## Introduction

In Bangladesh, spinal cord injuries (SCI) are commonly due to work-related accidents in young manual labourers [1,2,3,4,5], often with low socioeconomic backgrounds [6,7,8]. A similar pattern is seen across other low- and middle-income countries (LMIC) [9,10,11,12]. SCI causes hardship for those who are injured and often also for the families who are financially dependent on them [3, 13,14,15]. The loss of income and additional medical and ongoing costs associated with the injury place a large financial strain on families [16, 17]. Not surprisingly, therefore, SCI in Bangladesh and other LMIC often throws families into extreme poverty [18,19,20].

### Data analysis

The total income for each family was divided by the number of adults and children in the family to derive income per family member. Families were then defined as living below the poverty line or extreme poverty line on the basis of the average income per family member. The poverty line was defined as US$57 per person per month (~US$1.90 per person per day) and the extreme poverty line was defined as US$37.50 per person per month (~US$1.25 per person per day) as per the definitions of the United Nations and World Bank [21, 23,24,25].

The same calculations were repeated but with the income of the person with SCI removed to determine the average income for each adult and child within a family after loss of income. These later calculations assumed that the person with SCI had no income once injured and while in hospital, and that there was no change in the employment status of other family members since injury.

## Results

### Participants’ characteristics

The characteristics of participants and their families as well as their work status are shown in Tables 1 and 2. The median (IQR) age of participants was 33 years (25–43), 90% were male, 57% had paraplegia, 43% had tetraplegia, 69% were married and 71% had American Spinal Injuries Association (ASIA) Impairment Scale A lesions. In addition, 98% were from rural areas of Bangladesh and 65% had no or limited ability to read. The median (IQR) size of each family (including the person with SCI) was 5 (4–6) individuals with a median (IQR) 2 (1–3) adults and 2 (1–2) children per family. We did not ask participants to specify their relationship with other adults in the household but they could have been spouses, parents, siblings or children.

Prior to injury, 82% of participants were in full-time employment. Most participants (58%) were manual labourers or tradespersons (17%) prior to injury, and only nine participants (3%) had professional jobs. The others were either office workers, business workers or shopkeepers. Seventy-four percent were the main income earners for their families. In 50% of families, no other person worked. The incomes of participants and their family members before and after SCI are shown in Table 3. The monthly median (IQR) income of each family member prior to injury was US$30 (US$19–US$48) per person. This was sufficient to keep 33% of family members above the extreme poverty line. After injury, the monthly median (IQR) income of each family member was US$0 (US$0–US$18) per family member. This was sufficient to keep 9% of family member above the extreme poverty line. That is, after injury, 91% of family members were living below the extreme poverty line of US$37.50 per person per month. ## Discussion This study investigated work-related family incomes before and immediately after SCI in Bangladesh with the aim of determining the degree of impoverishment of people with SCI and their families. A notable finding was just how impoverished families were prior to injury: families had a median (IQR) income of US$30 (US$19–US$48) per family member per month. Thus 67% of family members were living below the extreme poverty line prior to injury. It was not surprising to see the median (IQR) income drop to US$0 (US$0–US$18) per month per family member following injury, because those injured where typically young males and 74% were the main income earners for their families. This drop of income placed 91% of families below the extreme poverty line. There are various ways of defining poverty with poverty lines often adjusted for the number of children under the age of 12 years living in a family unit. However, regardless of how poverty is defined, there can be little doubt that our findings highlight the dire financial implications that SCI can have on families in Bangladesh. Our data on participants’ incomes prior to injury are broadly consistent with data from our previous study of a similar cohort discharged from CRP in 2011 [26]. We found in our previous study that participants who were wheelchair dependent at discharge earnt a median (IQR) of US$65 (US$39–US$104) per month prior to injury. In the current cohort, participants earnt a median (IQR) of US$106 (US$60–US\$180) prior to injury. The higher income observed in the current study may reflect wage inflation over the intervening 4 years (estimated to be 5.75% per year) [27]. Nonetheless, the incomes of participants of both cohorts were lower than in the general population of Bangladesh. For example, while 67% and 83% of family members in the current cohort lived below the extreme poverty and poverty lines prior to injury, respectively, only 13% and 23% of the general population of Bangladesh are in the same financial situations [13, 28,29,30]. Similarly, the literacy levels of our cohort (only 35% had a good ability to read and write) were lower than the national literacy rate of 74% [31]. These factors together point to the social disadvantage of our cohort, and probably reflect that those most likely to sustain an SCI in Bangladesh are unskilled labourers from poor rural backgrounds [6, 32].

