size need to be performed to confirm our results and document treatment and outcome changes over time.

Acknowledgments

We would like to acknowledge the Leonard & Evelyn Lauder Foundation, Eric Javits Family Foundation, William Beecher Scoville Foundation, and the Colin McDonald Family, who have supported the Tanzania Neurosurgery Project at Weill Cornell.

Disclosures

Dr. Härtl: consulting fees from AOSpine, Brainlab, DePuy Synthes, and Lanx.

This publication will be included in the PhD dissertation of Andreas Leidinger, MD, at Universidad Autónoma de Barcelona, Barcelona, Spain.

Author Contributions

Conception and design: Leidinger, Navarro-Ramirez, Shabani. Acquisition of data: Leidinger, Kim, Msuya, Greving. Analysis and interpretation of data: Leidinger, Kim. Drafting the article: Leidinger. Critically revising the article: Härtl, Kim, Navarro-Ramirez, Rutabasibwa, Shabani. Reviewed submitted version of manuscript: Härtl, Leidinger, Navarro-Ramirez, Rutabasibwa, Shabani. Approved the final version of the manuscript on behalf of all authors: Härtl. Statistical analysis: Leidinger, Kim, Askin. Administrative/technical/material support: Härtl, Kim, Navarro-Ramirez. Study supervision: Härtl, Shabani.

References

  • 1

    Ametefe MK, Bankah PE, Yankey KP, Akoto H, Janney D, Dakurah TK: Spinal cord and spine trauma in a large teaching hospital in Ghana. Spinal Cord 54:11641168, 2016

  • 2

    Biluts H, Abebe M, Laeke T, Tirsit A, Belete A: Pattern of spine and spinal cord injuries in Tikur Anbessa Hospital, Ethiopia. Ethiop Med J 53:7582, 2015

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Cripps RA, Lee BB, Wing P, Weerts E, Mackay J, Brown D: A global map for traumatic spinal cord injury epidemiology: towards a living data repository for injury prevention. Spinal Cord 49:493501, 2011

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4

    Draulans N, Kiekens C, Roels E, Peers K: Etiology of spinal cord injuries in Sub-Saharan Africa. Spinal Cord 49:11481154, 2011

  • 5

    Edsberg LE, Black JM, Goldberg M, McNichol L, Moore L, Sieggreen M: Revised National Pressure Ulcer Advisory Panel Pressure Injury Staging System: Revised Pressure Injury Staging System. J Wound Ostomy Continence Nurs 43:585597, 2016

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6

    Hart C, Williams E: Epidemiology of spinal cord injuries: a reflection of changes in South African society. Paraplegia 32:709714, 1994

  • 7

    Igun GO, Obekpa OP, Ugwu BT, Nwadiaro HC: Spinal injuries in the Plateau State, Nigeria. East Afr Med J 76:7579, 1999

  • 8

    Joseph C, Delcarme A, Vlok I, Wahman K, Phillips J, Nilsson Wikmar L: Incidence and aetiology of traumatic spinal cord injury in Cape Town, South Africa: a prospective, population-based study. Spinal Cord 53:692696, 2015

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Lehre MA, Eriksen LM, Tirsit A, Bekele S, Petros S, Park KB, et al.: Outcome in patients undergoing surgery for spinal injury in an Ethiopian hospital. J Neurosurg Spine 23:772779, 2015

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 10

    Levy LF, Makarawo S, Madzivire D, Bhebhe E, Verbeek N, Parry O: Problems, struggles and some success with spinal cord injury in Zimbabwe. Spinal Cord 36:213218, 1998

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11

    Moshi H, Sundelin G, Sahlen KG, Sörlin A: Traumatic spinal cord injury in the north-east Tanzania—describing incidence, etiology and clinical outcomes retrospectively. Glob Health Action 10:1355604, 2017

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    National Bureau of Statistics (Tanzania): Employment and Labour Census 2012. Dar es Salaam, Tanzania: National Bureau of Statistics, Ministry of Finances, 2013 (http://www.nbs.go.tz/) [Accessed January 18, 2019]

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13

    Nwankwo OE, Uche EO: Epidemiological and treatment profiles of spinal cord injury in southeast Nigeria. Spinal Cord 51:448452, 2013

  • 14

    Nyberger K, Jumbam DT, Dahm J, Maongezi S, Makuwani A, Kapologwe NA, et al.: The situation of safe surgery and anaesthesia in Tanzania: a systematic review. World J Surg 43:2435, 2019

