6
Implementation and Effects of CDC Guidelines

Neither voluntary guidelines nor government regulations implement themselves. Given factors such as competition within and among organizations for scarce resources and disagreements about problems and priorities, implementation cannot be assumed. The literature reviewed in Chapter 5 has already pointed to departures from recommended tuberculosis control measures as likely contributors to outbreaks of tuberculosis in health care facilities in the late 1980s and early 1990s.

This chapter reviews what is known about the implementation and effects of the 1994 Center for Disease Control and Prevention (CDC) guidelines to prevent worker exposure to tuberculosis in hospitals, correctional facilities, and other work settings. The review also covers information about the implementation of earlier tuberculosis control guidelines including those that were recommended by CDC in 1990. Although some specific recommendations have changed and technologies have been evolving, the basic elements have remained constant enough that studies that predate the 1994 guideines are still useful. In fact, most of the published reports located by the committee describe steps taken before publication of the 1994 guidelines.

In addition to the CDC’s own efforts, recommendations and actions by other public and private agencies may also have influenced employer decisions about tuberculosis control measures in the 1990s. For example, in 1993 the Occupational Safety and Health Administration (OSHA) announced efforts to enforce the adoption of tuberculosis control measures under the agency’s general-duty clause and its 1987 respiratory protection standard. The next year it issued a notice of proposed rulemaking on



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 108
Tuberculosis in the Workplace 6 Implementation and Effects of CDC Guidelines Neither voluntary guidelines nor government regulations implement themselves. Given factors such as competition within and among organizations for scarce resources and disagreements about problems and priorities, implementation cannot be assumed. The literature reviewed in Chapter 5 has already pointed to departures from recommended tuberculosis control measures as likely contributors to outbreaks of tuberculosis in health care facilities in the late 1980s and early 1990s. This chapter reviews what is known about the implementation and effects of the 1994 Center for Disease Control and Prevention (CDC) guidelines to prevent worker exposure to tuberculosis in hospitals, correctional facilities, and other work settings. The review also covers information about the implementation of earlier tuberculosis control guidelines including those that were recommended by CDC in 1990. Although some specific recommendations have changed and technologies have been evolving, the basic elements have remained constant enough that studies that predate the 1994 guideines are still useful. In fact, most of the published reports located by the committee describe steps taken before publication of the 1994 guidelines. In addition to the CDC’s own efforts, recommendations and actions by other public and private agencies may also have influenced employer decisions about tuberculosis control measures in the 1990s. For example, in 1993 the Occupational Safety and Health Administration (OSHA) announced efforts to enforce the adoption of tuberculosis control measures under the agency’s general-duty clause and its 1987 respiratory protection standard. The next year it issued a notice of proposed rulemaking on

OCR for page 108
Tuberculosis in the Workplace occupational tuberculosis. In addition, as noted in Chapter 4, some state regulatory agencies and some accrediting organizations included tuberculosis control measures in their regulations or standards for health care and correctional facilities. The committee identified three general types of information on the implementation and effects of CDC guidelines: multi-institution surveys, multi-institution inspections, and reports on individual organizations. A few studies focus on the adherence of individuals (e.g., physicians and nurses) to recommended practices such as using personal respiratory protection devices. The literature review in Appendix D includes additional details, and Appendixes B and F also provide relevant information on two specific control measures: tuberculin skin testing and personal respiratory protections. The primary outcome measures reported are tuberculin skin test conversions and cases of active tuberculosis (including multidrugresistant disease). Again, nearly all the information that the committee located relates to hospitals. The committee found little on nursing homes, ambulatory care clinics, health units of correctional facilities, and other organizations covered by the 1994 CDC guidelines for health care facilities or by the 1996 CDC guidelines for correctional facilities.1 IMPLEMENTATION OF TUBERCULOSIS CONTROL GUIDELINES Broadly, implementation refers to the practical activities and interventions undertaken to turn guidelines or policies into desired results. Implementation of the tuberculosis controls measures recommended by CDC calls for a complex set of actions at both the organizational and the individual levels. The primary focus of the tuberculosis control measures is the organization as a whole rather than the individual. As described in Chapter 4, institutional responsibilities include the preparation and implementation of an overall tuberculosis control plan and record-keeping system; assessment of the tuberculosis risk in the facility; the development and application of written policies and protocols for the rapid identification, isolation, and treatment of individuals with infectious tuberculosis; the creation and maintenance of surveillance, education, and other programs for workers; the establishment and maintenance of appropriate engineering controls for negative-pressure isolation rooms and other areas; and the creation and monitoring of a respiratory protection program. 1   The CDC recommendations for preventing tuberculosis in correctional facilities describe core activities of screening, containment, and assessment but are generally similar to the guidelines for health care facilities, taking into account differences in the purposes and operation of correctional facilities and health care facilities.

OCR for page 108
Tuberculosis in the Workplace Policies are, of course, carried out by individuals. The personnel who provide clinical care, maintain engineering systems, and otherwise do the day-to-day work of the organization often have considerable discretion in following policies and recommended practices. In recent years, many health care and other organizations have attempted to design systems to minimize opportunities for unwanted variations in work practices. For example, some facilities have installed electronic monitoring systems that check whether doors to tuberculosis isolation rooms are closed, consistent with policy. Nonetheless, for many activities, universal standardization or monitoring of work practices would be viewed as impossible, offensive, counterproductive, or excessively expensive. Thus, individual adherence to organizational policies continues to be a concern. Surveys of Organizational Implementation of Tuberculosis Control Measures Mailed surveys are a relatively inexpensive way of collecting information about a large number of geographically dispersed institutions. When the surveyed institutions are familiar with both the surveying organization and the kinds of questions asked and when the topic is viewed as important, voluntary questionnaires can generate respectable response rates of 70 percent or more. The potential limitations of survey data are, however, familiar. If all members of a population are not surveyed and the sample of the population is not properly selected, the subset chosen may be unrepresentative of the population. This limits generalizations from the surveyed population to the larger population. Whether surveys are directed to a universe or a representative sample, the lower the response rate, the greater the concern that responses will be unrepresentative. In addition, survey questions may be deliberately or unintentionally biased or otherwise formulated in ways likely to produce inaccurate and unrepresentative responses. Even if the questions are sound, those who respond may intentionally or unintentionally provide inaccurate or insufficient information. Bias is a particular concern if those surveyed know that important policy decisions may hinge on the survey results. Organizational surveys may also be misdirected to and returned by individuals who lack the knowledge to respond accurately. Results from National Surveys of Hospitals In 1992, while increases in tuberculosis cases and case rates were still being recorded, CDC surveyed hospitals about their tuberculosis control practices. Questionnaires went to all 632 federal, state, and local public hospitals in the United States and to a 20 percent random sample

