National Surveillance System For Health Care Workers
Abstracts
4th Decennial International Conference on Nosocomial & Healthcare-Associated Infections

ESTIMATE OF THE ANNUAL NUMBER OF PERCUTANEOUS INJURIES IN U.S. HEALTH-CARE WORKERS. AL PANLILIO*, DM CARDO, S CAMPBELL, PU SRIVASTAVA, I WILLIAMS, J JAGGER, J ORELIEN, R COHN, NASH SURVEILLANCE GROUP, AND EPINET DATA SHARING NETWORK, Centers for Disease Control and Prevention, Analytical Sciences, Inc, University of Virginia.

Needlestick and other percutaneous injuries (PIs) pose the greatest risk of occupational transmission of bloodborne viruses to health-care workers (HCWs). The annual number of PIs sustained by U.S. HCWs have been estimated using a variety of methods and have ranged from 100,000-1,000,000. To construct a single representative result, we estimated the total number of PIs by combining data collected in 1997 and 1998 at 15 National Surveillance System for Health Care Workers (NaSH) and 45 Exposure Prevention Information Network (EPINet) hospitals. The combined data were used as a sample of all U.S. hospitals and adjusted for underreporting. Since the number of PIs has been correlated with various measures of hospital size, the estimate of the number of PIs nationwide was weighted to reflect the number of admissions in all U.S. hospitals relative to those in NaSH and EPINet. The estimated number of PIs sustained annually by hospital-based HCWs was 384,325, with a 95% confidence interval from 311,091 to 463,922. The number of PIs sustained by HCWs outside of the hospital setting was not estimated. Our estimate, based on combined NaSH and EPINet data, may be more widely generalizable than those based on either system alone due to the improved heterogeneity of the hospitals represented. NaSH hospitals tend to be larger than average and are more likely to be found in the Northeast. EPINet hospitals tend to be smaller than NaSH hospitals and are clustered in the West Coast and southeastern U.S. Although our estimate is smaller than some previously published estimates of PIs in HCWs, its magnitude remains a concern and emphasizes the urgent need to implement prevention strategies. In addition, improved surveillance is needed to monitor injury trends among HCWs in all health-care settings and to evaluate the impact of prevention interventions.

Percutaneous Injury Reporting in U.S. Hospitals, 1998 F Alvarado*, A Panlilio, D Cardo and the NaSH Surveillance Group Hospital Infections Program, Centers for Disease Control and Prevention Atlanta, Georgia

Of all occupational exposures, percutaneous injuries (PIs) pose the greatest risk for transmission of bloodborne pathogens. Health care workers (HCWs) do not report all PIs despite the availability of prophylaxis for some exposures. To determine the level of PI reporting and assess the effect of hospital characteristics and occupation on reporting rates (RRs), we analyzed data from HCW surveys at 12 hospitals participating in the National Surveillance System for Health Care Workers (NaSH) in 1998. In this survey, 14,215 HCWs indicated if they sustained a PI in the last 12 months, how many they reported, and their reason(s) for not reporting. RRs were stratified by hospital size, geographic location, HIV-inpatient days, and occupation. Of 1922 PIs sustained, 800 were reported for an overall RR of 42%. RR varied significantly by region: northeast, 54%; southeast, 38% (range 29-86%; relative risk=1.48; 95% CI 1.31, 1.67; p<0.01) and by hospital size: 200-750 beds, 52%; 751-1200 beds, 42% (range 29-86%; relative risk 1.25; 95% CI 1.12, 1.39; p<0.01). RRs did not vary with number of HIV-inpatient days per year: 150-700, 53%; 701-8350, 56%. Surgeons' RR was 27% vs. 48% for all other HCWs (range 46-53%; relative risk=0.57; 95% CI 0.49, 0.65; p<0.01). The most commonly cited reason for not reporting was an assessment that the injury or the source was low risk (51%). PI reporting appears to be influenced by hospital size, location, and occupation but not HIV prevalence. All hospitals should increase their efforts to facilitate and promote PI reporting.

