.-- i -.._ _-._.--.-.__--. INFECTION ~l~lll\lill~\ I !)'I0 I ' CONTROL VA Programs Are Comparable to Nonfederal I’rograrns but Can Be Enhanced - united State8 “.4 ,’‘,,::*i vd cA0 General Accountinyr Office Washington, D.C. 20548 ‘p; ! Human Beeourcee Division B-232863 January 31,199O The Honorable Frank H. Murkowski Ranking Minority Member Committee on Veterans’ Affairs United States Senate Dear Senator Murkowski: In responseto your request, this report discussesthe effectiveness of the Department of Veterans Affairs’, (VA'S) infection control programs and how the programs compare with those in nonfederal hospitals. To assessVA'S programs we developed basic elements of hospital infection control programs. Using these elements we found the content of VA medical centers’ and nonfederal hospitals’ programs to be similar. However, our visits to sevenVA medical centers indicated that many of the VA programs were understaffed and neededmore support from medical center management.Several recommendations are made to improve these conditions, all of which VA concurred with and indicated that action was being taken. Copies of this report are being sent to the Secretary of Veterans Affairs, appropriate congressionalcommittees, and other interested parties. This report was prepared under the direction of David P. Baine, Director, Federal Health Care Delivery Issues,who may be reached on (202) 276-6207.Other major contributors are listed in appendix VII. Sincerely yours, Lawrence H. Thompson Assistant Comptroller General ! Ekecutive Summary Purpose The Centers for DiseaseControl (CDC)estimates that some6 percent of the patients who enter a hospital contract an infection during their stay, With over 1.3 million inpatient admissions annually, some60,000 veter- ans could contract infections each year while being treated in hospitals operated by the Department of Veterans Affairs (VA). The incidence of hospital-acquired infections can be reduced, however, if hospitals oper- ate effective programs to control infections. The Ranking Minority Member of the SenateVeterans’ Affairs Commit- tee asked GAO to review and evaluate infection control programs at VA medical centers. Specifically, he requested that GAO determine . whether infection control data are used to help prevent future infections, . whether VA medical centers have adequately staffed and organized their infection control programs, and l how VA'S infection control programs compare with those of nonfederal hospitals. VA operates 169 medical centers throughout the United States, VA B$ckground requires each center to have an infection control program to identify infections and help prevent future ones. GAO neededcriteria outlining the basic elements of an effective infection control program in order to assessthe programs VA operates at its medi- cal centers. It found that the program guidance issued to the medical centers by VA'S central office was too broad to serve this purpose. It also found, however, that no other U.S. health care organization had up-to- date and specific guidance. In order to undertake this review, therefore, GAO worked with representatives of nine organizations and one other individual with expertise in infection control to develop a list of 66 basic elements of an effective program. The nine organizations included the Centers for DiseaseControl (CDC), the Joint Commissionon Accredita- tion of Healthcare Organizations, the Association for Practitioners in Infection Control and the Society of Hospital Epidemiologists of America. (Seeapp I.) These elements, referred to in this report as GAO'S elements, are applica- ble to infection control programs in any acute-carehospital with more than 60 beds. (Seeapp 11.) Page2 GAO/HRD-90-27VA'sInfection CmtrolPrograms / I ( 1 I , / Execntive Summary I GAO used these elements to examine the content of VA'S infection control programs. It collected information through visits to sevenmedical cen- ters and a questionnaire sent to all VA medical centers and a sample of 667 nonfederal hospitals. GAO used the questionnaire responsesto com- pare VA infection control programs with those in nonfederal hospitals, Both VA and nonfederal infection control programs are using most of GAO'S basic elements. Forty-four of the elements are widely used by VA infection control programs and 42 of the sameelements are widely used by nonfederal programs. In the VA medical centers GAO visited, the use of the elements was due to the individual initiative of the infection control practitioners. Most of GAO'S elements either are not included in VA guid- ance or are stated in such a general manner that VA'S guidance is of little use to VA practitioners. Becausethey had only limited guidance and direction from VA'S central office, VA practitioners sought current infec- tion control information and implemented many activities that went beyond their guidance and met the basic elements. Several infection control elements that GAO identified, however, should be used by more practitioners in both the public and private sector. These elements are generally more labor intensive than those that are receiving widespread acceptance. In addition, to be most effective, VA programs need managementatten- tion VA infection control programs are generally understaffed, not coor- dinated at the central office, and not adequately monitored by the regional offices. At the central office, six different units issue infection control guidance. There is no central focal point. At the sevenmedical centers GAO visited, regional office oversight of infection control pro- grams was limited. i Principal Findings VA Infection Control VA'S overall guidance on infection control programs was issued in 1978 Guidance Needs to Be and 1979, and has not been updated since. Of the 66 elements GAO believes should be basic to any program, VA'S guidance specifies clearly Updated r only 16. Seventeenof the elements are not mentioned in VA'S guidance and the other 23 are discussedin such vague terms that their inclusion is of little use to practitioners. At five of the seven centers GAO reviewed, Page 3 GAO/HRD-99-27VA’sInfection Control Programs ‘. - ExecutiveSummary - -’ the infection control practitioners took the initiative to take training, contact others, or read literature concerning infection control programs. As a result, they used activities in their programs that went beyond VA guidance and adhered to most of GAO'S basic elements.(Seepp. M-19.) Twelve of the elements identified by GAO are not being used as fre- quently by the medical centers as the other 44. These elements either are not included in VA'S guidance or are stated so vaguely that the guid- ance is of little use to practitioners. They include: assuring that physi- cian advisors have taken a course in hospital infection control programs, increasing the frequency of activities to identify certain infections, rou- tinely reporting surgical wound infection rates to practicing surgeons, and reporting ward-specific infection data to ward supervisors. (Seepp. 19 to 26. ) V& Infection Control In terms of the extent to which each are using the GAO elements, VA'S PrQgramsComparable infection control programs are comparable with those of nonfederal hos- pitals. VA practitioners generally use 44 of the elements, while W&h Those in Nonfederal nonfederal practitioners use 42. With the exception of one area, the spe- Hqspitals cific elements used are the sameand the utilization rates are similar. The specific elements that are being used less in VA are also being used less in nonfederal hospitals and, again, the utilization rates are similar, Minagement Issues Must VA'S guidance requires one full-time infection control practitioner for Be’Resolved every 200 to 260 occupied beds. Basedon this requirement, GAO'S ques- tionnaire results indicated that 66 percent of VA'S infection control pro- grams were understaffed in 1987. At three of the sevenmedical centers GAO reviewed, lack of resourcesled to the practitioners’ not accomplish- ing necessaryinfection control activities. (Seepp. 28 to 30.) GAO found that regional office monitoring was inadequate at four of the sevenmedical centers it visited. The VA regional office inspection team did not identify existing deficiencies, each of which was significant enough to hamper the effectiveness of the centers’ infection control pro- grams. This occurred becausethe regional inspection teams (1) did not always include someoneknowledgeable about infection control pro- grams and (2) were given guidance that was not sufficient to assessthe programs. (Seepp. 31-32.) Y Page4 GAO/HRD90-27VA'sInfectionControlPrograms J, .i i ‘I t \ j Executive summary VA does not have a specific headquarters unit to direct its infection con- trol programs. As a result, at least six different offices are issuing guid- ance on infection control related issues,and someof it is confusing and incorrect. In addition, no formal mechanism has been set up to allow infection control practitioners to share information on their programs. As a result, practitioners spend time developing educational programs, as well as policies and procedures that have already been established by other VA practitioners at other medical centers. (Seepp. 32 to 34.) GAO recommendsthat the Secretary of Veterans Affairs require the ReJcommendations Chief Medical Director to update VA guidance so that it includes componentssimilar to GAO'S basic elements (seep. 27), require medical center directors to reexamine the level of support given to their infection control programs and to provide additional support where appropriate, . incorporate procedures in regional office survey requirements to assure that each medical center’s infection control program is adequately reviewed, and . designate a unit in VA'S central office to direct and coordinate its infec- tion control programs. (Seep. 36.) VA concurred with GAO'S recommendations and stated that it had initi- Agency Comments ated actions to implement them. Y Page6 GAO/HRD-90-27 VA’sJnfectionControlProgran~s Executive Summary 2 Chapter 1 8 Introduction Importance of Infection Control Programs VA’s Infection Control Programs 8 9 Costs of Hospital-Acquired Infections 11 Objectives,Scope,and Methodology 12 Chkpter 2 16 In&eased Use of Our Basic Elements of an Infection Control Program 16 VA Staff Use Many Basic Elements on Their Own 18 Ba$c Elements Could Initiative Improve Infection SomeBasic Elements Are Not as Widely Used as Others 19 ’ Use of One Element Differs Significantly Between VA and 26 Cobtrol Programs Nonfederal Sector 6 Conclusion 26 Recommendation 27 Agency Comments 27 Chapter 3 28 Infection Control Medical Center ManagementDoesNot Emphasize 28 Infection Control Should Receive More Inadequate Regional Office Monitoring of Programs 31 Managerial Attention Fragmented Central Office Oversight 32 Conclusion 34 at &ll VA Levels Recommendations 36 Agency Comments 35 Aopendixes Appendix I: Methodology Used to Develop the Basic 36 Elements of an Infection Control Program Appendix II: Basic Elements of an Infection Control 37 Program Appendix III: Methodology for Sampling Nonfederal 46 Hospitals Appendix IV: VA Medical Centers Visited 47 Appendix V: VA Medical Centers’ and Nonfederal 48 Hospitals’ Use of GAO’s Basic Elements Appendix VI: CommentsFrom the Department of 52 Veterans Affairs Appendix VII: Major Contributors to This Report 54 Page 6 GAO/HRD-90-27VA’sinfection Control Progrmns Table 2.1: Hospitals That Do Not Have Physician 20 Consultants/Supervisors Trained in Infection Control Table 2,2: Hospitals Not Performing CaseFinding at Least 21 OnceEvery 3 Days Table 2.3: Hospitals Performing Bloodstream, Pneumonia, 23 and Urinary Tract Infection Analysis, but Not Reporting Results to Ward Supervisors Table 2.4: Hospitals That Perform Surgical Wound 24 Surveillance, Analyze the Data, and Do Not Report Results to Practicing Surgeons Table 2.6: Hospitals That Conduct Surveillance but Do 26 Not Have Baseline Rates Table II. 1: General Elements 38 Table 11.2:Bloodstream Infections 39 Table 11.3:Pneumonia 40 Table 11.4:Surgical Wound Infections 41 Table 11.6:Urinary Tract Infections 43 Table 111.1:Nonfederal Hospitals-Sample Sizesand 44 ResponseRates by Strata Table 111.2:Estimates and Corresponding Sampling Errors 46 for Nonfederal Hospitals Figure Figure 2.1: Main Categoriesof Elements in Infection 17 Control Programs Abbreviations AIDS acquired immunodeficiency syndrome AHA American Hospital Association CDC Centers for DiseaseControl GAO General Accounting Office Y IV intravenous SENIC Study on the Efficacy of Nosocomial Infection Control VA Department of Veterans Affairs Page7 GAO/HRD-90-27VA%Lnfection Control Program Chabter 1 I$roduction Hospital-acquired infections are a sizable health concern in the United States. The Centers for DiseaseControl (coo) estimates that 6 percent of all inpatients contract at least one hospital-acquired infection during their hospital stays. The Department of Veterans Affairs (VA) operates the largest health care delivery system in the United States and served 1,279,976inpatients in 1987. A S-percentrate of infection would suggest that about 64,000 of these patients could have acquired infections. Although these infections can be relatively minor, somecould be life threatening. Even if relatively few of these individuals choseto submit claims against VA for infections contracted while in medical centers, the cost to the government could be significant. Becauseof the importance of infection control, the Ranking Minority Member of the SenateVeter- ans’ Affairs Committee asked us to evaluate VA’S program to control such infections. , Many factors contribute to the prevalence of hospital-acquired infec- Wportance of tions. Hospitalized patients tend to be more susceptible to infection than Infection Control healthy individuals becausethey are often ill or injured when they enter Prpgranw the hospital. Others may becomemore susceptible as the result of sur- gery, insertion of catheters and tubes, or use of other equipment related to hospital care, such as ventilators. Further, patients admitted with infections could exposeother patients to those infections. Visitors and hospital staff also introduce disease-causingorganisms. Health care workers can reduce the spread of infections from one patient to another by following certain practices when caring for patients. These practices can be simple, such as washing their hands before and after providing care to each patient or using the proper technique to insert a needle intravenously. To minimize the incidence of infections, hospitals’ infection control pro- grams monitor and emphasizepatient care practices through two inter- related activities-surveillance and control. Surveillance activities involve (1) identifying patients with hospital-acquired infections and collecting appropriate data to calculate infection rates, (2) analyzing data about those patients and their infections to determine causes,and (3) reporting analyzed data to hospital management and other staff who can use it to identify the causesof infection and weaknessesin their patient care practices. Control activities consist of the specific actions taken to prevent and control infections, such as developing and revising Y hospital policies; teaching and reinforcing proper patient care practices; and implementing certain practices, such as the isolation of certain infected patients. Page 8 GAO/HRB90-27 VA’sInfection Control Programs .’ . . ’ ‘.I. ,,’ “&,,L. A.- chapter 1 Intmduction - In 1974, CDC began a major study on the efficacy of infection control programs in hospitals.’ The study results showed that when an effective program was present, hospital-acquired infections could be reduced sig- nificantly. The study, which was updated in 1983, covered the four major types of hospital-acquired infections: bloodstream, pneumonia, surgical wound, and urinary tract. Researchersconducting the study estimated that such infections constitute more than 80 percent of all hospital-acquired infections. cnc officials informed us that, generally, the findings of this study are as pertinent today as they were in 1974 and 1983. Specifically, if a hospital implements an effective infection control program, it will decreaseits hospital-acquired infection rate. I i Each VA medical center is required to maintain an infection control pro- E b’s Infection Control gram as part of its quality assuranceefforts. VA'S central office provides PI ~ograms general guidance as to how the programs should be conducted and gives medical center directors responsibility for the program. VA'S regional offices are responsible for the general oversight of the centers’ infection control programs. Medical Center The directors of VA'S 169 medical centers are responsible for developing A$ministration of and implementing their own infection control programs.2The directors receive general guidance from the central office, which includes the Piograms appropriate staffing level for the program and what should be included in surveillance and control activities. In somecenters, infection control may be a part-time responsibility of a single practitioner; in others, the staff may consist of several practitioners. Most of VA’S medical centers are large by hospital standards-61 have 400 or more acute-carebeds, and only 9 have fewer than 100 acute-care beds. VA guidance calls for each medical center to have at least one infec- tion control practitioner for every 200 to 250 beds. The duties of these practitioners-generally registered nurses-include l conducting surveillance to identify infections and potential epidemics, l analyzing infection data (e.g., trending and calculating infection rates), Y ’ CDC’s Study on the Efficacy of Nosocomial Infection Control (SENIC) evaluated the impact of sur- veillance and control activities on hospital-acquired infections in a sample of nonfederal hospitals. “A medical center consists of one or more hospitals and one or more outpatient clinics, and may also include a nursing home and a domiciliary. Page 9 GAO/HRJS90-27VA’sInfection Control Programs chapter 1 Introduction - . preparing written reports of findings for the centers’ infection control committees, l developing policies and procedures, and l educating patients and staff on infection control practices. VA guidance also requires each medical center to establish an infection control committee to be chaired by a physician specially trained in and/ or especially interested in infectious diseases3The committee represents the center’s departments of medicine, surgery, nursing, laboratory, pharmacy, supply, medical administration, building management,and dietetics. The committee also directs the activities of the medical center’s infection control staff. Redional Office Monitoring VA is divided into sevengeographic regions, each responsible for moni- of JV0gran-k toring from 21 to 28 medical centers. Regional directors exercise direct line supervision over medical centers within their region. They are also responsible for enforcing VA’S infection control guidance and evaluating the medical care and related services provided in individual centers. At the time of our review, the regions’ examination of medical care and compliance with VA guidance was conducted primarily through visits to each medical center by a team of health care and administrative person- nel. Central office officials told us that each center should be reviewed at least every 3 years. If deficiencies in the infection control program are found, the center is required to respond with an action plan indicat- ing the corrective steps that will be taken. Action plans are submitted to the region and to the central office for review. The region is responsible for ensuring that final action plans are adequate and that they are car- ried out. Central Office Role in No specific office or person in VA’S central office is directly responsible Infection Control Efforts for the infection control program. However, two central office program units-the infectious diseasesoffice and the nursing service-have assumedresponsibility for those parts of infection control programs that relate to their functional responsibilities. These two units have issued most of the VA guidance used by the medical centers on how infec- tion control programs should be conducted. In addition, the nursing ser- Y vice has assignedan infection control liaison function to one of its “Infectious disease physicians treat patients with infectious diseases, such as hepatitis. However, these physicians do not necessarily have training in hospital infection control. Page 10 GAO/HRDQO-2’7 VA’sInPection Control Programs Chapt8r 1 Introduction nurses in the central office. The liaison function is one of many responsi- bilities assignedto this individual, and no specific infection control pro- gram training is either provided or required for the position. In addition, the liaison only interacts with infection control practitioners who are nurses. In January 1988, VA’S central office appointed a program director for infectious diseasesand gave him responsibility for developing guidance to the medical centers on infectious diseaseissues.This individual is located at a medical center and, in addition to his central office role, is the chief of medicine for his medical center. He has not been allocated any additional staff to assist him with his central office responsibilities, The goal of infection control programs is to prevent infections so that Cbsts of Hospital- patients’ suffering and costs to the VA system are minimizedIn 1987, VA Abquired Infections paid out over $3 million to close 17 malpractice claims related to hospital-acquired infections. Sixteen of these claims were settled out of court and, therefore, do not represent judgments against VA. However, these claims represent only one type of cost to the VA system. Other costs include the extra days the patient is in the medical center being treated for an infection and the additional cost of diagnosing and treating the infection (e.g., extra laboratory test or intravenous antibiotics). Researchershave estimated that an effective infection control program more than pays for itself by preventing infections and thereby reducing patients’ length of stay and related costs of treating infections. The “cost” to the patient who has an infection varies. For example, uri- nary tract infections are generally the least serious and may causethe patient somediscomfort; bloodstream infections or pneumonia could be fatal if not addressedin a timely manner. In July 1988, the Department of Medicine and Surgery, now the Veter- ans Health Servicesand ResearchAdministration, identified infection control programs as a high-risk area under the Financial Integrity Act. High-risk areas identify potential risks in agency operations that require corrective action or further investigation and should be acted on during the first year they are identified. Page 11 GAO/HRD-90-27VA’sInfection Control Programs On November 18,1986, the former Chairman of the SenateVeterans’ Olqjektives,Scope,and Affairs Committee requested that we evaluate VA’S infection control pro- Mqthodology gram. Basedon the Senator’s request letter and discussionswith com- mittee staff, we agreed to determine . if VA medical centers have adequately staffed and organized their infec- tion control programs, . how infection control data are used to help prevent future infections, l the amount of financial lossesresulting from tort claims due to hospital- acquired infections, and l how the VA medical centers’ programs compare with similar nonfederal hospitals’ programs. We conducted this evaluation from June 1987 to June 1989 in accord- ance with generally acceptedgovernment auditing standards. At the inception of this review, we met several times with CDC officials to discuss,from a conceptual perspective, what a good basic infection control program should consist of (Le., a program that would consist only of fundamentals). Drawing from these discussions,we determined that there was no current generally acceptedwritten guidance available to the health care community on what constitutes the basic elements of an effective surveillance programe4Given the lack of guidance, we sought out eight additional organizations recognizedby the medical com- munity as having expertise in infection control, and one infection con- trol expert, to develop a comprehensivelist of elements that might be included in an infection control program in today’s medical environ- ment. Appendix I describeshow the basic elements were developed and gives the organizations and individual we contacted to assist us in this effort. The list developed as a result of these interactions contains 56 basic ele- ments that we believe represent the minimum requirements of an effec- tive infection control program (see app. II). Although the basic elements place emphasis on surveillance activities, control activities are vital to an effective infection control program. We included in our list only broad categoriesof control activities performed by infection control staff becausesomecontrol activities, unlike surveillance activities, are performed not only by the infection control staff but by other hospital 4Guidance on infection surveillance programs should be differentiated from guidance on patient care practices related to infection control. CHICguidelines detail recommendations on practices to be car- ried out by providers to prevent infections but specific guidance on surveillance procedures was not available. Page 12 GAO/~90-27 VA’sInfection Control Programs chapter 1 Introduction , personnel. Further, the control activities performed by the infection con- trol staff depend heavily on the surveillance findings and circumstances within the hospital. The basic elements are limited to those appropriate for acute-carehospitals with 60 or more beds becauseinfection control experts informed us that infection control needsand practices differ for hospitals providing long-term care and for hospitals with fewer than 60 beds. We prepared a questionnaire basedon the elements we developed to help evaluate the content of infection control programs in VA medical centers and to compare VA and nonfederal programs. The questionnaire was developed in close consultation with CDCinfection control experts and sent to all 169 VA medical centers. Responseswere received from all the centers.”However, one VA medical center indicated that it had fewer than 60 acute-carebeds and was excluded from our review. We also mailed questionnaires to a random sample of 567 nonfederal hospitals with 60 or more beds and received 443 usable responses.The data from these hospitals were used to make estimates about the universe of nonfederal hospitals with 60 or more acute-carebeds (estimated at 3,872). All of the data on nonfederal hospitals in this report are based on our sample and are subject to sampling error. Appendix III describes the methodology we used to select the nonfederal hospitals and presents the sampling errors associatedwith the estimates for the nonfederal hospitals. We promised respondents confidentiality to encouragethem to answer the questionnaire accurately. In