OCR for page 108
Tuberculosis in the Workplace (444 institutions) of hospitals with 100 beds or more that were listed in the American Hospital Association database (Manangan et al., 1998). The response rate was about 70 percent for each group (726 institutions total). (Note that Table 6-1, described in the next paragraph, covers only a subset of this survey’s respondents.) Half of the respondents reported that their institutions had admitted six or more patients with tuberculosis. One-quarter said that their institutions had admitted patients with multidrug-resistant disease. Nearly all reported some kind of tuberculin skin testing program. Just over 70 percent reported having isolation rooms that met the 1990 CDC criteria for the isolation of patients with tuberculosis, but 60 percent of that group reported that their institutions did not routinely check the airflow in isolation rooms. Of those that did routinely check, few (13 percent) checked it at least monthly. One of five institutions allowed patients out of isolation for other than medical reasons. Although nearly 90 percent reported that their policy was to keep TABLE 6-1. Comparison of Tuberculosis Control Measures for 103 Hospitals That Reported More than Six Admissions of Patients with Tuberculosis in 1992 CDC Survey and That Also Responded to 1996 CDC Survey   1992 No. (%) 1996 No. (%) Engineering Controls Isolation rooms meeting CDC criteria 59/92 (64) 99/103 (96) Routine check of negative air pressure 42/85 (49) 96/99 (97) Monthly check of negative air pressurea 5/35 (14) 76/90 (84) Respiratory Protectionb Nonfitted surgical mask 69/101 (68) 1/103 (1) Soft mask, molded or fitted 34/101 (34) NA Particulate respirator 8/101 (8) 40/103 (39) N95 NA 85/103 (83) Tuberculin Skin Testing Program Testing by Worker Category Nurses 103/103 (100) 103/103 (100) Respiratory therapists 102/103 (99) 103/103 (100) House staff 65/81 (69) 65/73 (89) Attending physicians 43/86 (50) 65/94 (69) Students 55/95 (58) 74/97 (76) Testing Elements After exposure incident 98/101 (97) 102/103 (99) Two-step testing NA 77/98 (79) Maintain yearly reports 64/98 (65) 93/98 (95) aWhen an isolation room is actually in use for a patient with suspected or confirmed tuberculosis, the 1994 CDC guidelines recommend that pressure be checked daily. bNumbers add to more than one hundred because facilities may use more than one type of mask. SOURCE: Manangan et al., 1998.

OCR for page 108
Tuberculosis in the Workplace the doors of isolation rooms closed, a near majority reported that staff left doors open some or all of the time. Few (11 percent) reported negative-pressure isolation facilities for their emergency departments that met CDC recommendations. In 1996, nearly 2 years after the 1994 CDC guidelines were released, CDC randomly selected and resurveyed half of the 272 hospitals that had reported six or more admissions of patients with tuberculosis in the 1992 survey (Manangan et al., 1998, 2000). Responses were received from 75 percent (103) of these facilities. Table 6-1 compares the 1992 and 1996 responses for these hospitals. The 1996 responses showed substantial improvement in all areas in which implementation of control measures had not already been near or at 100 percent. CDC initiated a new survey of hospital tuberculosis control practices in 2000 (Pugliese, 2000). Final results were not yet available when the committee finished its work. In 1993, the Society for Healthcare Epidemiology of America (SHEA) and the CDC surveyed members of SHEA to assess compliance with the 1990 CDC guidelines (Fridkin et al., 1995a,b). They obtained responses that were suitable for analysis for 210 hospitals, but not all respondents returned complete information. The researchers concluded that despite an increase in patients with multidrug-resistant tuberculosis, “TB infection control measures still did not meet the 1990 CDC guideline recommendations” (Fridkin et al., 1995b, p. 129). Another analysis focused on tuberculosis control measures in emergency departments in a randomly selected sample of the hospitals responding to the 1992 CDC survey described above (Moran et al., 1995). Of the institutions responding (305, or 68 percent of 446 facilities contacted), more than half (53 percent) reported seeing tuberculosis patients at least monthly. More than 90 percent reported giving surgical masks to patients with suspected tuberculosis. Although 76 percent reported written patient isolation criteria for the emergency department, only 56 percent had such criteria for triage or waiting areas. (Some institutions may not have had separate triage areas.) Respondents reported tuberculosis isolation rooms in only 20 percent of emergency departments and 2 percent of triage and waiting areas. Air recirculation measures were reported for approximately 80 percent of emergency departments. In 1992 and 1995 surveys of hospitals with 100 or more beds, Tokars and colleagues (1996) focused on the CDC recommendations for mycobacteriology laboratory methods and on rapid laboratory processing of smears and cultures. The 1992 survey of 1,076 institutions obtained a 70 percent response rate. In 1995, 20 percent of those responding to the earlier survey were surveyed again and 70 percent responded. Those responding to both surveys reported increased use of recommended testing procedures in 1995. They also reported drops in the median time for