Variations in Needlestick Injuries in the National Surveillance System for Healthcare Workers Over Time LA Chiarello*, D Cardo, and the National Surveillance System for Healthcare Workers (NaSH) Surveillance Group. Centers for Disease Control and Prevention, Atlanta, GA, USA

Surveillance data on percutaneous injuries (PIs) among healthcare workers (HCWs) are necessary to assess the impact of prevention interventions and identify emerging risks. Data on PIs, including type and purpose of device involved and occupation of injured HCW, reported from hospitals participating in NaSH were analyzed by year. A total of 5,178 PIs reported from 1/96 through 7/99 (study period) were analyzed, combining data from 1/98 through 7/99. There was no change over time in the distribution of occupations of HCWs sustaining PIs. However, the distribution of device types involved and purposes for which devices were used varied over time. PIs due to winged steel needles (WSN) decreased from 18.2% to10.5% (p<0.0001) while PIs due to needles attached to intravenous (IV) tubing increased from 1.7% to 4.5% (p<0.002) from 1996 to 1999. The proportion of PIs associated with percutaneous blood withdrawal decreased from 22.5% to 16.2% (p<0.0001). To examine whether the observed variations reflect a changing trend in these types of PIs, we analyzed data from 17 hospitals reporting at least 75 PIs since 1998, since the number of hospitals contributing data to NaSH increased during the period studied from five in 1996 to 31 in 1999. The proportion of injuries attributable to different devices and procedures varied considerably among hospitals. When data from five hospitals participating in NaSH for 2 or more years were analyzed, there was no significant variation over time in the proportion of PIs associated with blood withdrawal, WSNs, or needles attached to IV tubing. Interpretation of aggregated surveillance data on PIs must take into consideration changes in the number and characteristics of institutions contributing data. In addition, each healthcare organization must assess its own surveillance data when setting prevention priorities.

Multiple Blood Exposures Among Healthcare Workers. Sulis CA, Derridinger O. Boston Univ School of Medicine and Boston Med Center, Boston, MA; Boston Med Center, Boston, MA.

Boston Medical Center (BMC) is a 547 bed teaching hospital. Over the past 10 years several interventions have contributed to a reduction in employee (HCW) exposures. HCW may report a single exposure (SE), multiple exposures (ME), or fail to report. Our analysis is described below. Risk-reduction strategies are discussed during evaluation and treatment. Supplemental information is elicited from observational studies and anonymous surveys. NaSH software is used to assess trends and focus interventions. Between 1/97 and 7/99, 327 exposures were reported by 292 HCW. 11% reported ME (27 with 2, 4 with 3). Post exposure prophylaxis (PEP) was initiated for 51% of all HCW. A similar proportion began PEP following exposure top HIV+ source (63% for SE, 60% for ME), 38% declined. Median time between hire and first exposure was shorter for residents with ME (8 months) than for other HCW with ME (44 months). Most frequent cause of exposure was suturing (9), handling equipment/specimens (8), passing equipment (7), and manipulating needles (5). Of 18 exposures observed during 874 procedures, only 1 (6%) was reported. HCW surveys confirmed variable rates of under-reporting, but supplied no clues to a solution. We have failed to ascertain why certain HCW have multiple exposures, or why many HCW fail to report. Optimal strategies to achieve improvement are unknown. New initiatives planned for the next 12 months include development of a multidisciplinary hands-on skills lab to teach high-rick procedures to residents, standardization of equipment, ongoing evaluation of safer devices, and improvement of reporting procedures.

EPIDEMIOLOGY AND REPORTING OF NEEDLE-STICK INJURIES AT A TERTIARY CANCER CENTER. Abdel Malak S, Eagan J, Sepkowitz KA. Memorial Sloan-Kettering Cancer Center, New York, NY.

