Page 230 - DECISION-MAKING IN SEVERE TRAUMATIC BRAIN INJURY PATIENT OUTCOME, HOSPITAL COSTS, AND RESEARCH PRACTICE

Chapter 12
these measures for outcome valuation and their usefulness in the acute decision- making process of s-TBI patients remains disputed.
Patient mortality
The most frequently used and most straightforward outcome measure. Death is usually considered to be the worst possible outcome that should be prevented at any cost. 53 However, in s-TBI patients, survival with severe post-traumatic deficits can be a fate worse than death. 54-58 When considering the possibility of very severe cognitive, emotional, and physical disabilities, life and death are not necessarily equal to ‘acceptable’ and ‘unacceptable’. As such, acute treatment decisions should not solely be based on predicted mortality.
Functional outcome
The Glasgow Outcome Scale (GOS) is the most highly cited outcome measure in brain injury studies. 59-61 Its use as TBI outcome measure is recommended by many organizations. 60 It assesses multiple aspects of life to determine the impact of TBI on patient functional outcome with a focus on social recovery. It uses dichotomous endpoints, in which ‘favourable’ outcome (the ability to function independently, see Table 2), is usually considered to be the ‘acceptable’ outcome. The introduction of the Glasgow Outcome Scale Extended (GOSE) and the structured interview 61 have solved points of criticism on validity and lack of sensitivity in the higher functional end of the scale, but there are remaining issues. 52,60,61
The ‘favourable’/’unfavourable’ division remains arbitrary and ignores a patients’ or proxy’s perception of satisfaction with life. Patients with severe disability who are dependent in daily life (defined as ‘unfavourable’) can still judge their health status to be ‘acceptable’. 60 But the other way around is also possible. Some studies classify ‘upper severe disability’ (GOS-E) to be ‘favourable’, while probably most physicians, researchers and healthy individuals would classify this outcome as ‘unacceptable’ within their own social and cultural context. 50,62
Instead of using dichotomized outcome, sliding dichotomy or proportional odds methods are considered to be more informative. These methods are increasingly popular, but still have insufficient sensitivity to detect all changes. Subtle changes can be highly valuable for a patients’ wellbeing, without having a measurable impact on pre-defined categories. 60
228