OCR for page 108
Tuberculosis in the Workplace providing results from 2 days to 1 day for smear results, from 40 to 21 days for culture results, and from 45 to 35 days for drug sensitivity test results. Some of these improvements probably reflect the availability of better technologies. Appendix D reports limited information from some dental schools. It suggests limited implementation of tuberculin skin testing and, probably, other protective measures. Other Surveys of Health Care Facilities As background for continuing efforts to develop tuberculosis control policies, researchers in Minnesota surveyed a voluntary sample of 18 hospitals to determine hospital practices and analyze tuberculin skin test results for the period from 1989 to 1991. Although the survey documented variable compliance with recommended practices of the time, the researchers concluded that practices were “reasonably consistent with the critical elements in the 1990 CDC guidelines” (Van Drunen et al., 1996). The Maryland Hospital Association and the state of Maryland surveyed the state’s hospitals in 1992, 1993, and 1997 to assess tuberculosis control practices (Fuss et al., 2000). The 1992 survey, which obtained responses from nearly three-quarters of the hospitals, found that about half reported having a routine (at least annual) tuberculin skin testing program for employees. About half also reported that they routinely checked isolation rooms for negative pressure. Subsequent site visits found deficiencies in isolation room performance. Less than a quarter of the hospitals reported that they supplied workers with respirators. The 1997 Maryland survey also obtained responses from about three-quarters of the hospitals surveyed. This time 90 percent of the hospitals reported that they checked isolation rooms for negative pressure. All reported having a routine employee skin testing program and providing workers with respirators consistent with CDC recommendations. More than 90 percent had conducted a risk assessment consistent with CDC guidelines. Manangan and colleagues reported on 1992 and 1996 survey results for New Jersey hospitals (1999). Again, the reports showed improved compliance with tuberculosis precautions. The committee also located a 1992 convenience survey of Texas hospitals that concluded that many hospitals had policies and practices that were inconsistent with the CDC guidelines in place at the time (Manangan et al., 1997). Surveys of Correctional Facilities In 1992–1993, 1994–1995, and 1996–1997, CDC and the National Institute of Justice (NIJ) sent surveys on tuberculosis control practices to the

OCR for page 108
Tuberculosis in the Workplace Federal Bureau of Prisons, all 50 state systems, and a number of large local jail systems (37 for the first two surveys, 41 for the third survey) (NIJ, 1996, 1999). All of the federal and state systems responded to the surveys, as did approximately 80 percent of the local jail systems. The first two surveys predated the official publication of CDC guidelines for correctional facilities (CDC, 1996b), although earlier agency and other guidelines covered high-risk populations and tuberculosis control measures applicable to facilities treating or housing people with tuberculosis (ATS, 1992; NCCHC, 1992; CDC, 1990b, 1994b). The survey questions for correctional facilities differed somewhat from those for hospitals, so comparisons are not always possible. For example, the first two surveys apparently did not include questions about screening of correctional facility staff. Responses to the third survey indicated that more than 90 percent of federal and state systems and almost all local jail systems reported screening of new employees. Roughly three-quarters reported periodic retesting. For each survey, reported use of negative-pressure rooms (in infirmaries or community hospitals, or both) for the isolation of inmates with suspected or confirmed infectious tuberculosis increased: from approximately 30 percent (1992–1993) to approximately 65 percent (1994–1995) to nearly all (98 percent) of the federal and state systems and 85 percent of the local jail systems (1996–1997). The reported use of directly observed therapy for all inmates with active tuberculosis also increased from 77 to 94 to 98 percent for federal and state systems and from 84 to 90 to 95 percent for local jail systems, for the three surveys, respectively. The 1996–1997 survey included validation surveys of institutions within 13 systems. These surveys showed some differences between system-level and institution-level policies. For example, for systems with policies requiring four-drug initial treatment of active tuberculosis, only three-quarters of the individual institutions in those systems reported having the same policy. In the 1996–1997 survey, nearly one-third of the federal and state systems failed to report whether or not they had cases of tuberculosis. Reporting on tuberculin skin testing programs was even more incomplete, with more than half of the state and federal prison systems and more than a third of the jail systems failing to report conversion data. The authors suggest that cases of tuberculosis in prisons may be undercounted because reporting is incomplete. A separate survey of staff in Texas correctional facilities reported lack of knowledge of how tuberculosis is transmitted and how it can be prevented and treated (Woods et al., 1997). A survey of 225 health care workers in the Maryland Department of Corrections noted similar gaps in tuberculosis-related knowledge among frontline correctional health care workers (DeJoy et al., 1995). For example, 30 percent of the workers thought that a

OCR for page 108
Tuberculosis in the Workplace standard surgical mask would protect them from inhalation of aerosolized tuberculosis droplets, and 23 percent thought that correctional health care workers were not at risk for infection with multidrug-resistant tuberculosis. The researchers found that training on tuberculosis risk was inconsistent. Some prison facilities provided extensive training, whereas others provided almost no training. A third of all respondents said that they had received no workplace training at all on tuberculosis in the previous year. Eleven percent of the workers reported that they had not been offered tuberculin skin testing in the previous 12 months, which is in conflict with stated institutional policy. Almost a fifth of respondents reported that they had a positive skin test history, and roughly half of this group said that they had received some type of follow-up care. A report on the Cook County Jail in Chicago underscores the logistical challenges of implementing tuberculosis control measures to protect jail inmates and staff in a large facility (McAuley, 2000). This jail admits more than 100,000 people a year and houses about 10,000 per day on average, more than the facility was designed to handle. All those detained have a medical evaluation that includes a tuberculin skin test (read within 48 to 72 hours) and a chest radiograph (read within 18 hours). Persons with suspected tuberculosis identified during or after the evaluation are sent to the jail’s emergency room, which has negative-flow isolation rooms. Those who have a suspicious radiograph but are released before it is read are to be seen by a communicable disease investigator and brought to the tuberculosis clinic of the county health department. (An analysis of the experience with this system’s screening strategy is reported later in this chapter.) An article by Jones and colleagues (1999) about their experience at the Memphis city jail is also illuminating. From January 1995 through December 1998, the Memphis city jail admitted and discharged more than 173,000 individuals, an average of 159 a day. The median length of stay was 1 day, and four-fifths of those admitted had been incarcerated in the jail previously. Single cells held between 18 and 36 inmates, and mingling of inmates was extensive. Other Surveys In 1997, the American Federation of State, County, and Municipal Employees (AFSCME) developed separate but similar surveys to collect information about employer compliance with tuberculosis control recommendations affecting health care workers, law enforcement personnel, and social services workers (August, 1999).2 They received responses for 2   The survey went to approximately 100 district councils (which distributed them to local unions) and large unaffiliated local unions. Of the 170 responses, some came from employers, but most (145) came from workers. Reporting on skin test conversions and cases of tuberculosis was incomplete, but cases of disease were reported in all sectors.