PURPOSE: Our objectives were to determine the epidemiology of needle-stick and other sharp object injuries among Health Care Workers (HCWs) and to examine injury reporting behavior of HCWs at Memorial Sloan-Kettering Cancer Center. METHODS: The Infection Control Program(ICP) conducted an anonymous self-administered survey of the institution's HCWs during several ICP training sessions.1,423 surveys were distributed to surgical medical staff, non-surgical medical staff, nursing staff, building services staff, and other staff. 1,33 staff members completed the survey (response rate, 92%). The survey included the following questions: occupation, number of needle-stick injuries and/or injuries with other sharp objects within the past 12 months, number of injuries reported to the infection control program, employee health service, or urgent care center, reason for not reporting any injury, and number of phlebotomy procedures performed in a typical week. RESULTS: Respondents included 183 surgical medical staff, 137 non-surgical medical staff, 796 nursing staff, 167 building services staff, and 2other staff, 73% of respondents did not have an injury. The total number of injuries reported was 747. 18% of HCWs reported one injury, 1% reported two or more injuries. The HCWs who had two or more injuries accounted for 69% of all injuries. Surgical medical staff had the highest mean number of needle-stick injuries (1.22), followed by nursing staff (.6), other occupations (.35), building services (.22) and non-surgical medical staff (A). The total rate of reporting to the institution was 22%, lowest rate among MDs, nonsurgical medical staff (%) and surgical medical staff(4%);other occupational reporting rates included nursing staff(24%), building services (95%), other occupations (71%).The most common reasons for not reporting were that HCWs believed that the injury was low risk (32%), the patient appeared low risk for blood borne disease (23%), or the needle/object was sterile (24%). CONCLUSION: A high rate of recidivism in 1% of our institution's HCWs accounted for the majority of injuries (69%). Interventions to reduce the incidence of injuries due to needle-sticks or other sharp objects need to be targeted to specific sub-populations of HCWs. Only 22% of all injuries at our institution are reported. Anonymous surveys may provide a more accurate method of determining the incidence of needle-stick and other sharp object injuries.

Pregnant Health Care Workers Sustaining Occupational Blood Exposures F Alvarado*, A Panlilio, D Cardo and the NaSH Surveillance Group, Centers for Disease Control and Prevention, Atlanta GA

Women comprise 76% of hospital workers in the U.S., and at least 64% of these women are of child-bearing age. To characterize occupational blood exposures in pregnant health care workers (HCWs), we analyzed data collected from January 1998 to July 1999 by 25 hospitals participating in the National Surveillance System for Health Care Workers. Of 4144 exposures, 2252 (54%) occurred in women 18-45 years of age; 60 (3%) of these HCWs were pregnant. They sustained 45 (75%) percutaneous injuries (PIs), 10 (17%) mucous membrane exposures, four (7%) skin exposures and one (2%) bite. The exposures occurred in all trimesters: first 23 (38%), second 25 (42%), third 10 (17%). Three source patients were HIV-positive and seven were Hepatitis C Virus (Hepatitis C Virus) positive. Of thirty HCWs offered HIV postexposure prophylaxis (PEP), 4 accepted; one of three exposed to an HIV-positive source and three exposed to an HIV-negative or unknown source. The other two HCWs who were exposed to an HIV-positive source and did not take PEP sustained mucous membrane and/or skin exposures of short duration. Information on PEP is available for two of four HCWs who initiated a regimen. One, exposed to an unknown source, stopped after 5 days because of side effects. The other, exposed to an HIV-negative source, took PEP for 22 days. Both HCWs took zidovudine, lamivudine and indinavir. Of the 45 PIs in pregnant HCWs, 25 (56%) were potentially preventable because either the needle use was unnecessary, or there was a needle device with a safety feature and/or a work practice control that could have been used to prevent the injury. Pregnant HCWs sustain occupational blood exposures placing them at risk for infections. Because PEP for Hepatitis C Virus is not currently recommended and HIV PEP may have adverse effects on the HCW and/or her fetus, greater emphasis should be placed on preventing these exposures.

Preventability of Needlestick Injuries to Health Care Workers in the National Surveillance System for Healthcare Workers. SR CAMPBELL*, L CHIARELLO, P SRIVASTAVA, D CARDO, and the NaSH SURVEILLANCE GROUP. Centers for Disease Control and Prevention, Atlanta, Georgia, USA.

