Response to Questions From Hearing on Patient Safety and Quality of Care
at VA Facilities (Correspondence, 10/13/2000, GAO/GAO-01-123R).

This correspondence responds to follow- up questions to its July 27,
2000, testimony before Congress on patient safety and quality of care at
the Department of Veterans' Affairs (VA). The Patient Safety Centers of
Inquiry and VA's Office of Research and Development are not directly
linked organizationally. The four centers of Inquiry do not report to
and are not funded by the Office of Research and Development. The
Directors of the four centers report to either the Veterans Integrated
Service Networks or Medical Center manager where they are located. The
work at the Centers of Inquiry address some but not all of the known
principal adverse and sentinel events at VA medical facilities. The key
challenges to improving patient safety include setting goals, planning,
and communicating the priority of patient safety to its employees. VA's
patient safety program has not been fully implemented, because it is too
early to predict whether, in the final analysis, it will be a model for
other health care organizations.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  GAO-01-123R
     TITLE:  Response to Questions From Hearing on Patient Safety and
	     Quality of Care at VA Facilities
      DATE:  10/13/2000
   SUBJECT:  Health care programs
	     Veterans hospitals
	     Hospital care services
	     Patient care services
	     Safety standards
	     Reporting requirements
	     Health statistics
	     Management information systems
IDENTIFIER:  VA Combined Assessment Program
	     VA Veterans Integrated Service Network
	     VA Patient Safety Program

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************

GAO-01-123R

VA Patient Safety

United States General Accounting Office Washington, DC 20548

October 13, 2000 The Honorable Terry Everett Chairman Subcommittee on
Oversight

and Investigations Committee on Veterans' Affairs House of Representatives

Subject: Response to Questions From Hearing on Patient Safety and Quality of
Care at VA Facilities

Dear Mr. Chairman: The enclosed information responds to your follow- up
questions on our July 27, 2000, testimony for the Subcommittee on Oversight
and Investigations hearing concerning VA's Patient Safety and Quality of
Care. We will make copies of this correspondence available to other
interested parties upon request.

If you have any questions or would like to discuss this information further,
please contact me at (202) 512- 7101.

Sincerely yours, Cynthia A. Bascetta

Director, Health Care – Veterans' Health and Benefits Issues

Enclosure

ENCLOSURE ENCLOSURE GAO- 01- 123R VA Patient Safety 2 SUPPLEMENTAL
INFORMATION ON VA'S PATIENT SAFETY INITIATIVES

This enclosure presents your questions and our responses, which supplement
information in our testimony before the Subcommittee on Oversight and
Investigations, VA Patient Safety: Initiatives Promising but Continued
Progress Requires Culture Change (GAO/ T- HEHS- 00- 167, July 27, 2000).

1. Regarding VA's four Patient Safety Centers of Inquiry, how does their
role and research focus relate to the work conducted by VA's Office of
Research and Development? In what way does the work of the centers address
the principal adverse and sentinel events that were disclosed in VA's 1999
Office of the Medical Inspector's report?

The Patient Safety Centers of Inquiry and VA's Office of Research and
Development are not directly linked organizationally. The four Centers of
Inquiry do not report to and are not funded by the Office of Research and
Development. Rather, the Directors of the four centers report to either the
Veterans Integrated Service Networks (VISN) or Medical Center manager where
they are located. The Office of Research and Development staff assisted the
Under Secretary for Health in developing the Centers of Inquiry request for
proposals and subsequently helped in rating and ranking the proposals
submitted. The centers were then chosen by the Under Secretary for Health
and received $500,000 per year in funding from medical care funds. Some
staff associated with the Centers of Inquiry also have active projects
funded by Health Services Research and Development (HSR& D), which is part
of the Office of Research and Development. In addition to these projects,
HSR& D funds other patient safety research projects that are not directed by
the centers and is currently funding four projects that overlap with the
work of the centers. Two of these focus on selected aspects of adverse drug
reactions, and the other two address fall prevention in the hospital and
home environments. As the patient safety effort develops, VA told us that it
expects to articulate clear linkages between these efforts to maximize the
efficiency of this research effort.