addition, in the letter that accompanied our questionnaire to VA medical centers, we stated that we would ask some centers to provide documentation to support the accuracy of their responses.Later, we requested documentary support for nine questions from 32 VA centers (about 20 percent) randomly selectedfrom returned questionnaires. We were able to verify 94 percent of the responsesfor which we sought documentation. Three percent were incorrect and we could not determine the accuracy of the other three percent. We do not believe the incorrect and nonvalidated responsesaffect our conclusions in this report. We also visited sevenVA medical centers to review their infection control programs (see app, IV). They were selectedon the basis of size, geo- graphic location, and medical school affiliation. We selectedmedium to 6We also sent the questionnaire to 79 Department of Defense hospitals. We will provide the results of that analysis in a separate report on infection control activities in military hospitals. Page 13 GAO/H&D-90-27VA’sInfection Control Progrruns Chapter 1 Introduction large centers from a variety of geographic regions0 Our selection included five centers that were affiliated with medical schoolsand two that were not. At each center, we interviewed the director, chief of staff, infection control staff, service chiefs, department heads, and ward nurses, as well as officials responsible for the quality assurancepro- gram. We also reviewed pertinent files and records pertaining to quality assuranceand infection control surveillance and control activities, and accompaniedinfection control practitioners on ward rounds to observe infection control activities and practices within the center. We met with officials at three VA regional offices to discusstheir moni- toring of infection control programs and reviewed a selectedsample of regional survey reports, medical center corrective action plans, and regional follow-up reports. We also talked with officials from two addi- tional regional offices to discusstheir monitoring of center’s programs. At the VA central office, we determined how VA rated the risk level asso- ciated with infection control under the Financial Integrity Act; inter- viewed previous and current chiefs of the infectious diseasesprogram and officials within the offices of quality assurance,medical inspector, general counsel, nursing services,and building management;and reviewed files on tort claims closed in calendar year 1987 to identify those related to hospital-acquired infections. To help validate the accu- racy of our determinations that these claims were related to hospital- acquired infections, we asked VA’s medical inspector for verification. Throughout the review, if questions developed about the basic elements or about other matters related to our evaluation, we contacted CDCfor clarification or advice. “VA has only nine medical centers with fewer than 100 acute-care beds. Page 14 GAO/HRD-!30-27 VA’sInfection Control Programs Chapter 2 *creased Use of Our Basic ElementsCobld ’ prove Ink&ion Control Programs T The 66 elements we identified are applicable to both VA medical centers and nonfederal hospitals. Both groups are using many of these elements and the utilization rates for specific elements are similar. (Seeapp. V.) VA infection control staff generally used 44 of the 66 basic elements we developed,practitioners in the private sector generally used 42. Only 16 of the elements are clearly required by VA guidance. The elements that were not widely used by either VA or nonfederal prac- titioners include the frequency of surveillance activities for surgical wound infections, urinary tract infections, and pneumonia; the reporting of infections to ward supervisors and surgeons;and the development of baseline infection rates for pneumonia, Failure to utilize these elements hampers the effectiveness of an infection control program. At the inception of this review, we found that VA guidance on infection B&sic Elements of an control programs (1) had not been updated since 1979, and (2) included Infection Control requirements that were vague and open to interpretation by the practi- Ptiogram tioners. Further, the guidance on infection control programs that existed in the health care community in general was either outdated, did not cover all the major componentsof a program, or did not specifically identify the basic elements of a surveillance program. For example, the Joint Commission on Accreditation of Healthcare Organizations was in the processof rewriting its standards on infection control programs.1In addition, researchershad published studies showing the effectiveness of individual program activities in reducing the number of infections, but no one had studied the effectiveness of a comprehensiveprogram since CDC’S Study on the Efficiency of Nosocomial Infection Control (SENIC). Furthermore, an American Hospital Association infection control hand- book listed 16 surveillance and control activities and stated that someor all of the duties may be performed by the practitioner. The handbook did not differentiate between activities that ought to be performed as part of any basic program and activities that could be considered as part of an optimal program. As discussedon page 12, to fairly evaluate the content of VA’S infection control programs, we worked with several organizations knowledgeable in infection control to identify a set of current basic elements that are flexible enough to apply to different acute-carehospital environments2 ‘The revised standards on infection control took effect on January 1, 1990. ‘Hospital environments may differ in the types of clinical services offered, types and amount of available resources, and the number and risk of infections. Page 15 GAO/HRD99-27 VA’sInfection Control Programs I r , chapter 2 Increased Usesof Our Basic Elementa Could Improve Infection Control Program9 The result is a set of 66 elements in five categories:general elements, bloodstream infections, pneumonia, surgical wound infections, and uri- nary tract infections. Appendix II lists all these elements.The general category includes 14 elements that apply to all hospitals and address program structure, surveillance activities, and control activities. The other four categoriesrepresent surveillance activities specific to the major types of infections. The surveillance activities relate to identification of infections-what sourcesto use and how often to iden- tify infections, analysis of infections -what factors should be analyzed to determine the causesof an infection, and reporting of infections -which hospital officials should receive ana- lyzed infection data. Figure 2.1 cites specific examples of the elements that are included in each of the categories. Page 16 GAO/HRD-90-27VA’sInfection Control Programs brewed Uw of Our Bah Ellemente Could Improve Infection Control Program ire 2.1: Main Categories of Element8 hction Control Program8 caiogory Examples of Basic Elements General (14 elements) l The infection control committee meets at least every 2 months. 0 The hospital’s infection control program has written standardized criteria (definitions) for hospital-acquired infections at specific sites. l The hospital’s infection control program assists in identifying and developing topics for in-service training. ldentlflcatlon l A hospital’s infection control program uses at least one of the following case-finding approaches to identify bloodstream infections either in all patients or in a subset of patients: l review of patients’ blood cultures 4 review of patients’ charts l review of patients’ fever charts Pneumonla Surveillance ldentlflcatlon (11 elements) 0 During the surveillance period, a hospital’s infection control program performs an acceptable case- finding approach on an average of every 3 days. Surgical Wound Infectlon Analysis Surveillance (13 elements) 0 The infection control staff analyzes surgical wound infection data by surgeon. Reportlng 0 The infection control staff reports ward-specific summarized/analyzed data on hospital-acquired urinary tract infections to the ward supervisors or head nurses. Our basic elements allow for two different surveillance approaches: total surveillance or targeted surveillance. Under total surveillance, an infection control practitioner searchesfor all four major types of infec- tion in every hospital patient, on either a periodic or a continuous basis. Practitioners performing this type of surveillance would generally use Page 17 GAO/HRD-90-27VA’sInf’ection Control Programs a. chapter 2 Increased Ule of Our Basic Elementa Could Improve Infection Control Programn 61 of our 66 elements.3Under targeted surveillance, the areas of highest risk or concern are periodically identified, and the infection control practitioner focuseson those areas. For example, for a specified period of time a hospital may elect to search for one or more of the infection types in only high-risk patients, such as those in the intensive care unit. Under targeted surveillance, 14 general elements would always apply, and the remaining elements used would depend on the specific infection types (bloodstream, pneumonia, surgical wound, or urinary tract) being targeted in the high-risk patients. The type of surveillance used dependson the staff available and the special requirements or interests of the hospital or medical center. CDC officials stated that, in the 19709,medical authorities believed that all hospitals should conduct total surveillance. Sincethat time, experts have acknowledged that infection control program resourcesare better spent focusing on patients with a high risk of infection, such as those in the intensive care unit, where the impact from reducing infections is greater. Thus, targeted surveillance can be an acceptableapproach. Forty-four of the 66 basic elements were being used by at least 70 per- VA Staff Use Many cent of VA medical centers when such use was appropriate.4 Similarly, 42 Basic Elements on of the sameelements used by VA centers were being used by at least 70 Teeir Own Initiative percent of the nonfederal hospitals when appropriate. The other ele- ments were also used by someVA medical centers and nonfederal hospi- tals, but to a more limited extent. VA’s guidance on infection control programs clearly requires only 16 of the 66 basic elements we developed: 17 of our basic elements are not included in VA’S guidance and the other 23 are discussedin such vague terms that the guidance is of little use to the practitioner. For example, our basic elements indicate that ward supervisor@should receive data, analyzed by ward, on bloodstream infections, urinary tract infections, and pneumonia. However, VA guidance states only that the practitioner “The five remaining elements include activities that are applicable when targeting certain types of infections (e.g., analyzing hospital-acquired bloodstream infection data by whether or not the patient had an intravenous (IV) catheter). 4Ekcause not all centers and hospitals need to use all elements, when we calculated percentages we included only those centers in which the element was applicable. “We use the term “ward supervisors” to also refer to “head nurses.” Page 19 GAO/HRD-90-27VA’sInfection Control Programs chaPtm2 Increased the of Our Bdc Elements C!ould Improve InBction Control Program13 should prepare a report for “appropriate personnel” and allows infec- tion control staff to decide who is appropriate and what is to be included in the report. At five of the sevenVA medical centers we visited, the practitioners took the initiative to take training, contact others both inside and outside VA, or read literature concerning infection control programs. As a result, they included activities in their programs that went beyond VA guidance and adhered to many of our basic elements. Certain elements we identified were not used on a consistent basis by )me Basic Elements infection control programs in either VA or the nonfederal sector.” These re Not as Widely elements relate to sed as Others - . the availability of a trained physician consultant for the infection con- trol program, l the frequency with which surveillance activities should take place, l the submission of infection control analysesto personnel who can use them, and . the development of a baseline rate for pneumonia. While we did not ask respondentsto our questionnaire why these ele- ments were not used as frequently as the others, VA’S Chief of Infectious Diseasessaid that these elements require resourcesthat are not always available to the medical center. In addition, in the caseof VA, none of these elements were clearly required in its guidance. Trained Physician One of our elements calls for infection control programs to have, as a Consultants Needed for consultant or supervisor, a physician who has taken at least one training course in hospital infection control. This element has a relatively low Ibfection Control utilization rate in both VA medical centers and nonfederal hospitals when Programs compared with most other elements. According to VA practitioners who filled out our questionnaire, 166 VA medical centers have a physician available to the infection control staff as a consultant or supervisor. But 63 of these physicians had not received any training in hospital infec- tion control. “We define “consistent basis” as when 70 percent or more of the medical centers or nonfederal hospi- tals use the element. Page 19 GAO/HRDM-27 VA%Infection Control Programs I Chapter 2 Increased Uw of Our Bade Element* Could Improve Infection control Programs cnc’s SENIC found a correlation