Comparing health status in Belarus between 2001-10: a novel method using surveys with different response categories

Abstract

Background: Two population surveys were conducted in Belarus: The Living Conditions, Lifestyle and Health (LLH) in 2001 (n=2000) and The Health in Times of Transition (HITT) in 2010 (n=1800). Each survey included a question on health status. The LLH questionnaire provided a 4-point Verbal Response Scale, but the HITT questionnaire used a 5-point scale. When translated into Russian, only two response categories of these scales had identical wording. These differences made a direct comparison of self-reported health status between 2001 and 2010 difficult. Methods: We conducted a Health Category Response Scale (HCRS) survey in 2010 (n=570) using a 100ths graduated Visual Analogue Scale (VAS) to understand how the response categories of different scales are perceived by Russian speakers. We implemented the HCRS survey’s data to calculate the weighted health status (WHS) for each of the original surveys and to compare health status in Belarus between 2001 and 2010. Results: The WHS in Belarus showed a small, but statistically significant, improvement of 2.9 points on a 0–100 scale between 2001 and 2010 (56.2 vs. 59.1). Identical response categories were perceived differently on a 4-point and 5-point VAS. The category ‘good’ (‘Umomwee’) measured ~12 points higher, and the category ‘bad/poor’ (‘Ojmtme’) measured ~16 points lower, on the 4-point compared with the 5-point VAS. Conclusion: Our HCRS survey and novel method enabled a direct comparison of questions with different response options. When applied to the LLH and HITT projects, we concluded that health status in Belarus has improved between 2001 and 2010.