Needlestick injuries with hollow-bore needles (NIs) represent the most frequently reported type of exposure sustained by health care workers (HCWs) within hospitals participating in the National Surveillance System for Healthcare Workers (NaSH). To determine the proportion of potentially preventable NIs, we analyzed information on NIs reported by 31 NaSH hospitals. Variables assessed included needle type, procedure, and circumstances of injury. Preventability of NIs was defined hierarchically as 1) needle use was unnecessary for the procedure or 2) a "safer" needle device or 3) safer work practice may have been used. NIs were defined as non-preventable if they happened during use in the patient and/or no "safer" needle device was available. NIs that involved a device with a safety feature were assessed independently. From 6/95 to 10/99, 5,548 percutaneous injuries were reported; 3,410 (61%) were NIs. Of the 3,410 NIs, 2,029 (60%) were classified as preventable: in 663 (33%) needle use was unnecessary; 787 (39%) were preventable with a "safer" needle device; and 579 (29%), by a safer work practice. The proportion of preventable NIs varied by hospital ( = 64%, range 48% to 85%). The mean proportions of various preventability categories also varied by hospital. Of the remaining 1181 NIs, 672 (21%) were classified as non-preventable and for 509 NIs, preventability could not be determined based on data provided. An additional 200 NIs involved a "safety" device; in 17 (9%) use of the needle was unnecessary, and for the remaining 183 the NIs most commonly occurred either before activation was appropriate (43%), the user failed to activate the safety feature (22%), or the safety feature failed (3%). Most reported NIs are preventable by eliminating unnecessary needles, implementing devices with safety features, and ensuring compliance with recommended work practices. However, a large proportion of NIs are still considered non-preventable. Methods to prevent these NIs, including the use of devices with safety features that ensure needle protection throughout a procedure, are needed.

Using the National Surveillance for Hospital Healthcare Workers to Reduce Percutaneous Injuries. Trape M, Schenck P, Warren A. Univ of Connecticut Health Ctr., Farmington, CT.

The National Surveillance for Hospital Healthcare Workers (NaSH) data on percutaneous injuries collected over two years was used to improve a health center's infection control program in two ways: (1) improved surveillance with increased reports of injuries; and (2) targeted interventions to reduce injuries. The NaSH surveys over the 1997-98 and 1998-99 supplemented the employee health infection control surveillance program. Reports of percutaneous injuries with blood and body fluid exposure (BBFE) increased from 82/5220 HCW (1.5%) the year before the NaSH, to 155/5305 HCW (2.9%) and 189/5422 HCW (3.4%) during the two years using the NaSH database. The reports likely reflect improved awareness of the importance of evaluation and treatment after an incident rather than increased problem practices. The NaSH data was used to characterize BBFE injuries and identify higher risk groups and activities. NaSH information was reviewed on: occupation of HCW; where the incident happened; HIV, hepatitis B and C status of the source patient; visible blood on the sharp; how the injury occurred; whether through gloves or other clothing; and depth and body site of injury. Educational programs on available safety devices and protective protocols were disseminated and interactive computer safety training was improved and targeted at higher risk groups. The largest group with BBFE was the resident physicians who compromised 39% of the exposures in 1997-98. Percutaneous injury was reduced by 12% from 60 to 51 in 1998-99. Because residents and students go to various affiliated hospitals, each with a unique programs in place, additional educational efforts are planned that will use further analysis of the NaSH data. The challenge is to decrease the total number of BBFE and at the same time to encourage reporting of all possible exposures. Activities are planned: inter-hospital interactive tele-video conferences; training the trainer programs to nursing staff coordinated with infection control staff; health fairs with displays of available safety devices and of data collected from BBFE over the years.

Using NaSH (National Surveillance System for Hospital Healthcare Workers) for Designing Programs to Reduce Percutaneous Injuries in A Univ Hospital. Fisher M, Rogers A, Kahkoo R, Capodieci J, Sabo L, Buterbaugh A, Hortsman P. Robert C. Byrd Health Sciences Center, Morgantown, WV; Ruby Memorial Hosp, Morgantown, WV.

Healthcare workers (HCWs) have an increased risk of exposure to bloodborne pathogens (BBP). Monitoring trends of percutaneous (PI) and other injuries to HCWs is facilitated by a comprehensive computerized program. We have used the software program NaSH developed by the CDC since Jan 1998 to record data on BBP exposures at a Univ hospital (370 beds), associated outpatient facilities, and health sciences center. From Jan through Oct 1999 there were 235 exposures to BBP; 198 PI and 37 non-percutaneous injuries (Non-PI) for a total of 5700 HCWs. The NaSH program allowed us to report data readily to individual units and identify a high risk location (operating room) where 36 (18%) of PI occurred. Further analysis of PIs in the OR during this 10-month period showed that the residents had the highest rate of Pi with 18 (50%). During the same ten-month period in 1998, the rate of Pi for surgical technicians (ST) was 12 (28%). A targeted intervention to reduce the PIs in ST began in Dec 1998. From Jan through Oct 1999, the proportion of PIs in STs dropped from 18% to 11% (p=.12) Further analysis will be performed to determine the proportion of PIs that were potentially preventable among STs in order to determine the effectiveness of the educational intervention. Using NaSH to compare the total number of PIs from the first ten months in 1998 versus 1999, the number of reported PIs in the same outpatient surgical units increased from 1% to 8% (p<.01). These variations in the number of PIs as demonstrated by NaSH underscore the need for continuous, comprehensive monitoring. Interventional programs are being implemented based on the results of NaSH data targeting high-risk groups and locations. We conclude that the NaSH software program provides an efficient tool for tracking PIs in HCWs. Data generated are useful to Employee Health for individual follow-up, finding trends in exposures, and planning specific educational programs that will decrease the risk of exposure to BBP.

