oversight

VA Health Care: Actions in Response to VA's 1989 Mortality Study

Published by the Government Accountability Office on 1990-11-27.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                                                                                                                    (:ornmit,tee on
                                                                                                            , U.S. Senate


-___._ l_l_,__._ -...____...........      I- _-.-_ __--               . ..-.. - .-------.   ^ ---__.

,Yovc~rrrt,t~r


                                                                                                   VA HEALTH CARE
                                 I!)!)0




                                                                                                   Actions in Response
                                                                                                   to VA’s 1989
                                                                                                   Mortality Study


                                                                                                                      llIlllllIlll
                                                                                                                            ll
                                                                                                                         142716




  -----                                                   __..-----pI---


  (;Ao,/‘IIHI)-!)I-~~i
I   -.--.   “. .*l*“““l”l.“._.--_-_.-   ..-,---
                 Humau    Resources   Divisiou

                 B-241910

                 November 27,199O

                 The Honorable Alan Cranston
                 Chairman, Committee on Veterans’ Affairs
                 United States Senate
                 Dear Mr. Chairman:
                 As requested in your June 16, 1989, letter, we have reviewed the actions
                 that the Department of Veterans Affairs (VA) has taken to address
                 quality-of-care problems associatedwith deaths in several VA medical
                 centers during fiscal year 1986. Specifically, in a June 1989 report, VA
                 stated that 44 of its 172 medical centers had higher-than-expected mor-
                 tality rates in one or more diagnostic categories(for example, cancer or
                 severeheart disease)in that period. Further, “likely” quality-of-care
                 problems were found in 90 casesin which deaths occurred. VA said it
                 would conduct follow-up activities to determine to what extent such
                 problems exist in its medical centers and what actions will be taken to
                 correct them.
                 You requested that we monitor and review VA’S follow-up actions, specif-
                 ically determining whether
             . the validation methodology for the follow-up is appropriate,
             l the follow-up is completed as described in the mortality study,
             l the actions taken as a result of the follow-up are proper, and
             l other actions may be neededto ensure that the probable causesof death
               resulting from quality-of-care problems are identified and corrective
               actions taken.
                 We briefed your staff on the results of our work on July 9, 1990. This
                 report summarizesthat information and provides an update of certain
                 VA activities through October 31, 1990. Our work’s scopeand method-
                 ology are discussedin appendix I.

                 In June 1989, at the direction of the Chief Medical Director, VA’S Office
Background       of Quality Assurance published a report concerning deaths in VA medical
                 centers during fiscal year 1986. The report showed that 44 of VA’S 172
                 medical centers had higher-than-expected mortality rates in one or more




                 Page 1                                  GAO/HRB91-26 VA’s 1999 Mortdty   Study
                       B-241910




                       diagnostic categories.’ In the aggregate, 3,050 deaths occurred in these
                       medical centers2 in those diagnostic categories, whereas only 2,098
                       deaths would have been anticipated. Staff from VA’s Medical District Ini-
                       tiated Peer Review Organizations (MEDIPROS) reviewed 1,77 13of these
                       cases according to preestablished criteria and referred 473 cases from
                       43 hospitals to MEDIPRO physicians for review. These physicians deter-
                       mined that 90 cases had “likely” quality-of-care problems. Almost half
                       of these cases (38 of 90) occurred in six primarily psychiatric hospitals.
                       The remaining were identified as having occurred in 16 medical centers
                       providing primarily medical and surgical acute care.

                       In its report, VA stated that, among other things, the following actions
                       would be taken to assess the significance of the aforementioned data:

                   . Require each medical center to comment on MEDIPRO decisions and pro-
                     vide both the central office and MEDIPRO a summary of corrective actions
                     to be taken to address issues raised.
                   l Require MEDIPRO and central office personnel to review and approve the
                     medical centers’ responses and corrective action plans.
                   . Analyze mortality cases not available at the time the original study was
                     published and report in a supplement the findings and corrective actions
                     taken.
                   . Validate the original study methodology.
                   . Monitor the care in the medical and surgical wards of primarily psychi-
                     atric facilities to determine if there are any significant differences
                     between this care and that provided at other VA medical facilities.

                       A complete statement of the actions VA planned to take to address the
                       study findings is contained in appendix II.