The work at the Centers of Inquiry address some but not all of the known
principal adverse and sentinel events at VA medical facilities. In December
1999, a report by VA's Medical Inspector disclosed that between June 1997
and December 1998, the top five categories of adverse events reported by VA
facilities were patient falls, suicide and attempted suicide, other
unplanned occurrences, and patient abuse (not in rank order). Adverse events
in these five categories account for 69 percent of reported events during
the timeframe reviewed by the Medical Inspector. The Patient Safety Center
of Inquiry at the Tampa VA medical center is studying fall prevention and
injury reduction related to patient falls in high- risk populations that
have mobility problems. The Center in White River Junction is currently
working to reduce medication errors (ranked seventh on the Medical
Inspector's list) using a collaborative technique. Not targeted by the

ENCLOSURE ENCLOSURE GAO- 01- 123R VA Patient Safety 3 centers are unplanned
occurrences, patient abuse, and suicide and suicidal

behavior. Although none of the centers is currently studying suicide
prevention, VA has undertaken a variety of suicide prevention efforts since
the 1970s, including most recently its March 1, 2000, suicide summit called
"Suicide: Recognizing Risks Across Treatment Settings." 1 This program was
designed to encourage caregivers to screen patients for risk of suicide;
share resources regarding suicide awareness, prevention, and treatment; and
recommend proper treatment and referral. Suicide risk assessment pocket
cards were also printed and distributed to clinicians to use in their daily
activities. VA has also implemented depression screening in nonmental health
care settings to identify veterans at risk for suicide and to get them help
from a mental health professional. There are also 12 VA facilities with
active projects related to improved prevention and management of suicidal
behavior.

“Unplanned occurrences” is a general category for events that do
not fit neatly into one of the other categories used by VA. To work in this
area, the events would need to be further categorized. VA is no longer
handling cases of alleged patient abuse under the patient safety program and
instead has mandated that facilities take immediate administrative steps to
investigate these incidents.

2. Given GAO's early assessment of VA's efforts to improve patient safety,
what obstacles or other problems might VA have to overcome to ensure
continued progress?

As we testified, VA's initiatives to improve patient safety mirror some of
those suggested by the Institute of Medicine (IOM), but VA faces significant
challenges to ensure the success of its patient safety efforts. The key
challenges include setting goals, planning, and communicating the priority
of patient safety to its employees. Beyond that, VA needs to overcome
obstacles that impede the move from a “blame and shame” way of
handling adverse events to a culture of safety that looks openly at how and
why adverse events occur and how systems can be improved to prevent them in
the future. To make a major change in patient safety, VA must convey the
message to all its employees that patient safety is everyone's
responsibility and then put this belief into practice by making it an
integral part of everyday work. Safety experts agree that management
officialsï¿½many of whom have been trained to look for the person responsible
for an error- need to focus instead on the systems that allowed the adverse
event to occur. They must also be able to create a nonpunitive work
environment in which employees feel safe enough to report and investigate
adverse events and, even more important, close calls (situations in which an
adverse event could have occurred but did not). Because successful culture
change takes years, we believe that patience and sustained commitment in the
event of leadership changes should also be a top priority. Also, once the
culture change has taken root, VA leadership at all levels

1 In VA, a summit is a series of educational and networking events designed
to share best practices across the VA health care system.

ENCLOSURE ENCLOSURE GAO- 01- 123R VA Patient Safety 4 of the organization
must remain committed to continuous improvement in the

effort to drive the medical error rate to zero. It is also important to
point out that, in the short run, if the culture change is successful, we
would expect to see an increase in the number of reported adverse events,
and we should view this increase as a positive result of VA's efforts. As
reports of adverse events increase and more errors come to light, managing
the potential negative reaction of skeptics and highlighting the fact that
staff are able to learn from these errors will be critical to ensuring that
the patient safety program stays on track.

3. Much has been said about VA's patient safety program and some people
point to it as a model for the private sector to follow. Does GAO believe
that it is a model program?

