Back

Meta-analysis of studies on the association between physician characteristics and patient outcomes

Haslam, A.; Miller, S.; Prasad, V.

2025-01-23 health systems and quality improvement
10.1101/2025.01.22.25320956 medRxiv
Show abstract

ObjectivePhysicians have a large role in patient outcomes, and some have questioned whether some physician characteristics lead to better outcomes for the patient. Multiple studies have been conducted on the topic, leading to a variety of conclusions. To determine how frequently studies find favorable patient outcomes, by physician characteristics and the pooled estimated effect. MethodsBy searching PubMed, Web of Science, and Google Scholar, we identified studies that reported the association between physician characteristics (age, experience, sex/gender, race/ethnicity, nationality, and patient/physician characteristic concordance) and patient outcomes (patient outcomes, such as survival/mortality, hospitalization, readmissions, length of stay, and morbidity). We calculated the frequency of studies reporting on each type of physician characteristics and each type of patient outcome, risk of bias in included studies, and pooled estimates of association between physician characteristics and patient outcomes. ResultsOf the 32 included studies, the most common outcome was mortality. There were no differences in outcomes for physicians with more experience, compared with less experience (pooled odds ratio [OR]: 0.99; 95% confidence interval: 0.98 to 1.01; I2: 56.5%), between older and younger physicians (pooled OR: 1.01; 95% confidence interval: 0.97 to 1.05; I2: 68.1%), but there were significantly better outcomes for patients with a female physician, compared to patients with a male physician, although this difference was small and biased (pooled OR: 0.97; 95% confidence interval: 0.96 to 0.98; I2: 35.5%). Selection bias, unaccounted measurement in physician and patient characteristics, and not accounting for unintended exposures limit the findings of these types of studies. ConclusionStrategies to improve patient outcomes should focus on other factors. Moreover, when seeking care, patients should be more concerned about physician track record than non-modifiable characteristics.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.