Clinical context as a source of bias in self-rated health: evidence of examination-induced anchoring

Background

Self-rated health (SRH) is widely used in public health, yet evidence suggests psychological and contextual factors may systematically bias its responses. The Korea Community Health Survey (KCHS) and Korea National Health and Nutrition Examination Survey (KNHANES) are both nationally representative health surveys conducted through identical procedures, except that KNHANES includes clinical examination before SRH assessment. This study examined whether this context introduces bias (anchoring effect whereby salient reference points shift subsequent judgements) in SRH responses as a key public health indicator.

Methods

Using 2024 KCHS data (n = 215,536), we built a weighted ordinal logistic regression model predicting SRH (1=excellent to 5=poor) from sociodemographic and health-related variables (c-statistic=0.703). Coefficients were applied to KNHANES respondents (n = 4,826). Residuals (observed minus predicted) quantified response bias. Stratified and multivariate analyses identified subgroups and predictors.

Results

KCHS internal validation yielded a near-zero residual (−0.005, p = 0.149), confirming baseline stability. KNHANES respondents showed a positive residual (+0.151, p<.0001), indicating more conservative SRH responses than objectively predicted. This bias peaked in adults in their 30s (+0.378) and declined with age. Stratified by diagnosis status, bias was significant only without hypertension (+0.195, p<.0001) or diabetes (+0.177, p<.0001) while diagnosed patients showed no effect, suggesting those perceiving themselves as healthy are most susceptible to this bias. Age was the strongest predictor of bias magnitude (β = −0.011, p<.0001).

Conclusions

Clinical examination context anchors SRH responses, notably among younger and non-diagnosed populations, reflecting health self-concept disruption. These findings suggest that SRH as a public health indicator is not absolute, and comparisons may be confounded when survey contexts differ.

Key messages

• SRH is a trusted public health indicator, yet its validity is not context-free. This study demonstrates that clinical examination settings can introduce systematic bias in SRH responses.

• When comparing SRH across groups or nations, survey context must be standardised. Ignoring measurement environment risks drawing false conclusions from otherwise reliable health data.