                          has used an appropriate methodology to identify and follow up on
Results in Brief       VA
                       deaths associated with quality-of-care problems. Further, most of the
                       actions it planned to take to assess the significance of the mortality

                       1These diagnostic categories were severe heart disease; metabolic, electrolyte disorder; cancer; ortho-
                       pedic conditions; cerebravascular disease; gastrointestinal disease; low-risk heart disease; pulmonary
                       disease; renal and urologic disease; and follow-up and after-care.
                       2Although the June 1989 mortality report identified 44 hospitals with higher than expected mortality
                       rates, the report included data from only 43. Data from the 44th hospital were analyzed as part of the
                       follow-up actions.
                       “VA selected 2,417 of the 3,050 mortality cases for MEDIPRO review. Of the cases selected, 273 were
                       found to involve deaths after discharge or erroneous coding and were eliminated from the study, and
                       373 pertinent records were unavailable; thus, 1,771 cases were ultimately reviewed.



                       Page 2                                                   GAO/HRD-91-26 VA’s 1989 Mortality       Study
---
            B-241910




--..-
            study findings have been completed. But VA is still analyzing deaths that
            occurred in psychiatric centers in fiscal year 1989, to determine if there
            are any significant differences between the quality of care provided in
            psychiatric facilities and that provided at other VA medical centers.

            VA’S completed actions are as follows:

        . Each medical center has commentedon the MEDIPROdecisionsand has
          provided a summary of any corrective actions taken to resolve the
          issuesraised.
        . MEDIPRO and central office personnel have reviewed and commentedon
          each medical center’s corrective action plan.
        . MEDIPROpersonnel have reviewed mortality casesnot available for
          review when the original study was conducted.
            Actions yet to be completed are:

        l   The original study methodology is being examined, and a final report is
            expected to be iss,uedin March 1991.
        l   MEDIPRO staff are reviewing deaths that occurred in fiscal year 1989 in
            seven4psychiatric hospitals that had high mortality rates in 1986 to
            determine the extent to which quality-of-care problems are occurring.
            Preliminary data indicate that at least four of the sevenpsychiatric hos-
            pitals continue to have deaths with possible quality-of-care problems at
            rates higher than the comparable mortality rate in medical and surgical
            acute care hospitals. Assessmentsof the quality of care provided in the
            psychiatric facilities involved are in process.
            Although VA has taken specific actions to follow up on its mortality
            study, it has not used the information it obtained from individual med-
            ical centers to improve systemwide operations. For example, at least
            five medical centers were known to have problems implementing their
            “Do Not Resuscitate” procedures,6and each developedits own revisions
            to correct the situations encountered,This issue may have systemwide
            applicability. But the central office did not disseminate the data to all of
            its medical centers, The result may be duplication of effort and a lost


            ‘The review of previously unavailable casesidentified anotherthree “likely” problemcasesin a SW-
            enth psychiatric hospital.
            %ocedures written to clarify matterssuch asthe extent to which emergencymeasureswill be insti-
            tuted if a patient’s condition warrants, who hasthe authority to makesuchdecisions,and what
            factors must enter into thosedecisions.



            Page 8                                                GAO/HiUbW26      VA’s 1888 Mortality   Study
-
                          5241910




                          opportunity to share data that could prevent similar problems from
                          occurring at other medical centers.

                          The Committee should consider discussing with VA the results of actions
                          that are still underway and the need to disseminate information that has
                          systemwide applicability.


                          VA used physician peers (MEDIPRO) from outside the involved medical
VA Used a Generally       centers to examine each mortality case selected for review, determine
Accepted Methodology      potential quality-of-care problems, and evaluate the appropriateness of
to Identify Quality-of-   corrective actions taken. This peer review technique is consistent with
                          methods used in the private sector to evaluate medical care. It is based
Care Problems             on a fundamental medical community premise that physicians are best
                          qualified to review and judge the clinical activities of other physicians.

                          VA initiated MEDIPRO in 1985 on a limited basis and, by 1986, had
                          expanded it to include all districts in the system. Its purpose was to pro-
                          vide a mechanism for physician peers to evaluate the quality of care and
                          utilization of resources in VA medical centers. Clinically active VA physi-
                          cians from hospitals within each of VA'S 27 districts were selected to
                          serve on district boards, which analyzed data to identify potential
                          quality-of-care problems in VA medical centers.” VA used MEDIPRO for the
                          mortality study and its follow-up because it had experience in con-
                          ducting multi-facility medical record reviews and was external to the
                          facilities being reviewed.

