Objective. To determine the cost-effectiveness and cost-benefit of an infection control program to reduce nosocomial respiratory syncytial virus (RSV) transmission in a large pediatric hospital.
Design. RSV nosocomial infection (NI) was studied for 8 years, before and after intervention with a targeted infection control program. The cost-effectiveness of the intervention was calculated, and cost-benefit was estimated by a case–control comparison.
Setting. Children's Hospital of Philadelphia, a 304-bed pediatric hospital.
Patients. All inpatients with RSV infection, both community- and hospital-acquired.
Intervention. Consisted of early recognition of patients with respiratory symptoms, confirmation of RSV infection by laboratory testing, establishing cohorts of patients and nursing staff, gown and glove barrier precautions, and monitoring and education of staff.
Outcome Measures. The incidence density of RSV NI before and after the intervention was calculated as the rate per 1000 patient days-at-risk for infection. Intervention costs included laboratory testing, isolation, and administration of the program. The cost of RSV NI was estimated by comparing hospital charges for 30 cases and matched uninfected controls.
Results. A total of 148 patients acquired NI (88 before and 60 after the intervention). The Mantel-Haenszel stratified relative risk for NI in the period before the infection control program, compared with the postintervention period, was .61 (95% confidence interval: .53–.69). By applying the preintervention stratum-specific rates of infection to the days-at-risk in the postintervention period, an estimated 100 NIs would have been expected, which in comparison to the 60 NIs observed, yielded an estimated program effectiveness of 10 RSV NIs prevented per season. The total cost of the program per season was $15 627 or $1563/NI prevented. In comparison, the mean cost to the hospital was $9419/case of RSV NI, resulting in a cost-benefit ratio of 1:6.
Conclusions. A targeted infection control intervention was cost-effective in reducing the rate of RSV NI. For every dollar spent on the program, approximately $6 was saved.
Respiratory syncytial virus (RSV) infection, the most important cause of bronchiolitis and viral pneumonia in young children, causes an estimated 90 000 hospitalizations and 4500 deaths each year from lower respiratory tract disease.1 Numerous RSV-infected children occupy pediatric hospitals during the winter peak months of infection, when RSV is the leading cause of nosocomial respiratory illness. Transmission rates approach at least 45% of contacts in the absence of infection control measures.2Because RSV infection does not induce long-term immunity, repeated symptomatic and asymptomatic infections may be acquired in both adults and children.3 The virus is spread directly by large-particle droplets or by contaminated secretions carried on infected patients, health care workers, or fomites.3,4Transmission by aerosolized droplets is minimal. Infection control interventions suggested to reduce nosocomial RSV have included isolation of patients in single-bed rooms5; assigning cohorts of patients to the same rooms6–8 or assigning cohorts of nursing staff to care for infected patients8–10; contact isolation techniques such as gowns,8–13 gloves,8,9,11,13 eye–nose goggles,14 and masks15; admission screening7,13; and visitor restrictions.8,9However, it is unclear which combination of these interventions is most effective, and hospital infection control practices vary widely.11 The broad elements of a successful infection control program are likely to be early recognition of patients with RSV infection, contact isolation precautions, and compliance with these precautions by health care workers.5
In the United States, bloodstream nosocomial infections (NIs), urinary tract infections, and pneumonia are monitored by the National Nosocomial Infection Surveillance program. A national program to monitor the incidence of nosocomial RSV infection does not exist. Efficacy and cost analyses of NI control practices in adult patients have been performed,16–21 including the original nationwide Study on the Efficacy of Nosocomial Infection Control, which established the basis for understanding the economic impact of NI.16 However, the economic impact of reducing nosocomial RSV infection has not previously been assessed. Although reducing NIs may be one of the only proven methods for reducing resource utilization while improving patient care, it is increasingly important to demonstrate that such interventions accomplish protection of the patient and health care worker in a cost-effective manner.22,23
This study was undertaken as a quality improvement initiative of Children's Hospital of Philadelphia, with the goal of reducing the hospital-wide incidence of RSV NI. A targeted infection control program that utilized multiple strategies to reduce RSV transmission was implemented simultaneously throughout the hospital, and its impact was assessed over the subsequent four RSV seasons. The success of the intervention allowed us to estimate both the cost-effectiveness and the cost-benefit of preventing nosocomial RSV in this setting.