OCR for page 108
Tuberculosis in the Workplace 170 workplaces including 94 health care facilities, 48 correctional and law enforcement facilities, and 28 social service agencies. The results reported for the 16 responding acute-care hospitals were, overall, the most consistent with the CDC guidelines. Just over half of the correctional and law enforcement facilities were reported to have a written tuberculosis control plan. Of the social service agencies (which were not covered by the 1994 CDC guidelines for health care facilities), only one respondent reported a written tuberculosis control plan or a worker training program. For all organizations, the lowest levels of practice consistent with the 1994 CDC guidelines were reported for respiratory protection programs. Half of the 16 hospitals, less than 10 percent of the 23 long-term facilities for the elderly, 20 percent of the 28 mental health facilities, and 20 percent of 48 correctional and law enforcement facilities reported such programs. The responses to the AFSCME survey come from a very small, nonrandom set of respondents and must be viewed with considerable caution. They do, however, help explain organized labor’s continuing concern about the protections being offered workers, particularly those outside hospitals. In 1997, researchers from Johns Hopkins University asked attendees at a national funeral director’s convention to complete a risk assessment questionnaire and undergo tuberculin skin testing (Gershon, 1998). Approximately 800 funeral home employees completed the survey and consented to a tuberculin skin test. This group included 500 embalmers, who have the highest risk of exposure. Only 16 percent of the embalmers reported consistently wearing any kind of face mask during embalming procedures. About half reported some kind of training about tuberculosis during their career; less than 20 percent reported such training in the preceding 12 months. Nonetheless, the researchers concluded that most were reasonably knowledgeable about the disease. Overall, these data supported findings from a smaller pilot study of 123 Maryland embalmers that also showed limited adoption of measures for the prevention of transmission of M. tuberculosis (Gershon et al., 1995b). Taken together, survey results suggest, at a minimum, two conclusions. First, institutional departures from recommended tuberculosis control policies and procedures were common, if not the norm, in the late 1980s and early 1990s. Second, institutions—at least hospitals and correctional facilities—were taking tuberculosis control measures more seriously and reporting substantially higher rates of implementation of recommended measures in later years. As discussed below, written policies may not be consistent with routine practices. Facility Inspections or Visits Although limited in some respects by the lack of a specific standard on occupational tuberculosis, OSHA can inspect health care and other

OCR for page 108
Tuberculosis in the Workplace facilities under its general duty (which provides that employers maintain a safe workplace) and its respiratory protection standard.3 In addition, many state governments and private agencies periodically inspect health care and other facilities to determine compliance with regulations or voluntary standards. As described in Chapter 4, some of these regulations and standards include provisions related to tuberculosis control measures, although the committee found no overall summary of state regulatory requirements or accreditation requirements. Committee members were aware of institutions that had been cited or questioned by state agencies or accrediting organizations about tuberculosis control measures during visits by the state licensure agencies and the Joint Commission on the Accreditation of Healthcare Organizations. In addition to such routine inspections, inspections may also be prompted by complaints by patients, families, health care workers, or others. The facilities involved in these kinds of complaint-generated inspections may not be representative. In addition, a few on-site inspections were specifically prompted by state concern about facility readiness to cope with the increasing rates of tuberculosis seen in the late 1980s and early 1990s. Although inspectors often rely on responses to written questions and written records, they may have the opportunity to conduct more flexible, open-ended interviews with facility personnel and to view or test the physical plant, equipment, and work practices. Such inspections are laborintensive and expensive, which limits their number and scope. OSHA Inspections Between May 1992 and October 1994, OSHA inspected 272 health care, correctional, and other facilities to assess compliance with the tuberculosis control measures that were described first in a May 1992 OSHA Region 2 directive and then in a nationwide enforcement policy (McDiarmid et al., 1996). Inspections in New York and New Jersey accounted for a substantial proportion of the total. Worker or union complaints prompted most inspections (71 percent). Hospitals accounted for almost half of the workplaces inspected. Basic citation data were available for nearly all the facilities, but detailed questionnaire data were available for only 149 facilities. Inspectors found compliance with recommended tuberculosis control measures to be quite variable. It was best, overall, for administrative controls. For example, annual tuberculin skin testing was reported for better than three-quarters of hospitals, prisons, shelters, and nursing homes. 3   As described in Chapter 4, OSHA revised its 1987 respiratory protection standard in 1998. Pending publication of the standard on occupational tuberculosis, the 1998 general standard did not cover tuberculosis, which instead was covered by special interim regulations.

OCR for page 108
Tuberculosis in the Workplace Negative-pressure isolation rooms were reported for 33 of 56 hospitals (59 percent), 15 of 35 prisons (43 percent), and 4 of 9 nursing homes (44 percent) but no shelters. Respiratory protection was reported for nearly 60 percent of hospitals but less than 20 percent of prisons, shelters, and nursing homes. Overall, 42 percent of facilities received citations, most for noncompliance with respiratory protection requirements. (The inspections occurred before the National Institute for Occupational Safety and Health [NIOSH] had certified the N95 respirators, which were less expensive and generally more convenient and comfortable than the devices previously certified.) Again, because inspections were generally prompted by complaints, the results may reflect a negative bias. State and Other Inspections New York State officials examined tuberculosis isolation procedures in 22 New York City hospitals in 1992, 1993, and 1994 (Stricof et al., 1998). They reviewed medical and laboratory records to collect information about patient risk factors and history, signs and symptoms, length of time in the emergency department, turnaround time for laboratory reports, timing of isolation and treatment, and other information. They also directly observed and evaluated isolation rooms and isolation practices. From 1992 to 1994, they found that hospitals made substantial progress in correcting deficits in tuberculosis control measures. The percentage of isolation rooms with negative pressure increased from 51 to 80 percent. The number of patients with active tuberculosis sharing rooms dropped from 13 percent to zero, and the percentage of patients with suspected or diagnosed tuberculosis isolated upon admission increased from 75 to 84 percent. The number of facilities able to process smears 7 days a week increased from 40 to 95 percent. Despite improvements, the inspections also revealed continuing problems in some areas, including open doors and windows for isolation rooms and isolation rooms without negative pressure. In addition to any state-specific requirements, states must survey and inspect nursing homes annually to assess compliance with Medicaid certification requirements set by the U.S. Health Care Financing Administration (HFCA). HCFA requires that nursing homes have an infection control program. Recent data (June 2000) showed that states had cited 10.8 percent of facilities for deficiencies in their infection control programs, 0.9 percent for deficiencies related to isolation of residents, 0.1 percent for deficiencies related to employees with (any) communicable disease, and 6.4 percent for hand-washing and infection control deficiencies (AHCA, 2000). One on-site study (supported by NIOSH and the California Department of Health) compared written tuberculosis control policies with actual