                          MEDIPROSserved a function similar to that played by Peer Review Orga-
                          nizations (PROS) in the private sector. PROSare congressionally mandated
                          private organizations established to ensure that services furnished
                          through Medicare are necessary, appropriate, and of high quality.


                          VA medical centers have completed all of the follow-up efforts they were
VA Actions Are Being      required to take as a result of the June 1989 study. VA'S central office
Conducted as Planned      and MEDIPROSare in the process of conducting their intended follow-up
                          efforts. MEDIPROScompleted their reviews of cases that were not avail-
                          able when the original study was performed; every VA medical center
                          that had deaths attributed to quality-of-care problems commented on
         ”                MEDIPRO decisions and provided a summary of any proposed corrective


                          'As of April 1, 1990,VA abolishedits district offices and, as of October 1990, had not determined
                          what peer review structure will be employed.



                          Page4
                       B241910




                       actions to the cognizant MEDIPRO and to the central office; and the med-
                       ical centers’ corrective action plans were reviewed and evaluated by
                       cognizant MEDIPROS. Disagreements bet,ween medical centers and
                       MEDIPROS about the quality of care provided in individual cases were
                       referred to panels appointed by central office officials.

                       VA’S central office, however, has not completed its planned study to
                       determine whether there are differences between the quality of care
                       furnished in medical and surgical wards of primarily psychiatric facili-
                       ties and that provided to medical and surgical patients in other VA med-
                       ical centers. MEDIPROS are reviewing deaths that occurred in fiscal year
                        1989 in psychiatric hospitals previously found to have quality-of-care
                       problems to determine if such problems continue to occur at rates higher
                       than that experienced in acute care VA facilities in 1986 (3.7 percent).
                       These MEDIPROS will also report their conclusions about the quality of
                       care provided at locations in which deaths occurred that had related
                       quality-of-care problems. VA hopes to complete this process by December
                       15, 1990.

                       The following sections discuss what                 VA   is doing in each area in which
                       follow-up action was planned.


Patient Records        From January to November 1989, MEDIPRO staff reviewed patient
Unavailable During     records in 627 mortality cases that had been identified in the original
                       statistical study as occurring at higher-than-expected rates, but, for
Original Survey Were   various reasons, were not in the final report.7 Of these cases, 104 did not
Reviewed               meet the staff’s screening criteria and were referred to MEDIPRO physi-
                       cians for further review. These physicians concluded that 21 cases had
                       “likely” quality-of-care problems, thus increasing the total number of
                       problem cases to 111. VA intends to publish a supplement to the mor-
                       tality study to describe these additional findings and the results of its
                       review of deaths that occurred in the seven psychiatric hospitals in
                       fiscal year 1989. This supplement is expected to be published by March
                       1991.




                       7These follow-up steps were begun at the time the published study report was being finalized. Of the
                       patient records not reviewed in time for the report’s publication, 267 were located in one hospital that
                       was delayed in starting its review; 94 were from two hospitals that expanded the scope of their
                       review; and the other 266 were from various hospitals that did not have the patient records available
                       immediately.



                       Page 5                                                   GAO/H&D91-26 VA’s1989Mortality Study
                             B241910




Medical Centers              Every medical center in which MEDIPRO alleged that a quality-of-care
Commented on MEDIPRO         problem existed commented on the MEDIPRO findings. Not all agreed with
                             the MEDIPRO’S findings; therefore, some took no corrective action. Of the
Findings, but Not All Took   111 cases involving “likely” quality-of-care problems, the medical cen-
Corrective Actions           ters agreed with MEDIPRO findings in 78 cases and disagreed in 33. The
                             latter were sent for review to VA’S central office, which upheld the
                             MEDIPRO’S decisions in 22 cases.

                             Medical centers submitted corrective action plans, which, in most cases,
                             the cognizant MEDIPRO approved, to address problems identified in 81 of
                             the 111 cases. In the remaining 30 cases, medical centers took no correc-
                             tive actions: in 15 cases, officials disagreed with the MEDIPRO’S findings;
                             in 11 cases, all of the involved physicians had left the medical center
                             and center officials believed no further action was necessary; and in 4
                             cases, the hospital had no evidence that the recommended actions had
                             been taken. In cases where there was disagreement, VA central office
                             officials told us that no corrective action by the medical center will be
                             required. Specifically, they stated that staff in each of the affected
                             facilities have already reviewed the case data several times to determine
                             if any corrective action was necessary. In the opinion of these officials,
                             any requirement to review the cases again would be superfluous.

