There is low variation in readmission rates among primary care physicians (PCPs), according to a study recently published in the Annals of Internal Medicine. This may lead to ineffectiveness in programs aiming to hold PCPs accountable for their readmission rates.

This retrospective cohort study included 4230 PCPs and 565,579 hospital admissions between 2012 and 2015 in Texas. Eligibility criteria included patients who were discharged alive, had a PCP in the previous year, and had a PCP with at least 50 admissions between January 2008 and November 2015. In addition to the 2012 to 2015 data, 3408 PCPs also had data from 2008 to 2011 available. The two datasets were compared to estimate stability in the PCP-adjusted readmission rates. The primary outcome of the study was 30-day readmission after discharge, while 7-day follow-up visits with a PCP were also recorded. Multilevel logistic regression and a generalized estimating equations model were used to predict 30-day readmission rates per patient characteristic.

The 2012 to 2015 mean rate of 30-day readmissions after standardization for risk was 12.9%. Among the entire cohort, only 1 PCP had a readmission rate significantly above the mean, and there were no PCPs with significantly lower readmission rates. PCP readmission rates were remarkably consistent, with a 10th percentile of 12.4% and a 90th percentile of 13.4% — each just 0.5% from the mean. The lowest readmission rate was 11.2%, whereas the maximum was 15.3%. The rate of 7-day post-discharge follow-up visits was 20.4% among 413,527 individuals with 3603 PCPs.

Continue Reading

Related Articles

Limitations to this study include a lack of control for patient features outside of the Medicare database, a potential lack of generalizability to younger individuals, a lack of classification of avoidable vs unavoidable readmissions, a lower rate of readmission than national figures, and the exclusive use of participants in Texas.

The study researchers concluded that “this study shows negligible variation in readmission rates among PCPs when patient characteristics are controlled for. PCPs do not generate sufficient admissions to expect reasonable power to detect differences by PCPs or by most group practices. Therefore, pay-for-performance programs to reduce readmissions on the basis of variation in readmission rates among PCPs may not be effective.”


S Singh reports personal fees from AstraZeneca.


Singh S, Goodwin JS, Zhou J, Kuo Y, Nattinger AB. Variation among primary care physicians in 30-day readmissions [published online May 21, 2019]. Ann Intern Med. doi: 10.7326/M18-2526