Prevention of Needlestick Injuries in Healthcare Workers: 27 Month Experience with a Resheathable "Safety" Winged Steel Needle Using CDC NaSH Database. Chen LBY, Bailey E, Kogan G, Finkelstein LE, Mendelson MH. Mount Sinai Med Center, New York, NY.

NIS from WSNs are considered high-risk for bloodborne pathogen transmission. We evaluated a safety WSN (SafetyLok, BD) at an 1,100-bed hospital , previously reporting a 50% reduction in WSN related NIs by using a safety WSN. Subsequent to this trial the safety WSN was evaluated during a 16 month (6/1/98-9/30/99) post study period (total 27 month experience with this safety device). NIs were tracked using the NaSH exposure form; a survey of sharps disposal boxes was performed to assess usage and activation rates. The non-safety baseline period I (9/1/95-3/31/97) WSN NI rate was 13.41/100,000 WSNs (86 NIs/641.282 WSNs); the study period II (7/1/98-5/31/98) WSN NI rate was 6.87/100,000 WSNs (30 NIs/ 436, 180 safety WSNs); and the post study period III (6/1/98-9/30/99) WSN NI rate was 5.5/100,000 WSNs (39 NIs/710,652 safety WSNs). The post study WSN NI rate was 59% lower than the baseline period ( p<0.01). Analysis of post study safety NIs by procedure: 27 percutaneous venous puncture, 8 arterial puncture, 3 to insert a peripheral I.V. before disposal, 5 during or after disposal, 1 before use of the item. 23 occurred before mechanism activation was appropriate, the safety mechanism was not activated in 8, 5 occurred during the activation process. A survey of 627 disposed WSNs during period III revealed 627 (100%) safety WSNs, activation rate 71% (444/627). In conclusion, the Safety Lok (BD) WSN has remained consistently effective in reducing WSN related NIs for 27 months at our institution. Use of the Safety Lok WSN should prevent bloodborne pathogen transmission to HCWs. Compliance with proper activation procedures needs to be routinely stressed.

Evaluation of a Safety IV Catheter (Insyte Autogurad, Becton Dickinson) Using the Centers for Disease Control and Prevention (CDC) National Surveillance System for Hospital Healthcare Workers Database. Mendelson MH, Chen LBY, Finkelstein LBY, Bailey E, Kogan G. Mount Sinai Med Center. New York, NY.

A safety IV catheter (Insyte Autoguard, Becton Dickinson) was evaluated at an 1,100 bed Univ affiliated medical center to determine efficacy in reducing needlestick injuries (NIs). A baseline period I (pre-safety trials) from 6/1/93-8/31/96 (27 months) was compared to a study period II (safety IV catheter, two-month training, 2-3/99 and six month pilot, 4-9/99; 8 months data thus far, study ongoing). The interim between the baseline and the study periods was inclusive of an evaluation of Protectiv® Plus Catheter (Johnson and Johnson). Training included model practice insertions for IV catheter users. NI data was analyzed utilizing the National Surveillance System for Hospital Healthcare Workers (NaSH) data collection tool and database. A survey of sharps disposal boxes was performed to assess usage and activation rates. An 89% reduction in IV stylet related NIs was demonstrated comparing the baseline period injury rate of 6.6/100,000 IV stylets (56 injuries/848,958 stylets) to the training and pilot periods (8 months) injury rate of 0.7/100,000 IV stylets (1 injury/152,952 safety IV stylets) (p<0.01). The period II injury occurred while the stylet was being withdrawn from the patient and the healthcare worker (HCW) failed to activate the safety mechanism. A survey of 495 disposed IV stylets during the pilot period revealed 495 (100%) safety IV stylets with an activation rate of 85% (420/495). In conclusion, the safety IV catheter (Insyte Autoguard) resulted in a marked and significant reduction in IV stylet-related injuries during the training and pilot periods with an overall compliance with activation of 85%. Although the Insyte Autoguard require activation by the user, the simplicity of the activation process should promote user compliance and therefore reduction in injuries. In that IV stylet-related injuries are high risk ( hollow-bore needle, inserted directly into vein or artery) if reduction of injuries continues during the study period, usage of this safety device should result in decreased blood-borne pathogen transmission to HCWs.

