The FIM score at baseline was median 86.0 (IQR 47) and significantly increased (Wilcoxon Signed Ranks test p < 0.001) at 5 years post discharge to median 111.0 (IQR 56.5). At baseline scores for the SF-36/PCS were mean 31.6 (SE 0.99) and significantly increase over 5 years to mean 35.9 (SE 1.11) (paired t-test p = 0.002). For SF-36/MCS no significant difference was measured between baseline median 54.6 (IQR 17.0) and after 5 years median 53.5 (IQR 15.5) (Wilcoxon Signed Ranks test, p = 0.60). Table II shows there was a significant difference in PCS scores after 5 years between the two groups (t-test, p = 0.04) with the compensation group having less recovery from disability.

Predictors of outcomes for brain injury cohort

In the univariate analysis with total DRS scores ≥ 4 points at 5 years as dependent dichotomous variable, we found that not having a high educational level (started or finished university), lower baseline total SPRS scores, higher baseline total DRS scores, longer length of PTA, and receiving financial compensation were associated with moderate to severe disability at 5 years. We performed a backward multivariate logistic regression analysis with all the relevant variables from the univariate analysis and compensation status as predictor variables and DRS scores ≥ 4 points at 5 years as dependent variable. The final model explained 48% of the variance in total DRS scores at 5 years, with financial compensation explaining 3%. In this model length of PTA was excluded as this variable was no longer significant in multivariate analysis (Table III). Further analysis showed that the association between DRS scores ≥ 4 points at 5 years and receiving financial compensation was caused by the ‘other’ forms of financial compensation and not with CTP or workers compensation (B = 1.61, SE = 0.67, Exp (B) 5.00, 95% CI 1.34 to 18.66).

Similar baseline variables were also associated with total SPRS scores ≥ 40 points as dependent dichotomous variable at 5 years. A similar backward multivariate logistic regression analysis was performed with these variables including compensation status as the predictor variable. We found a model (high educational level, high baseline total SPRS scores, low baseline total DRS scores) explaining 42% of the variance in high SPRS scores at 5 years (SPRS scores ≥ 40 points) without the variable length of PTA. In this model receiving financial compensation was not a significant predictor variable (B = –0.73, SE = 0.44, Exp (B) 0.48, 95% CI 0.21 to 1.13).

Predictors of outcomes for the spinal cord injury cohort

For the SCI cohort we found that having a high impairment and low baseline FIM score were the only baseline variables significantly associated with low total FIM scores (FIM scores<113 points) at 5 years. Receiving financial compensation was not associated with total FIM scores 5 years. In the backward multivariate logistic regression analysis only baseline FIM score was significantly predictive for FIM scores ≥ 113 points explaining 73% of the variance at 5 years (B = 0.12, SE = 0.08, Beta = 0.91, CI 95% 1.03 to 1.33).

We found that receiving financial compensation was associated with lower SF-36/PCS scores at 5 years, together with having a high impairment and low baseline FIM scores for the SCI cohort. In contrast, baseline SF-36/PCS scores were not associated with SF-36/PCS scores at 5 years. In the backward multivariate linear regression analysis we found that receiving financial compensation and the baseline FIM score were significant predictors for SF-36/PCS at 5 years, explaining 18% of the variance, with financial compensation explaining 7% (Table IV). Further analysis showed that the association between SF-36/PCS scores at five years and receiving financial compensation was again caused by the ‘other’ forms of financial compensation (B = –7.68, SE = 3.47, Beta = –0.29, 95% CI –14.63 to –0.73).

For SF-36/MCS scores ≥ 50 at 5 years for the SCI cohort we found that only age was marginally associated (B = –0.04, SE = 0.02, Exp (B) = 0.96, 95% CI 0.09 to 1.00). None of the other variables including receiving financial compensation showed an association.

Discussion

Literature has so far speculated a great deal on the relationship between financial incentives and recovery from various types of injuries (2, 3, 6). A substantial number of these studies reported that financial compensation could have a negative effect on recovery (5, 18). However, the majority of these studies included patients with only mild or moderate types of injury. In this current study we analysed the influence on recovery from receiving financial compensation in people experiencing major trauma. For people with a severe TBI we found a significant difference in disability measured with the DRS at 5 years between those who received financial compensation and those who did not. Furthermore, receiving financial compensation was a small but relevant predictor, next to baseline disability and total years of education, in the multivariate logistic regression analysis with DRS scores ≥ 4 points as dependent dichotomous variable. For people with SCI SF-36/PCS scores were significantly lower at 5 years for those receiving financial compensation compared with those who did not. Receiving financial compensation and baseline FIM score were both small but significant predictors for SF-36/PCS scores after 5 years.

