File
Hossain M1, Islam S2, Rahman A2, Herbert R3, Glinsky J1, Harvey L1
1The University of Sydney, John Walsh Centre for Rehabilitation Research, Sydney, Australia, 2Centre for the Rehabilitation of the Paralysed, Physiotherapy, Dhaka, Bangladesh, 3Neuroscience Research Australia, Neuroscience, Sydney, Australia
Background: Survival following spinal cord injury (SCI) is a major problem in Bangladesh and other low- and middle-income countries (LMIC) - many people dying prematurely after they are discharged from a hospital. However, there are no accurate estimates of the magnitude of the problem. Our study is unique because we followed a cohort for 5 years from a specialized hospital in Bangladesh from the time of discharge.
Purpose: The purpose of this study was to determine the survival of people with SCI in Bangladesh 5 years after discharge from hospital and to develop a prediction model to identify those at risk of premature death.
Methods: Patients discharged from a hospital in Bangladesh in 2011 were identified using medical records. Participants or their family members were then contacted 5 years after discharge to determine vital status or date of death. Survival from time of discharge was estimated with Kaplan-Meier curves. A bootstrap variable selection was used to identify a parsimonious set of predictors from 5 candidate predictors (age, sex, mobility, type of SCI, the cause of SCI). The retained predictors were included in the final logistic model of the odds of death within 5 years of discharge. The model was internally validated.
Results: Three hundred and forty-five people were identified from the medical records and discharged in 2011. Five years later, 342 (99%) were contacted. Of these, 74 (22%) had died (survival=78%; 95% CI 74% to 82%). Only two predictors were retained in the final prediction model (age and mobility). The odds of dying increased by a factor of 1.6 (95% CI, 1.3 to 2.0) with every decade of age and by a factor of 12.6 (95% CI, 4.8 to 32.9) if wheelchair-dependent. The final multivariate model included mobility at discharge and age as the predictors. The model had good calibration and discrimination.
Conclusion(s): The risk of dying after discharge from hospital with SCI in Bangladesh is high, especially amongst older and wheelchair-dependent people. A simple prediction model discriminates those at risk of death 5 years following discharge.
Implications: This study provides accurate estimates of survival following SCI in Bangladesh. The data are important because until now there were no estimates of survival and the extent of the problem was not known. With our data, we are in a better position to lobby for better services to support people with SCI following discharge in an effort to reduce the mortality rates. In addition, our prediction model has important implications for identifying at the time of discharge those at most risk of premature death. These patients can be targeted for additional community-based follow-up and support. This sort of targeted approach is critical in countries where health resources are scarce.
Keywords: Spinal cord injury, survival, prediction model
Funding acknowledgements: Funded by the University of Sydney, Australia
Purpose: The purpose of this study was to determine the survival of people with SCI in Bangladesh 5 years after discharge from hospital and to develop a prediction model to identify those at risk of premature death.
Methods: Patients discharged from a hospital in Bangladesh in 2011 were identified using medical records. Participants or their family members were then contacted 5 years after discharge to determine vital status or date of death. Survival from time of discharge was estimated with Kaplan-Meier curves. A bootstrap variable selection was used to identify a parsimonious set of predictors from 5 candidate predictors (age, sex, mobility, type of SCI, the cause of SCI). The retained predictors were included in the final logistic model of the odds of death within 5 years of discharge. The model was internally validated.
Results: Three hundred and forty-five people were identified from the medical records and discharged in 2011. Five years later, 342 (99%) were contacted. Of these, 74 (22%) had died (survival=78%; 95% CI 74% to 82%). Only two predictors were retained in the final prediction model (age and mobility). The odds of dying increased by a factor of 1.6 (95% CI, 1.3 to 2.0) with every decade of age and by a factor of 12.6 (95% CI, 4.8 to 32.9) if wheelchair-dependent. The final multivariate model included mobility at discharge and age as the predictors. The model had good calibration and discrimination.
Conclusion(s): The risk of dying after discharge from hospital with SCI in Bangladesh is high, especially amongst older and wheelchair-dependent people. A simple prediction model discriminates those at risk of death 5 years following discharge.
Implications: This study provides accurate estimates of survival following SCI in Bangladesh. The data are important because until now there were no estimates of survival and the extent of the problem was not known. With our data, we are in a better position to lobby for better services to support people with SCI following discharge in an effort to reduce the mortality rates. In addition, our prediction model has important implications for identifying at the time of discharge those at most risk of premature death. These patients can be targeted for additional community-based follow-up and support. This sort of targeted approach is critical in countries where health resources are scarce.
Keywords: Spinal cord injury, survival, prediction model
Funding acknowledgements: Funded by the University of Sydney, Australia
Topic: Disability & rehabilitation; Service delivery/emerging roles; Neurology: spinal cord injury
Ethics approval required: Yes
Institution: Centre for the Rehabilitation of the Paralysed(CRP, Bangladesh
Ethics committee: Centre for the Rehabilitation of the paralysed(CRP)-Ethics Committee
Ethics number: CRP-R&E-0401- 218
All authors, affiliations and abstracts have been published as submitted.