Study Population and Setting
All patients admitted to Children's Hospital of Philadelphia, a 304-bed pediatric hospital, for 8 consecutive RSV seasons from 1988–1989 to 1995–1996 were included in the study. During this time, the hospital consisted of general medical and surgical wards, and neonatal, cardiothoracic, and general intensive care units. In most wards the rooms had 4 to 6 patients per room; few single-bed rooms existed. The only changes to the hospital wards during this time consisted of the merger of 2 neonatal units and the separation of the cardiothoracic from the general intensive care unit in the final year of the study. The characteristics of patients admitted to the hospital remained consistent throughout the study period. Specifically, the percentage of medical versus surgical admissions did not change significantly (medical:surgical ratio was 1.8 in the preintervention period and 2.2 in the later period), the percentage of patients with medical assistance remained constant (18.5% vs 19.4%), and the percentage of patients transferred to the hospital remained the same. In addition, the percentage occupancy of hospital beds did not differ significantly throughout the study period (83% occupancy in the preintervention period, compared with 78% in the later period).
The RSV season was defined as the time from the first hospitalized case of laboratory-confirmed RSV infection, usually early November, through the end of April (∼6 months). Patients with RSV infection were prospectively identified throughout the study period by tracking laboratory specimens. Nasopharyngeal aspirate specimens were collected from all patients who had symptoms of a viral upper or lower respiratory tract infection (including rhinorrhea, cough, wheeze, tachypnea, or apnea)3 either on admission or during the course of their hospital stay.5 Testing for RSV was performed by enzyme immunoassay (EIA [Abbott Laboratories, Abbott Park, IL]) in the hospital's Clinical Virology Laboratory with a same-day turnover time, 7 days per week. For the purposes of the infection control intervention, EIA for RSV was performed as the screening test for identification of RSV cases and had a sensitivity of >90%.24 Conventional viral culture of nasopharyngeal aspirate specimens was also performed occasionally; however, <10% of all RSV infections were diagnosed by this method. Testing for RSV infection by viral culture was not mandated by the intervention because the longer test turnaround time (5–14 days) rarely impacted decisions regarding isolation and cohorting. Children of any age who had symptoms compatible with RSV infection and a positive RSV test result were considered as RSV cases and were included in the study. Because the average incubation period for RSV is 5 days, community-acquired RSV infection was defined as the development of symptoms within the first 5 days of admission.3 RSV infection was considered nosocomial when symptoms developed on or after the sixth hospital day. Patients were also considered to have RSV NI if they were readmitted with RSV within 5 days after discharge from the hospital (when the first admission was for an unrelated illness). Postdischarge surveillance was not performed.
RSV Infection Control Intervention
At the beginning of the 1992–1993 RSV season, a quality improvement intervention was launched with the aim of decreasing the incidence of hospital-acquired RSV infection. Before the intervention, methods used for prevention of RSV NI varied among patient care units. Screening of patients for RSV infection occurred; however, the use of barrier methods for isolation and cohorting of patients and nursing staff was inconsistent.
The interdisciplinary intervention consisted of the following components:
Formal education of nursing, medical, and paramedical staff on the epidemiology and control of RSV infection, before and throughout each RSV season.
High index of suspicion in suspected cases of RSV infection, with laboratory confirmation of infection in all patients with respiratory symptoms.
Contact precautions for all patients with symptoms of viral respiratory tract infection, consisting of handwashing before and after contact and the use of gloves and cotton cover gown by all staff for any physical interaction with a patient or their environment. Neither masks nor eye–nose goggles were used. Gown and glove use for visitors was not required.