These present findings contradict our hypothesis expecting to find no association between receiving financial compensation and recovery for severely injured patients. Our hypothesis was supported by Wood et al. (19) who examined recovery in people with severe brain injuries after 10 years and found no difference between those who were litigant and those who were not. In line with this, a meta-analysis of studies reporting outcomes in patients with traumatic brain injuries by Binder et al. (3) suggested that in patients with more severe injuries less effect of financial compensation could be demonstrated. For patients with SCI no studies were found concerning recovery and financial compensation. However, there were many studies on whiplash or chronic pain and financial compensation (2, 4, 20).

The results of the current study showed that financial compensation seems to have an association with functional recovery, even for patients with a severe traumatic injury. The actual predictive effect on the variance in DRS scores ≥ 4 points, or in SF-36/PCS scores after 5 years was relatively small (3% and 7%, respectively), but may have some clinical value. We have no good explanation for this result. Why, in particular only a small effect was found for “other forms” of financial compensation needs further research. These forms of compensation are sports injury compensation, crime victim’s compensation or personal injury compensation. The difference between these types of financial compensation and workers compensation or CTP insurance is possibly that people with catastrophic injuries who have access to CTP and workers compensation insurance have more extensive and effective treatment than people with the other types of compensation. However, it is acknowledged that this explanation is speculative. In general, the potential relationship between poor outcome and seeking compensation is still not resolved. Some authors have reported that receiving compensation is a predictor for poor outcome; in contrast others could not demonstrate such an association (2).

Results showed that for both types of injury there was a significant improvement in the level of disability and functioning 5 years after injury. DRS improved significantly for the TBI group, FIM increased significantly for the SCI group. The psychosocial reintegration for the TBI group as measured with the SPRS decreased significantly over the 5 years, mainly as a result of the domain of interpersonal relationships. This finding will also need further research. Overall, the distribution of scores over the 3 domains was comparable to a similar group of TBI patients who were scored after 10 years (14). A limitation in this study is that a SCI specific disability measure was not included in the available dataset. It is possible that such a measure might have responded differently to the included baseline variables such as receiving financial compensation. The reason that in the SCI group we only found an association between financial compensation and the SF-36/PCS scores after 5 years, but not for the total FIM scores after 5 years, may be explained by ceiling effects of FIM and/or concepts, such as pain and general health, which are broader than simply activity limitation and mobility being captured by SF-36/PCS. It could also be that by dichotomizing the outcome variables, except for SF-36/PCS, for the multivariate regression analysis information was lost. The cut-off points selected were the median scores after 5 years, and it is arguable whether other cut-offs scores may have been more appropriate.

Strengths of this study were the inclusion of physical and socioeconomic factors for two different cohorts of people with severe traumatic injuries that have major long term effects on health. The questionnaires used in the study were all validated and well used in these types of trauma. However, a weakness of the study was that the same outcome measures were not available for the two groups. Future studies need to realize that different health outcome measures may influence the measured association between financial compensation and recovery. In this study, the focus was on health status at 5 years and not on work status or financial situation at 5 years. Although, there is increasing literature on long term return to work, social benefits and financial situation after serious illness or injuries, there is still a high need for good qualitative longitudinal studies to evaluate what determines these outcome measures.

Another potential weakness of this study is the relatively small sample size per injury group. Because patients with the level of injury severity as occurred in the present cohorts are fortunately relatively rare, this will probably remain a problem in future studies. In this study, lack of statistical power may also be a reason for no baseline differences between the two compensation groups. Further, no differentiation was made between those who pursued compensation using a lawyer and those who received compensation without use of a lawyer or involvement with any compensation related factors. Engagement of a lawyer has previously been associated with poor outcome after trauma (5, 21, 22). Although, it is likely that the two groups of patients in this study for the most part have had lawyer involvement, future studies may need to differentiate for this covariate.

In conclusion, even for severely injured patients receiving financial compensation may have some small relation with recovery after 5 years. More research is necessary to confirm this finding, and to further analyse what form of financial compensation may have an association with long term health status for different types of patients.

References