To the Editor:

Herzig et al. are to be commended on their large prospective multicenter study of readmissions1—a problem which is common, costly, complex (multifactorial), and often preventable.2 However, one of their main findings, that patient factors have the major role in readmissions as judged by physicians’ perceptions, may not be as sound as it seems.

The all too human tendency to assign flaws to others while having an exaggerated self-opinion and failing to acknowledge weaknesses in their own field may have skewed the results, with significant implications. This is strongly supported when noting that primary physicians perceived “problems with the index admission” (i.e. hospital-associated quality of care [QOC] issues) in over a third of the patients, much more frequently than the discharging physicians were critical of their own performance regarding the 993 readmitted patients evaluated (36.5 vs. 14%, χ2 = 67.68, P < 0.01)Footnote 1

Indeed, when a mix of 271 consecutive unplanned readmissions to a general medical service and matched controls who were not readmitted was blindly evaluated in our prospective study, 90 readmitted patients (33%) had quality of care problems vs. 6% among controls (P < 0.001).3 All these readmissions were deemed preventable. Other investigators have also identified a significantly increased quota of preventable readmissions associated with well-defined QOC deficiencies.4,5,6, – 7 As a result, readmission to the hospital shortly after discharge is increasingly recognized as a marker of inpatient QOC.5

This important aspect is underplayed in the current study. Efforts to reduce readmissions should focus not only on patient-related factors, but on identifying and correcting essential care and communication issues (including contextual problems) in the realm of the hospital providers. These can be identified only through a large-sample peer review process based on expert opinion.