Estimates of a patient’s quality of life are increasingly used to determine which treatments are worth funding and which are not [2]. This is particularly true for new drugs and new treatment options. In an ideal world, all drugs and treatment options should be funded and made available to appropriate patients, but, in reality, health budgets are stretched and health providers worldwide face difficult choices when deciding which drugs and treatment options merit the costs that they incur. What criteria can be used to make this decision? Clearly a drug that can cure a common disease will always be a strong contender for funding. However, in many cases, particularly in the case of stroke, a cure is not available. Available treatments for stroke may at best reduce the likelihood of further strokes, slow down or stop the further development of neuropathology, reduce neurological symptoms, or allow patients to compensate more effectively for acquired neurological disabilities. A crude way of assessing the impact of the various treatment options on the course of a chronic disorder, such as stroke, is to compare the number of life years saved by administering a given treatment. However, this is too crude. Most people would agree that a treatment that promises to provide 5 more years of normal, symptom- and pain-free life is clearly superior to a treatment that can extend life by the same amount but only in a pain-ridden, disabled status. This example shows that quantity alone is not enough. The quality of the life saved or extended needs to be taken into account as well. This insight has led to the development of the quality-adjusted life-year (QALY) measure [3]. To calculate QALYs we take the number of years saved or extended and multiply those with an estimate for the quality of that life. These estimates for the quality of life score can vary between 0 (equals death) and 1 (equals perfect health). In the following we will give a hypothetical example of how QALYs can be used to compare the value of two treatments.

Let’s say there are two drugs—drug A and drug B. Drug A increases life expectancy by 4 years but the life saved will be associated with pain and disability, accordingly its quality of life is low and has a score of 0.4. Drug B increases life expectancy by only 3 years but the life saved can be lived in normal health; accordingly the quality of life score is high (i.e., 1). We can now calculate the QALYs for both drugs:

By taking quality of life into account, drug B appears to be the more valuable drug despite the fact that it increases life expectancy by 1 year less than drug A. In this case the health provider will possibly decide to fund drug B but not drug A. Clearly with so much at stake it is important that we use reliable methods to estimate the quality of life associated with a given medical condition.

Before measuring quality of life, it has to be defined, which is problematic. There is some agreement that health-related quality of life (HRQoL) is related to the concepts of disability and handicap as outlined in the International Classification of Impairments, Disabilities and Handicaps [4] and it is argued that HRQoL lies beyond both concepts [5]. There is further agreement that HRQoL is a multidimensional concept presumably including physical, mental, and social aspects of life [6], but there is little agreement about anything else. For this reason it is attractive to find ways of measuring HRQoL that do not require its explicit definition but rely on the implicit understanding of patients and other respondents.

There are several ways to produce a number for a patient’s HRQoL without having to resort to a formal definition of quality of life [2]. The resulting measures are called utility measures and are primarily used to calculate QALYs and, thus, inform health-economic decisions. The standard-gamble method will ask the respondent to state the odds of a gamble between death and perfect health that they regard as equivalent to accepting a given health state. For example, a respondent might indicate that a gamble where there is an 80 % chance of perfect health and a 20 % chance of death is equivalent to accepting a minor stroke. In this case we assign a utility index of 80 % or 0.8 to minor stroke. The time-tradeoff method uses a similar approach. The respondent is asked to trade lifetime against health. For example, if a respondent states that living for only 90 % of their remaining life span is equivalent to living the total of their remaining life but with a minor stroke, then this is taken as an indication that they value life with a minor stroke at 90 % (or 0.9) of the value of life without stroke. Finally respondents can be simply asked to assign to a given health state (e.g., stroke) a value between 0 (death) and 1 (perfect health). The same assignment can be made by an expert or a panel of experts.

Utility measures have the advantage of presenting a single number that characterizes the value of a person’s life. This may be sufficient to calculate QALYs and thereby inform health-economic decisions. However, this approach ignores the fact that quality of life is a subjective property and patients suffering from the same disease vary significantly in their subjective experience of disease. It is therefore important to understand how and why patients with the same health status differ in their assessment of their life’s quality. Health profiles, which are based on standardized questionnaires, provide much more detailed information about how a given health status affects a patient’s life. Respondents are typically asked to indicate how their health affects their ability to fulfill their work or household duties, how it affects their mood or their ability to enjoy specific pastimes, and how it affects their capacity for communication and participation in social events. Responses to these scales will provide a detailed picture of how a disease impacts on somebody’s life but they can also be used to calculate a single score, which can then be plugged into the formula for calculating QALYs.