                             In many instances, medical centers took more than one action to correct,
                             a problem: 63 actions involved medical center policy and procedure
                             changes, 34 related to increased use of occurrence screening! 48
                             involved additional training or education, 13 involved individual coun-
                             seling with the provider, and 12 involved physician reassignment. In 9
                             of the 81 cases, hospital officials believe that the corrective action taken
                             ivas not effective, and further problem-solving activity is in progress. In
                             10 cases, the medical centers took various corrective actions even
                             though the involved physicians had left the medical centers. Appendix
                             IV summarizes the corrective actions taken at each medical center.




                             sOne of several elements in a VA medical center’s quality assurance program. It involves a review of
                             patient records by trained personnel, who use designated criteria to identify occurrences that
                             represent deviations from normal procedures or expected outcomes. Once identified during the
                             review of a medical record, the occurrence is evaluated through a peer review of physicians, who
                             determine whether the care given was appropriate and met acceptable medical standards.



                             Page 6                                                  GAO/HRB91-26 VA’s 1989 Mortality Study
                            B241910




VA Is Validating the        The chief of staff at VA’S Hines Medical Center is examining the original
Methodology Used in the     study methodology through a VA central office-funded research project.
                            He has submitted several status reports to the central office covering
Original Mortality Study    the period July 1989 to August 1990 and expects to issue a final report
                            to the central office in March 1991.


Psychiatric Case            In its June 1989 report, VA stated that it would monitor the care pro-
Monitoring Is in Progress   vided in medical and surgical wards in primarily psychiatric facilities to
                            determine whether it differs significantly from that provided at other VA
                            medical facilities. As of October 1990, this action had not been com-
                            pleted, but VA officials hope to complete the analysis by December 15,
                            1990.

                            To meet its follow-up commitment, VA is examining all deaths that
                            occurred in fiscal year 1989 at the seven psychiatric facilities that were
                            determined to have quality-of-care problems in 1986. Cases with such
                            problems are being identified using a process similar to that employed
                            by MEDIPRO to review deaths that occurred in fiscal year 1986, and a rate
                            of cases with quality-of-care problems is being computed. As the case
                            analysis is completed, the 1989 data are being compared to aggregate
                            data developed by VA for all VA acute care hospitals (3.7 percent). Prelim-
                            inary data from six of the seven psychiatric facilities examined show
                            that at one center, 14 percent of the mortalities have related quality-of-
                            care problems. Three centers have quality-of-care problems ranging
                            from 6.7 to 6.9 percent of the death cases reviewed;g one psychiatric
                            center has a rate comparable to acute medical and surgical facilities, 3.9
                            percent; and one center has a rate lower than such facilities, 2.9 percent.

                            VA  regional offices are analyzing data describing the corrective actions
                            taken by the medical centers to resolve any problems identified. Fur-
                            ther, each MEDIPRO board that analyzed mortality cases has been tasked
                            with drawing conclusions about the quality of care provided at the facil-
                            ities and making recommendations regarding the need for further moni-
                            toring and other activities.

                            This approach is consistent with a recommendation made in December
                            1989 by a committee of central office and medical center personnel
                            appointed by the former Associate Deputy Chief Medical Director to

                            gDeath cases at two medical centers have been reviewed by one physician. VA central office officials
                            have directed the cognizant region to assign a second physician to review the problem cases. When
                            the second review is complete, the problem rates may be lower than the preliminary data indicate
                            (0.0 and 6.3 percent).



                            Page 7                                                  GAO/HRB91-26 VA’s 1999 Mortality Study
                         B-241910




                         review mortality cases in psychiatric hospitals that had “likely” quality-
                         of-care problems. The committee reviewed 36 cases MEDIPRO identified as
                         having such problems in six psychiatric medical centerslo and deter-
                         mined that death was preventable in 12 cases. It also concluded that 30
                         cases had one or more quality-of-care problems: 25 cases had delays in
                         diagnosis; 15, drug usage problems; and 8, a delay in or omission of
                         referral or consultation. The committee recommended that VA conduct
                         future studies to compare the quality of care in psychiatric facilities
                         with that in other VA facilities.