Maintenance of contact precautions for 2 weeks for all patients with confirmed RSV infection (or for the duration of respiratory symptoms if RSV testing was negative and no other cause for their respiratory illness was found).
Placement of patients in single-bed rooms, when available, or cohorting of patients on RSV contact precautions to the same rooms.
Cohorting nursing staff to care for isolated patients.
Discouraging staff with symptoms of an acute viral respiratory tract infection from caring for intensive care unit and immunocompromised patients and encouraging them to wear a mask only to prevent nose/mouth contact when caring for other patients. RSV testing was not performed on staff members.5,25
Restriction of visits by family members with acute respiratory symptoms.
Regular surveillance by infection control staff to monitor compliance, verbal feedback to nursing unit managers in the event of a nosocomial RSV case, and monthly written performance reports to all wards.
The risk of acquiring RSV NI is related to the intensity of exposure of susceptible patients to all hospitalized patients shedding RSV.13 Because both children and adults may repeatedly acquire RSV infections, all uninfected children hospitalized during the RSV season were considered susceptible to RSV NI.3,13Because all infected hospitalized patients can act as reservoirs for transmission of the virus, and because our objective was to decrease the incidence of RSV NI throughout the hospital, we implemented and assessed the intervention simultaneously in all patient care areas.
To account for variations in intensity of exposure to the virus, 5 exposure strata were calculated, based on the proportion of all patient hospital days in each month of the study period that were accounted for by patients shedding RSV (using the total number of patients with a diagnosis of RSV infection, International Classification of Diseases, Ninth Revision, Clinical Modification[ICD-9-CM] 466.1 or 480.1, and the estimate that each patient shed virus for 7 days). The total number of patient days-at-risk in each month was calculated from records of hospital admissions. Because RSV NI acquired in the last 4 days of hospitalization would not have been symptomatic during that stay (and thus not detected by our in-hospital surveillance system), all but the last 4 days of the admission of uninfected patients were counted in the number of days-at-risk. For patients who acquired RSV NI, hospital days occurring subsequent to acquiring the infection were excluded from the days-at-risk. All study months were then grouped into 1 of the 5 exposure strata, and incidence density rates of RSV NI were calculated for each exposure stratum as the rate of NI per 1000 patient days-at-risk.
Incidence rates of nosocomial RSV infection were compared before and after the intervention using crude relative risk and Mantel-Haenszel stratified relative risk. To estimate the number of RSV NI prevented, preintervention stratum-specific rates of infection were applied to the number of days-at-risk in the postintervention period.
The cost of the infection control intervention per season was calculated from the perspective of the hospital, and all costs were estimated in US dollars based on the currency value in 1996. The 3 major cost components of the infection control intervention were gown and gloves for contact isolation, additional RSV testing performed because of the intervention, and administrative costs.
The cost of each use of a reusable cotton gown included the purchase price and laundering cost. The mean number of gowns and gloves used per RSV patient day was obtained after observation of 10 isolated patients over a 24-hour period in various wards, and was multiplied by the number of isolated patient days per season. These observations were made during a season in the postintervention period (1996), when the intervention and compliance with gown and glove use was well established. However, because some utilization of gowns and gloves was made in the preintervention period, only a proportion of the observed gown and glove use was attributed to the intervention. In a study by Leclair et al,13 compliance by staff with wearing gowns and gloves increased from 38.5% of patient contacts to 95% of patient contacts after an intervention comparable to ours. We assumed that a similar 60% increase in gown and glove utilization occurred in our study setting and adjusted the cost accordingly. In addition, estimates that assumed either 40% or 80% increase in utilization of gloves and gowns were evaluated for sensitivity analysis.
