Scholarlink

Background <p>Monitoring of healthcare quality is typically focused on differences between group averages in relation to a desirable benchmark. However, we need to consider (<i>i</i>) the existence of interconnected socioeconomic axes of inequality like age, sex, income, and country of birth and (<i>ii</i>) individual heterogeneity around group averages. Additionally, (<i>iii</i>) we need clear criteria to quantify group differences. By applying the framework <i>analysis of individual heterogeneity and discriminatory accuracy</i> (AIHDA) on an established quality indicator (potentially inappropriate medication (PIM)), we illustrate how to achieve these improvements and how to avoid both unnecessary group stigmatization and false expectations.</p> Methods <p>We analyzed 731,339 individuals, <i>≥</i> 75-year-old belonging to 36 socioeconomic strata defined by the intersection of age, sex, income, and country of birth, who were alive and residing in the 21 regions Swedish during 2011. We calculated PIM prevalences and evaluate the discriminatory accuracy (DA) of the socioeconomic and geographical group differences using the area under the ROC curve (AUC). The benchmark value was defined as a prevalence of 19%.</p> Results <p>In Sweden, the prevalence of PIM was 24% among <i>≥</i> 75-year-olds and regionally it ranged between 21% and 27%. Immigrant 80–84-year-old women with low income had the highest prevalence (29%). All strata including women had higher prevalence than those including men. However, the regional (AUC = 0.520) and socioeconomic (AUC = 0.544) differences were very small. For instance, in the five socioeconomic strata with the lowest prevalence there were about 8,000 more cases of PIM than in the five strata with the highest prevalence of PIM.</p> Conclusion <p>The prevalence of PIM was higher than the desired benchmark value. There were disparities between group averages, but overall, the regional and socioeconomic differences were very small as informed by their low AUC values. Therefore, interventions to reduce PIM in Sweden should be universal rather than only targeted at the regions and socioeconomic strata with the highest PIM prevalence.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Geographic and socioeconomic differences in potentially inappropriate medication among older adults – applying a simplified analysis of individual heterogeneity and discriminatory accuracy (AIHDA) for basic comparisons of healthcare quality

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月28日

查看全文
  • Johan Öberg,
  • Kani Khalaf,
  • Raquel Perez-Vicente,
  • Kristina Johnell,
  • Johan Fastbom,
  • Juan Merlo

摘要

Background

Monitoring of healthcare quality is typically focused on differences between group averages in relation to a desirable benchmark. However, we need to consider (i) the existence of interconnected socioeconomic axes of inequality like age, sex, income, and country of birth and (ii) individual heterogeneity around group averages. Additionally, (iii) we need clear criteria to quantify group differences. By applying the framework analysis of individual heterogeneity and discriminatory accuracy (AIHDA) on an established quality indicator (potentially inappropriate medication (PIM)), we illustrate how to achieve these improvements and how to avoid both unnecessary group stigmatization and false expectations.

Methods

We analyzed 731,339 individuals, ≥ 75-year-old belonging to 36 socioeconomic strata defined by the intersection of age, sex, income, and country of birth, who were alive and residing in the 21 regions Swedish during 2011. We calculated PIM prevalences and evaluate the discriminatory accuracy (DA) of the socioeconomic and geographical group differences using the area under the ROC curve (AUC). The benchmark value was defined as a prevalence of 19%.

Results

In Sweden, the prevalence of PIM was 24% among ≥ 75-year-olds and regionally it ranged between 21% and 27%. Immigrant 80–84-year-old women with low income had the highest prevalence (29%). All strata including women had higher prevalence than those including men. However, the regional (AUC = 0.520) and socioeconomic (AUC = 0.544) differences were very small. For instance, in the five socioeconomic strata with the lowest prevalence there were about 8,000 more cases of PIM than in the five strata with the highest prevalence of PIM.

Conclusion

The prevalence of PIM was higher than the desired benchmark value. There were disparities between group averages, but overall, the regional and socioeconomic differences were very small as informed by their low AUC values. Therefore, interventions to reduce PIM in Sweden should be universal rather than only targeted at the regions and socioeconomic strata with the highest PIM prevalence.