                             medical centers have identified quality-of-care problems and initiated
VA Needs to Better       VA
                         corrective actions that could have had systemwide applicability. But
Disseminate              VA'S central office did not use the experience gained in these centers as a
Information to Medical   means of improving care throughout the system. According to central
                         office officials, medical centers provided information about the
Centers                  problems and proposed corrective actions. However, data were not suffi-
                         ciently detailed to allow a determination to be made as to whether the
                         problems may exist systemwide and whether the actions taken to
                         address them could have universal applicability. Central office per-
                         sonnel also stated that no effort was made to obtain additional informa-
                         tion from the medical centers because, at the time, no consideration was
                         being given to the potential systemwide applicability of the data.

                         For example, a medical center pharmacist team developed a drug usage
                         evaluation plan for several therapeutic drugs that, among other things,
                         specified procedures for when and how to use theophyllinell for thera-
                         peutic use in the emergency room. This and other drug usage protocols
                         were presented at a national pharmacists’ meeting, and many attendees
                         requested copies. However, these guidelines were not communicated to
                         other VA medical centers.

                         VA traditionally has delegated operational decision-making to its medical
                         center directors. As a result, central office personnel are generally reluc-
                         tant to act in ways that might suggest they are attempting to direct the
                         medical centers’ activities. Thus, in discussing the potential systemwide
                         applicability of some of the changes made at individual medical centers,
                         central office quality assurance officials stated that they had not consid-
                         ered the possibility of wider applicability of these changes. The officials

                         loThe special review had already begun before the follow-up identified a seventh psychiatric hospital
                         with problem cases.
                         * lA drug used for treating certain bronchial conditions.



                         Page 8                                                      GAO/HRD-91-26 VA’s 1989 Mortality   Study
                           &241910




                           stated, however, that in the future they will consider disseminating data
                           with potential systemwide applicability.


                           Patients in VA psychiatric facilities should benefit if VA (1) completes, as
Conclusions                planned, its comparison of the differences between care provided on
                           medical and surgical units in primarily psychiatric hospitals with that
                           furnished in medical and surgical acute care centers and (2) takes
                           appropriate action to resolve any problems identified. Statistical data
                           derived from the original study indicated that, in 1986, mortality rates
                           and quality-of-care problems were higher in psychiatric facilities than in
                           other acute care hospitals. Preliminary data on 1989 deaths in psychi-
                           atric facilities indicate that these problems still exist. Until the MEDIPROS
                           complete their analysis, serious questions about the quality of care pro-
                           vided on medical and surgical units of psychiatric facilities will remain.

                           VA's central office could have made more extensive use of the informa-
                           tion that MEDIPROSand medical centers developed as part of the mor-
                           tality study and its follow-up. MEDIPROS' review of mortality cases
                           identified some quality-of-care problems with possible systemwide
                           applicability. If these problems and their corresponding corrective
                           actions had been communicated throughout the VA medical system, indi-
                           vidual medical centers might have identified similar problems and used
                           effective corrective responses developed elsewhere. More importantly,
                           initiating these corrective actions systemwide could have helped assure
                           more uniform care for all VApatients.


                           As part of its continuing oversight of VA health care issues, the Com-
Matters for                mittee may wish to discuss with the Secretary
Consideration by the
Committee              l the results of VA'S efforts to compare the quality of care provided in
                         medical and surgical units in primarily psychiatric hospitals with that
                         provided in medical and surgical acute care hospitals and
                       . the need for central office staff to review the results of future quality
                         assurance studies for possible systemwide applications and disseminate
                         appropriate information to all medical centers.


                           As agreed with your office, we did not obtain written comments on this
                           report from VA, but did discuss its contents with Office of Quality Assur-
                           ance officials. Their comments have been incorporated, where
                           appropriate.


                           Page 9                                    GAO/HRD91-26 VA’s 1989 Mortality Study
We are distributing this report to VA and to interested congressional com-
mittees and members. We will also make.copies available to others upon
request.

If we can provide any further assistance, please call me at
(202) 276-6207. Other major contributors to this report are listed in
appendix V.