OCR for page 108
Tuberculosis in the Workplace tuberculosis control measures (Blumberg, 1999). A multivariate analysis found no association between tuberculin skin test conversions and patient contact (frequent contact versus no contact). The analysis did, however, show an association between skin conversions and bacille Calmette-Guérin (BCG) vaccination, lower salary levels, and shorter time of employment. This suggests that community exposure was likely important for lower-salary workers, who probably come from parts of Atlanta with high rates of active tuberculosis. Jackson Memorial Hospital (Miami) From 1988 to 1990, Jackson Memorial Hospital in Miami experienced an outbreak of multidrug-resistant tuberculosis related to patient-to-patient transmission on an HIV ward (Beck-Sague et al., 1992; Fischl et al., 1992; Wenger et al., 1995). After reviewing their infection control policies and work practices, hospital managers implemented a series of more stringent tuberculosis control measures (Wenger et al., 1995). The first control measures, which were implemented in March 1990, included a four-drug initial treatment regimen and more rigorous isolation policies on the ward (i.e., stricter isolation criteria for HIV infected patients, stricter criteria for discontinuing isolation; stricter enforcement of policies that infectious patients stay in their rooms unless medically necessary and wear surgical mask when out of their rooms, and restriction of sputum induction procedures to isolation rooms). In April 1990, the hospital repaired improperly functioning isolation rooms and improved the ventilation in other rooms. In June 1990, the hospital instituted a policy that aerosolized pentamidine would be administered only in isolation rooms. In the following months the hospital switched respiratory protections for health care workers from a surgical mask to a submicron mask (September 1990), established and staffed a separate unit for tuberculosis control (October 1990), added laboratory staff to improve turnaround times for specimen results (December 1990), required isolation for all patients with multidrug-resistant tuberculosis (February 1991), and switched to dust-mist respirators (April 1992). NIOSH checked the ventilation in the isolation rooms, and hospital staff checked negative pressure daily. Implementation of other practices (e.g., keeping doors to isolation rooms closed, and wearing of respirators) was checked by observation. The effects of these changes were monitored for three time periods: January through May 1990 (which overlaps the first interventions), June 1990 through February 1991 (which overlaps most of the remaining interventions), and March 1991 through June 1992. The investigators found that all patients with multidrug-resistant tuberculosis who were admitted during the first monitoring period had been exposed to other such patients while on the HIV ward. In contrast, none of the patients with multidrug-resistant disease admitted during the subsequent monitoring

OCR for page 108
Tuberculosis in the Workplace periods had infection traceable to their stay on the ward during those periods. When workers on the HIV ward were compared to workers on a control ward that did not admit HIV infected patients, the former group had significantly higher skin test conversions rates during the outbreak period from January 1988 through January 1990 (Beck-Sague et al., 1992). Rates for the two groups of health care workers did not differ significantly for the period from June 1990 through June 1992 (Wenger et al., 1995). The study authors note that although the “density” of patients with multidrug-resistant disease declined after the initial monitoring period, “infectious patients were still present and the potential for transmission still existed” (p. 239). Indeed, two of the three skin test conversions in the follow-up period occurred in workers who were exposed to a patient who had previously been diagnosed with tuberculosis but who was thought to be no longer infectious. This led to a requirement that all patients with multidrug-resistant tuberculosis be isolated upon admission. Following the implementation of this policy, no further tuberculin skin test conversions were reported among health care workers. The study authors concluded that most of the effect of the controls came before complete implementation of the engineering controls and respiratory protections and, thus, were likely due to administrative controls. Cabrini Medical Center (New York City) Another report following an outbreak of multidrug-resistant tuberculosis tracked the sequential adoption of tuberculosis control measures from June through October 1991 (Maloney et al., 1995). The control measures included stricter isolation criteria and use of molded surgical masks for employees (June), improved laboratory services (July), increase from no isolation rooms (0 of 10) with negative pressure to a majority of rooms (16 of 27) with negative pressure (September), and use of an isolation chamber for sputum induction and administration of inhaled pentamidine (October). The initial assessment of worker tuberculin skin test conversions found similar conversions rates during the preintervention period (January 1990 to June 1991) and the intervention period (July 1991 through August 1992). When the analysts categorized workers by job category and ward location, however, they found higher conversion rates during the 18-month preintervention period for workers on wards serving tuberculosis patients than for workers on other wards (16.7 versus 2.8 percent). In contrast, during the 13-month intervention period, rates for the comparison groups differed little (5.1 versus 4.0 percent). When analysts categorized workers by whether or not they had direct patient contact, the difference in rates was smaller for the preintervention period (6.4 percent for those with patient contact and 1.0 percent for those without patient contact) and the change during the intervention period was less (4.7 percent