The total number of RSV EIAs performed was recorded per season, and the cost of each test calculated from the purchase price of the test kit, labor costs for laboratory personnel, and hospital overheads. Because the intervention mandated RSV testing for all symptomatic patients, additional RSV tests may have been performed in the postintervention period because of the infection control program. To estimate the number of intervention-related RSV tests, we assumed that ordering of RSV tests by physicians was proportional to the number of patients with the clinical diagnosis of bronchiolitis. The number of tests ordered in the pre- and postintervention periods was compared with the number of cases of bronchiolitis (by ICD-9 coding). RSV tests performed in the postintervention period that exceeded the number of tests performed in the preintervention period (per diagnosis of bronchiolitis) were attributed to the infection control program. Sensitivity analysis was performed using 2 additional assumptions regarding test utilization. The first analysis assumed that physician ordering of RSV tests was proportional to the number of patients with a positive test (that is, the rate of positive tests remained constant over time). Thus, only additional tests beyond the number expected after adjusting for the observed increase in the rate of test positivity were attributed to the intervention. The second analysis assumed that the absolute number of RSV tests would otherwise have remained constant, so that all additional tests in the postintervention period were attributed to the program. The cost of testing for RSV by viral culture was not included in the economic analysis because viral culture was not mandated by the infection control intervention and had minimal impact on decisions to isolate and cohort patients.
The administrative costs of the program consisted of the cost of salaried time dedicated by infection control and nursing staff and the cost of administrative supplies (which were estimated directly). The estimated increase in time directed toward the intervention, compared with the staff's preintervention activity, was multiplied by each participant's full-time salary with benefits and adjusted to 1996 dollars.
The financial burden of a RSV NI was estimated by comparing the hospital charges of 30 randomly selected RSV NI cases with 30 matched inpatients who did not have RSV infection (controls). Controls were chosen from computer-generated lists of patients who were admitted during the same RSV season as the patients with RSV NI, had the same principal discharge diagnosis by ICD-9-CM coding, had the same number of secondary diagnoses (where possible), and were the same approximate age as the case patients. To account for the risk of acquiring RSV NI, a control was selected only if he or she was hospitalized for at least as many days as it took the corresponding case patient to develop their RSV NI. When multiple controls existed, the patient with a length of stay (LOS) closest to the median LOS for the group was selected. RSV NI and control patients were excluded if their hospital stay was >180 days (ie, spanning 2 RSV seasons). Hospital charges were adjusted to 1996 dollars using the hospital and related services price index of the Health Care Financing Administration. The mean difference in hospital charges between the 2 groups was attributed to RSV NI. The difference in hospital charges was divided by the hospital charge to cost ratio in 1996, to obtain an estimate of the cost of a RSV NI. Cost-effectiveness was calculated by dividing the estimated number of RSV NIs prevented by the estimated cost of the infection control intervention. Comparing the cost per RSV NI to the estimated cost of preventing an infection derived the cost-benefit ratio of the intervention.
Rates of RSV NI Before and After the Intervention
Overall, in the 4 RSV seasons before the intervention, 88 RSV NIs occurred in a total of 90 174 patient days-at-risk. In the 4 seasons after the intervention, 60 RSV NIs occurred in a total of 82 196 patient days-at-risk. The total number of patients hospitalized with community-acquired RSV was 1604 in the preintervention period, compared with 2065 patients in the later period. The median age of the patients who acquired RSV NI was 1 year. Fifty-two percent of the RSV NI patients had preexisting illnesses that predispose to greater morbidity from RSV infection (20%, congenital heart disease; 17%, chronic lung disease; 10%, immunodeficiency; and 5%, prematurity). The number of patients with underlying illnesses did not differ significantly between the pre- and postintervention periods (51% vs 54%, respectively).
The number of RSV NIs that occurred in each of the 5 risk strata is shown in Fig 1. The stratum-specific rate of RSV NI increased nearly linearly as the RSV infection risk increased, in both the pre- and postintervention periods. This is consistent with the assumption that the incidence of RSV NI depends on the number of hospitalized patients shedding virus at a given time. Comparison of rates of infection before and after the intervention showed a decrease in the incidence of RSV NI across all 5 risk strata. The nosocomial RSV infection rate in the preintervention period was .098 (or .98 cases per 1000 hospital days-at-risk) compared with .073 (or .73 cases per 1000 hospital days at risk) in the postintervention period.
