Introduction

Traumatic spinal cord injury (TSCI) is a catastrophic event worldwide often leading to permanent disabilities [1]. Patients with TSCI may suffer physically and emotionally as well as having financial difficulties after the injury [2–4]. Presently there is no available treatment enhancing functional recovery of the paraplegic/tetraplegic patients following a TSCI [5]. This demonstrates the importance of the development of preventing strategies worldwide. Most of the SCI are due to trauma; [6] commonly motor vehicle accidents, falls, violations, and sports injuries. Although there are publications on incidence and prevalence of TSCI from different countries of the world as well as systematic reviews on incidence and prevalence of TSCI in city(s) [7], or at national [8], regional [5], and global level [9], yet the literature lacks an overview on the extent of SCI in single countries from parts of the world where data are of low quality or even not available. The present study aims to gather available data of the current situation of TSCI worldwide, through extensive systematic search strategies.

Methods

Data collection

In order to obtain all available information, different search methods including strategic searching, reference checking, searching for grey literatures, contacting registries, authors, and organisations requesting for unpublished data, and browsing-related websites and journals were employed. Details of following steps have been reported in “Appendix”. In search process, there was no language, country, or date limitation for inclusion of the reports.

Development of search strategies

All abstracts from the bibliographic database used in the study by Rahimi-Movaghar et al. [1] on the epidemiology of spinal cord injury in developing countries including records up to July 16th 2012, were reviewed. The search included the terms ‘spinal cord injury’, ‘incidence’, ‘prevalence’, and ‘epidemiology’ [1]. A librarian performed monthly search from search the engine EMBASE via Ovid SP and PubMed (including MEDLINE and PubMed Central). All available full text papers from the search were obtained (April 29, 2013). During full-text screening, references of the related papers were checked manually for identifying other related papers. To retrieve non-indexed reports, we used Google search engine combining ‘spinal cord’ with ‘incidence’, ‘prevalence’, ‘epidemiology’, ‘registry’, or ‘database’. Along with retrieval of relevant reports, we identified registries from different countries. The identified resources were added to the references list in the report of International Spinal Cord Society (ISCoS) team for global mapping of TSCI [9].

Websites/grey literature

By contacting authors of related websites identified through the links of registries and checking references manually we acquired more related information. The websites were checked by three members of the team. Data available to public were found from four registries: Australia, Germany, Israel, and USA. By searching the resources of grey literature, related reports and theses were found. Reports from the National Spinal Cord Injury Statistical Center (https://www.nscisc.uab.edu) and International Spinal Cord Society (https://www.iscos.org.uk) websites were also reviewed.

Communications for unpublished data

When data were not available online, we contacted the authors/websites/registries/organisations via available email addresses or contact forms from the websites. We included a list of registries provided by O’Reilly et al. [10, 11]. Potential authors, websites, and organisations were contacted by two members of the team using an email template. By communication with authors of related papers and authors of the book “Epidemiology of Spinal Cord Injuries” [12] we retrieved unpublished data. In total, 114 individual authors were contacted. Three researchers corresponded with the authors; the communications were focused on three requests:

  1. 1.

    Provision of raw data.

  2. 2.

    Access to any unpublished data in a local registry or locally published journal.

  3. 3.

    Translation of parts of non-English reports.

In case that no reply was received, we emailed a reminder 1 week later. This direct contact with authors led us to new information regarding TSCI in New Zealand, Austria, and Denmark.

Hand-searching of journals

The retrieved records were kept in a bibliographic database using EndNote X5. A list of highly focused journals on Spine, Injury and Trauma which published most of the papers was created. All issues of these journals were checked manually to obtain possible missed papers.

Resource selection and data extraction

A database containing the full-text papers, registry reports, conference proceeding and books were generated, including the following data: type of study, extent of study (national vs. subnational level), coverage years of study, extracting raw data of study if available, number of cases, age and incidence report, and type of SCI (traumatic vs. non-traumatic), to report incidence rate of TSCI in form of raw or age-adjusted rates or number of cases. Papers with unavailable incidence figures, cumulative report of traumatic and non-traumatic incidence rate without traumatic versus non-traumatic ratio, as well as reports of prevalence figures only, were excluded. A template of a uniform excel database was designed to enter data of available countries based on WHO classification of countries. In papers reporting average incidence in timespans of more than one year, data in the paper were used to compute/estimate incidence rate for single years of the study, when available.

