The Validation of Conducting Retrospective Chart Reviews by Nurses to Detect Adverse Events

Abstract

The purpose of this study was to confirm the validity of conducting chart reviews by nurses in Japan to detect one or more of 18 explicit criteria described in “The Quality in Australian Health Care Study” as indicators of potential adverse events.

A sample of 200 records of hospitalized patients from a population of non-psychiatric patients discharged from two private hospitals was used in the study. Initially, the charts were screened by trained RNs. To test the validity of the screening process conducted by the RNs, all the records were subsequently reviewed by physicians, who also determined whether an adverse event had occurred. In Japan, we adopted the methodology that a nurse supervisor examines only cases screened as negative after the trained nurses conducted retrospective chart review in the first stage.

The sensitivity and specificity of the screening process conducted by the nurses were 100.0% and 60.2%, respectively ; this specificity was lower than that of the screening process conducted by the Australian nurses in the study mentioned above. One reason for this discrepancy is that the operational definition of intensive medical treatment and care-to-cure patient injuries caused by health care was unclear ; thus, nurses had difficulty screening for potential adverse events.

In conclusion, the specificity of the chart screening process conducted by nurses must be improved by presenting actual cases to clarify the definitions of adverse events, while maintaining the high sensitivity.