The crude relative risk of acquiring a RSV NI in the postintervention period compared with the earlier period was .75 (95% confidence interval: .54–1.04). The Mantel-Haenszel stratified relative risk was .61 (95% confidence interval: .53–.69). This represents a statistically significant reduction of 39% in the rate of RSV NI in the post- compared with the preintervention period. By applying the preintervention rates of infection to the number of days-at-risk in the postintervention period, 40 cases of RSV NI, or 10 infections per season, were estimated to have been prevented by the program (Table 1).
Cost of RSV Infection Control Intervention
The costs of the components of the RSV infection control intervention are shown in Table 2. Each use of a cotton cover gown was estimated to cost $.39, based on a purchase price of $3.83 per gown, assuming each gown was laundered 65 times (at a cost of $.25 per laundering). The cost of 1 glove was $.038. The total estimated cost of contact isolation was $11 094 per season (sensitivity analysis range: $7396–$14 791).
The mean increase in personnel and administrative costs attributed to the intervention was $2022/RSV season (Table 2). In the preintervention period, the salary with benefits of the infection control nurse was $21.42/hour ($1996), and the time dedicated to RSV infection control was ∼8 hours per week for each of the 4 seasons (assuming 24 weeks/season). The infection control intervention was implemented and maintained by a registered nurse and the medical director of the infection control department, whose salaries with benefits were $22/hour and $60/hour ($1996), respectively. In the first 2 seasons, ∼16 hours/week of infection control nursing time and 1 hour/week of medical director time were directed toward the intervention. In the subsequent 2 seasons, the time required by the infection control nurse for the RSV program activities was decreased to 4 hours/week. The infrastructure had been well established, allowing the nurse's time to focus on program support and maintenance. During this period, medical director time was negligible. No additional nursing staff was required for cohorting patients, and additional salaried nursing time was not required for education or other activities (which were incorporated into routine nursing meetings).
The cost of a RSV EIA test was $19.17, which included the test-kit cost of $9.14 per test, and labor and overhead costs. The number of RSV tests performed increased from a mean of 2850 tests per season in the first 4 seasons to 3255 tests per season in the postintervention period, and the number of positive EIA tests increased by 19%, from a mean of 368 per season to 432 per season. There was a 9.6% increase in the number of patients admitted with bronchiolitis in the postintervention period. After adjusting for the increase in bronchiolitis, 131 additional RSV EIA tests per season were attributed to the intervention at a cost of $2511 per season. When the assumption was made that the rate of RSV tests ordered should remain the same, compared with the number of positive tests, the resulting cost estimate was $0 per season. Alternatively, if all the extra RSV tests in the postintervention period were attributed to the program, the cost was $7764 per season (Table 2).
The total cost of the intervention was $15 627 per season (sensitivity analysis range: $9418–$24 577). Because 10 RSV infections were prevented by the intervention per season, the cost of preventing a single RSV NI was estimated at $1563.
Cost of Nosocomial RSV Infection by Case–Control Comparison
Thirty RSV NI patients were randomly selected from all 8 RSV seasons in the study and matched with 30 control patients (Table 3). The mean time to acquire RSV NI was 14.9 days. The mean age of the 30 RSV NI patients was 1.8 years, compared with 2.0 years for control patients (and 2.1 years for all RSV NI patients throughout the study period). The mean excess in hospital charges between the RSV NI patients and controls was $45 335 per patient, after adjustment for inflation. The mean difference in LOS was an additional 10.7 days per RSV NI patient. However, 2 of the patient–control pairs had differences in LOS (50 and 54 days, respectively) that exceeded 39 days, the longest median duration of stay for high-risk patients in a study of community-acquired RSV.26 This may have overestimated the differences in hospital charges. After excluding these 2 patients, the mean difference in hospital charges for the remaining 28 case–control pairs was $20 721, with a mean difference in LOS of 7.8 days. Further elimination of patients based on differences in LOS did not alter the results of cost comparison between the 2 groups but reduced the mean difference in LOS (Table 3).