between fewer infections and programs with a physician supervisor trained in infection control. While this study is now dated, cnc officials still believe that trained physician con- sultants or supervisors are an important factor for an effective infection control program. VA’s guidance doesnot require an infection control advisor to have infec- tion control training -it is optional. Specifically, VA guidance states that the chairman of the infection control committee should be a physician who is specially trained in or especially interested in infectious diseases. Many nonfederal hospitalsalso had infection control physician consul- tants or supervisors who had not taken at least one training course in hospital infection control. Further, large VA centers and nonfederal hos- pitals were more likely to have a trained physician than their smaller counterparts. Table 2.1 provides a comparative analysis of the utiliza- tion of this element by VA and nonfederal hospitals. Phyqiclan Conrultants/Supewiron Number of Hoapitala without a trained Traitled In Infection Control applicable infection control physician Hospital8 hOspitEd Number Percent Under 400 beds: VA 97 47 48 Nonfederal 3.363 1.817 54 400 or more beds: VA 61 19 31 Nonfederal 509 233 46 N&d for Frequent Our basic elements call for an infection control practitioner to take steps Surveillance at least every 3 days to (1) determine which hospital patients may have infections and (2) identify infections before a patient’s discharge. This activity-called casefinding-can be done by reviewing laboratory cul- tures or patient charts, and/or asking nurses about patients with signs or symptoms of infection. These activities are time consuming and require the practitioners continual attention, but performing them every 3 days allows the infection control staff to identify problems and take corrective action before the infections get out of control. CDC officials believe that a frequency of at least every 3 days is necessarybecause hospitals are emphasizing shorter lengths of stay for patients. Thus, if 0 casefinding is not done within this time frame, patients with an infec- tion may not be identified by the program before discharge. Page 20 GAO/HRD-90-27VA'sInfection Control Programs chapter 2 Increaeed Ule of Our Baeic Elemente Could Improve Infection Control Progrtun# With the exception of surgical wound infection surveillance, current VA guidance doesnot addressthe frequency with which surveillance should be performed. For surgical wound infections, a surgical services direc- tive states that such surveillance should be performed daily by surgical service personnel. The guidance doesnot mention the infection control staff’s responsibility in this area. Our questionnaire results indicated that infection control practitioners at most VA medical centers were performing casefinding at least every 3 days for bloodstream surveillance but almost half were not doing so as frequently for surgical wound, pneumonia, and urinary tract surveil- lance. Many infection control practitioners in nonfederal hospitals indi- cated a similar pattern of casefinding. Table 2.2 comparesthe utilization by VA and nonfederal hospitals of our case-finding elements pertaining to surgical wound, pneumonia, and urinary tract infections. Pinging at Leabt once Every 3 Daya” Number of hO8pitai5 to which iiO8Pitai8 not adhering to element Category/hospital element applies Number Percent Cabe finding for surgical wound infections: VA 131 55 42 Nonfederal 3.732 1.671 45 Case finding for pneumonia: VA 141 62 44 Nonfederal 3,546 1,419 40 Cabe finding for urinary tract infections: VA 141 66 47 Nonfederal 3,586 1,631 45 % this and other tables to follow, when we report the results of our analysis for the universe of VA medical centers and the projected universe of nonfederal hospitals, it is because we did not find signifi- cant differences between VA and nonfederal responses by hospital size. In contrast to the aforementioned utilization rates, only 26 (17 percent) of the 162 VA medical centers and 606 (17 percent) of 3,643 nonfederal hospitals performing bloodstream surveillance did not perform case finding for bloodstream infections every 3 days. One possible explana- tion for the higher use rate is that the case-finding activities associated with bloodstream infections are less time consuming than those required for other infections. Case-finding for bloodstream infections can be done through a review of laboratory results, whereas identifying the other three types of infections require more time-consuming methods. These methods include a review of laboratory results coupled with discussions with nurses about signs and symptoms of infection in patients or an Page 21 GAO/HID-90-27 VA’sInfection Control Progmns I ,- Chapter 2 Increased Use of Our Basic Element8 Could Improve Inhction Control Program13 examination of patients’ medical records. VA’S program director for infectious diseasessaid that these elements were not being used in VA becausethe activities are labor intensive and the infection control pro- grams are understaffed. Inf ’ ction Control Data Infection control experts indicate that pertinent data on infections Sh uld Be Reported to should be reported to hospital personnel who can take action to prevent infections. Thirteen of our basic elements identify specific personnel to Pei onnel Who Can Take whom certain data analyses should be provided (e.g., urinary tract Preiventive Actions infection data should be analyzed by ward and given to the ward super- visor). VA guidance requires only that the infection control staff “pre- pare a written report for the Infection Control Committee and other appropriate personnel.” The guidance doesnot elaborate on what the written report should contain, and the determination of who are “appro- priate personnel” is left to the discretion of the infection control staff. Despite the lack of guidance on who should receive infection control reports, nearly all VA medical centers generally used 6 of these 13 ele- ments. Specifically, at least 90 percent of the VA medical centers reported appropriate data on infections to the infection control commit- tee and/or surgical wound infection data to the chief of surgery and operating room supervisor. The other seven elements were used, but to a lesser extent. The elements with relatively low utilization rates involve reporting to ward supervisors, practicing surgeons,and various other hospital groups, such as respiratory therapy teams. Reporting to Ward Supervisors Our basic elements state that medical facilities that perform blood- stream, pneumonia, or urinary tract infection surveillance, should ana- lyze their data by ward and report that information to ward supervisors. Our questionnaire results reveal that over 95 percent of the VA medical centers that perform surveillance on each of these types of infections did analyze by ward. But many did not report this infection data to ward supervisors. Hospitals in the nonfederal sector that ana- lyzed infection data by ward reported this data to ward supervisors with about the same frequency as their VA counterparts. Table 2.3 com- pares the utilization of these elements by VA and nonfederal hospitals. Y Page 22 GAO/HRD-90-27VA’sInfection Control Programs i . , chapter 2 Increaeed Ueeof Our Be&c Elements Could Improve Infection Control Progrema Tab e 2.3: Horpltalr Performlng 810 detresm, Pneumonla, and Urinary Number of Tra t lnfectlon Analy818, but Not hospital8 Ho8pltal8 not reporting analyrk Re ortlng Rerults to Ward Supervlrorr performln results to ward 8UPerVl8Or8 Category/hospital analysis by war 3 Number Percent i Bloodstream rurvelllance: VA 148 52 35 Nonfederal 2,988 962 32 Pneumonia 8urvelllance: VA 135 47 35 Nonfederal 2,928 789 27 Urinary tract surveillance: VA 139 44 32 Nonfederal 3,051 708 23 Reporting to Practicing Surgeons Studies by various infection control researchersindicate that reporting surgical wound infection rates to the surgeonswho perform the opera- tions has a major effect on reducing infections. Several infection control experts argue that reporting surgeon-specificrates is essential if an infection control program is to have any successin preventing surgical wound infections. Our basic elements call for the analysis of surgical wound infection rates by surgeon and the reporting of the surgeons’ specific infection rates back to them. Thus, each surgeon would receive information about the infections contracted by his or her patients. VA’S program director for infectious diseasesstated that analysis of surgical wound data by sur- geon is time consuming and requires adequate staffing to be done cor- rectly. But once the analysis is done, reporting the data to surgeons should not be time consuming. However, he also indicated that surgeons may not be receptive to the infection control staff’s analysis of surgical wound data becausethey generally believe that without their involve- ment in the collection and interpretation of these data, the data could be misinterpreted. Table 2.4 compares utilization rates on these elements of VA medical cen- ters and nonfederal hospitals. Page 23 GAO/HRD9@27VA’sInfection Control Programs chapter 2 Increased Use of Our Basic Elements Could Improve Infection Control Programs Horpltalr That Perform tiO8pltal8 IlOt Hospitals reporting Percentage not performlng HO8pitalS Ofl8ly8i8 reporting surgical wound analyzing data results to an8ly8i8 Horpltal surveillance by surgeon surgeon results VA 131 93 72 77 Nonfederal 3.732 2.973 2.234 75 Re rting to Other Groups Our basic elements include the reporting of pneumonia infection rates to the respiratory therapy department and bloodstream infection rates to the intravenous therapy team. In both instances, these personnel pro- vide direct patient care and would benefit from knowing such infection rates. Questionnaire results indicated that of the 66 VA medical centers with respiratory therapy teams, 36 did not have infection control pro- grams that reported the results of pneumonia surveillance back to the teams, Only nine VA medical centers doing bloodstream surveillance have intravenous therapy teams, and sevenof these teams did not receive data from the infection control program. In the nonfederal sector, of the 1,064 infection control programs at hos- pitals with respiratory therapy teams, 426 did not report the results of pneumonia surveillance back to the teams. Of the 494 nonfederal hospi- tals doing bloodstream surveillance that have intravenous therapy teams, 214 teams did not receive data from the infection control program, Our basic elements also require infection control program personnel to provide surgical wound infection data to the surgical complications com- mittee, which is responsible for reviewing surgical complications, includ- ing infections. The purpose of this element is to assurethat this committee receives complete and accurate information on surgical wound infections since they discusssurgical complications and actions to prevent them in the future. Further, in our opinion, the discussions can be enhancedif information collected by the infection control staff is analyzed before it is submitted to the committee so that trends can be identified. At the present time, the committee usually relies on physi- cians to report on complications causedby surgical procedures; research indicates that self-reporting doesnot provide complete information. Questionnaire results revealed that the infection control staff at 39 (42 percent) of the 92 VA medical centers that monitored surgical wound infections with surgical complications committees, did not share surgical Page 24 GAO/HRD-90-27VA’sInfection Control Programs Chapter 2 Increlllled ,Um of Our LbaaicElements Could Improve In&ction Ckmtrol Programs wound surveillance data with the committee, In the nonfederal sector, infection control staff in 746 (38 percent) of the 1,976 hospitals that monitor surgical wound infections with surgical complications commit- tees did not share pertinent data with the committee. ed for Baseline Rates by Baseline infection rates give a hospital an indication of its “normal” pe of Infection level of infection. These rates represent the frequency with which a spe- cific type of infection occurs within a targeted population in a particular hospital based on past surveillance. For example, surveillance conducted on bloodstream infections at a given hospital over the last 2 years may show a l-percent bloodstream infection rate. This percentagebecomes the baseline from which future rates will be measured. Current VA guidance states that baseline rates should be established but does not specify how they should be calculated. Our basic elements spe- cifically require baseline rates to be developed by each of four types of infection-surgical wound, urinary tract, bloodstream, and pneumo- nia-because the risks and causesof infections vary by type. Although VA medical centers generally calculated baseline rates for surgical wound, bloodstream, and/or urinary tract infections, 31 percent of the centers did not calculate rates for pneumonia. VA’S program director for infectious diseasesbelieves that the data for calculating such baseline rates are readily available but noted that VA guidance should more clearly stress the importance of developing baseline rates for each type of infection. As shown in table 2.6, with the exception of baseline rates for surgical wound surveillance, VA medical centers and nonfederal hospitals were similar in their use of our elements concerning baseline rates. Page 26 GAO/Hl?D9O-27VA’sMe&ion Control