HIV Postexposure Prophylaxis: 1996-98. Koll B, Raucher B, Nadig R. Beth Israel Med Ctr (as part of the NaSH Surveillance Group), New York, NY

Beth Israel Med Ctr-Petrie Division, an 850-bed hospital located on the lower East Side of Manhattan, has used the National Surveillance System for Hosp Health Care Workers (NaSH) since 1994. In 1996, it implemented the revised CDC guidelines for HIV postexposure prophylaxis (PEP) with zidovudine, 3TC, and indinavir. To assess the impact of our PEP program, a review of NaSH data was conducted on all percutaneous injuries (PIs) reported to the Employee Health Service from Jan 1996-Dec 1998. In 1996, there were 106 blood and body fluid exposures. 92 (87%) were due to PIs. Hollow bore devices accounted for 68 (74%) of the PIs. Physicians and nurses accounted for the majority of the PIs. 46 source patients had known HIV serostatus or were asked to consent to HIV testing. 17 (37%) were infected with HIV. 81 HCWs (88%) were offered, 19 (23%) began and 17 (89%) completed PEP. Over the next two years, a change was observed. In 1998, there were 134 blood and body fluid exposures. 110 (82%) were due to PIs. Hollow bore devices accounted for 82 (75%) of the PIs. Physicians and nurses still accounted for the majority of exposures but there was a significant increase seen in physicians. 93 source patients had known HIV serostatus or were asked to consent to HIV testing. 12 (13%) were infected with HIV. 107 HCWs (97%) were offered, 47 (44%) began, and 21 (45%) completed PEP. Underreporting surveys were done to ensure that there was adequate reporting of PIs among a broad spectrum of HCWs. Since the introduction of the PEP program, the reported number of exposures to blood and body fluids has increased, but the rate of PIs has trended downward. Hollow bore devices continue to account for a majority of PIs and a significant increase was seen among physicians. There was also a significant increase in HCWs offered and beginning PEP, but a decrease in those completing PEP. In the past this was due to side effects of the medications used for PEP, but now is due to improved source patient follow-up with a significant decrease in source patients documented to be infected with HIV.

The impact of a Rapid HIV Test to Limit Unnecessary Post Exposure Prophylaxis Following Occupational Exposures. Veeder AV, McErlean M, Putnam K, Caldwell WC, Venezia RA. Albany Med Ctr, Albany, NY.

Post exposure prophylaxis (PEP) is recommended for healthcare workers (HCWs) following high risk occupational exposure. Since the toxicity and side effects of PEP are significant, timely HIV results on the source patient are essential to limit days on PEP when the source is HIV negative. In 1999, a rapid HIV test (SUDS®, MUREX) was introduced in an effort to limit unnecessary PEP. Our purpose was to compare the duration and cost of PEP between the Enzyme Immunoassay (EIS) and the HIV rapid test. The average time until results were available in our institution was 4 days for EIA and 1 day for the rapid test. The data on occupational injuries were obtained from the National Surveillance System for Hospital Health Care Workers (NaSH). From Jan 1-Oct 31, 1999, 180 HCWs reported exposures to blood or other body fluids. For the purposes of this study, HCWs were excluded if the source patients were known HIV positive, could not be identified, or consent to test source patients could not be obtained. Forty-two (42) HCWs (23% of all reported exposures) were placed on PEP pending source patient HIV results. The 26 HCWs whose source patients were tested with EIA stayed on PEP a total of 101 days (median 4 days, range 1-8). Eleven (11) stopped PEP prior to HIV results due to side effects. The average cost per HCW, including cost of test and drugs, was $123. The 16 HCWs whose source patients were tested using the rapid test remained on PEP a total of 23 days (median 1 day; range 1-3). Only 2 HCWs stopped PEP in the first 24 hours due to side effects. The average cost of test and drugs per HCW for these patients was $69. Based on 42 HCWs requiring PEP during the first 10 months of 1999, we estimate annual institutional savings of $2,700 if the rapid test is used for all source patient testing.