For moderate to severe stroke the HRQoL scores range between 0.45 and −0.02. Even if we take the best score, quality of life after stroke is still less than half of what it was prior to the insult. Is this dramatic drop in life quality an inevitable consequence of this neurological disease or are there additional factors at work? Stroke severity certainly goes a long way in explaining the reduced HRQoL. Most studies on HRQoL after stroke picked out stroke severity (typically operationalized as the modified Rankin Scale) as a significant predictor [2]. The role of stroke severity or neurological disability can be seen by comparing utility estimates for patient groups with different patterns of neurological syndromes; whereas for patients with non-severe hemiparesis a utility index of 0.50 is reported, this measure drops to 0.45 in the case of severe hemiparesis, to 0.28 for hemiplegia, and to 0.06 in the case of global aphasia [2, 28, 29]. Thus, stroke severity is clearly an important predictor of HRQoL, but even patients who are considered to be fully recovered frequently show severely depressed HRQoL scores [30]. Other factors must be at work. Work-status, family problems, depression, anxiety, posttraumatic stress disorder (PTSD), and coping strategies are among some of the factors that can also have detrimental effects on post-stroke HRQoL, even in patients who otherwise suffer little neurological disability [30–32].

Another cause of reduced HRQoL is pain. About a third of all stroke patients develop shoulder pain [33] and roughly 20 % still have moderate to severe pain 1 year after stroke [34]. Sexual dysfunction and dissatisfaction with sexual life are found in roughly 70 % of all pairs where one partner has suffered a stroke. Post-stroke impotence is found in 50 % of all male stroke patients. Problems in sexual functioning are linked to the onset of post-stroke depression. Other social and psychosocial factors also exert a strong influence on sexual satisfaction after stroke [35, 36]. Unsurprisingly, sexual function has a significant impact on HRQoL after stroke. A surprising but consistent finding is that female patients have a significantly worse HRQoL after stroke than male patients [37, 38]. A multitude of factors are probably responsible for this finding. On average female stroke patients are more disabled [23]. Female patients are also less likely to be discharged home, as male partners are less likely to care for their spouses [23]. Moreover, differences in coping strategies and the prevalence of depression in male and female patients might also contribute to the lower HRQoL in female patients [37]. Coping strategies and depression play an important role in the long-term HRQoL, not just of female patients. In the initial stage, a feeling of uncontrollability prevails, frequently triggering the onset of depression. Later, positive coping strategies allow patients to regain some control of the situation leading to improved HRQoL [39]. Strategies such as tenacious goal pursuit and flexible goal adjustment are linked to a positive development [40]. This link between HRQoL, coping strategies, and depression might suggest that depression is another important predictor of HRQoL after stroke (see Chap.​ 12). This expectation is certainly confirmed by empirical data. Depression is a common consequence of stroke [41] and has a significant impact on HRQoL [37, 42]. The relationship between depression and HRQoL is quite complex. Depression affects functional recovery, cognitive function, and healthcare use, and leads to a worse functional outcome [43]. We can therefore assume that the reduced HRQoL found in stroke patients with depression reflects both a worse objective situation and a negatively biased subjective appraisal of the patient’s current situation. Given that appraisal and coping strategies change over time with more positive coping strategies observed 6 months after stroke [44], one might also expect to see parallel changes in the occurrence of depression. However, a recent large-scale (more than 8,000 patients), 5-year longitudinal study on post-stroke depression found stable rates of depression across all 5 years [41].

In younger stroke patients, who are still at work and raising a family, the effect of a stroke can have a particularly detrimental impact on their life. In a recent systematic review on the social consequences of stroke for working-aged adults, it was found that only 44 % of patients return to work after stroke [45]. Financial problems will therefore be frequent in households where one of the members has suffered a stroke (24–33 %), increasing tensions within the family and ultimately causing severe family problems (such as separation or divorce). The frequency with which these problems are found varies considerably between studies, with one study reporting a deterioration of spousal relationship in only 5 % of their sample [46] and another study reporting a prevalence for partnership conflicts of 38 % [47]. Despite this variability there is no doubt that such psychosocial factors have a significant impact on post-stroke HRQoL.