Sincerely yours,




David P. Baine
Director, Federal Health
  Care Delivery Issues




Page 10                                  GAO/HRD-91.26VA’s 1989Mortality Study
P8ge 11   GAO/HRD-91-28VA’s1989Mortality Study
contents


Letter                                                                                                  1

Appendix I                                                                                             14
Scopeand
Methodology
Appendix II                                                                                            15
Review of Mortality in
VA Medical Centers-
Planned Actions
Appendix III                                                                                          16
Results of Death Cases
Examined by
MEDIPROsand
Central Office
Reviewers
Appendix IV                                                                                           18
Medical Center
Corrective Actions
Appendix V                                                                                            20
Major Contributors to
This Report


                         Abbreviations

                         MEDIPRO   Medical District Initiated Peer Review Organization
                         PRO       Peer Review Organization
                         VA        Department of Veterans Affairs
                         VHS&RA    Veterans Health Services and Research Administration


                         Page 12                                GAO/HRD-91-26 VA’s 1989 Mortality   Study
Page 13   GAO/HRD-91-26VA’s1989Mortality Study
Appendix I

Scopeand Methodology


             In this monitoring effort, we relied on the work of district MEDIPROSand
             did not evaluate their decisions or conclusions. Our objective was to
             determine whether problems were addressed, corrective actions were
             taken, and any lessons were learned that could be applied to medical
             centers systemwide.

             We periodically met with quality assurance staff in VA'S central office to
             obtain an update on their follow-up activities. We also reviewed VA med-
             ical center correspondence that discussed follow-up actions, examined
             data aggregated by central office quality assurance staff about correc-
             tive actions taken in each medical center, and analyzed questionnaires
             completed by medical center officials describing corrective actions taken
             and submitted to MEDIPRO and central office staff. To determine the ade-
             quacy of hospitals’ corrective actions, we talked with medical center
             directors, quality assurance coordinators, chiefs of staff, or other appro-
             priate VA personnel and obtained documents that included the policy
             revisions made and described education or training provided. Our work
             was performed between August 1989 and October 1990 at VA headquar-
             ters in Washington, DC., in accordance with generally accepted goven-
             ment auditing standards.




             Page14                                   GAO/HRDBl-26 VA’8 19S9 Mortal&y Study
Apbendix II
Review of Mortality in VA Medical Centers-
Plmed Actions

                  1, Medical centers will report the following information for each case
                  that was assessed by two peer reviewers as having a likely quality-of-
                  care problem:

              l Whether the case was previously reviewed by the facility’s quality
                assurance program and, if so, the outcome of that review.
              l Whether the medical center agrees with the findings of the review and,
                if not, the reasons for its disagreement.
              . The proposed corrective actions to address the issues raised by the case.

                  2. The medical centers’ responses will be reviewed by the MEDIPROboards
                  and VA'S central office.

                  3. Medical record reviews of a number of cases that were targeted by the
                  mortality rate analysis were not completed in time to be included in the
                  report. A supplement will present the results of these medical record
                  reviews and a description of all corrective actions resulting from them.

                  4. Monitoring of the care provided in the medical and surgical wards of
                  primarily psychiatric facilities will be conducted through a number of
                  Veterans Health Services and Research Administration (VHS&RA) quality
                  assurance mechanisms, such as MEDIPROand, where appropriate, site
                  visits by central office clinical and administrative officials. This moni-
                  toring will determine:

              . Whether there are significant differences between the quality of medical
                and surgical care provided at primarily psychiatric medical facilities
                and that provided at other VA medical facilities.
              l If so, what actions should be taken to improve the care at primarily psy-
                chiatric facilities.

                  5. A VHS&RA study will be conducted to determine the validity of the
                  statistical methodology described in part I of VA'S mortality study as a
                  mechanism for identifying possible quality-of-care problems. Upon com-
                  pletion of this study, VHS&RA will decide whether regularly scheduled
                  mortality rate analyses would be a useful addition to existing quality
                  assurance activities.