Postexposure Prophylaxis Use Among Health-Care Workers Who Were Exposed to HIV-Negative Source Patients. SE Critchley*, PU Srivastava, SR Campbell, DM Cardo and the NaSH Surveillance Group. Centers for Disease Control and Prevention, Atlanta, GA.

The U.S. Public Health Service recommends the use of antiretroviral drugs after certain occupational exposures to human immunodeficiency virus (HIV). To assess the use of postexposure prophylaxis (PEP) by health-care workers (HCWs) who were exposed to HIV-negative source patients (SPs), we analyzed data collected on occupational exposures to blood/body fluids reported from 21 National Surveillance System for Health-Care Workers (NaSH) hospitals. From June 1995 through September 1999, 1142 HCWs from these hospitals initiated PEP following a blood/body fluid exposure. Information on PEP usage was available for 405 HCWs who had an exposure to an HIV-negative SP. The types of exposures sustained by these HCWs were 368 (91%) percutaneous injuries (PIs), 25 (6%) mucous membrane exposures, 9 (2%) skin exposures, and 3 (1%) bites. PEP regimens taken were as follows: a single drug, 35 (9%); 2-drug combinations, 221 (55%); and 3-drug combinations, 149 (37%). Use a of a particular regimen did not vary by the type of exposure. The duration of PEP regimens taken by HCWs ranged from 1-43 days: 291 (72%) for 1-5 days; 72 (18%), 6-10 days; 22 (5%), 11-19 days; and 20 (5%), 20 days. The duration of PEP was not influenced by the type of exposure sustained by HCWs who took PEP < 20 days. The only type of exposure sustained by workers who took PEP 20 days were PI (19) or bite (1). Information on symptoms was available for 51 HCWs who took PEP and reported one or more symptoms. The most commonly reported symptoms were fatigue or malaise, 21%; nausea, 20%; emotional distress, 12%; and headache, 9%. These findings suggest that strategies such as use of a rapid HIV antibody assay, SP evaluation for risk of HIV infection, and follow-up counseling could improve the management of exposed HCWs.

Hepatitis C Virus Infection After Occupational Exposures SR CAMPBELL*, P SRIVASTAVA, I WILLIAMS, M ALTER, D CARDO, and the NATIONAL SURVEILLANCE SYSTEM FOR HEALTH CARE WORKERS (NaSH) SURVEILLANCE GROUP. Centers for Disease Control and Prevention, Atlanta, Georgia, USA.

Occupational transmission of hepatitis C virus (Hepatitis C Virus) is a continuing concern for health care workers (HCWs). We describe exposures to Hepatitis C Virus sustained by HCWs and infections resulting from those exposures within 24 hospitals participating in the NaSH Surveillance Group. From 6/95 to 2/99, 5,538 exposures to blood/body fluids were reported; 524 (9%) involved a source infected with Hepatitis C Virus (154 [29%] were co-infected with human immunodeficiency virus [HIV], 43 [8%] had unknown HIV serostatus). Of 524 exposures to Hepatitis C Virus, 435 (83%) involved blood or bloody fluids; 341 were percutaneous and 94 were mucocutaneous exposures. HCW follow-up rates were low: 187 (43%) completed only 3 months of follow-up and 122 (28%) completed 6 months of follow-up. Five HCWs became anti-Hepatitis C Virus positive after a percutaneous exposure, and none become positive after a mucocutaneous exposure; all five infected HCWs became anti-Hepatitis C Virus positive within 6 months of exposure. Hepatitis C Virus RNA was detected in all five HCWs; two were tested 4 weeks after exposure and both were Hepatitis C Virus RNA positive. ALT elevation was observed in all five HCWs (median peak ALT=870). In four, the elevation was noted at the time of the first positive Hepatitis C Virus RNA test, and in one it was noted before a positive test was obtained. Signs/symptoms of acute viral hepatitis were reported for three of the five HCWs. Devices involved in transmission were 4 hollow-bore needles used for venous access and 1 scalpel blade. Four of the five HCWs were exposed to sources co-infected with HIV; all four took two or three HIV post-exposure prophylaxis drugs for 14-28 days. One of the four was HIV positive 13 months after exposure, but was HIV negative at 6 months. HCWs are at risk of acquiring Hepatitis C Virus infection after occupational exposure. Exposures to source patients co-infected with HIV and Hepatitis C Virus require further study.