Because VA's patient safety program has not been fully implemented, it is
too early to predict whether, in the final analysis, it will be a model for
other health care organizations. While some of VA's patient safety
initiatives are clearly exemplary, such as removing concentrated potassium
chloride from wards and bar- coding medications, these initiatives preceded
the broader patient safety program that VA is now trying to put in place. In
other words, more fundamental than such stand- alone initiatives is VA's
effort to create a systemwide culture of safety. In this key endeavor, VA is
not yet a model. While VA leads the rest of the health care sector in
adopting the right concepts and consulting with appropriate experts, we
believe that it is too early to predict if VA will be successful in creating
a culture that promotes patient safety. Until VA demonstrates results
attributable to the processes it is putting in place, others will be unable
to emulate them. In our view, therefore, calling VA a model would be
premature.

4. VA's National Center for Patient Safety promises to have a pivotal role
in the development of new systems that will be used to analyze and report on
sentinel and adverse events. If the Center's new processes are successful,
what will be the likely impact on the number of patient safety events that
are reported?

In the short run, if the processes instituted by the National Center for
Patient Safety are successful, there should be a rise in the number of
reported adverse events at VA facilities. We believe an increase in reported
events will reflect a willingness on the part of employees to report events
that have been occurring all along, not a real increase in the number of
adverse or sentinel events. In fact, we would be suspicious that the system
is not working well if the number of events is too low. The magnitude of
this potential increase is unknown but could be steep. Increased reporting
will offer employees a chance to learn from adverse events, especially close
calls, and will provide an opportunity to propose changes that can prevent
such events from happening again in the future. In the long run, more
reports should lead to a drop in preventable adverse events as proposed
solutions and action plans are disseminated systemwide to reduce or prevent
occurrences.

ENCLOSURE ENCLOSURE GAO- 01- 123R VA Patient Safety 5

5. In addition to more definitive goal setting and measurement,
comprehensive planning, and communicating the importance of the program to
all its employees, are there other actions VA can take to ensure the success
of its patient safety program?

Reporting adverse eventsï¿½one of the most important steps VA is taking to
improve patient safetyï¿½depends on the cooperation and participation of VA
employees. However, a reporting system is not enough. In its training for
employees on VA's new method for reporting and analyzing the root causes of
patient safety problems, the National Center for Patient Safety emphasizes
the importance of feedback to all employees who are participating in the
processï¿½from the individual who reports the adverse event or close call to
the team of employees selected to perform the root cause analysis or
aggregate review. We believe that motivating employees by involving them in
these processes will enhance the success of VA's efforts even more than
setting goals and communicating patient safety as a top priority. The
importance of providing feedback to teams and disseminating findings
throughout VA cannot be overemphasized. Managers must be willing not only to
find time for employees to participate in the root cause analysis but must
also support the implementation of recommendations made by the teams. We
believe that only when these processes run smoothly and employees
participate fully as part of their everyday work will the full benefits of
the culture of safety be realized.

6. The VA Office of Inspector General's Combined Assessment Program (CAP)
reports often identify critical nursing staff shortages. Can you comment on
how these shortages have an impact on patient safety?

The nursing shortageï¿½a national problem affecting VA and private sector
hospitals alikeï¿½is projected to worsen in the coming years. GAO has not
conducted an evaluation of the adequacy of nursing staff ratios in VA. We
therefore cannot comment directly on the potential effect that any nursing
shortage may have on patient safety. However, in 1998, VA and Kaiser
Permanente jointly sponsored a public/ private sector focus group of health
care professionals to identify perceived barriers to patient safety. These
care providers, including VA participants, identified inadequate staffing as
the largest barrier to patient safety. Not surprisingly, they reported that
inadequate staffing leads to employee fatigue and frustration. Other
industries, such as aviation, have established strict rules that prevent
flight crews, for instance, from working without adequate rest periods. VA
is currently doing research in this area to determine the applicability of
this preventive measure in health care delivery situations.

7. The VA points to its bar- coding system for medications as one of its
great success stories. I understand the implementation of this system has
encountered problems in operating rooms and in interfacing with VA's
computerized patient record system. Would you please elaborate on these
difficulties?