Programs chapter 2 lncreasod Urn of Our Bmic Element4 could Improve InfectIon Control Program9 , 5: Hpapltala That Conduct Surveil ante but Do Not Have Baclollne Number of Rate8 hoapltals conducting Hospitals without baseline rate Category/hoapltal clurveillance Number Percent Bloodstream surveillance: VA 152 41 27 Nonfederal 3,648 1,327 56 Surgical wound rurveillance: VA 131 16 12 Nonfederal 3,732 970 26 Pneumonia crurveillance: VA 141 44 31 Nonfederal 3,546 1,150 32 Urinary tract surveillance: VA 141 38 27 Nonfederal 3,586 1,026 29 One basic element was used significantly less in the nonfederal sector Us4 of One Element than it was in VA medical centers. This element requires that infection Differs Significantly control practitioners and registered nurses have authority to implement EktkveenVA and isolation procedures in an emergencywithout a physician’s order; this authority should be in writing. Questionnaire results indicated that only Nonfederal Sector 19 percent of VA centers with 100 to 399 beds and 23 percent of those with over 400 beds did not have written authority. Conversely, about 42 percent of nonfederal hospitals with 100 to 399 beds and 46 percent of hospitals with over 400 beds did not have written authority. The basic elements we developed are fundamental, generally accepted Coklusion by infection control practitioners in both the public and private sectors, and supported by organizations knowledgeable about infection control. Cumulatively, they form the basis for an effective infection control pro- gram whether in VA or nonfederal hospitals. Infection control programs in both VA and the nonfederal sector are using most of the elements we developed. This is happening becausethe practitioners are taking the initiative to determine what elements should be used in an effective infection control program. But VA'S written guid- ance on the subject lags well behind the practitioners’ activities and, in many cases,is so general that it is of questionable value. We believe that Page 26 GAO/HRIMO-27VA’sInf’ection Control Programs I 1’ i F ” ’ chapter2 tncreaeed Urn ofOur Bade Element.4Cimld improve Infection Control Program the use of the elements we identified, included in a basic infection con- trol program, could improve the effectiveness of VA’S programs. I We recommendthat the Secretary of Veterans Affairs direct the Chief R&commendation Medical Director to update VA infection control guidance. At a minimum, the guidance should require components similar to those in our basic elements. I I By letter dated November 21,1989, the Secretary of Veterans Affairs Alpmcy Comments concurred with our recommendation and indicated that VA infection con- / trol guidance would be revised to incorporate our basic elements. Page 27 GAO/HRD-90-27 VA’sInfeztion Cbntrol Pmgrama Chap& 3 Infection Control Should ReceiveMore ’ M&mgerid Attention at All VA Levels Although our questionnaire showed that VA infection control programs used most of our basic elements (seech. 2), a questionnaire cannot cap- ture how well these programs were integrated and supported through- out the VA system. Lack of managementsupport and attention can undercut the programs’ effectiveness. Infection control programs at several of the medical centers we visited were understaffed and not well directed or monitored by the responsible VA offices. Of the sevenmedical centers we visited, four of the infection control programs were understaffed. At the central office level, no sin- gle office is responsible for directing the infection control programs and coordinating guidance. Instead, guidance is developed and issued by six different offices. This has resulted in confusing or inaccurate guidance. Further, regional office surveys conducted at six of the sevenmedical centers did not always identify existing deficiencies. As a result of these situations, infection control programs were not as effective as they could have been. Infection control is one of many medical center programs competing for Medical Center relatively scarceresources.It is also a program that affects many center Mtiagement Does Not activities. At five of the sevenmedical centers we visited, center direc- Embhasize Infection tors had either not given the program adequate resourcesto accomplish its objectives or not enforced participation in the program by all clinical Co&r01 services of the center. This reduced the program’s effectiveness. Medical Center In calendar year 1987,82 of VA’s168 infection control programs were Mariagement Did Not understaffed. Our criteria for determining the adequacy of staffing is based on VA guidance, which calls for one infection control practitioner All&ate Adequate for every 200 to 260 occupied beds. Of the sevenmedical centers we Resourcesto the Infection reviewed, four did not meet this staffing standard, and the practitioners Control Program at these centers stated that they did not have time to perform someof the essential infection control activities. For example, at one of the cen- ters, the practitioner collected all the necessarydata on surgical wound infections but had no time to analyze them and, thus, did nothing with them. The directors of two of the understaffed medical centers said that they were each planning to hire another nurse as an infection control practi- tioner. A third medical center director told us that he could not allocate another nurse to infection control becausenurses were in short supply. Page 28 GAO/HRD-90-27VA’sInfection Control Programs chapter 8 ‘r Infiwtlon Cbntrol Slmld ltecdve Mom Managerial Attention at All VA bveb : But adding staff is not the only solution available, For example, the infection control program at the fourth understaffed medical center (the center had one 40-hour-per-week practitioner for 316 occupied beds), met all our basic elements by assigning specific responsibility for parts of the infection control program to medical center services (e.g., nursing, surgery, and medicine). Somecenter serviceswere responsible for con- ducting training classeson new infection control policies while others were responsible for monitoring the staff’s patient care practices and documenting their findings. This gave the practitioner time to collect and analyze data on infections and prepare reports to the infection con- trol committee and others. We believe this coordination worked at this center becausethe chief of staff and center director, through their actions, emphasizedinfection control. Further, at this center, the chief of staff chaired the infection control committee. In our opinion, VA medical centers should have a designated person who can perform certain infection control activities when the infection con- trol practitioner is absent. This designated person should be trained to perform basic infection control activities so that the program doesnot stop while the practitioner is away. Four of the seven centers we visited had only one practitioner and no one to perform this function when he or she was absent. At one of these medical centers, during the week the practitioner was away, the center had a number of patients who devel- oped a bacterial infection. When the practitioner returned, she found that the infection started with two patients in the medical intensive care unit who were moved to the surgical intensive care unit. This infection had spread to sevenother patients by the time the practitioner returned. The practitioner speculated that had she been in the center or had a backup, the source of the infection would have been identified earlier, the patients who had contracted it would have been treated earlier, and perhaps someof the spread prevented. In addition to staffing, other resourcesare neededto carry out an effec- tive infection control program. Since at least 1984, infection control practitioners throughout VA have asked the central office for computer support, and central office officials have said that infection control should be given priority for computer time within a center. But the prac- titioners were also told by central office personnel that such support is a local policy issue and that they would have to compete for such resourcesat the center level. The need for computer support apparently Y still exists at 65 medical centers becausethose practitioners indicated through their questionnaire responsesthat they need more computer support (e.g., personal computers, software, and training). Further, in Page 29 GAO/HRD-99-27VA’sInfection Control Programs I Chapter 3 Infection Control Should ReceiveMore Managerial Attention at AU VALevel -?- the sevencenters we visited, only two practitioners were using a com- puter to perform data analysis on infections. Practitioners that were not using computers cited the need for such equipment and indicated that a computer would also help easethe burden of the understaffing because analysis by computer is faster. At one medical center we visited, the chairperson of the infection con- trol committee delayed surveillance on surgical wound infections for 1 year because,without a computer, the practitioners would, not have time to analyze the data collected. Two days before our arrival at this center, the infection control practitioners started their surveillance of surgical wound infections becausethey had just obtained accessto a computer and, therefore, could perform timely analysis of the data collected. Three months later, the physician who headed the program indicated that, as expected, the computer analysis showed that the surgical infec- tion rate for clean surgeries was higher than reported in previous years.’ Medical Center To foster an effective infection control program, all serviceswithin a Mahagement Gives medical center must actively support and participate in it. At five of the seven medical centers visited, it appeared that infection control was not Infhtion Control considered a centerwide program, and somemedical center components Probrams Low Priority gave it limited attention. For example, we found center directors, medi- cal service chiefs, and other key center officials who (1) did not assure that their representatives attended infection control committee meet- ings, (2) indicated that infection control surveillance was unimportant, and (3) were unaware of infection control issueswithin their centers. At one center, the chief of medicine relied on his “intuitive insights” to identify problems from infections rather than on reviews of the infec- tion control practitioner’s information. Infection control programs can be enhancedwhen there is interest and teamwork within a center, starting with center management.For exam- ple in early 1986, the infection control practitioner at one center we reviewed noticed that a large number of vascular surgery patients who should have had a low risk of infection (their surgery would be classi- fied as clean) had wound infections. She began investigating this but, becauseof concernswith infections occurring in cardiac surgery, had to stop and focus her efforts on cardiac surgery. During this period, the Y ‘Surgical wounds can be classified into four wound categories (clean, clean-contaminated, contmi- nated, and dirty) according to their risk of infection. Clean wounds have the lowest risk of infection and dirty wounds have the highest. Page 30 GAO/HID-99-27 VA’sInfection Control Programe . Chapter 8 Infection CmWol Should hcelve More Manngeti Attention at All VA Levels chief of vascular surgery left VA and a new chief was appointed. Upon arrival at the medical center the new chief noted that infection control practices were being neglected by his residents and staff. Within 2 T--- months, with input from the practitioners, he developed a policy manual for his department. It included infection control practices that he enforced. At the time we visited (March 1988), his department had had no infections on clean surgeries since September 1986. At the time of our review, regional offices were required to conduct a In dequate Regional general survey of the effectiveness of the medical centers and their 0 $fice Monitoring of quality assuranceactivities at least every 3 years. During this survey, P~ograxns an examination was to be made of how infection control programs were being implemented. Our review showed that regional offices may not always perform this function. Further, the teams formed by the regional offices to conduct the general survey did not always include personnel knowledgeable in infection control, and the guidance given to teams to review a medical center’s performance may not have been appropriate to measure performance accurately. In 1986 and 1987, the regional offices had surveyed six of the seven medical centers we reviewed. Three of the regional office survey reports made no mention of infection control. But our examination of infection control programs at these centers revealed that at the time the regional office survey was made deficiencies that should have been identified and reported were present in two programs. For example, at one center we found that infection control practitioners were identifying surgical wound infections by using culture results. But officials knowledgeable in infection control told us that many surgical wound infections are not cultured. Therefore, another identification method, such as asking ward nurses about infections, should also be used. In addition, the practition- ers were not aware that the chief of surgery was discouraging his staff from culturing obviously infected surgical wounds. The chief of surgery believed that as a result of this miscommunication practitioners were