OCR for page 108
Tuberculosis in the Workplace compared with 2.3 percent). Analysts examined worker age, race, and BCG vaccination status and concluded that these factors could not account for the differences in skin test conversion rates. They also found no clustering of conversions by the worker’s zip code of residence. The analysts concluded that “the combination of source and environmental controls together with the use of molded surgical masks were all effective in reducing tuberculin skin test conversions among health care workers” (p. 94). The individual effects of different measures could not be isolated. Other New York City Studies Three additional studies in different New York City hospitals also suggest the effectiveness of tuberculosis control measures. The first report (from Roosevelt Hospital) had insufficient data to analyze skin test conversions for workers (Stroud et al., 1995). For patients, it concluded that the implementation of stricter isolation policies was associated with reduced delays in initiating isolation and reduced rates of patient-to-patient transmission of tuberculosis. In a second study at St. Clare’s Hospital, analysts reported decreases in skin test conversions for medical house staff concurrent with the adoption of more stringent isolation policies, the initial installation of negative-pressure isolation rooms, and adoption of a new kind of respirator. The conversion rates fell from 20.7 percent for the 6-month preintervention period to 7 percent during the next 6 months (Fella et al., 1995). Subsequent adoption of particulate and then dust-mist-fume respirators was not associated with any further consistent pattern of decreases in conversion rates. The report did not include information on employee demographics. The study’s authors commented that they had a steady inflow of new workers with negative tuberculin skin tests, so “we do not think that our decrease in [tuberculin skin test] conversions is simply the result of an exhaustion of susceptible persons” (Fella et al., 1995, p. 355). However, given inadequacies in the previous skin testing program it was possible that there was “a backlog of 2 years of conversions” in the 20.7 percent rate reported for the first period studied (Fella et al., 1995, p. 355). A third study at Columbia-Presbyterian Medical Center of the sequential adoption of stricter tuberculosis control measures examined skin test conversions for medical house staff from June 1992 to June 1994 (Bangsberg et al., 1997). The largest drop (from 5.1 to 0 conversions per 100 person-years) occurred after the adoption of a more rigorous isolation policy (administrative controls) and the construction of isolation rooms in the emergency department (engineering controls). This drop occurred before the adoption of new respiratory protections. The number of tuberculosis patients seen remained steady. The authors conclude that stricter isolation policies contribute the most to decreases in skin test conversion rates. They did not report information on employee demographics.

OCR for page 108
Tuberculosis in the Workplace Summary of Individual Reports Notwithstanding their limitations, taken together, the studies reviewed above suggest that implementation of tuberculosis control measures can help end outbreaks and prevent new transmission of M. tuberculosis. They support the logic of CDC’s emphasis on the primacy of administrative controls, in particular, rigorous respiratory isolation policies to reduce exposure opportunities by promptly identifying, evaluating, and isolating people with signs and symptoms suspicious for tuberculosis. The studies suggest some positive effects from engineering controls, which come second in CDC’s hierarchy of controls, but their contributions are hard to disentangle from the effects of previously or simultaneously adopted administrative controls. Personal respirators did not appear to play a significant role in ending outbreaks of tuberculosis. In recent years, hospitals may also have benefited from changes in the treatment patterns including both a shift from inpatient to outpatient treatment for people with infectious tuberculosis and the availability of more effective treatments for AIDS that have reduced the need for inpatient care. Continued reports of outbreaks in correctional facilities in South Carolina and Pennsylvania (see Chapter 5) suggest the need for better information on surveillance programs and other tuberculosis control measures in these settings. Other Studies and Reports In addition to the reports reviewed above, the committee found some additional relevant studies that involved mostly low-risk or stable settings.5 One study reviewed tuberculosis control measures in 13 midwestern hospitals, all but one of which were categorized as low or very low risk for transmission of tuberculosis. The researchers did not find an association between the kinds of tuberculosis control measures adopted and worker skin test conversion rates (Woeltje et al., 1997). The study did not examine isolation policies, and the authors noted that compliance with written policies for other measures was imperfect. In one of the few studies examining a single control measure, Behrman and Shofer (1998) report on an emergency department that adopted improved engineering controls while leaving isolation and respiratory pro- 5   Appendix D reviews several surveys that asked questions about the implementation of tuberculosis control measures and about results of worker skin testing programs. Analyses of the association between control measures and conversion rates produced inconsistent results. Given the variations in response rates, the limited detail possible in survey responses, and similar concerns, the committee did not find that these analyses contributed to its understanding of the effects of the CDC guidelines.

OCR for page 108
Tuberculosis in the Workplace tection protocols unchanged.6 The controls included installation of four isolation rooms, improved general-area ventilation, and installation of Plexiglas shields for registration personnel. The researchers compared skin test conversion rates for emergency department personnel and other hospital personnel. (The two groups did not differ significantly in age, ethnicity, foreign birth, county of residence, BCG vaccination status, or initial tuberculin skin test status.) The department implemented tuberculosis control measures at the end of the first year for which conversion rates were compared. In that year, emergency department personnel had significantly higher conversion rates than other hospital personnel (12 versus 2 percent). In the year after the measures were adopted, the rates did not differ significantly (0.0 percent for emergency department personnel compared with 1.2 percent for other workers). In one of the few studies in a correctional facility, Puisis and colleagues (1996) reported on the introduction of radiographic screening in the Cook County Jail as part of the intake medical evaluation process. The new technology reduced the time from jail entry to isolation to 2.3 days from 17.5 days. During the period from March 1992 through December 1997, jail health care staff screened more than 445,000 inmates and found 206 cases of active tuberculosis (46.2 per 100,000 population) from the radiographic screening alone (McAuley, 2000). Staff concluded that screening had been cost-effective but noted that decreasing tuberculosis case rates could change the picture in the future. Finally, a few studies have attempted to model the contributions of engineering controls and respiratory protections to preventing transmission of tuberculosis.7 Summarizing the results of one such effort, Fennelly and Nardell (1998) suggest that the “risk of occupational tuberculosis probably can be lowered considerably by using relatively simple respirators combined with modest room ventilation rates for the infectious aero- 6   Menzies and colleagues, in an article published after the committee concluded its analyses, reported a cross-sectional study of 17 Canadian hospitals. It showed skin test conversions “strongly associated with inadequate ventilation in general patient rooms [and bronchoscopy rooms] and with type and duration of work, but not with ventilation of isolation rooms” (Menzies et al., 2000, p. 779). The authors suggest that this association reflects the “exposure in nonisolation rooms of undiagnosed patients…[who] are known to pose the greatest risk to hospital workers” (p. 788). In the higher risk hospitals, the room changes per hour in the negative-pressure isolation rooms averaged between 6.1 and 9.4. The 1994 CDC guidelines recommend a minimum of 6 air changes per hour for negative-pressure isolation rooms and 12 air changes per hour where feasible (CDC, 1994b). 7   Modeling studies are important for the assessment of risk in a number of situations, for example, when levels of a hazard are low, slow to produce observable effects, or difficult to measure directly. They may likewise be useful when the effect of an intervention is expected to be small. In such situations, clinical studies may be impractical or ethically dubious, and epidemiologic studies may be of limited use because they can not detect effects of intervention without very large numbers of subjects or very long periods of time, or both.