Psychosocial factors are of increased relevance in the case of patients with less severe forms of stroke and in particular when comparatively young age and reduced severity come together, as is the case for subarachnoid hemorrhage (SAH) (see Chap.​ 10). SAH is a relatively rare type of stroke with a worldwide incidence of 9/100,000 per year [48]. Its economic and social impact is out of proportion to its incidence, as it predominately affects younger people who are still working and may have small children [49]. The average age at which people suffer SAH is 52–55 years as compared to an average of 73–75 years for stroke patients [50–52]. Half of those patients who experience a SAH will stop working, work shorter hours, or accept a position with less responsibility [53]. Consequently, HRQoL is significantly reduced post-SAH. Up to 55 % of SAH patients experience substantial reductions in their HRQoL [54, 55]. However, as we found in a recent meta-analysis, traditional predictors for post-stroke HRQoL—such as severity of hemorrhage, gender, physical and cognitive disabilities, and age—do not predict HRQoL very well [56]. The only factor that proved significant was physical disability. Physical disability could account for 40 % of physical QoL and 10 % mental QoL. This means that in the case of mental QoL 90 % of the variability is unaccounted for. Two candidates deserve our attention: hormonal dysfunction [55, 57] and PTSD [56, 58, 59]. Up to 47 % of SAH patients can experience hypopituitarism [57]—a dysfunction that impacts both mental and physical HRQoL [55]. PTSD should also be taken into consideration. SAH is a traumatic experience. Half of all patients die; those who survive will often undergo surgery [60]. In fact, PTSD—which is characterized by intrusive thoughts and memories, emotional numbing, and persistent overarousal—can be found in about a third of all patients (as compared to a lifetime prevalence of 1–8 % in the general population) [56, 58, 59, 61]. We recently found in a prospective sample of 105 patients, that PTSD is by far the best predictor of the patients’ mental HRQoL, explaining up to 48 % of its variance [61]. Importantly, PTSD is not restricted to SAH but can also occur after other types of stroke [31].

Identifying predictors of post-stroke HRQoL is important because these factors might be modulated by prevention or intervention programs, thereby reducing their detrimental effect on the patient’s HRQoL. Post-stroke depression is an obvious target for treatment and in fact a few randomized placebo-controlled studies have demonstrated the benefit of antidepressant medication for patients with post-stroke depression [62, 63]. PTSD in SAH patients is another candidate. We found that it is typically those events and experiences, which occur in the hours and days after the onset of the SAH (such as the waking up in a neurological ward, facing up to a life with disabilities, or learning about the diagnosis), that are experienced as most traumatic by SAH patients [49]. This suggests that supervision and counseling of patients during that period might reduce the likelihood of later developing PTSD. Furthermore stress-coping skills have been identified as the prime determinant of the later development of PTSD in SAH patients [61]. Again training in coping skills (already available for other groups of patients with PTSD) might also be beneficial for SAH patients. Currently these suggestions are speculative but they offer specific opportunities for interventions that should be explored in future studies.

On discharge from the hospital, many stroke victims require continued care. This care is overwhelmingly provided by informal caregivers such as family members [64]. The majority of caregivers are female (typically the spouses of male stroke victims) and they tend to be younger than the patients in their care (on average between 50 and 60 years old) [65]. Since patients can require support for many years [64, 66], the care provided by family members constitutes a considerable commitment in personal and economic terms. It has been estimated that in 23 % of cases, informal caregivers lose up to 30 weeks from work, and in 77 % of cases, they lose up to 16 weeks [67]. Less quantifiable, but more problematic, are the emotional costs of caring for a stroke patient. Anxiety and depression are common in caregivers of stroke patients, with an estimated prevalence of 20 % [68]. As a consequence of their role as a caregiver they often feel emotionally drained, socially isolated, and stressed [35]. It is not surprising, therefore, that the caregiver’s HRQoL is significantly reduced [69–73]. Furthermore, poor psychological health in caregivers translates into a worse functional outcome for the patients in their care [74]. Unsupported caregivers might in the long term have to withdraw from their role as caregivers and stop providing a service that saves society substantial economic costs [75].