Who should assess quality of life

The options for collecting QL lie between the patient themselves and a proxy (the doctor, nurse, or the so-called significant other (spouse, son, daughter, sibling, partner, etc.)) and a number of studies have investigated the differences obtained if the same questions are asked of the patients and a proxy.

Although two studies showed that there was about a 70 per cent complete agreement between patients and proxies in terms of recording the severity of a symptom [14,15], most of the disagreement was a result of proxies consistently under-estimating the severity of symptoms [16]. However, patient/proxy comparisons can only be made on patients who are capable of completing QL forms, whereas the real questions are whether proxies can substitute for patients (a) throughout the trial, or (b) when patients are unable to complete forms (for example if they have cognitive impairment, communication deficiencies, are experiencing severe symptoms, or find the forms physically or emotionally too burdensome).

■ Physical condition (patient)

□ Overall condition (physician)

Improved Same Worse

Change at one month compared to presentation

Fig. 6.1 Changes in overall life quality and physical condition related by the patients together with changes in MRC grade for general condition rated by their physicians over one month following radiotherapy (adapted from Regan et al., [18]).

Sprangers and Aaronson [17] reviewed the role of proxies and summarized it as follows:

♦ health care providers (doctors, nurses, etc.) and significant others (spouses, siblings, etc.) tend to under-estimate (down-grade) the patient's QL,

♦ health care providers tend to evaluate the patient's QL with a similar level of (in)accuracy as significant others,

♦ health care providers tend to under-estimate the pain severity of their patients,

♦ proxy ratings appear to be more accurate when the information sought is concrete and observable (for example physical symptoms rather than psychological distress).

Thus on the surface the role of proxies appears minimal. Nevertheless, although doctors appear to be poor judges of QL, they do seem able to identify improvement or deterioration over time (Fig. 6.1) [18]. Similarly, although Sneeuw et al. [16] showed statistically significant differences between patients and proxies for mean scores in nine of the 15 EORTC QLQ-C30 domains (proxies nearly always reporting worse functioning or increased severity of symptoms), they also showed that data from patients and proxies were equally responsive to change over time. Therefore, if proxies are used throughout the trial to assess QL, similar between-treatment differences can often be observed.

0 0

Post a comment