missing about 20 percent of the wound infections. Of the three regional office survey reports that mentioned infection con- trol, two did not discussproblems that we found during our review. Spe- cifically, one report, dated May 1987, mentioned infection control only to the extent that surgical wound infections were not being categorized by type of wound; that is, clean, contaminated, or dirty wounds. But we found that the practitioner at this center was not performing surgical wound infection surveillance. The acting chief of surgery was tracking Page 31 GAO/HRD-90-27VA’sInfection Control Programs r Chapter 3 Infection Control Should Receive More Managerial Attention at All VA Levels wound infections on certain surgeries becausethe infection control pro- gram was understaffed and could not monitor such infections. In the acting chief’s opinion, however, surveillance by the infection control practitioner was still needed.This should have been noted in the regional office report. VA has not given any one office responsibility for providing infection Fragmented Central control guidance to its medical centers. As a result, at least six program Off ice Oversight offices within the central office have issued guidance to the medical cen- ters on infection control and this guidance is sometimesconfusing or / inaccurate. In addition, practitioners do not know whom to contact when they have questions regarding infection control, and the central office has no mechanismto share medical centers’ information related to infection control. Therefore, practitioners spend time developing poli- ties, procedures, and educational programs that have already been developed by practitioners at other VA medical centers. Infection Control Guidance Six different program offices in VA’S central office have issued guidance Is Not Coordinated by the relating to infection control. As a result, such guidance sent to the medi- cal centers can be confusing or inaccurate. The need for someonein VA’S Ceritral Office central office to coordinate such guidance was recognizedas early as 1984 at a conferenceof infection control practitioners. The two central office officials at the meeting, the nursing services’ infection control liai- son and chief of infectious diseases,said that they would look into the concernsand write a proposal. However, this was not done. Guidance that has not been coordinated can causeconfusion among practitioners. For example, in 1984, VA’S engineering servicesissued guidance to the medical centers on infectious waste that required need- les to be capped while being carried. Central office officials told us that they received numerous calls from medical center nurses indicating that the engineering guidance was wrong. The guidance implied that needles should be recapped after use whereas CM:recommendedthat needles should not be recapped. CDCmade this recommendation to decreasethe incidence of nurses sticking themselves when recapping a used needle, which could causeblood-borne infection, such as acquired immu- nodeficiency syndrome (AIDS) or hepatitis. To clear up the matter, the central office nursing service issued a clarification of this circular to all Y center personnel indicating that needlesshould only be capped before being used on a patient and that after use they should be discarded in a puncture-proof container. Page 32 GAO/HRD-90-27VA’sInfection Control Programs Infection Control Should Receive More Managerial Attmtlon at AU VA Levels The fact that different offices issue infection control guidance has also causedcommunication problems. This was recognizedin 1987 by the AIDSprogram director (former chief of infectious diseases)at VA’S central office. At that time, she noted that many practitioners had not received appropriate bulletins becauseno one program office distributes infec- tion control guidance. For example, medical centers rely on the infection control practitioners to handle AIDSissues,But the practitioners were not receiving all the Ams-relatedguidance. Thus, at a meeting of the practitioners, the AIDS program director listed all the guidance address- ing AIDS issues.This was the first time many of the practitioners had a complete set of AIDS-relatedguidance. I&ection Control VA has not implemented a formal, structured mechanism for practition- Pkactitioners Duplicate ers to share ideas and disseminate information among themselves. As a result, practitioners spend time developing program activities that may Efforts already have been developed at another center. At three of the seven centers we reviewed, practitioners commentedthat they were frustrated over “reinventing the wheel” in their program. In their opinion, other practitioners within VA might have developed procedures, policies, or training programs that could be used in their own programs, This con- cern was also raised in meetings of VA practitioners held in 1984, 1986, and 1987. For example, at the 1984 meeting, VA practitioners stated that a compilation of infection control procedures and policies would be help- ful to practitioners. If such a document were available, they would not have to write new policies and procedures but could modify others for their center. As of July 1989, this had not been done. In responseto practitioners’ requests, vAcentral office officials sug- gestedthat they consider developing a communication network among centers. However, the practitioners had already set up an informal net- work to share information among themselves. This network encourages practitioners to participate in the annual conferenceof the Association of Practitioners in Infection Control, where VA practitioners discuss areas of mutual interest. At the initiative of members of the informal network, such meetings were held annually during the 1984 to 1988 con- ferences. The practitioners involved in the network have identified a number of efforts that would help make their programs more efficient. For example, someVApractitioners have developed videotapes to edu- cate staff on certain practices to prevent infections. The network was Y trying to identify those centers with educational tapes and set up a mechanism by which other centers could borrow or copy the tapes. Page 33 GAO/IiRD-90.27 VA’sInfection Control Program8 Chapter 8 InfectIon Control Should RecedveMore Managerial Attention at All VA Levels Many of the practitioners we talked with stated that the informal net- work is very beneficial, but they believe the central office should under- take a more formal coordination of infection control procedures and policies. For example, at one medical center we visited, a recently appointed infection control practitioner found out about the network only becausea friend at another VA center mentioned it. The practition- ers that help with the network are volunteers, and their time comesout of either their own program or their personal time. As a result, many of the initiatives that are conceivedby the network are delayed or never completed. Infection control is an important aspect of hospital operations, and per- Corklusion sonnel assignedto conduct such programs should be supported by man- agement at all levels, The fact that six different units provide guidance to infection control practitioners is an indication that the program is not well defined or directed by the central office. Lack of management attention to a number of areas undercuts the effec- tiveness of infection control programs. For example, VA has provided limited or no responseto continued requests for (1) a formal mechanism to share information, (2) coordination of existing infection control guid- ance, and (3) computer support for analysis purposes. This gives a sig- nal to the practitioners that their program may not be as important to VA as other programs. Moreover, a failure of medical center managementto properly recognize infection control as a centerwide effort and a priority can hamper the program’s effectiveness. Finally, a shortage of infection control staff affects the program’s ability to accomplish its objectives. With a minimum of time and money, VA management at the medical center, regional, and central office levels can greatly enhanceboth the efficiency and effectiveness of infection control programs. Specifically, all infection control guidance should be reviewed and coordinated by one central unit. This will send a messageto VA staff at all levels that the program is important, understood, and well directed. The unit selected to perform this function should serve as a mechanism to facilitate com- munication between all practitioners in the VA system and should assure that developments in the field of infection control, both inside and outside VA, are widely disseminated. At the medical centers, directors must recognizethat infection control is important to all center staff- not just the practitioners -and should, through both words and actions, Page 34 GAO/IUD-90-27 VA’sInfection Control Programs chrpt%r 9 Infectlola confrol Should ReceiveMore Manage&l Attantlon at All VA Levels I get this messageout to all staff. The words should indicate strong sup- port; the actions should be the provision of appropriate resourcesto get the job done. We recommend that the Secretary of Veterans Affairs direct the Chief Recommendations Medical Director to . designate a single unit in VA’S central office to overseeits infection con- trol programs, including (1) coordinating the central office’s policies and procedures and (2) implementing a mechanism for sharing information among practitioners; . incorporate procedures in regional office survey requirements to assure that each medical center’s infection control program is adequately reviewed; and 9 require center directors to reexamine the level of managerial support and resourcesgiven to their infection control programs and to provide additional support where appropriate. In his November 21, 1989, letter, the Secretary of Veterans Affairs con- Agency Comments curred with each of our recommendations.The Secretary stated that an office has been designated to overseeVA’S infection control program and this office will work with VA’S infectious diseasefield advisory group in formulating infection control policies and procedures. The regions will work with these program officials to develop appropriate criteria to assessthe programs. The Secretary also stated that the Chief Medical Director will require medical center directors to reexamine their infec- tion control program resourcesand, basedon justifications and sys- temwide priorities, additional support will be provided. Y Page 35 GAO/klRD90-27 VA’sInfection Cmdrol Programs Apperdix I . To develop the basic elements, we first consulted with officials of the organizations and the individual listed below: American Hospital Association. Association for Practitioners in Infection Control. Centers for DiseaseControl (CDC). Joint Commission on Accreditation of Healthcare Organizations. The Society of Hospital Epidemiologists of America. Robert W, Haley, M.D., who directed CDC'S SENIC study. We discussedthe activities they thought were necessaryfor an effective infection control program and, working with CDC officials, compiled a comprehensive list of elements. The list was sent to the above organiza- tions (except CDC) as well as the organizations listed below: American Public Health Association. The Association of Operating Room Nurses, Inc. Association of State and Territorial Health Officials. Surgical Infection Society. We asked each to indicate which elements could be considered minimum requirements for an effective infection control program. From their responses,we developed a list of the elements that six or more agreed were minimum requirements and subsequently discussedthese with CDC infection control experts. Using existing Joint Commission standards, published studies demon- strating the effectiveness of an element, and CDC’S judgment as to whether the element would be widely supported by infection control experts, we arrived at a final list of 56 elements. CDC officials believe these elements represent a good basic infection control program. Page 36 GAO/HRD-90-27VA’sInfection Control Programs i. wndix II ‘I+sic Elementsof an Infection Control The basic elements are divided into five groups: general elements,blood- stream infections, pneumonia, surgical wound infections, and urinary tract infections. The general group includes 14 elements that apply to all hospitals and addressprogram structure, applicable surveillance activi- ties, and control activities. The remaining four groups of elements are organized by the four major types of infection and address surveillance activities (identification, analysis, and reporting) specific to each type. The elements allow for both total and target surveillance. A program performing total surveillance monitors all four major types of infection in every hospital patient on either a periodic or continuous basis. A pro- gram that targets surveillance identifies the areas of highest infection risk or concern and focusesits attention on such areas; for example, patients in the intensive care unit or all bloodstream infections. The spe- cific elements applicable to an infection control program depend on whether the program usestotal or targeted surveillance: Total surveillance-all five groups of elements would be used, however, a few individual elements within the groups may not be applicable (see the basic elements). Targeted surveillance-the general elements would be used, and, depending on the type(s) of infections being targeted in the high risk groups, other appropriate groups of