Limits...
Are patients' judgments of health status really different from the general population?

Krabbe PF, Tromp N, Ruers TJ, van Riel PL - Health Qual Life Outcomes (2011)

Bottom Line: Many studies have found discrepancies in valuations for health states between the general population (healthy people) and people who actually experience illness (patients).In addition, effect of being member of one of the two patient groups and the effect of the assessment of the patients' own health status was analyzed.Except for some moderate divergence, no differences were found between patients and healthy people for the ranking task or for the VAS.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Epidemiology, Unit Health Technology Assessment, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands. p.f.m.krabbe@epi.umcg.nl

ABSTRACT

Background: Many studies have found discrepancies in valuations for health states between the general population (healthy people) and people who actually experience illness (patients). Such differences may be explained by referring to various cognitive mechanisms. However, more likely most of these observed differences may be attributable to the methods used to measure these health states. We explored in an experimental setting whether such discrepancies in values for health states exist. It was hypothesized that the more the measurement strategy was incorporated in measurement theory, the more similar the responses of patients and healthy people would be.

Methods: A sample of the general population and two patient groups (cancer, rheumatoid arthritis) were included. All three study groups judged the same 17 hypothetical EQ-5D health states, each state comprising the same five health domains. The patients did not know that apart from these 17 states their own health status was also included in the set of states they were assessing. Three different measurement strategies were applied: 1) ranking of the health states; 2) placing all the health states simultaneously on a visual analogue scale (VAS); 3) separately assessing the health states with the time trade-off (TTO) technique. Regression analyses were performed to determine whether differences in the VAS and TTO can be ascribed to specific health domains. In addition, effect of being member of one of the two patient groups and the effect of the assessment of the patients' own health status was analyzed.

Results: Except for some moderate divergence, no differences were found between patients and healthy people for the ranking task or for the VAS. For the time trade-off technique, however, large differences were observed between patients and healthy people. The regression analyses for the effect of belonging to one of the patient groups and the effect of the value assigned to the patients' own health state showed that only for the TTO these patient-specific parameters did offer some additional information in explaining the 17 hypothetical EQ-5D states.

Conclusions: Patients' assessment of health states is similar to that of the general population when the judgments are made under conditions that are defended by modern measurement theory.

Show MeSH

Related in: MedlinePlus

Mean scores (added with standard error of means) of the set of EuroQol-5D health states derived by three different measurement methods (ranking, VAS, TTO) presented for the general population and for the two patient groups (For the VAS and the TTO the EuroQol-5D state '11111' is set to 1.0 and the condition 'dead' to 0.0 by definition).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3113921&req=5

Figure 1: Mean scores (added with standard error of means) of the set of EuroQol-5D health states derived by three different measurement methods (ranking, VAS, TTO) presented for the general population and for the two patient groups (For the VAS and the TTO the EuroQol-5D state '11111' is set to 1.0 and the condition 'dead' to 0.0 by definition).

Mentions: We found almost parallel lines between the three study groups for the mean ranking scores of the assessed hypothetical health states (Figure 1A). The patients' own state was ranked as less severe than state '11312' by cancer patients and as almost comparable to this state by the RA group. It is also clear that cancer patients and RA patients ranked state '21111' (some mobility problems) as less severe than healthy people did. In the comparison of the VAS values, RA patients show a pattern closely resembling the general population (Figure 1B). For the states with only one domain at level 2, however, it seems that RA patients assign slightly higher values to these states. Compared with the general population, cancer patients seem to respond more negatively to health states associated with problems in the domains of pain/discomfort and anxiety/depression. Apart from the deviation shown by the cancer group, a gradient decline can be observed over the 17 EQ-5D states. The TTO values (Figure 1C) show higher patient values for almost all health states. Differences among the three study groups are substantially greater for the TTO data than for the rank and VAS data. Furthermore, the TTO values for the EQ-5D health states cannot be described as a gradient decline; the plot looks more like a step function.