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Ojo OA, Poluyi EO, Owolabi BS, Kanu OO, Popoola MO: Surgical decompression for traumatic spinal cord injury in a tertiary center. Niger J Clin Pract 20:14551460, 2017

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16

    Parizel PM, van der Zijden T, Gaudino S, Spaepen M, Voormolen MH, Venstermans C, et al.: Trauma of the spine and spinal cord: imaging strategies. Eur Spine J 19 (Suppl 1):S8S17, 2010

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 17

    Rajan D: Situational analysis of the health sector, in Schmets G, Rajan D (eds): Strategizing National Health in the 21st Century: A Handbook. Geneva: World Health Organization, 2016 (http://www.who.int/healthsystems/publications/nhpsp-handbook-ch3/en/) [Accessed January 18, 2019]

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 18

    Rashid SM, Jusabani MA, Mandari FN, Dekker MCJ: The characteristics of traumatic spinal cord injuries at a referral hospital in Northern Tanzania. Spinal Cord Ser Cases 3:17021, 2017

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 19

    Sangowawa AO, Owoaje ET, Ekanem SE, Faseru B, Adekunle BJ: Economic costs of motorcycle injury among crash-involved commercial motorcyclists in Oyo State, Nigeria. Afr J Med Med Sci 40:385391, 2011

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 20

    Silberstein M, Tress BM, Hennessy O: A comparison between M.R.I. and C.T. in acute spinal trauma. Australas Radiol 36:192197, 1992

  • 21

    Umaru H, Ahidjo A: Pattern of spinal cord injury in Maiduguri, North Eastern Nigeria. Niger J Med 14:276278, 2005

  • 22

    Velmahos GC, Degiannis E, Hart K, Souter I, Saadia R: Changing profiles in spinal cord injuries and risk factors influencing recovery after penetrating injuries. J Trauma 38:334337, 1995

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 23

    Wesson HK, Boikhutso N, Bachani AM, Hofman KJ, Hyder AA: The cost of injury and trauma care in low- and middle-income countries: a review of economic evidence. Health Policy Plan 29:795808, 2014

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 24

    Winkler AS, Tluway A, Slottje D, Schmutzhard E, Härtl R: The pattern of neurosurgical disorders in rural northern Tanzania: a prospective hospital-based study. World Neurosurg 73:264269, 2010

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25

    The World Factbook (USA): Tanzania. Washington, DC: Central Intelligence Agency, 2018 (https://www.cia.gov/library/publications/the-world-factbook/geos/print_tz.html) [Accessed January 18, 2019]

    • PubMed
    • Search Google Scholar
    • Export Citation
  • Collapse
  • Expand
  • FIG. 1.

    Geographical and chronological information from trauma to treatment. A: Geographical distribution of trauma sites of the patients in our sample. Circles are proportional to the number of patients. Radial proximity to the hospital is denoted by color gradients. B: Mean delay in number of days at different time points posttrauma. C: Tabulation of values in panel B. Range is reported in parentheses. Map data from Google Maps, accessed May 31, 2018. Figure is available in color online only.

  • FIG. 2.

    Venn diagram showing number of patients by type of imaging done. Two patients did not undergo imaging.

  • FIG. 3.

    AIS outcome grades of surgical patients and nonsurgical patients. Figure is available in color online only.

  • 1

    Ametefe MK, Bankah PE, Yankey KP, Akoto H, Janney D, Dakurah TK: Spinal cord and spine trauma in a large teaching hospital in Ghana. Spinal Cord 54:11641168, 2016

  • 2

    Biluts H, Abebe M, Laeke T, Tirsit A, Belete A: Pattern of spine and spinal cord injuries in Tikur Anbessa Hospital, Ethiopia. Ethiop Med J 53:7582, 2015

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Cripps RA, Lee BB, Wing P, Weerts E, Mackay J, Brown D: A global map for traumatic spinal cord injury epidemiology: towards a living data repository for injury prevention. Spinal Cord 49:493501, 2011

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4

    Draulans N, Kiekens C, Roels E, Peers K: Etiology of spinal cord injuries in Sub-Saharan Africa. Spinal Cord 49:11481154, 2011

  • 5

    Edsberg LE, Black JM, Goldberg M, McNichol L, Moore L, Sieggreen M: Revised National Pressure Ulcer Advisory Panel Pressure Injury Staging System: Revised Pressure Injury Staging System. J Wound Ostomy Continence Nurs 43:585597, 2016

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6

    Hart C, Williams E: Epidemiology of spinal cord injuries: a reflection of changes in South African society. Paraplegia 32:709714, 1994

