US researchers have trialled a practical solution to burnout among pregnant and postpartum doctors in training, with smart bassinets, wearable breast pumps, round-the-clock virtual perinatal support and formal faculty mentorship all on offer.
A leading expert has urged hospitals to move beyond mere “accommodation” and implement similar evidence-based support models following the trial, in which high burnout rates fell by almost 17 percentage points in the intervention group while rising by more than 47 percentage points in those receiving usual care (P = .004).
The pragmatic randomised controlled trial, published in JAMA [link here] enrolled 156 pregnant residents and fellows across seven training institutions in the northeastern United States, randomising them to either the four-part parental support package or usual care.
The intervention group received a six-month lease of a Snoo smart bassinet, a Willow wearable breast pump, 24/7 virtual perinatal support via the Maven Clinic app, and a specialty-matched faculty mentor with experience of motherhood. The control group received standard program accommodations and $200 in gift cards.
The results, drawn from 143 participants who completed the trial, were stark. Mean burnout scores on the Stanford Professional Fulfillment Index rose from 3.13 to 3.79 in the usual care group over the 24 weeks to postpartum follow-up, but barely moved in the intervention group, from 2.96 to 3.03.
The adjusted between-group difference in change was -0.58 (95% CI, -1.10 to -0.07; P = .03), giving an effect size (Cohen d) of 0.65, which the authors said exceeded their prespecified threshold for a clinically meaningful result.
In adjusted analyses, the support package was linked to substantially lower odds of high burnout (OR, 0.19; 95% CI, 0.06-0.65).
The benefit was driven largely by improvements in interpersonal disengagement rather than emotional exhaustion, which the authors said pointed to a possible mechanism.
“Disengagement responds to community, shared purpose, and values alignment,” they wrote, noting that only 17% of usual care participants independently organised meetings with a faculty mentor.
Participants given the package also reported less strain on their personal relationships than those in usual care (adjusted difference, -0.90; 95% CI, -1.77 to -0.03; P = .04).
The authors estimated the package cost about $2300 per participant, covering the bassinet lease, breast pump and app subscription. They argued this was modest set against the estimated $7600 a year in turnover and lost productivity costs attributed to burnout in each employed physician.
“Childbearing trainees are a high-opportunity target for enterprise-level burnout reduction,” the authors wrote, adding that prior research in the field had “extensively documented challenges without testing solutions.”
They also noted the package’s effect size compared favourably with the broader literature, given that “most trials of physician burnout interventions report small or nonsignificant effects.”
Editorial urges caution on mentor reliance
In an accompanying editorial [link here], JAMA deputy editor Dr Linda Brubaker welcomed the findings as “a foundational step”.
She noted the trial could not isolate which of the four components drove the benefit, since “many intervention features were available to the control group as well.” More than half of usual care participants independently obtained at least one component of the package, although the virtual support app was used almost exclusively by the intervention group.
Dr Brubaker suggested this overlap likely worked against the intervention. “It is likely that the study results would have been even stronger without contamination of the control group by independent use of intervention components,” she wrote.
She also questioned whether unpaid mentorship was a sustainable model for wider rollout. “Reliance on volunteer faculty mentors is likely not sufficient for teaching hospitals seeking to improve parental support,” she wrote, arguing mentors should be compensated and supported with training.
Dr Brubaker pointed to a finding that a quarter of participants reported their training program had no written parental support policy, calling this “an opportunity to develop and ensure awareness of parental support policies” that hospitals could act on immediately.
She closed with a warning to training programs that change was overdue. “The clock for parental supportive action in graduate medical education is ticking—loudly,” she wrote.
The authors acknowledged their own limitations, including that the trial was conducted only in the northeastern US, that participants and staff were aware of group allocation, and that adverse events were not formally collected given what they described as the low-risk nature of the intervention. They called for a larger, more diverse national trial before broader implementation.