Brady Keri J S, Sheldrick R Christopher, Ni Pengsheng, Trockel Mickey T, Shanafelt Tait D, Rowe Susannah G, Kazis Lewis E
Department of Health Law, Policy & Management, Boston University School of Public Health, Boston, MA, 02118, USA.
Biostatistics & Epidemiology Data Analytic Center, Boston University School of Public Health, Boston, MA, USA.
J Patient Rep Outcomes. 2021 Jun 5;5(1):43. doi: 10.1186/s41687-021-00312-2.
Disparities in US physician burnout rates across age, gender, and specialty groups as measured by the Maslach Burnout Inventory-Human Services Survey for Medical Personnel (MBI) are well documented. We evaluated whether disparities in US physician burnout are explained by differences in the MBI's functioning across physician age, gender, and specialty groups.
We assessed the measurement equivalence of the MBI across age, gender, and specialty groups in multi-group item response theory- (IRT-) based differential item functioning (DIF) analyses using secondary, cross-sectional survey data from US physicians (n = 6577). We detected DIF using two IRT-based methods and assessed its impact by estimating the overall average difference in groups' subscale scores attributable to DIF. We assessed DIF's practical significance by comparing differences in individuals' subscale scores and burnout prevalence estimates from models unadjusted and adjusted for DIF.
We detected statistically significant age-, gender-, and specialty- DIF in all but one MBI item. However, in all cases, average differences in expected subscale-level scores due to DIF were < 0.10 SD on each subscale. Differences in physicians' individual-level subscale scores and burnout symptom prevalence estimates across DIF- adjusted and unadjusted IRT models were also small (in all cases, mean absolute differences in individual subscale scores were < 0.04 z-score units; prevalence estimates differed by < 0.70%).
Age-, gender-, and specialty-related disparities in US physician burnout are not explained by differences in the MBI's functioning across these demographic groups. Our findings support the use of the MBI as a valid tool to assess age-, gender-, and specialty-related disparities in US physician burnout.
通过针对医务人员的马氏职业倦怠量表 - 人类服务调查（MBI）衡量，美国医生职业倦怠率在年龄、性别和专业组之间的差异已有充分记录。我们评估了美国医生职业倦怠的差异是否可由MBI在医生年龄、性别和专业组中的功能差异来解释。
我们使用来自美国医生的二次横断面调查数据（n = 6577），在基于多组项目反应理论（IRT）的差异项目功能（DIF）分析中评估MBI在年龄、性别和专业组之间的测量等效性。我们使用两种基于IRT的方法检测DIF，并通过估计归因于DIF的组分量表分数的总体平均差异来评估其影响。我们通过比较个体量表分数的差异以及未调整和调整DIF的模型中的倦怠患病率估计值，评估DIF的实际意义。
除一个MBI项目外，我们在所有项目中均检测到具有统计学意义的年龄、性别和专业DIF。然而，在所有情况下，由于DIF导致的预期分量表水平分数的平均差异在每个分量表上均<0.10标准差。经DIF调整和未调整的IRT模型之间，医生个体水平的分量表分数和倦怠症状患病率估计值的差异也很小（在所有情况下，个体分量表分数的平均绝对差异<0.04 z分数单位；患病率估计值相差<0.70%）。
美国医生职业倦怠中与年龄、性别和专业相关的差异不能由MBI在这些人口统计学组中的功能差异来解释。我们的研究结果支持将MBI用作评估美国医生职业倦怠中与年龄、性别和专业相关差异的有效工具。
