Home - Patient Safety

Never-Event Reporting

The HHS 2010 Report

Hospitals do not report adverse events 93% of the time. When they do, they do not report accurately. This is according to the March 2010 Report from the Office of the Inspector General of the US Department of Health and Human Services. Below are two summaries of the report - this site's and theirs.

Adverse Events in Hospitals:  Methods for Identifying Events (OEI-06-08-00221)

This Site's Summary

Even though 98 percent of hospitals claim to have adverse event reporting systems, the hospitals reviewed did not generate incident reports for 93 percent of adverse events, including events involving death or disability. During the 7% of the time when they did generate incident reports, patient diagnosis codes were inaccurate or absent 63% of the time. That means that accurate reports were filed on adverse events only slightly more than 2% of the time.

Those inaccuracies and absences of information could prevent Medicare from identifying hospital acquired conditions and result in Medicare overpayments. That makes billing data an inaccurate means of monitoring quality in hospitals. Hospitals appear safer than they are. And the existing record is an unreliable source of information about how to fix and survive health care.

The solution offered in this report is for the Association of Healthcare Research and Quality (AHRQ) to convey to hospitals the importance of having strong internal incident-reporting procedures. Given the importance that everyone in medicine places on avoiding liability by not reporting, that is like conveying to bank robbers the importance of obeying the law.

They are not the only ones who believe that a little furniture polish on rotten wood will solve the problem. The Centers for Medicare & Medicaid Services (CMS) agrees with the recommendation. That neither they nor HHS knows better suggests that they are unworthy caretakers of the safety of patients. Especially since, even after this report, they continue to rely on hospital records without demonstrating awareness of how many adverse events cannot be discovered in such records. For instance, when would existing hospital records give evidence of the abuse of a child, or anyone else? We know of cases the likes of which they never will know because of how good the medical profession is at covering up. To find out about them, they would have to go to patients.              [p. 9]

This isn't the first time these low report rates have been determined (see medical reporting). And it won't be the last. But if anyone in government or health care starts incorporating this important information into how health care should be run, it will be for the first time. Instead we get people like Linda Fishman, a Senior Vice President in charge of Public Policy Analysis and Development  at the American Hospital Association, saying that actually there is quite a bit of information available to patients to help them evaluate their health care. She points out the availability of medical records in California and how little patients have made use of them. These would be the records in which 2% of adverse events are reported, the most important information patients need to know.

And she points to Hospital Compare as another good source of information for patients, apparently without being aware of what a poor and limited source of information it is (see Hospital Compare), in part because its information is based on records in which only 2% of adverse events are reported, but also because most care happens outside of hospitals. Even within hospitals it does not tell patients the success rates of individual surgeons or radiologists or any others to whom patients might be sent. So what if the record makes it appear that a specific hospital has a stellar reputation. Is the surgeon to which the patient is being sent someone with a history of being under the influence on the job and a string of unnecessarily disabled patients in his or her wake? If you think there would be evidence of that in the record, you are getting too much information from people with vested interests and too little from injured patients. For instance, see blacklisting patients. Covering up is what they do. The HHS report is careful and polite about it, but confirms it none the less.

To read the full report, click the link:


*        *        *

Health and Human Service's Summary

Overall, the methods that we reviewed were useful for identifying events that harmed Medicare beneficiaries in hospitals.  These methods include:  nurse reviews of medical records, interviews of Medicare beneficiaries, two types of analysis of hospital billing data, and reviews of internal hospital incident reports.  For hospitalizations with possible events identified by the five screening methods, physicians reviewed medical records to determine whether actual events occurred.  However, physician reviewers determined that 62 percent of the possible events identified by the five screening methods were not associated with actual events. 

This report is one in a series to fulfill the requirements of the Tax Relief and Health Care Act of 2006 (the Act), which requires that the Office of Inspector General (OIG) report to Congress regarding never events among Medicare beneficiaries.  The term “never events” refers to a specific list of 28 serious reportable events developed by the National Quality Forum.  For the series of reports, we expand beyond never events to address “adverse events,” defined as “harm experienced by a patient as a result of medical care.”  This report provides an in-depth examination of the five methods used in a two-county case study for identifying possible adverse events experienced by Medicare beneficiaries. 

We also found that shortcomings in two of the methods have implications for Medicare payments and Federal initiatives to identify, track, and monitor events. First, patient diagnosis codes were inaccurate or absent for 7 of the 11 Medicare hospital-acquired conditions (HAC) identified by physician reviewers.  These problems would prevent Medicare from identifying HACs, result in Medicare overpayments, and inhibit use of billing data to monitor quality of care in hospitals.  Second, reviewed hospitals did not generate incident reports for 93 percent of the events, including some of the most serious events involving death or permanent disability to the patient.  The lack of such reports could prevent hospitals from tracking events as required by Federal regulation and suggests that hospital incident-reporting systems may be an unreliable source of information for Patient Safety Organizations (PSO), entities that aggregate and analyze hospital data about events.

We recommend that CMS and AHRQ explore opportunities to identify adverse events when conducting medical record reviews for other purposes.  Additionally, CMS should (1) ensure that hospitals code claims accurately and completely to allow for identification of Medicare HACs and (2) provide guidelines for State survey agencies that assess hospital compliance with requirements to track and monitor adverse events.  Finally, AHRQ should inform PSOs that internal hospital incident-reporting systems may be insufficient for providing information about events to PSOs.  AHRQ concurred with the report as written.  CMS agreed with each recommendation addressed to CMS.

To read the Health and Human Services pdf of the full report, click the link:


Hit your back button to return to where you were