OCR for page 108
Tuberculosis in the Workplace sols likely to be present in isolation rooms of newly diagnosed patients” (p. 754). For workers involved in cough-inducing procedures for infectious patients, more sophisticated respirators may be needed to protect workers adequately. The benefit to workers of using respirators is probably minimal if patients are being properly treated in properly ventilated isolation rooms. Another modeling exercise reported by Barnhart and colleagues (1997) came to generally similar conclusions but placed greater emphasis on cumulative risk over a worker’s lifetime. The authors concluded that higher-level respiratory protection (more than a disposable mask respirator) was reasonable for workers in higher-risk situations (e.g., those performing bronchoscopies or those treating highly infectious patients or patients with multidrug-resistant disease). A Chicago study suggests the ineffectiveness of personal respirators when adequate administrative and engineering controls are lacking. Kenyon and colleagues (1997) reported an outbreak of multidrug-resistant tuberculosis in a facility that provided and fit tested workers with high-efficiency particulate respirators but that had no isolation rooms that met CDC criteria. Three of the 11 previously skin test-negative workers whose tuberculin skin test result converted to positivity (including a ward secretary with no patient care responsibilities) had no contact with the source case patients. The authors conclude that a respiratory protection program alone cannot protect all workers. In the absence of appropriate isolation rooms, air that escapes from rooms housing infectious patients can infect those outside the room. Delays in recognizing and treating infectious patients also contributed to the outbreak. (Appendix F presents additional background on personal respiratory protection as a tuberculosis control measure.) COMMITTEE CONCLUSIONS When the resurgence of tuberculosis began in the mid-1980s in the United States, communities and workplaces were generally not prepared. After years of effective treatment and declining tuberculosis case rates, tuberculosis control measures—including those recommended by CDC in 1983—were not priorities for either public or occupational health programs. The epidemic of HIV infection and AIDS and public health and medical responses to the epidemic were still emerging issues, and the interaction of that epidemic with tuberculosis was not well documented or understood. Similarly, the threat of multidrug-resistant disease was not yet clearly appreciated. Much has happened in the past 15 years. Certainly, the epidemiology of tuberculosis has changed, with case rates again in decline since 1993. Virtually all states have shown decline, although relatively high rates of

OCR for page 108
Tuberculosis in the Workplace tuberculosis persist in a number of states and communities. The special vulnerability of people with suppressed immune systems is now recognized, and the threat of multidrug-resistant disease and the conditions that give rise to it (primarily, incomplete and inadequate treatment of tuberculosis) are clearly understood. The tuberculosis control components of community health programs are better funded and better focused on measures that prevent spread of the disease including directly observed therapy for patients with active tuberculosis. For most hospitals, prisons, and other facilities, these external changes have decreased the likelihood that employees will see someone with diagnosed or undiagnosed active tuberculosis. These changes have also raised the visibility and understanding of the disease. Nonetheless, with more than 17,000 cases reported nationally in 1999, tuberculosis remains a threat. Inattention to community and workplace measures to control and prevent transmission of M. tuberculosis could lead to another, potentially more serious resurgence of tuberculosis. Thus, it is important to assess how workplace tuberculosis control measures are being implemented and how well they are working. The changing environment for workplaces makes it difficult, however, to assess the effects of workplace tuberculosis control programs. This difficulty is compounded by the practical problems of conducting rigorous, well-controlled research on these programs, which have often implemented multiple measures simultaneously. Nonetheless, after reviewing the literature (including theoretical arguments and mathematical models), considering discussions during the committee’s public meetings, and drawing on its members’ experiences and judgments, the committee reached several conclusions about, first, the implementation and, second, the probable effects of workplace tuberculosis control measures. Whether regulations may be needed to sustain or increase rates of compliance with tuberculosis control measures is considered in Chapter 7. Implementation of Tuberculosis Control Measures The information base for the following conclusions applies mainly to hospitals and to a lesser extent to prisons. The committee expects that the consistent, correct implementation of control measures may be more difficult in other institutions such as jails and homeless shelters, which generally lack the resources, oversight, and expertise available to hospitals. These workplaces may also differ in the degree to which managers and workers understand and accept tuberculosis as a risk and tuberculosis control measures as necessary. Most reports reviewed by the committee predate the 1994 guidelines, but the basic measures recommended have remained reasonably stable.

OCR for page 108
Tuberculosis in the Workplace The conclusions below relate only to the implementation of tuberculosis control measures; the following section considers their effects. Institutional departures from recommended tuberculosis control policies and procedures were common, if not the norm, in the late 1980s and the early 1990s. In large measure, the neglect that characterized community tuberculosis control programs (IOM, 2000) appears to have been duplicated in hospitals, correctional facilities, and, probably, but with less documentation, other facilities that serve people at increased risk of the disease. Even after public health authorities and newspapers were describing the resurgence of tuberculosis in the latter half of the 1980s, surveys in the early 1990s suggested that hospitals and prisons were neglecting recommended surveillance, isolation, and other measures that had been reinforced in 1990 CDC guidelines. Reports of tuberculosis outbreaks in hospitals also document lapses in infection control measures. Hospitals and correctional facilities reported increased implementation of tuberculosis control measures by the mid-1990s. By 1996, for hospitals and correctional facilities, responses to national surveys and some other studies were showing much more complete and consistent reported compliance with recommended tuberculosis control measures. The hospitals experiencing outbreaks in the early 1990s clearly had a stimulus to implement control measures earlier. For other institutions, increased implementation likely reflects the impacts of further and more complete reports on workplace outbreaks of tuberculosis, the CDC’s increased effort to educate health care managers and clinicians about tuberculosis and tuberculosis control measures, the pressure for action exerted by unions on both employers and public agencies, and the initiation by OSHA of enforcement procedures and rulemaking processes for occupational tuberculosis. Data do not allow the committee to draw conclusions about trends for other settings. Implementation appears to be most complete for administrative controls including respiratory isolation policies. For engineering controls, the installation of negative-pressure isolation rooms has increased, but ventilation performance and performance monitoring may still fall short of recommendations. Information about organizational implementation of the various elements of personal respiratory protection programs is limited. Most studies suggest that most employers have been providing some kind of protection (surgical masks or respirators) and that they have changed the devices provided as new options, such as N95 respirators, have been certified. Written policies have not necessarily been translated into routine practice. High levels of compliance with control measures, as reported in surveys, may not be matched by high compliance on a day-to-day basis. Although on-site reviews that match hospital policies to actual practices are limited,