Our data are not necessarily reflective of all people who sustain an SCI in Bangladesh. Bangladesh has a population of 164 million. While there are no accurate data on the incidence of SCI in Bangladesh, it is likely that the incidence is between 20 and 40 per million [8, 33]. This equates to between 3280 and 6560 people with a new SCI each year. The CRP only admits 400 people with recent SCI per year. So, clearly, many people who sustain an SCI are not admitted to CRP and these people may be different to those that are. For example, those from higher socioeconomic backgrounds may be admitted to private hospitals or hospitals in other countries. In contrast, those from even poorer backgrounds than our cohort may never get to CRP or never get to any hospital because of economic and social disadvantage.

Our calculations may underestimate the extent of the financial strain placed on families because we did not consider additional costs incurred by families for healthcare services. The medical costs for our participants were largely covered by CRP but nonetheless some participants may have incurred health-related costs before they were admitted to CRP. The situation may be quite different in other LMICs or other parts of Bangladesh. For example, one study from Nigeria found that 41% of participants used 50% of their annual income to meet the acute medical costs of an injured person [17]. This does not include the ongoing costs incurred after discharge from hospital.

Financial hardship following SCI affects all aspects of life and contributes to premature death in both high-income countries and LMIC [34, 35]. Our studies in Bangladesh indicate that 32% of participants who are wheelchair dependent at discharge die within 5 years, primarily from pressure ulcers [36]. The causal links between financial hardship, pressure ulcers and premature death is probably complex but it is plausible that poverty plays a major role.

A reduction in poverty needs to be a major focus of future initiatives to improve the lives of people with SCI and their families in LMIC [16, 37]. The data provided in this study could be used to highlight the financial hardship of people with SCI and their families, and to encourage governments to provide some financial support. In addition, attention needs to continue to be directed at getting those injured back to work. For this purpose, vocational training needs to be at the centre of any rehabilitation program [13, 17]. A study from Nepal found that less than half of those with an SCI had any income many years after discharge from hospital [38]. Our previous study  from Bangladesh showed that only 37% of people who were wheelchair dependent on discharge from hospital were employed 2 years after discharge [3] and only 42% were employed 6 years after discharge [26]. Some evidence suggests that those who do gain employment have much lower salaries than they had prior to injury [38].

Vocational training is provided at CRP but unfortunately many barriers to employment need to be overcome. These include wheelchair-inaccessible environments, and the low literacy rates and skill levels of people who are typically injured (see Tables 1 and 2). In addition, discrimination and poor societal attitudes to those with disabilities are still widespread. Often, the best employment option for people with SCI and limited skills is to set up small businesses or shops [16]. This may require a small amount of initial financial assistance. It is also important that young students who are injured are supported to return to their education. In addition, attention could be directed at vocational training for the wives of injured men. The women may require more support than in most countries because in Bangladesh women do not typically work outside the home [6]. However, the employment of wives could help relieve the financial pressures on families.

There are several limitations to this study. The main limitation, as discussed above, is that we cannot generalise our results to all people who sustain SCI in Bangladesh or other LMIC. The second limitation is that while our cohort is largely reflective of those admitted to CRP who survive to discharge and are wheelchair dependent, we were unable to recruit 33 potentially eligible participants (9% of the potentially eligible cohort). This may have introduced a small selection bias. Thirdly, we did not verify what participants told us at the time of discharge about their and their families’ income and employment status prior to injury. Participants may not have accurately remembered, or may not have known, or may have perceived that it was within their interests to underestimate their family’s incomes. In addition, we did not capture people’s assets or money given to people by extended family members or friends. We did not explore families’ capacities to be self-sufficient through their own crops and animals. Nor did we ask participants about other potential sources of income from insurance, government, prior savings or other sources, although these sources of income are uncommon in Bangladesh particularly for those working in unskilled jobs. We know from anecdotal evidence that many people sell property and animals or take loans to support their families. The results of this study should therefore be seen within these limitations. Future studies could better explore how families survive on such little income.

In summary, SCI have profound financial implications for individuals and their families, and cause extreme poverty in Bangladesh. This is because those most often injured are young people from low socioeconomic backgrounds who are the main income earners for their families.