Tuberculosis Contact Investigations in Seven NaSH Hospitals. PR Robertson*, E McCray, AL Panlilio, DM Cardo, PU Srivastava, and the NaSH Surveillance Group, Centers for Disease Control and Prevention, Atlanta, GA, USA

Despite heightened awareness of tuberculosis (TB), exposures of health care workers (HCWs) and patients continue to occur. To characterize TB contact investigations (CIs), we analyzed information from the National Surveillance System for Health Care Workers (NaSH). From February 1996 to August 1999, eight NaSH hospitals reported 39 CIs (1-12 per hospital). The median duration of exposure was 3 days (range 1-163 days). The mean number of HCWs exposed per CI was 36 (range 2-280); two CIs also involved 135 patients combined. Sources of exposure were both infectious TB patients and an infectious HCW. The sites of infection were pulmonary in 35 (90%); larynx and skin in one each (5%); and missing in two (5%). Of those with test results available, 84% (31/37) had positive sputum smears for acid-fast bacilli; 97% (36/37), positive cultures for |Mycobacterium tuberculosis|; and 14% (5/37), drug-resistant TB (streptomycin [2], isoniazid [2], and ethambutol [1]). Reasons for exposures were reported for 28 CIs; 16 occurred because patients with TB were asymptomatic or had symptoms that were unrecognized or not recorded and seven occurred because either TB isolation was not ordered or respiratory protection was not used properly. Among 328 HCWs tested following exposure, 158 (48%) received a single (follow-up) tuberculin skin test (TST), 82 (25%) received a baseline and follow-up TST, 88 (27%) received a TST less than 12 weeks after the reported exposure date. In addition, 70 patients with exposures had results reported for two TSTs. TST conversions were documented for three HCWs, and one patient with baseline and follow-up TSTs. All HCWs were offered preventive therapy; two accepted, and the other declined. As TB admissions fall, nosocomial exposures to and transmissions of TB still occur, highlighting the importance of rapid identification, isolation, diagnostic evaluation, and treatment of persons likely to have TB.

Detection and Prevention of Influenza in Health Care Workers. MJ Kuehnert*, CB Bridges, RM Strikas, PS McKibben, SR Campbell, K Fukuda, DM Cardo, and the NaSH Surveillance Group, CDC, Atlanta, GA.

Influenza outbreaks in hospitals often affect health care workers (HCWs), and infected HCWs have been implicated as important vectors of influenza transmission to patients. Guidelines for influenza infection control in health care facilities recommend HCW vaccination and outbreak investigation to reduce transmission. To assess influenza vaccination rates, we surveyed 24,736 HCWs from seven hospitals participating in the National Surveillance System for Health Care Workers (NaSH) were surveyed in 1996 or 1997. Overall, 6,903 (27.9%) were vaccinated (range 19.6-44.0%). Physicians or physician assistants were most likely and technicians or clerical staff least likely to be vaccinated (40 vs 22%, p<0.001). To assess institutional practices regarding influenza surveillance, we conducted a survey during a NaSH training course in 1999. Representatives from 34 hospitals located in 20 states and the District of Columbia (mean bed size 433 beds, range 120-1,120) were surveyed. Although 17 (50%) participants reported that rapid diagnostic testing was available at their facility, only 9 (27%) routinely conducted exposure investigations when influenza was suspected; availability of rapid testing was associated with investigation (p=0.05). Reasons given for not conducting investigations included lack of awareness that influenza was a significant problem, lack of expertise for investigation, disease reporting not required, or logistic difficulty (e.g., lack of staff, time, or resources). Few NaSH hospitals surveyed have policies for either surveillance or epidemic control of influenza, and adherence to recommendations for HCW vaccination is poor. Additional guidance and improved dissemination of existing information are needed for effective implementation of influenza prevention measures in acute-care facilities.