                  Page 15                                   GAO/HRD-91-26 VA’s 1989 Mortality   Study
Appendix III

Results of Death Casesl!kunined by MEDIPROs
and Central Office Reviewers


                                                                                             Number of cases
                                                                                    Original     Foltfyi                              Number of
                                                                                  -   study                                            hospitals
                                                                                                                           Total -_-_II__-
Cases reviewed bv MEDIPRO staff                                                        1,771             627               2,398               44
Cases referred to MEDIPRO physicians for invdepth review                                 473             104                 577               44
Cases
_ .__... that MEDIPRO believes have “likely” quality-of-care problems
       - ..-._-_-__----                                               --                  90              21                 111             --.-27
Cases that medical centers agreed have quality-of-care problems __--                                ---.                      78
                                                                                                                               l_----          23
Cases referred tocentral office because medical centers disputed MEDIPRO
findings                                                                        __l_-_-                                        33         --.--- 16
Cases for which central office reviewers concurred with MEDIPROfindings
that
 _.-_.-quality-of-care
        ..--.---______ problems existed                                                                ._I_                    22                14
Remaindera                                                                                                                     11                 6
                                            ‘In eight cases,the panels agreed with medical centers that no quality-of-care problem had occurred; in
                                            two cases, records were unavailable; and in one case, the panel could not determine from the records if
                                            a problem had occurred.




                                            Page 16                                                  GAO/HRD-91-26VA’s 1989Mortality Study
Page 17   GAO/HRD91-26VA’s1989Mortality Study
Appendix IV

Medical Center Corrective Actions



                                                                      Corrective action taken
                             Number of                 Occurrence       Individual          Addltional          Policy
Facility                        cases                   screening      counseling             training         change                     Othep
Albany, NY         .-._.._--         1                                                                                1            ----.-..-.. 1
Asheville, NC                        2                                                               2
Atlanta, GA                          3                            2              1                                               --.__    ~-.._..~
Battle Creek, Ml     ..-.---        12                           12                                 12                  12                    12
Chillicothe,OH~ .’                   2                                                               2                   2                      2
Columbra.SC                          1                                                                                                          1
Dayton; OH                           6                            5              2                       1              5      ----..-.-        6
                                                                                                                          ----     -.-- ._.-.~.
Denver, CO                                         8                             3                                      3                       2

Durham, NC                                         2

Fayetteville, NC    ..      -_-..                  1              1              1
Houston, TX                                       10                                                 9                  9


Iowa Crty, IA                                      1                                                                    1
Leavenworth, KS                                    3                             3                   3                  2
Lebanon: PA      ._-. --.-                         3              3                                  3
Loma Linda, CA                                    10              2                                                     IO
Mar&, IN -.                                        5                                                 5                                             5
Mountain Home, TN                                  1
New Orleans, LA                                    3                                                 3                                             2
                                                                                                                        --_-      ---. ~-... ..-
Oklahoma City, OK -- -~ .--.---.
                  .- -..-_-                        1              1              1                                       1
Phoenrx, AZ .-                                     1                                                               -1

Saksbury, NC
San Juan, PR
St. Louis, MC-

Tuskegee, AL                                       6              6                                  6                 6
                                                                                                              __l_----.-.                -- -.
Waco, TX                                           5

Walla Walla, WA
                         - ___      --l____._~                                                                           _----
Washington, DC                                     4

                    _._-                         111             34             13                  48               53 ____-.-           -.. .~~35
Totals




                                                       Page 18                                    GAO/H&D91-26 VA’s 1989Mortality Study
                                                             Appendix     lv
                                                             Medleal Center Chrective Actions