ENCLOSURE ENCLOSURE GAO- 01- 123R VA Patient Safety 6 VA reports that it has
implemented bar code medication administration (BCMA)

in over 60 percent of its inpatient care areas. For BCMA to work, the
physician order entry package, which is part of the Computerized Patient
Record System (CPRS), must be functioning. One reason for the slippage in
the BCMA schedule is that CPRS is not uniformly available in all facilities,
even though VA tells us that the software has been installed at all VA
facilities. We are unaware of any specific interface problems between BCMA
and CPRS, but VA has experienced hardware and training problems in some
locations that have prevented full implementation of CPRS.

VA did not indicate that problems in operating rooms were occurring.
However, there are some problems in intensive care units (ICUs), where the
situation is more complex. VA told us that ICUs, in which about 70 percent
of medications are administered intravenously, are further behind other
inpatient care areas in implementing BCMA. Specifically, only about 40 ICUs
had implemented BCMA, as of the end of June 2000. According to VA officials,
the original BCMA computer package was intended primarily for administration
of oral medications and not for intravenous therapy. VA tells us that the
Version 2 upgrade of the software, scheduled for 2001, will resolve this
problem.

8. What evidence have you seen of VA's senior management participation or
commitment to the ambitious safety training programs being conducted
throughout the VA system?

There has been little participation by VA senior management in its safety
training programs. We asked the National Center for Patient Safety to
provide a breakout of the job titles of those employees who attended one of
the 3- day patient safety improvement training sessions. On July 19, 2000,
VA told us that, until then, no VISN or medical center director had
participated in the training sessions. The highest level managers who did
attend included 41 chiefs of staff, 49 service chiefs, and 83 associate
directors for patient care services or nurse executives. According to VA, of
the nearly 600 employees trained, the majority were facility risk managers
or quality managers.

9. Has VA identified the highest priority areas for medical errors, and have
they developed a standardized system for measuring the reduction of these
errors?

VA can use its adverse event registry to categorize adverse events that
occur most frequently and that would merit priority attention, such as
falls, suicides, and medication errors. VA has not yet developed a
standardized system for measuring a reduction in medical errors, and we are
not aware of any VA plans to do so. However, before VA can target the
patient safety problems that most need attention, it will have to put in
place a well- functioning reporting system and establish an accurate
baseline from which to measure change systemwide. Until VA's new system is
fully in place and operating for some time, VA will use the root

ENCLOSURE ENCLOSURE GAO- 01- 123R VA Patient Safety 7 cause analysis process
to provide a standardized tool for assessing the causes of

errors and to compare analyses across all VA facilities.

10.The IOM report uses the Harvard Medical Practice Study in New York, which
states that adverse events occurred in 2. 9 percent of hospitalizations.
This is a widely respected peer reviewed study. The VA reported 2,927
adverse events in a 19- month period. The July 17, 2000, U. S. News and
World Reportcited the Johns Hopkins Hospital in Baltimore, Maryland, as the
highest rated hospital in America for several years running. Johns Hopkins
had 69,603 inpatients in FY 1999. If you use IOM's 2. 9 percent times
69,603, Johns Hopkins would have had just over 2,000 adverse events. This is
just one hospital. How can the entire VA hospital system only report 2,927
adverse events? Using the same formula, one would predict 21,802 possible
adverse events. Can GAO try to explain this phenomenal discrepancy?

We do not know if Johns Hopkins has a reporting system, or, if it does, the
number of adverse events it contains. Nevertheless, this example underscores
the growing consensus in the health care industry, including the VA system,
that underreporting of adverse events is a serious problem. The percentage
reported by the IOM is based on limited research in two states and
represents a national estimate not intended for application to individual
health care systems. Moreover, other researchers have made compelling but
contradictory arguments that the IOM report overstates and understates the
extent of underreporting. Nevertheless, experts agree that underreporting is
a problem and that there is inadequate data to estimate the degree of
underreporting with much precision. So, whether 22,000 adverse events is the
“right” number or not is impossible to say. During fiscal year
2000, VA will be conducting a survey to establish a baseline measurement of
how employees feel about reporting adverse events. If employees report that
they do not feel safe enough to report adverse events, then we can assume
that underreporting will continue to be a problem. However, determining the
magnitude of this problem will continue to be difficult.

(406208)
*** End of document. ***