OCR for page 108
Tuberculosis in the Workplace they suggest the need for some caution in accepting survey responses as conclusive. Departures from guidelines occur at both the institutional level (e.g., provision of respirators and installation of negative-pressure isolation rooms) and the individual level (e.g., use of respirators and closing doors of isolation rooms). Whether less than total compliance makes a practical difference in preventing workplace transmission of tuberculosis is a separate question. Effects of Tuberculosis Control Guidelines The caveats cited for implementation also apply to the following conclusions about the effects of tuberculosis control measures. In addition, the committee could not readily disentangle the effects of the CDC guidelines from environmental influences including the effects of community public health measures, regulatory actions by OSHA and others, and the changing epidemiology of the disease. Furthermore, because control measures were often introduced simultaneously or close in time, the relative contribution of individual measures is difficult to distinguish. Finally, the committee could reach no conclusions about what level of compliance with different measures might be sufficient to prevent transmission of M. tuberculosis under different workplace conditions. Again, the conclusions presented below apply primarily to hospitals. The picture for other workplaces is less clear. Overall, the measures recommended by CDC for prevention of the transmission of tuberculosis in health care facilities have contributed to the ending of outbreaks of tuberculosis and the prevention of new outbreaks. This conclusion rests primarily on several outbreak reports and on information from institutions that did not report outbreaks but reduced skin test conversion rates after implementing control measures. Although each report has its limitations, taken together they show consistent results. The hierarchy of control measures recommended by CDC is supported by studies of tuberculosis outbreaks in hospitals as well as by logic and biologic plausibility. Outbreak studies support CDC’s stress on administrative controls, in particular, application of protocols to reduce opportunities for worker or patient exposure to M. tuberculosis through prompt identification and isolation of people with signs and symptoms suspicious for infectious tuberculosis. Outbreak studies and modeling exercises suggest that engineering controls also make a contribution in limiting the transmission of tuberculosis. Although outbreak studies suggest that most of the benefit of control measures comes from administrative and engineering controls, modeling exercises support the tailoring of personal respiratory protections to the level of risk faced by workers—that is, more stringent protection for those in high-risk situations and less stringent measures for others.

OCR for page 108
Tuberculosis in the Workplace The Limits of Control Measures Tuberculosis control measures cannot be expected to prevent all worker exposure to tuberculosis, especially in areas with moderate to high rates of tuberculosis. Although control measures have helped end workplace outbreaks of the disease and prevent transmission of M. tuberculosis, they cannot be expected to prevent all exposures. Not all individuals with infectious tuberculosis have evident symptoms or signs of the disease, so workers may be exposed to them for some time before tuberculosis is suspected and a diagnosis is made. In addition, opportunities will exist for exposure in emergency departments and elsewhere before infectious individuals are recognized and isolation protocols can be applied and completed. Conscientious implementation of guidelines does not guarantee that transmission will never occur, but it appears to reduce risk significantly, especially in high-prevalence areas. In communities with little or no tuberculosis, the effectiveness of control measures is necessarily limited. If success in community control of tuberculosis continues, more communities can be expected to join this low-prevalence group. Nonetheless, given the mobility of the U.S. population including immigrants from high-prevalence countries, it can be expected that people with infectious tuberculosis will occasionally appear in low-prevalence communities and their health care facilities. For example, in 1997, just 2 percent of the U.S. population lived in counties that had had no reported cases of tuberculosis in 5 years (Geiter, 1999).8 As noted in Chapter 5, workers in low-prevalence areas who encounter someone with infectious tuberculosis may be at higher risk of exposure than their colleagues in high-prevalence areas. They are less likely to be familiar with and alert to the disease’s signs and symptoms and may be less likely to have protective engineering controls in place in the emergency departments and other areas where such individuals are first encountered. SUMMARY In the late 1980s and early 1990s, institutional departures from recommended tuberculosis control policies and procedures were widespread. 8   By using CDC data and a definition of tuberculosis elimination as no cases in 5 years, 587 of 3,142 (19 percent) counties in the United States could be considered tuberculosis free as of 1997 (Geiter, 1999). Of the counties with no tuberculosis from 1993 to 1995, 75 percent had no cases in the next 2 years and an additional 19 percent only had one case in either 1996 or 1997. Another definition of tuberculosis elimination is a case rate of less than 1 per 1 million population. Starting from a case rate of 74 per 1 million population in 1997, it would take 50 years to reach the elimination target if case rates were declining at an average annual rate of 5 percent and 41 years to reach the elimination target if case rates were declining at an average annual rate of decline of 10 percent. The average yearly rate of decline in recent years has been about 7 percent.

OCR for page 108
Tuberculosis in the Workplace By the mid-1990s, hospitals, correctional facilities, and possibly other facilities began to report higher levels of adherence to CDC recommendations. On-site inspections and other data suggest the need for caution in assuming that written tuberculosis control policies represent routine institutional or worker practice. Implementation of tuberculosis control measures appears to have contributed to ending outbreaks of tuberculosis and preventing new ones. Outbreak studies as well as logic, biologic plausibility, and modeling exercises support CDC’s hierarchy of tuberculosis control measures. That hierarchy stresses administrative controls (in particular, rigorous application of protocols to promptly identify and isolate people with signs and symptoms suspicious for infectious tuberculosis), followed by engineering controls and, finally, by personal respiratory protections. Especially in high-prevalence areas, occasional worker exposure to patients with infectious tuberculosis can still be expected, despite the implementation of generally effective protocols for respiratory isolation.