Long-term changes of patient-reported quality of life after major trauma: The importance of the time elapsed after injury

Background: Numerous studies have identified various risk factors for a poor health-related quality of life (HRQOL) after severe trauma. The relative importance of the time elapsed after injury, however, is unknown and results of clinical studies have been conflicting. Methods: A cross-sectional study was performed in two trauma centres using data from the German TraumaRegister DGU (R), which contained prospectively collected information on the type and severity of the injury, on critical care, and on outcome. To evaluate HRQOL in patients surviving more than 500 days after the injury, we used a self-rating instrument, the EQ-5D which contains a visual analogue scale (EQ-VAS), and which allows the calculation of a global outcome indicator, the EQ-D5 index value. Complex statistical models were used to evaluate independent associations between the time elapsed after injury and a poor HRQOL. Results: Of 380 contacted patients, follow-up assessments could be obtained in 168 patients (44.2%) 3.6 + 1.6 (SD) years after the injury. There was a linear association between the time elapsed after the injury and the% of contacted patients not participating in the study (p = 0.013). In participating subjects, average EQ-5D index value was 0.599 +/- 0.299, and average EQ-VAS rating 67.8 +/- 22.0. A very poor quality of life (EQ-5D index value < 0.6, EQ-VAS rating <= 50) could be found in 43.5% and 28.0% of the patients, respectively. After adjusting for multiple confounders, the number of days elapsed after injury showed a complex non-linear and independent association with a poor HRQOL (low EQ-5D index value: p = 0.027; low EQ-VAS rating: p = 0.008). Frequencies of a poor HRQOL reached their minimum about four to five years after the injury and increased thereafter. Conclusions: There is an independent, U-shaped association between the frequency of extreme values of HRQOL and the time elapsed after injury. Time patterns of HRQOL may be sensitive to increasing rates of attrition since patients with a good outcome are less likely to respond to questionnaires. Time from injury should be incorporated into all future cross sectional studies trying to identify predictors of HRQOL. (c) 2017 Elsevier Ltd. All rights reserved.