Results

A total of 1,261 records from PubMed and 1,621 records from EMBASE were identified and included in the EndNote X5 database. After excluding 1,153 duplicate records, titles and abstracts of 1,729 records were screened by two members of the team. After exclusion of 1,245 irrelevant records, 484 papers were screened and assessed using the inclusion criteria. This process limited the numbers to 258 papers of which 145 were excluded due to unavailable records of raw data or specific reports on incidence figures and full texts of 113 papers were retrieved. After assessing the full text, 101 papers [1, 7, 8, 13–110] were selected (Fig. 1). We identified 17 reports from national registries [111–127], five books [12, 128–131], five conference abstracts [132–136], two theses [137, 138] as well as data from New Zealand and Austria through personal communications [139, 140]. The trauma registries from Germany [125], Israel [127] and Ontario (Canada) [124] were excluded due to lack of detailed data on SCI. Two of the papers contained valuable information regarding the incidence of TSCI in Ireland [91] and USA [37], but lacked information on the precise years of the study. Data of TSCI incidence in Iceland were reported by Knutsdottir et al. [56, 57, 134] in a period between 1975 and 2009 in three separated resources. Overall, we identified incidence data for 41 countries worldwide, from current literature of 2013. Extracted data of incidence rates are shown in Table 1.

Fig. 1
figure 1

Flow diagram of studies based on the PRISMA statement

Full size image
Table 1 Data of traumatic spinal cord injuries retrieved from search strategies of the study
Full size table

American region (AMR)

American region includes 28 countries within two sub regions of AMR A and AMR B. There were 36 reports on TSCI incidence from US, Canada and Brazil [7, 13, 18–20, 23, 29, 34, 37, 38, 43, 45–47, 52, 69, 73, 77, 82–85, 97–100, 106, 111, 122–124, 126, 128–130, 138], mainly from US and Canada (35 out of 36 records). Although there are incidence data of TSCI in the US since 1935 [46], the reports are mainly limited to registries of single or multiple states [7, 13, 18, 20, 23, 37, 43, 46, 52, 77, 85, 97, 98, 100, 105, 106, 122, 123, 126, 128] and there are no cumulative data regarding the incidence of TSCI. In Canada there are multiple reports at subnational levels [29, 34, 38, 69, 82–84, 99, 111, 129, 130, 138]. The incidence of TSCI in Canadian ranges from 3.6 to 52.5 cases per million in different years and regions. TSCIs in geriatrics has decreased over time from 83.2 to 30.0 cases per million during 1991–2001 as noted in a thesis by Aditya Sharma [138]. However, the recent survey in Canada has estimated about 53.0 cases of TSCI per million [73].

Brazil is the only country with available data of TSCI in AMR B. Incidence of SCI ranges from 17.0 to 71.0 cases per million, with a mean of 17.3 cases per million in the period 1986–2007 [45].

African region (AFR)

Three papers were identified from Africa: Sierra Leone, South Africa and Zimbabwe [44, 62, 103]. Only two papers reported data on incidence of TSCI [62, 103]: one paper from late 80 s and early 90 s and one paper from early 2000 [44]. The incidence of 11.7 cases per million is estimated from national rehabilitation centre of trauma patients between 1988 and 1994 in Zimbabwe. Estimates of 48.5 cases of TSCIs in South Africa are based on a single-centre report of patients with traumatic spine injuries between 1988 and 1992 in Johannesburg [103]. The only available study in Sierra Leone is a follow-up study of treated patients of a single hospital. The incidence in Sierra Leone is estimated around 3.4 cases per million.

East-Midland region (EMR)

Six papers from countries in East-Midland region (EMR) including Iran, Jordan, Kuwait, Qatar, Saudi Arabia and Pakistan were identified [78, 87–90, 108, 137]. The incidence figure from Pakistan (5.1 per million) is based on a national survey of medical records of patients with head and spine trauma hospitalised between 1995 and 1999 [90]. Jordan, Kuwait and Qatar figures are estimated from single centres while Iran and Saudi Arabia figures are from population-based studies [88, 137]. The incidence in these countries ranges from 5.1 cases per million in Pakistan to 72.4 cases per million in Iran. None of the countries in this region have available reports from trauma registries, though most cases of TSCI in these countries are believed to be due to traffic accidents.

Europe region (EUR)

Data are available for 22 of these countries. European countries including Austria, Bulgaria, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Israel, Italy, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, England, Romania, Turkey, Estonia, and Russia have published the major part of available data in literature with 40 separate records [8, 14, 15, 17, 22, 24, 32, 36, 40–42, 48, 49, 53–59, 64, 65, 67, 71, 75, 79–81, 86, 91, 94–96, 101, 102, 131, 132, 134, 135, 140]. The incidence of TSCI in Europe ranges from 5.5 to 195.4 cases per million in Norway and Ireland, respectively.

Western pacific region (WPR)

A total of 23 papers from Australia, Japan, New Zealand, China, Fiji Islands, Malaysia, Vietnam and Taiwan were found on TSCI incidence [33, 50, 51, 61, 63, 66, 70, 72, 76, 92, 93, 104, 112–121, 136]. However, there are 20 countries in this region without publications on the epidemiology of SCI. The incidence in WPR ranges from 10.0 to 60.6 cases per million in Fiji Islands and China, respectively. Australia is the leading country worldwide in registration of SCI. Annual reports of SCI in Australia have been published since 1986 and the registry is updated each year. The incidence of TSCI in Australia has decreased from 20.7 in 1986 to 15.0 cases per million in 2007.