Using the 1996 hospital charge to cost ratio, and the conservative estimate that the excess hospital charge for a nosocomial RSV infection was $20 721 per patient, the cost to the hospital was estimated to be $9419 per infection. Because the estimated cost of preventing a RSV NI was $1563, the cost-benefit ratio of the infection control intervention was 1:6. For every dollar spent on this RSV infection control measure, an estimated $6.00 was saved.
The incidence of RSV NI throughout our hospital was reduced 39% by a targeted infection control intervention. This outcome is comparable to the 37% reduction in RSV NI achieved in a study by Leclair et al,13 in an infant ward over a shorter period. Accurate comparison of RSV NI rates in this study was achieved by calculating the incidence density of NI, using risk strata that accounted for the number of patients excreting RSV and the number of days of exposure of susceptible patients. This method is preferable to other techniques, such as the RSV nosocomial ratio, to determine rates of NI for external reporting.11,13 Other studies have analyzed various measures to decrease nosocomial RSV and have shown statistically significant reductions in transmission rates; however, differences in study methods preclude direct comparison.6–9,14,15
To our knowledge, this study is the first to determine the cost-effectiveness of a hospital-based infection control program to reduce nosocomial RSV transmission. An estimated 10 RSV infections were prevented per RSV season, resulting in a cost-effectiveness of $1563 per infection prevented (sensitivity analysis range: $942–$2458). There are limited data available on the cost of infection control programs in general, and no published values for comparison of the cost of RSV prevention programs. However, in an analysis of the economics of hospital infections, Wenzel21 liberally estimated the cost of preventing a NI in an adult hospital to be $2100 per infection, an estimate comparable to that found in this study. We did not address the relative cost-effectiveness of individual components of the intervention. The most recent Centers for Disease Control and Prevention guidelines for the prevention of RSV NI recommend handwashing and wearing gloves and a gown for patient contact, with the additional measures of isolation of patients and cohorting of staff and patients suggested as prudent.5 Guidelines in place until 1994 suggested the use of gowns (but not gloves) and single-room isolation, which is often impractical in hospitals with multibed rooms. Because our hospital had very few single-bed rooms over the entire study period, it seems that a significant reduction in RSV transmission rates can occur without isolating patients to single rooms.
Using a case–control comparison, we determined the LOS attributed to RSV NI to be 7.8 days (sensitivity analysis: 3.5–10.7 days). In a study of RSV NI in Canadian pediatric hospitals, the median LOS attributable to RSV NI was double that for community-acquired illness (10 vs 5 days, respectively).11 Children with RSV NI were significantly more likely than patients with community-acquired RSV to have preexisting factors for severe disease (66% vs 17%, respectively).11 Similarly, we found that 52% of patients who acquired RSV NI had preexisting conditions for severe RSV infection (congenital heart disease, chronic lung disease, prematurity, or immunocompromise). Such patients are known to have increased morbidity and mortality from RSV infection, with LOS for community-acquired illness ranging from 5 to 39 days.26
The mean cost of a RSV NI was estimated by our analysis to be $9419 (range: $9249–$20 721), resulting in a cost-benefit ratio of 1:6 for the intervention. In an analysis of the cost-effectiveness of RSV prophylactic agents in preterm infants, the mean cost of hospitalization for RSV infection was $8502 (range: $5207–$13 518).27 Our analysis was limited to determining direct medical costs.22,23,28 We did not account for additional monetary factors, such as delayed return to work of caregivers, long-term disability, loss of life, or nonmonetary outcomes such as quality of life savings, patient satisfaction, legal considerations, and negative publicity.22,28 Such factors would have likely enhanced the cost-benefit ratio of 1:6.