elements would be used (i.e., blood- stream, pneumonia, surgical wound, or urinary tract). Although the basic elements place emphasis on surveillance activities, an infection control program cannot be effective without control activi- ties. We included in our list only the broad categoriesof control func- tions becausecontrol activities, unlike surveillance activities, are performed not only by the infection control staff but by other hospital personnel. CDC has published detailed recommendations on procedures to be followed by providers to prevent infections. Further, control activ- ities that (1) are carried out by the program and (2) are beneficial to all hospitals are difficult to specify becausethe appropriate control activi- ties depend heavily on the surveillance findings and the circumstances within the individual hospital. The list of elements that follows is not all inclusive, the elements in the tables form a basic rather than an optimal program. Therefore, the ele- Y ments should be used in conjunction with other standards, such as the Joint Commission’s accreditation standards and m-c’s guidelines. Page 37 GAO/HRD-90-27VA’sInfection Control Programs AppendbK rr. I> Bsele BlentWe of an Infection Cmtrbl Progrtun I,, km s Tab1 11.1:CMnenl Element8 (Thew appl to all hospitala) structure Gl* The hospital has at least a part-time infection control practitioner, G2. t The hospital has a physician who supervises or consults in the infection control program and has taken at least one training course in hospital infection control. G3. The hospital has a multidisciplinary infection control committee. G4. Permanent membership on the committee includes representation from the following: *hospital administration, *microbiology laboratory (if one exists), *medical staff, and *nursing service. G5. The committee meets at least every 2 months. Surveillance Activities G6. The infection control program performs surveillance for at least one of the four major infection sites (bloodstream, pneumonia, surgical wound, and urinary tract). G7. The hospital’s infection control program has written standardized criteria (definitions) for nosocomial Infections at specific sites. G8. The infection control program has a system to detect and control outbreaks of infections. Control Activities G9. The hospital’s infection control program assists in developing and revising hospital departments’ policies aTiV.Q!RFcedures as they relate to infection control issues, GlO. The hospital’s infection control program assists in developing a system for reporting infections or infectioTMTf5Bsuresof employees. Gil. The hospital’s infection control pro ram assists in identifying and developing infection control topics Bor orMV3ITBnclasses, G12. The hospital’s infection control pro ram assists in identifying and develoorna infection control topics 9or in-~ trainino. G13. The hospital’s infection control program monitors or assists in monitoring the hospital staff’s compliance with specific patient care practices, such as aseptic techniques during intravenous catheter insertion and maintenance of insertion sites. (continued) Page 38 GAO/mm27 VA’sJnfection Control Progrm Appendix II Bar4icElement3 of an Infection Control Program Control Activities G14. Infection control practitioners and registered nurses on hospitals units have written authority to implement isolation procedures in an emergency without a physician’s order. Tal, lo 11.2:Bloodstream Infections Applicability to surveillance type ldentlflcation Total Target 91. X X A hospital’s infection control program uses at least one of the following case-findin approaches to identify bloodstream infections either in alPpatients or in a subset of patients: *Review results of blood cultures in all patients in target population. *Review all patients’ charts in target population. +Ieview all patients’ fever charts in target population. 82. X X During the surveillance period, a hospital’s infection control program performs an acceptable case-finding approach (previous criterion) on an average of every 3 daysa 83. X X In their case-confirmation effort, infection control staff perform at least one of the following activities if they do not review all patients’ charts in target population as a case-finding activity: *Review results of blood cultures in patients in target population, identified through case finding (if they do not review results of blood cultures in all patients in target population as a case-finding activity). *Review patients’ charts in target population, identified through case finding. Analyslsb 04. X X The infection control program has developed initial baseline rates for hospital-acquired bloodstream infections within the hospital. 85. X X Infection control staff analyze hospital-acquired bloodstream infection data by pathogen, B6. X Infection control staff analyze data on hospital-acquired bloodstream infections by whether or not patient had peripheral and/or central IV cannulation. 87. X X Infection control staff analyze hospital-acquired bloodstream infection data by ward. Y Page 39 GAO/IiRD-90-27VA’sInfection Control Programs APpbndix II, Basic Ehmenta of an Infection Control Program Appllcablllty to surveillance type Reporting Total Target 88. X X Infection control staff report summarized/analyzed data on hospital-acquired bloodstream infections to the infection control committee. 59. X Infection control staff report summarized/analyzed data on hospital-acquired bloodstream infections to the supervisor of the IV therapy team, if one exists. 610. X X Infection control staff report summarized/analyzed data to the ward supervisors or head nurses. Tablie 11.3:Pneumonia Applicability to surveillance type ldentlflcation Total Target Pi. X X A hospital’s infection control program uses at least one of the followrng case-finding approaches to identify pneumonia either in all patients or in a subset of patients: *Review all patients’ Kardexes in target population. *Ask nurses about signs or symptoms of a respiratory infection in all patients in target population. *Review all patients’ charts in target population. P2. X X During the surveillance period, a hospital’s infection control program performs an acceptable case-finding approach (previous criterion) on an average of every 3 daysa P3, X X In their case-confirmation effort, infection control staff perform at least one of the following activities if they do not review all patients’ charts in target population as a case-finding activity: *Review lab and X-ray results for evidence of pneumonia in patients in target population, identified through case finding (if they do not review lab and X-ray results in all patients in target population as a case-finding activity). *Review patients’ charts in target population, identified through case finding. Analyslsb P4. X The infection control program has developed initial baseline rates for hosoital-acauired pneumonia within the hospital. P5. ’ . X X Infection control staff analyze data on hospital-acquired pneumonia bv oathoaen. P6. X Infection control staff analyze data on hospital-acquired pneumonia by whether or not patient was on a ventilator, if target population includes ventilator patients, (continued) Page 40 GAO/HRD-90-27VA’sInfection Control Programs I I i Appendix II Basic Elementa of an Infection control Program Applicablllty to surveillance type Analysisb Total Target P7. X X Infection control staff analyze data on hospital-acquired pneumonia by ward. P8. X X Infection control staff analyze data on hospitaLacquired pneumonia by whether or not patient had surgery, if target population includes surgical patients. Reporting P9. X X Infection control staff report summarized/analyzed data on hospital-acquired pneumonia to the infection control committee. PlO. X If target population includes ventilator patients, infection control staff report summarized/analyzed data on hospital- acquired pneumonia to the respiratory therapy department, if one exists, Pll. X X Infection control staff report summarized/analyzed data on hospital-acquired pneumonia to the ward supervisors or head nurses. Table 11.4:Surgical Wound Infections Applicability to swelllance type ldentlficatlon Total Target Si. X X A hospital’s infection control program uses at least one of the following case-findin approaches to identify surgical wound infections either in alYsurgical patients or in a subset of surgical patients: *Review results of gram stains and cultures of wounds in all patients in target population, and ask nurses about signs or symptoms of surgical wound infections in all patients In target population, *Review all surgical patients’ Kardexes in target population. *Review all surgical patients’ charts in target population. s2. X X During the surveillance period, a hospital’s infection control program conducts case finding using an acceptable approach (previous criterion) on an average of every 3 days.” (continued) Page 41 GAO/IiRhftO~27 VA’s Infection Control Programs Appendix II Basic Elements of an Infection Control Program Applicablllty to surveillance type ldentlflcatlon Total Target s3. X X In their case-confirmation effort, infection control staff perform at least one of the following activities if they do not review all surgical patients’ charts in target population as a case-finding activity: *Review results of gram stains and wound cultures for patients in target population identified through case finding (if they do not review gram stains and wound cultures for all patients in target population as a case-finding activity). @Reviewsurgical patients’ charts in target population identified through case finding. *Ask nurses about signs or symptoms of surgical wound infections in patients in target population identified through case finding (if they do not ask nurses about signs or symptoms of surcical wound infections in all patients in target population-as a case-finding activity). ’ Analysisb s4. X X The infection control program has developed initial baseline rates for surgical wound Infections in the hospital. s5. X X Infection control staff analyze surgical wound infection data by surgeon. S6. X Infection control staff analyze surgical infection data by type of wound classification (e.g., clean clean-contaminated, contaminated, and dirty). s7. X X Infection control staff analyze surgical wound infection data bv oathoaen. S8. X X Infection control staff analyze surgical wound infection data bv ward. Reportina s9. X X Infection control staff report summarized/analyzed data on surcical wound infections to the infection control committee. SlO. X X Infection control staff report summarized/analyzed data on surgical wound infections to the surgical complications committee, if one exists. Sll. X X Infection control staff report summarized/analyzed data on surgical wound infections to the chief of the surgical service. s12. X X Infection control staff report summarized/analyzed data on surgical wound infections to the operating room supervisor. s13. X X Practicina suraeons receive suraeon-specific infection rates. Page 42 GAO/HRD-90-27VA’sInfection Control Programs &we* Il Basic Element8 oi an Inrection Control Program 11.5: Urinary Tract InfectIons Appllcablllty to surveillance type ldentltlcation Total Target Ul. X X A hospital’s infection control program uses at least one of the following case-fin&n approaches to identify urinary tract infections either in al$ patients or in a subset of patients: *Review all patients’ Kardexes in target population. *Review results of urine cultures in all patients in target population, and ask nurses about signs or symptoms of a urinary tract infection in all patients in target population. *Review all patients’ charts in target population. u2. X X During the surveillance period, a hospital’s infection control program performs an acceptable case-finding approach (previous criterion) on an average of every 3 days8 u3. X X In their case-confirmation effort, infection control staff perform at least one of the following activities if they do not review all patients’ charts in target population as a case-finding activity: *Review results of urine cultures in patients in target population, identified through case finding (if they do not review lab results of urine cultures in all patients in target population as a case- finding activity). *Review patients’ charts in target population, identified through case finding. Analyslsb u4. X X The infection control program has developed initial baseline rates for hospital-acquired urinary tract infections within the hospital. u5. X X Infection control staff analyze data on hospital-acquired urinary tract infections by pathogen. U6. X X Infection control staff analyze data on hospital-acquired urinary tract infections by ward. Repoftlng u7. X X Infection control staff report summarized/analyzed data on hospital-acquired urinary tract infections to the infection control committee. U8. X X Infection control staff report summarized/analyzed data on hospital-acquired urinary tract infections to the ward supervisors or head nurses. almportant infection control problems may require case finding more frequently than every 3 days bThe proper analysis of infection data requires calculation of infection rates in specific patient risk groups, as well as frequency distributions and line listings of the infections. If infection rates are to be useful for estimating infection risks in patient groups, appropriate data should be collected. For exam. ple, if bloodstream infections caused by intravenous catheters are being analyzed, then both the number of patients with intravenous catheters and the