Examples of corrective actions and other comments                                         ____-.-
Pokey changed to require all car&%a~e~s
                                .I. . ..__.--_--__ and-dehstobe-reviewed.
                                                                 -_-._---      --..--..----.          A respirator was placed in the emergency room.
                                                                                                  ___--
Staff trained to better.document mortality case review.                                                                                                  .______ __-.-       .----~-~
The medtcal center disagreed with   .-. MEDIPRO’s
                                            ..--~                  decision in one case.
The medical center and MEDIPRO now review all deaths concurrently. Laboratory and radiology are staffed 24 hours.
Nurses were instructed to prepare better documentation, The “Do Not Resuscitate” ~__-.-                              policy was revised.
Physicians presented case to the mortality and morbidity committee and conducted a literature review on infections.
Medical center implemented new “Do-Not%&us%ate”                               po~y%a~ty&sursnc&taff?now                       reviews all autopsy reports. In one case, the
chief of surgery discussed the need for adequate documentation with surgeons.                                                    ____-.
One case is now the subject of a lawsuit, Policy involved: following up on use of anticoagulation drugs, monitoring of blood pressure drops
during surgery, and monrtoring of oxygen tank gauges.                                                                                            ----~
No actions taken In response to these cases-becausethemedicalcenter                                    disagreed with MEDIPRO’s conclusions. Further, VA review
substantiated the medical center’s position   ..-....- ^...- that
                                                             I__^___no____._._
                                                                        problem
                                                                          - ..-- had       occurred. -. .----.~----.~--~---.
                                                                                      ..__~.__-.--..                                                      .....--- ~-__ ----~~-
The need for adequate documentation was discussed.
Medical center noted that by the time of the follow-up review, all the involved residents had left. However, the medical center admissions
  okcy was changed to requrre an attending physician to see all new admissions within 24 hours of admission, seven days a week. Also,
R rgh dosages of theophylline are now monttored; a strong drug usage education program was started; and surgery patients now have
cardiac workup.
X-ray results are now reported more promptly.                                                                                                           --.
Use of four antibiotics is now monitored and discussed in staff training.
Medical staff received refresher training in medical center’s “Do Not Resuscitate” policy.                                                         -..- _____~---.--.-----.-
Medical center rewrote its “Do Not Resuscitate” policy.
Medical staff recerved instruction in treating severe cardiac cases.
No correctrve actions were taken because.medical cknterbidnotasreethataquaiity-of-care                                          problem had occurred.
Continuing education now given for nurses’ documentation of emergency situations and physicians’ documentation of progress notes
before surgery. Also, glucometers were placed in wards.
Medical center now reviews all autopsy reports.                                                                                                        __.______~       --......-.. --~
Continuing education was furnished on how to recognize acute medical problems in debilitated elderly patients, particularly at time of
transfer from other facilities.
No action-taken because the medical staff involved, including the chief of staff, had left the medical center.
Staff education on case workup, management, treatment, and need for better documentation enforced.
Medrcal staff were Instructed in improved dia nosis and treatment of ulcer surgery and were directed to improve their consultation with
surgical staff. Results of test reports are now .~~~           8.lled more    promptly. -- _~.~.
                                                                       _~~_.....
The medical center developed new “Do Not Resuscitate” policy and improved its reporting of lab and X-ray test results to physicians.
Also, weekly mortality and morbidity reviews ..~-~             have been
                                                                       _      improved.
                                                                              ..-       -~~~.-.....-.
Medical center documentation of mortality review was improved. No other actions were taken because medical center disagreed with
MEDIPRO in all five cases                                                    _~~~~~~~   ~~~
Critical lab results are now reported by telephone. A consulting spectatist~providedstaff                                 withc<minuingeducation         in diagnosis and
treatment of stroke patients.                                          ,.
Medical center cauld not locate documentation indicating-/f actions were~&%enTG<ne case%%oEas                                                  taken in remaining three cases
because medical center disagreed with MEDIPRO findings.


                                                             aOther corrective actions taken include increased use of resources, such as patient mclitoring equip.
                                                             ment; improved or faster procedures, such as reporting lab results; and modified quality assurance
                                                             procedures, such as more thorough reviews of all mortality cases. Some of these examples are
                                                             described in the last column of this table.



                                                              Page 19                                                            GAO/HRD-91-26VA’s1989Mortality Study
k&pendix V                                                                               ’

Major Contributors to This Report


                  JamesA. Carlan, Assistant Director, (202) 708-4228
Human Resources   Lawrence L. Moore, Evaluator-in-Charge
Division,
Washin&on, DC,




             Y




(101359)          Page 20                               GAO/HID-91-26 VA’s 1989Mortality Study
![ .^..
     I. ..--... -_.---.-..--.._
                             -__“-.---________
                                      Orttt~ritlg         Illf’orrtl;it,iort




                                      I I.S. (h*nt~rid Ac~oitrit.ing Ofl’ict~
                                      I’.(). IhJX M)l5
                                      (;ait ht~rslmrg, MI) 20877

                                      Orttt~rs      may       also btb ptac~~d by calling   (202) 275-6241.
                                                                                                                                                                                                                                 r.
__,,, _., -___-”   ,..._--_.               .__.__ .^___ _ __                ._   .___ _ ,“. .,.             ,.                       -_-..-             --..-          ..-.   ..--.---...-......   I .l.““l-*   .I- --I_-   --




___...   --.-      l”.“l..-_-.-._-““l-.-           . .   --.--   ..--.-.-        -_..-.--         .._....        .._,_.   -._._.“m       I.”   1.1-.1”....-..-..----