  • 7

    Igun GO, Obekpa OP, Ugwu BT, Nwadiaro HC: Spinal injuries in the Plateau State, Nigeria. East Afr Med J 76:7579, 1999

  • 8

    Joseph C, Delcarme A, Vlok I, Wahman K, Phillips J, Nilsson Wikmar L: Incidence and aetiology of traumatic spinal cord injury in Cape Town, South Africa: a prospective, population-based study. Spinal Cord 53:692696, 2015

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Lehre MA, Eriksen LM, Tirsit A, Bekele S, Petros S, Park KB, et al.: Outcome in patients undergoing surgery for spinal injury in an Ethiopian hospital. J Neurosurg Spine 23:772779, 2015

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 10

    Levy LF, Makarawo S, Madzivire D, Bhebhe E, Verbeek N, Parry O: Problems, struggles and some success with spinal cord injury in Zimbabwe. Spinal Cord 36:213218, 1998

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11

    Moshi H, Sundelin G, Sahlen KG, Sörlin A: Traumatic spinal cord injury in the north-east Tanzania—describing incidence, etiology and clinical outcomes retrospectively. Glob Health Action 10:1355604, 2017

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    National Bureau of Statistics (Tanzania): Employment and Labour Census 2012. Dar es Salaam, Tanzania: National Bureau of Statistics, Ministry of Finances, 2013 (http://www.nbs.go.tz/) [Accessed January 18, 2019]

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13

    Nwankwo OE, Uche EO: Epidemiological and treatment profiles of spinal cord injury in southeast Nigeria. Spinal Cord 51:448452, 2013

  • 14

    Nyberger K, Jumbam DT, Dahm J, Maongezi S, Makuwani A, Kapologwe NA, et al.: The situation of safe surgery and anaesthesia in Tanzania: a systematic review. World J Surg 43:2435, 2019

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Ojo OA, Poluyi EO, Owolabi BS, Kanu OO, Popoola MO: Surgical decompression for traumatic spinal cord injury in a tertiary center. Niger J Clin Pract 20:14551460, 2017

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16

    Parizel PM, van der Zijden T, Gaudino S, Spaepen M, Voormolen MH, Venstermans C, et al.: Trauma of the spine and spinal cord: imaging strategies. Eur Spine J 19 (Suppl 1):S8S17, 2010

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 17

    Rajan D: Situational analysis of the health sector, in Schmets G, Rajan D (eds): Strategizing National Health in the 21st Century: A Handbook. Geneva: World Health Organization, 2016 (http://www.who.int/healthsystems/publications/nhpsp-handbook-ch3/en/) [Accessed January 18, 2019]

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 18

    Rashid SM, Jusabani MA, Mandari FN, Dekker MCJ: The characteristics of traumatic spinal cord injuries at a referral hospital in Northern Tanzania. Spinal Cord Ser Cases 3:17021, 2017

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 19

    Sangowawa AO, Owoaje ET, Ekanem SE, Faseru B, Adekunle BJ: Economic costs of motorcycle injury among crash-involved commercial motorcyclists in Oyo State, Nigeria. Afr J Med Med Sci 40:385391, 2011

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 20

    Silberstein M, Tress BM, Hennessy O: A comparison between M.R.I. and C.T. in acute spinal trauma. Australas Radiol 36:192197, 1992

  • 21

    Umaru H, Ahidjo A: Pattern of spinal cord injury in Maiduguri, North Eastern Nigeria. Niger J Med 14:276278, 2005

  • 22

    Velmahos GC, Degiannis E, Hart K, Souter I, Saadia R: Changing profiles in spinal cord injuries and risk factors influencing recovery after penetrating injuries. J Trauma 38:334337, 1995

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 23

    Wesson HK, Boikhutso N, Bachani AM, Hofman KJ, Hyder AA: The cost of injury and trauma care in low- and middle-income countries: a review of economic evidence. Health Policy Plan 29:795808, 2014

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 24

    Winkler AS, Tluway A, Slottje D, Schmutzhard E, Härtl R: The pattern of neurosurgical disorders in rural northern Tanzania: a prospective hospital-based study. World Neurosurg 73:264269, 2010

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25

    The World Factbook (USA): Tanzania. Washington, DC: Central Intelligence Agency, 2018 (https://www.cia.gov/library/publications/the-world-factbook/geos/print_tz.html) [Accessed January 18, 2019]

    • PubMed
    • Search Google Scholar
    • Export Citation

Metrics

All Time Past Year Past 30 Days
Abstract Views 3595 381 0
Full Text Views 537 240 70
PDF Downloads 555 197 38
EPUB Downloads 0 0 0