number of those patients who develop blood- stream infections are needed. Page 43 GAO/HRDBO-27VA’sInfection Control Programs / I Aphndix III IbQthodologyfor SamplingNonfederal Hospit& The sample of nonfederal hospitals for this study was selectedfrom the 1986 listing of American Hospital Association (MIA) memberships. Since our questionnaire on infection control was not relevant to very small hospitals, we excluded memberships with fewer than 60 beds before selecting the sample. Of the 4,411 memberships with 60 or more beds, we selecteda stratified sample of 660. To insure statistically reliable estimates for small, medium, and large hospitals, as well as for teaching and nonteaching hospitals, the sample was selectedfrom five subgroups (strata) of hospitals (seetable 111.1). TabId 111.1:Nonfederal Hospitals- Samble Sizes and Response Rates by Number of Strati 1905 AHA Initial Adjusted responses/ Bed rlze/afflllatlon memberships sample sample0 (response rate) 50-99 beds 1,350 110 110 84 (76%) loo-399 beds/nonteaching 2,133 110 111 92 (83%) loo-399 beds/teaching 376 110 110 95 (86%) 400+ bedsfnonteaching 141 90 102 93 (91%) 400-k beds/teaching 411 130 134 116 (87%) Total 4,411 550 567 480 (85%) aAdjusted sample reflects number of hospitals included in survey after adding extra hospitals identified as part of a group membership. Becausewe were aware that a small proportion of AWA memberships covered more than one hospital, we used the 1986 AHA guide to identify which of the sampled memberships may have represented multiple hos- pitals, We then made telephone calls to determine whether those mem- berships in fact represented more than one hospital and, if so, to obtain addressesfor each such hospital. As a result of this effort, we discov- ered in the sample 12 multiple memberships covering 29 hospitals. Con- sequently, an additional 17 hospitals were added to the sample, resulting in an adjusted sample of 667 hospitals. We obtained responsesfrom 86 percent (480 of 667) of the hospitals to which we mailed questionnaires. Among the sampled subgroups, the responserate ranged from 76 to 91 percent (seetable 111.1). While the initial sample of hospital memberships was stratified accord- ing to the AHA information on total number of beds for the membership, our results are presented according to the number of acute-carebeds reported by the hospital. We excluded 37 hospitals that reported having fewer than 60 acute-carebeds. Consequently, our results are basedon the 443 hospitals that reported having 60 or more acute-carebeds. Page 44 GAO/HRD-90-27VA’sInfection Control Programs Since data from 443 hospitals are used to make estimates about the uni- verse of nonfederal hospitals with 50 or more beds (estimated number is 3,872 hospitals), all data in this report on nonfederal hospitals are sub- ject to sampling error. The size of the sampling error reflects the preci- sion of the estimate; the smaller the sampling error, the more precise the estimate. Sampling errors for reported estimates about nonfederal hospitals are presented in table III.2 at the 95-percent confidence level. This means that the chancesare about 19 out of 20 that the actual number or per- centagebeing estimated falls within the range defined by our estimate plus or minus the sampling error. Table 111.2:Eatlmaks and Correapondlng San)pllng Error. for Nonfederal Horpltala Sam ling error ! !PC-percent Estimated number confidence level of applicable Percent that Element hosoltals comolv G2. Trained physician consultant (50 to 399 bed hospitals) 3,363 45 7 (400 or more bed hospitals) 509 54 6 G14. Practitioners and registered nurses have isolation authority (100 to 399 bed hospitals) 2,184 58 9 (400 or more bed hosdtals) ~I I 509 55 6 Bloodstream infections 82. Case finding every 3 days 3,640 83 5 EM Baseline rates 3,648 63 6 87. Analyze data by ward 3,648 82 5 B9. Report data to IV team 494 57 17 BIO. Renort data to ward SuDervisors 3,648 56 7 Pneumonia P2. Case finding every 3 days 3,546 60 7 P4. Baseline rates 3,546 68 6 P7, Analyze data by ward 3,546 83 5 (continued) Page 46 GAO/HRD-90-27VA’sInfection Control Programs Appendix III Methadology for Sampling Nonfederal H~pitala - Sam ling error 8 5-percent Estimated number confidence level of a plicable Percent that Element Ii ospitals comply Pneumonia PlO. Report data to respiratory therapv department 1.054 60 11 Pll.” ’ Report data to ward supervisors 3,546 60 7 Surgical wound Infections Qr) $se finding every 3 days 3,732 55 6 s4. Baseline rates 3,732 74 -5 S5. Analyze data by surgeon 3,732 80 5 SlO. Report data to surgical complications committee 1,975 62 8 s13. Practicing surgeons receive surgeon-specific data 3,732 20 5 Urinary tract infections u2. Case finding every 3 days 3,586 55 7 u4. Baseline rates 3,586 71 6 U6. Analyze data by ward 3,586 85 5 U80 Report data to ward supervisors 3,586 65 6 Page 46 GAO/IiRIHO-27 VA’sInfection Control Programs Miroed with a medical Number of acute- Medlcal center care beds Coatesville, PA YeS 352 Kansas City, MO Yes 315 Lebanon, PA Memphis, TN Ye!3 55u New Orleans, LA Yes 327 Portland, OR No 366 Washington, DC Yes 371 Page 47 GAO/HRD-90-27VA’sInfection Control Program I Apbendix V v Medical Centers’and Nonfederal Hospitals’ u”,seof GAO’sBasic Elements Nonfederal ho@pltals VA centers (estimated) Number Percent Number Percent Barlc element applicable using applIcablea using General elements: Gl. 158 100 3.872 100 Hospital has an infection control Dractitioner G2. ‘- 158 58 3.872 46 Trained physician consultant G3. 158 100 3.872 99 Multidisciplinary infection control committee G4. 158 88 3,872 91 Appropriate permanent committee membership G5. 158 98 3,872 94 Meets at least every 2 months G6. 158 99 3,872 98 Surveillance for at least one of the four major infection sites G7. 158 99 3,872 97 Written criteria for infections at specific sites G8. 158 99 3,872 97 System to detect and control outbreaks of infections G9. 158 100 3,872 100 Help to develop hospital infection control policies and procedures GlO: 158 99 3,872 97 Help to develp a system for ;~~~~~mWee Gil. 158 91 3.872 88 Help to develop infection control topics for orientation classes G12. 158 95 3,872 96 Help to develop infection control tooics for in-service training ’ G13. 158 88 3,872 85 Monitor compliance with sDecific patient care ljractices (continued) Page 48 GAO/HRD-90-27VA’sInfection Control Programe A~pendlx V VA Medicnl tinterr’ snd Nonfederal Hoopitab’ Use of GAO’sBasic Elementa VA center8 Number Percent Number Percent Basic element applicable using applIcablea using G14. 158 78 3,872 63 Practitioners and registered nurses have isolation authority Bloodstream Infections: 61. 152 100 3,648 99 Appropriate case-finding aooroach 82. 152 83 3,648 83 Case finding every 3 days 83. 152 99 3,648 100 Appropriate case- confirmation approach 84. 152 73 3,648 63 Baseline rates 85. 152 99 3,648 97 Analvze data bv oathoaen 66. 80 90 1,177 90 Analyze data by whether patient had IV B7. 152 97 3,648 82 Analyze data by ward 68. 152 99 3,648 99 Report data to infection control committee B9. 9 22 494 57 Report data to IV team BlO. 152 63 3,648 56 Report data to ward supervisors Pneumonia: Pl. 141 88 3,546 93 Appropriate case-finding aporoach P2. 141 56 3,546 60 Case finding every 3 days P3. 141 98 3,546 100 Appropriate case- confirmation approach P4. 141 69 3,546 68 Baseline rates P5. 141 98 3,546 97 Analyze data by pathogen P6. 65 86 976 90 Analyze data by whether or not patient was on a ventilator (continued) Page 49 GAO/HRD99-27VA’sInfection Control Programs - --7 Appendix V VA Medical 6entere’ and Nonfederal Hoepitab’ Ueeof GAO%Basic Elements - Nonfederal hosdtala VA center8 (ertlmated~ Number Percent Number Percent Baric element applicable using applicable0 urlng P7. 141 96 3,546 83 Analyze data by ward P8. 122 88 3.485 93 Analyze data by whether or not patient had surgery P9. 141 99 3,546 99 Report data to infection control committee PlO. 65 46 1.054 59 Report data to respiratory therapy department Pll. 141 62 3,546 60 Report data to ward supervisors Burglcal would infections: Sl. 131 95 3,732 93 Appropriate case-finding approach %se finding every 3 days 131 58 3,732 55 s3. 131 99 3,732 99 Appropriate case confirmation approach s4. 131 88 3,732 74 Baseline rates s5. 131 71 3,732 80 Analyze infection data by surgeon S6. 71 93 1,262 91 po;lyy data by type of s7. 131 95 3,732 96 Analyze data by pathogen S8. 131 89 3,732 80 Analyze data by ward s9. 131 100 3,732 99 Report data to infection control committee SlO. 92 58 1,975 62 Report data to surgical complications committee Sll. 131 97 3,732 80 Report data to chief of surgery s12. 131 90 3,732 91 P Report data to operating room supervisor (continued) Page 50 GAO/HRD-90-27VA’sInfection Control Programs _ .- Appendix V VA Medical Centers’ and Nonfixleral Hospital& Use of GAO’sBasic Elements VA center8 Number Percent Number Percent Baric element applicable uring applicablea uring s13. 131 16 3,732 20 Practicing surgeons receive surgeon-specific infection rates Urinary tract infections: Ul. 141 83 3,586 90 Appropriate case-finding approach u2. 141 53 3,586 55 Case finding every 3 days u3. 141 97 3,586 98 Appropriate case- confirmation approach u4. 141 73 3,586 71 Baseline rates u5. 141 99 3,586 98 Analyze data by pathogen U6. 141 99 3,586 85 Analyze data by ward u7. 141 99 3,586 99 Report data to infection control committee U8. 141 67 3,586 65 Report data to ward supervisors aThe number of nonfederal hospitals applicable represents an estimate for the universe of all nonfederal hospitals, based upon responses to our questionnaire. Sampling errors must be taken into account when drawing conclusions from the percentages (see app. 111). In addition, because VA has a higher percentage of large medical centers than the nonfederal sector, caution should be used when compar- ing the estimated percentage of all nonfederal hospitals to VA centers using the same elements. Page 51 GAO/IiRIHO-27 VA’sInf’ection Control Programs i 1 Apbendix VI Qrnrnents From the Department of V/eterans I Affairs THE SECRETARY OF VETERANS AFFAIRS WASHINGTON Mr. Lawrence H. Thompson Assistant Comptroller General Human Resources Division U. S. General Accounting Office Waehington, DC 20548 Dear Mr. Thompson, am pleased to provide the enclosed detailed comments on your I draft report aON C-L. . VA Proarwre -able to al prow But Can Be Enhanced (GAO/HRD-89-1461, dated October 6, 1989. The Department is concurring with each of the four recommendations, and corrective action is in process to improve our infection control program. We appreciate your assistance in this matter. . Bnclosure Page52 lrppencUrrM Csnuneatsl%omthsDepartmentof Veteran6 Affdrm Enclosure DEPARTWENT OF VETERANSAFFAIRS COMMENTS ON THE OCTOBER 6, 1989, GAO REPORT aON . VA PROCRAMs CONTROL. BUT CAN BE m %0 Q 27 GAO reOOPrPaOnd8that I require the Chief Hodioal Dirootor to update VA infeotion control guidanoo. At a minimum, the guidanoo 8hould reguiro oomponants airilar to tho8e in GAOga bamio l lemonts. We concur w$th the recommendation. Newly published infection control guidance incorporates GAO’s basic elements. This guidance will soon be incorporated into a VA manual chapter. QAO al80 rooonmond8 that I direot the Chief Yedioal Direotor to: -- De8ignate a 8inglo unit in VA*8 oentral offioe to Over8ee See p itr infeotion oontrol progr8m8, inoluding (1) ooordinatingthe oentral offioel8 polioie8 an4 prooedures and (2) implementing a meohanism for 8haring information among praotitioners. We concur with the recommendation. The Veterans Health Services and Research Administration Director, Medical Service is designated as the central focus for overseeing VA's infection control program. He will work in concert with the Infectious Disease Field Advisory Group in fomulating policies and procedures for the Departments's infection control program. See p. 35. a- Inaorporate proaedure8 in regional offiaa8 survey requirement8 to asnure that l aoh medical aenter's infeotion oontrol program ir adequately reviewed. We concur with the recommendation. The regions will assure that each medical center's infection control program is adequately reviewed and will work with the infection control program officials in VA Central Office to develop appropriate review criteria. See p. 35. am Require aenter direotor8 to reeramine the level of managerial nupport and re8ouroe8 given to their infection control program8 and to provide additional WAgpOrt where appropriate. We concur with the recommendation. The Chief Medical Director will require his medical center directors to reexamine resources allocated to their infection control programs. Additional support for these programs will be provided based on justified need and systemwide priorities. Pagelt Ap&ndix VII , Major Contributors to This Report P James A. Carlan, Assistant Director HqknanResources Michelle L. Roman, Assignment Manager Dibision, Mary Ann Curran, Evaluator Wdshin&on, DC. I Randall B. Williamson, Regional ManagementRepresentative tle Regional Office Walter R. Eichner, Evaluator-in-Charge / Lori D. Pang, Evaluator I SusanK. Hoffman, Evaluator : optional Security and Michael J. Morgan, Evaluator 1 International Affairs Dikision (101321) Page 64 GAO/IiRD-90-27 VA’s Infection Control Programs - _ -- .., - “, ,, ,,.,- ~. r= r- *. - -_ ; f z. IT -- _ -, -1 --- -,A. 2. r, - % =i,- ., - c 2 ; = f A I --=; -- I *A_-z -. ii = qr-
Infection Control: VA Programs Are Comparable to Nonfederal Programs but Can Be Enhanced
Published by the Government Accountability Office on 1990-01-31.
Below is a raw (and likely hideous) rendition of the original report. (PDF)