In a recently released article in Pediatrics (10.1542/peds.2017-1571), Jeremy Feng and colleagues from Boston Children’s Hospital examine the rate of 30 day readmission for mental health related diagnoses compared to all other diagnoses. This is a profoundly important paper that gives us information that goes beyond surprising to shocking, and I believe every pediatrician will benefit from learning what the authors have carefully and clearly presented. Although it often feels like pediatric medical floors have become de facto holding units for distressed youth who have attempted suicide and are awaiting psychiatric placement, the numbers Feng et al report are nonetheless overwhelming. Can it possibly be true that 18.7% of pediatric hospital stays are now for mental health diagnoses, and that in the last two decades mood disorders have overtaken asthma as the most common reason for pediatric hospitalization? The authors examined rates of hospitalization and readmission for all diagnoses, and by mental health versus non-mental health diagnoses, for 5-18 year olds in 22 states during the period January - November 2014, which included 253,309 index readmissions (planned readmissions were excluded).
I hope I have set you up sufficiently that you are not surprised that the Feng et al study found that the rate of mental health related readmissions (8.0%) was significantly higher than that for non-mental health diagnoses (6.2%, p<0.001). And of those readmitted within 30 days with a mental health diagnosis almost all (98.4%) were readmitted for a primary mental health diagnosis. But I also hope you are eager to read the study for details, many of which are not readily predicted. I would have guessed that hospitals that have a higher rate of mental health-related admissions would have a lower rate of re-admissions for mental health diagnoses, but I was wrong. While rates of hospital readmission for transcatheter aortic valve replacement (TAVR), for example, are inversely correlated with hospital volume, suggesting that greater experience with the procedure and underlying illness leads to greater expertise, which benefits patients (Khera et al JAMA Cardiol 2017), this was not true for pediatric mental health conditions. Hospitals with high volumes or high proportions of mental health-related admissions did not have significantly lower readmission rates for these conditions than hospitals with low volumes or low proportions of mental health-related admissions. In fact adjusted readmission rates varied significantly between hospitals even after taking case mix into account. And unfortunately, the authors found evidence of race/ethnicity-related disparities in readmission rates that were not readily explained by other factors.
The good news that I can take home from this otherwise highly disconcerting study is that there is a lot of room for improvement, hopefully with “low hanging fruit” ready to be addressed. While there is no evidence that hospital-level experience (high volumes or high proportions of mental health-related admissions) helps patients, there are also unexplained variations between hospitals. This suggests that any reasonable quality improvement (QI) initiative would be likely to have a meaningful impact – for example, simply having an outpatient appointment made prior to discharge. The beauty of system-wide QI initiatives is that these could potentially also impact prevailing racial disparities and inequities, since by definition QI serves all that are eligible. We have a long way to go in order to provide each affected youth with the mental health care he or she needs. Articles like this one are both a “wake up call” about the magnitude of the problem we face and also give us ideas for a path forward.