The number of physicians working part time in the United States has increased by nearly 11% since 1993, and as more physicians opt for part-time work, quality of care deserves further study, the investigators wrote in. Most studies comparing outcomes for patients treated by full-timers and part-timers have focused on outpatient care settings, where mortality is low and the potential for confounding is high, according to the study authors Hirotaka Kato, PhD, of Keio University in Tokyo, and colleagues. The new study, in contrast, is based on data from nearly 400,000 hospitalizations.
The researchers conducted a cross-sectional analysis on a 20% random sample of Medicare patients aged 65 years and older who were treated by a hospitalist for an emergency medical condition between 2011 and 2016. They examined associations between the number of days per year worked by hospitalists and they 30-day mortality rates among the patients they treated. The researchers analyzed a total of 392,797 hospitalizations in which patients were treated by 19,170 hospitalists. The mean age of the hospitalists was 41 years; 39% were female. Clinician work days were divided into quartiles.
Overall, the 30-day mortality was significantly higher among patients treated by clinicians in the bottom quartile with the fewest number of days worked, compared with those treated by clinicians in the top quartile with the most days worked (10.5% vs. 9.6%). The rates were similar in the second and third quartiles (10.0% and 9.5%).
The average number of days worked clinically per year was 57.6 in the lowest quartile versus 163.3 in the highest quartile, a 65% difference. No significant associations were noted between days worked and patient outcomes with regard to physician age, gender, or hospital teaching status.
Hospital 30-day readmission rates were examined as a secondary outcome, but there was no association between patient readmission and the number of days worked by the clinician. The adjusted 30-day readmission rate for clinicians in the bottom quartile of days worked, compared with those in the top quartile, was 15.3% versus 15.2% (P = .61).
The researchers found no difference in patients’ severity of illness (defined by expected mortality) or reason for admission between physicians in the different quartiles of days worked. They eliminated confounding from hospital-level differences by comparing outcomes of patients between physicians in the same hospital.
Possible explanations for worse patient outcomes
“As the number of physicians who engage in part-time clinical work continues to increase, these findings should lead to careful consideration by health systems to reevaluate preventive measures to address potential unintended patient harm,” the researchers wrote.
The researchers proposed several reasons for the association between fewer clinical work days and worse patient outcomes. First, physicians putting in less clinical time may be less updated on the latest guidelines, their skills may decline with less frequent patient care, and they may be less familiar with the nurses, medical assistants, and support staff, which may contribute to poor teamwork. The researchers also stated that some part-time physicians may need to balance nonclinical responsibilities, such as research or administrative tasks, concurrently with inpatient care. “It is also possible that physicians with less clinical knowledge or skills select to become part-time physicians, whereas physicians with higher clinical performance decide to work full time,” they noted.
The study findings were limited by several factors including the observational design and potential for unmeasured confounding variables, and the results may not generalize to younger patients or surgical patients, the researchers noted. Also, the study did not include care by hospitalists that was not billed, days in which clinicians treated non-Medicare patients or patients not part of the Medicare sample, or information about the reasons for clinicians’ part-time work.
However, the results were strengthened by the large sample size, and suggest the need for better institutional support to maintain the clinical performance of physicians who may be balancing a range of obligations, they concluded.