Use of a case–control comparison is considered a valid method to determine the cost of NIs; however, it is subject to limitations.22 Matching patients by first diagnosis alone greatly overestimates costs. In this analysis we accounted for differences in severity of illness by matching, using the number of patient discharge diagnoses.22 We also sought to control for the risk of acquiring RSV NI, by selecting uninfected patients who were hospitalized for at least as long as the case patients took to acquire their infection. Controversy exists over whether to divide charges by departmental charge-cost ratios or the overall hospital ratio (because patient's utilization of different services and departmental charge-cost ratios may vary widely).20,22 In this study, we chose to use the overall hospital charge-cost ratio because the patients studied were drawn from throughout the hospital, we had a moderately large sample size, and a prospective study comparing the 2 methods yielded little difference in outcome.20 However, it is important to note that dividing by the charge-cost ratio estimates total costs (both fixed and marginal costs), which may inflate the estimated cost of a RSV NI, because the fixed costs of care may not be affected by extended hospital stays.21,22
This analysis was limited to assessing the impact of the infection control intervention across the entire hospital inpatient population, consistent with our goal of decreasing hospital-wide rates of RSV NI. It is likely that there was a variability of effect of the intervention, not reflected in the overall risk reduction estimate provided. Because RSV infection causes more serious illness in infants and patients with preexisting conditions such as chronic lung disease, heart disease, and prematurity, analysis of the intervention's impact based on patient age or underlying diagnosis would likely have enhanced the results of this study. In addition, it would have been preferable to examine the changes in rates of RSV NI over time by a time-series analysis. This was problematic because of the great deal of variability in RSV disease activity and the relatively small number of NIs, making year-to-year fluctuations difficult to interpret. Data from a reliable time-series analysis may have added to establishing a causal relationship between the intervention and the reduction in rates of RSV NI. Despite our attempts to control for numerous factors affecting fluctuations in nosocomial transmission rates of RSV infection, factors other than the intervention itself may have been responsible for the reduction in RSV NI seen in this study.
From the perspective of a hospital, it has been estimated that only between 1% and 5% of the costs of treating NIs are reimbursed under a prospective payment system based on diagnostic-related groups.17 In this study at least 95% of the estimated $7856 savings per RSV NI prevented may have represented a financial gain to the hospital. Studies of the economic impact of preventing RSV NI should continue as newly available RSV prophylactic agents are increasingly used in high-risk patients.27
We gratefully thank Michael J. Barbella and the staff of Health Information Management, and the Admissions Department of Children's Hospital of Philadelphia for providing assistance with medical records and hospital admission information.
We also thank all infection control personnel who collected data during the study years, and the staff of the Clinical Virology Laboratory for their technical contributions.
- Received November 23, 1999.
- Accepted February 1, 2000.
Reprint requests to (K.K.M.) Division of Immunologic and Infectious Diseases, Abramson Bldg, 12th Floor, 34th St and Civic Center Blvd, Philadelphia, PA 19104. E-mail:
- RSV =
- respiratory syncytial virus •
- NI =
- nosocomial infection •
- EIA =
- enzyme immunoassay •
- ICD-9-CM =
- International Classification of Diseases, Ninth Revision, Clinical Modification •
- LOS =
- length of stay
- Collins PL, McIntosh K, Chanock RM. Respiratory syncytial virus. In: Fields BN, Knipe DM, Howley PM, et al, eds. Fields Virology. 3rd ed. Philadelphia, PA: Lippincott-Raven Publishers; 1996:1313–1351
- Centers for Disease Control and Prevention. Guidelines for prevention of nosocomial pneumonia. MMWR Morb Mortal Wkly Rep. 1997;46(RR-1):1–79
- Isaacs D,
- Dickinson H,
- O'Callaghan C,
- Sheaves R,
- Winter A,
- Moxon ER
- Langley JM,
- LeBlanc JC,
- Wang EEL,
- et al.
- Haley RW,
- Culver DH,
- White JW,
- et al.
- Welliver RC
- Bolyard EA,
- Tablan OC,
- Williams WW,
- Pearson ML,
- Shapiro CN,
- Deitchman SD,
- and the Hosptial Infection Control Practices Advisory Committee
- Joffe S,
- Ray GT,
- Escobar GJ,
- Black SB,
- Lieu TA
- Copyright © 2000 American Academy of Pediatrics