Only a small portion of readmissions within 30 days are truly preventable. Rather, readmissions depend more on community factors outside the hospital's control, they wrote. One study showed that fewer than 20% of readmissions are preventable.
Safety net hospitals have higher rates of readmission and will be penalized more under the readmission penalty rule. "We know that some of the most important drivers of readmissions are mental illness, poor social support, and poverty, which are often deeply ingrained," Joynt and Jha wrote. Thus, "the current scheme to penalize hospitals with high readmission rates is likely to disproportionately affect institutions that care for poor or minority populations or those with a high burden of mental illness."
Rather than reflecting poor quality of care, a readmission may in fact be a life-saving action, providing "good access to hospital care." Referencing research by Eiran Gorodeski, MD and colleagues at the Cleveland Clinic, published in the New England Journal of Medicine, Joynt and Jha wrote, hospitals with low heart failure mortality rates have higher readmission rates, "presumably because they keep their sickest patients alive, and those patients are subsequently more likely to be readmitted."
Too much emphasis on reducing readmissions may be a misdirection of efforts. When resource-poor hospitals expend so much effort reducing readmissions, "they have probably forgone quality-improvement efforts related to more urgent issues, such as patient safety."
"Many of these deaths are preventable, yet we are focusing tremendous resources on preventing rehospitalizations for three conditions that account for approximately 10% of all hospital admissions in the Medicare population," they wrote. "Even if CMS expands its readmission penalties to include more conditions, the policy will fail to address what patients care about most."
Of course, preventing readmissions is a worthy goal and it will save a ton of money, $7.1 billion over 10 years according to the Congressional Budget Office.