Are patients' judgments of health status really different from the general population?

Krabbe PF, Tromp N, Ruers TJ, van Riel PL - Health Qual Life Outcomes (2011)

Mean scores (added with standard error of means) of the set of EuroQol-5D health states derived by three different measurement methods (ranking, VAS, TTO) presented for the general population and for the two patient groups (For the VAS and the TTO the EuroQol-5D state '11111' is set to 1.0 and the condition 'dead' to 0.0 by definition).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3113921&req=5

Figure 1: Mean scores (added with standard error of means) of the set of EuroQol-5D health states derived by three different measurement methods (ranking, VAS, TTO) presented for the general population and for the two patient groups (For the VAS and the TTO the EuroQol-5D state '11111' is set to 1.0 and the condition 'dead' to 0.0 by definition).
Mentions: We found almost parallel lines between the three study groups for the mean ranking scores of the assessed hypothetical health states (Figure 1A). The patients' own state was ranked as less severe than state '11312' by cancer patients and as almost comparable to this state by the RA group. It is also clear that cancer patients and RA patients ranked state '21111' (some mobility problems) as less severe than healthy people did. In the comparison of the VAS values, RA patients show a pattern closely resembling the general population (Figure 1B). For the states with only one domain at level 2, however, it seems that RA patients assign slightly higher values to these states. Compared with the general population, cancer patients seem to respond more negatively to health states associated with problems in the domains of pain/discomfort and anxiety/depression. Apart from the deviation shown by the cancer group, a gradient decline can be observed over the 17 EQ-5D states. The TTO values (Figure 1C) show higher patient values for almost all health states. Differences among the three study groups are substantially greater for the TTO data than for the rank and VAS data. Furthermore, the TTO values for the EQ-5D health states cannot be described as a gradient decline; the plot looks more like a step function.

Bottom Line: Many studies have found discrepancies in valuations for health states between the general population (healthy people) and people who actually experience illness (patients).In addition, effect of being member of one of the two patient groups and the effect of the assessment of the patients' own health status was analyzed.Except for some moderate divergence, no differences were found between patients and healthy people for the ranking task or for the VAS.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Epidemiology, Unit Health Technology Assessment, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands. p.f.m.krabbe@epi.umcg.nl

ABSTRACT

Background: Many studies have found discrepancies in valuations for health states between the general population (healthy people) and people who actually experience illness (patients). Such differences may be explained by referring to various cognitive mechanisms. However, more likely most of these observed differences may be attributable to the methods used to measure these health states. We explored in an experimental setting whether such discrepancies in values for health states exist. It was hypothesized that the more the measurement strategy was incorporated in measurement theory, the more similar the responses of patients and healthy people would be.

Methods: A sample of the general population and two patient groups (cancer, rheumatoid arthritis) were included. All three study groups judged the same 17 hypothetical EQ-5D health states, each state comprising the same five health domains. The patients did not know that apart from these 17 states their own health status was also included in the set of states they were assessing. Three different measurement strategies were applied: 1) ranking of the health states; 2) placing all the health states simultaneously on a visual analogue scale (VAS); 3) separately assessing the health states with the time trade-off (TTO) technique. Regression analyses were performed to determine whether differences in the VAS and TTO can be ascribed to specific health domains. In addition, effect of being member of one of the two patient groups and the effect of the assessment of the patients' own health status was analyzed.

Results: Except for some moderate divergence, no differences were found between patients and healthy people for the ranking task or for the VAS. For the time trade-off technique, however, large differences were observed between patients and healthy people. The regression analyses for the effect of belonging to one of the patient groups and the effect of the value assigned to the patients' own health state showed that only for the TTO these patient-specific parameters did offer some additional information in explaining the 17 hypothetical EQ-5D states.

Conclusions: Patients' assessment of health states is similar to that of the general population when the judgments are made under conditions that are defended by modern measurement theory.

Show MeSH
Related in: MedlinePlus