South East Asia (SEAR) and Taiwan

The only available data for SEAR are the reports of TSCI in Thailand. In Thailand one paper estimated the incidence of TSCI between 1985 and 1991 to be 23 cases per million [60]. A paper from a single hospital reported 219 patients with TSCI were hospitalised in Thailand (incidence 5.8 per million) in the period 1989–1994 [109]. Taiwan is not classified in WHO regional distribution of countries; however, there are 3 papers which include the incidence of TSCI in Taiwan, between 14.6 and 63.2 cases per million [25–27, 61, 107].

Discussion

Spinal cord injury is a rare but debilitating health condition worldwide. Most of the SCI cases are due to trauma including traffic accidents and falls. Epidemiological data on TSCI have not been studied in many countries worldwide. In this review an extensive search was made to retrieve and access all available information regarding the incidence of TSCI in every country of the world. The searches resulted in finding of TSCI incidence data of 41 countries. Most of the available data are found in developed countries, where the incidence of TSCI in recent years has decreased or been stable due to preventive strategies probably based on knowledge of epidemiology of SCI [9]. However, SCIs are increasing in developing and middle to low-income countries, where there is lack of information regarding the epidemiology of SCI. Although SCI is not a common condition with an estimated overall incidence of 23.0 cases per million worldwide [9], the results of SCI as well as the costs, have made a growing concern in both developed and developing countries around the world.

The incidence of TSCI in this review ranged from 3.6 per million in Canada [99] to 195.4 per million in Ireland [91]. A decreasing/stable incidence of TSCI is seen in Australia, Austria, Canada, France, Iceland, New Zealand, Turkey, and US. Nevertheless, the incidence of TSCI increases in Ireland, Italy, Norway, Russia, Saudi Arabia, Spain, and Taiwan. The increase in incidence of TSCI may be due to improved recognition, registration, and health care system in these countries. However, there is insufficient evidence from other countries around the world to define trends of spinal cord injury.

According to registries across the US, the incidence of SCI is estimated to be 30.0–40.0 cases per million in the US [141]. Nonetheless, this incidence number includes 8.5 % of cases with unknown origin, which might be due to non-traumatic etiologies like tumours and infections. TSCI data in different states of Canada are available [3, 29, 34, 73, 82–84, 99, 111, 124, 130, 138]. Rick Hansen Spinal Cord Injury Register data are soon becoming public and these data are the most up-to-date data on TSCI in Canada. In Brazil, the incidence of 17.3 cases per million of TSCI based on retrospective surveys might be an underestimation in a country with more than 200 million residents and a growing concern about traffic accidents [12]. Data are not available for other 25 countries in AMR. There are 52 European countries, and seven (Austria, Finland, France, Germany, Ireland, Spain and Sweden) have published data on SCI from trauma registries. Although there might be more trauma registries across the Europe, e.g. in Norway, no publication on SCI is available through public access websites. Data are available for 22 countries across Europe. Publications from high-income countries (EUR A) form the major part of the present available data.

The African continent has yet the least available information on SCI epidemiology. There are 46 countries in African region according to WHO classification of countries. However, data are available only for three countries (Sierra Leone, South Africa, Zimbabwe). Moreover, the incidence figures reported in African region are inconclusive; they are mostly derived from single hospital-based or follow-up studies of treated patients.

Today developed countries are better informed and have effectively gathered data on TSCI; however, the definition of SCI is used differently or broadly. The reporting of those who die at the scene of the injury is included in some studies or data collection and excluded in others. These discrepancies in data collection have made it difficult to identify the cases and estimate the incidence of TSCI globally. There is an ongoing global effort to standardise registries and data management as well as effectively utilise the timing of surgical intervention to drive the translation of clinical practice on early intervention to prevent neurological damage, reduce secondary complications and improve outcomes resulting in better HRQOL in the community [142–144].

The new evidence will enable us to collaborate with policy makers to introduce new measures of prevention strategies for motor vehicle accidents and falls and regulations on safety for prevention of TSCI.

Due to the poor reporting and recording of TSCI in developing countries, there is no real urgency to drive the policy decision for prevention programmes. The obligation lies on the developing countries to prioritise care, bring awareness and look to developing countries for knowledge translation from their best practice programmes.

Conclusion

In this review the data retrieved from available data, both published and unpublished, demonstrate large variations in classification of age and aetiology of TSCI as well as the computation of incidence. The literature still lacks accurate epidemiological data on TSCI in many countries, most prominently in Africa. An extensive work is done by ISCoS-WHO collaboration to standardise the reporting of TSCI. This will improve future epidemiological studies on TSCI and promote the establishment of preventive strategies worldwide.