Nosocomial and community-acquired infection rates of patients treated by prehospital advanced life support compared with other admitted patients

Transcription

1 American Journal of Emergency Medicine (2011) 29, Original Contribution Nosocomial and community-acquired infection rates of patients treated by prehospital advanced life support compared with other admitted patients Scott M. Alter BS, EMT-P a, Mark A. Merlin DO, EMT-P b,c, a Department of Medical Education, University of Medicine and Dentistry of New Jersey-Robert Wood Johnson Medical School, Piscataway, NJ 08854, USA b Department of Emergency Medicine, University of Medicine and Dentistry of New Jersey-Robert Wood Johnson Medical School, New Brunswick, NJ 08901, USA c Emergency Department, Robert Wood Johnson University Hospital, New Brunswick, NJ 08903, USA Received 25 April 2009; revised 22 July 2009; accepted 23 July 2009 Abstract Objectives: Nosocomial infections are a large burden to both patients and health care organizations, causing hospitals to take measures in an attempt to reduce microorganism transmission. Patients treated by emergency medical services are one population that has not been studied regarding infection rates. This study examines admitted patients treated by advanced life support (ALS) and their likelihood of having community-acquired and nosocomial infections. Methods: A retrospective cohort study was conducted of admitted patients between 2003 and Subjects identified as having either community-acquired or nosocomial infections were grouped based on infection type and ALS treatment. The proportion of infected patients among total hospital admissions in each of these groups was calculated and compared using odds ratios (ORs). Results: A total of 5418 patients had at least 1 infection while admitted (3653 nosocomial, 1765 community). The probability of an ALS patient getting a nosocomial infection was 3.20% versus 2.28% for non-als patients (OR, 1.42; 95% confidence interval [CI], ). There was no significant difference in community-acquired infections between ALS and non ALS-treated groups (1.22% vs 1.14%; OR, 1.08; 95% CI, ). Conclusions: Despite having similar rates of community-acquired infections, patients admitted after ALS treatment had significantly greater risk for nosocomial infections. Because causality is not established, it remains unknown whether paramedic interventions contributed to the increased rate. Quite possibly, these patients are more susceptible to virulent organisms; however, prospective research is needed to identify causal relationships. Thus, treatment by ALS can be used as an identifier of patients at an increased risk of acquiring nosocomial infections Elsevier Inc. All rights reserved. This study received no grants or financial support. It has not been presented at any meeting or prior submission to another journal. No authors have conflicts of interest. Dr Merlin has an American Heart Association Grant that is not related to the study. Corresponding author. Department of Emergency Medicine, UMDNJ-Robert Wood Johnson Medical School, New Brunswick, NJ 08901, USA. Tel.: address: merlinma@umdnj.edu (M.A. Merlin) /$ see front matter 2011 Elsevier Inc. All rights reserved. doi: /j.ajem

2 58 S.M. Alter, M.A. Merlin 1. Introduction 1.1. Background Infection prevention has become a priority for health care organizations because nosocomial infections are a large burden to patients and health care organizations alike. Infected patients require a prolonged hospital stay with possible life-threatening complications, which costs health care organizations billions of dollars each year [1]. Since 2007, the Centers for Medicare and Medicaid Services no longer pays for nosocomial infection-related care, so the burden of cost is now on the hospitals [2]. Although different types of infections have different criterion to be considered nosocomial, these infections typically appear in patients after 48 hours of hospitalization [3]. Nationwide, in an attempt to reduce transmission of microorganisms, hospitals have adopted guidelines from the Centers for Disease Control and Prevention [4-12]. However, there have been no comprehensive studies examining the impact of emergency medical services (EMS) on the rates of infection. Previous research has shown the presence of bacteria in ambulances that cause nosocomial infections but has not examined the pathogens' effects on patients [13]. One study noted an increased number of phlebitis cases associated with prehospital intravenous (IV) cannulation and a significantly higher infection rate during a 2-month period than that for IVs started in the emergency department (ED) [14]. In contrast, another study found that there was no significant difference in infection rate between prehospital or in-hospital initiated IVs [15]. To elucidate this discrepancy and examine the impact on other infection sites, the present study provides additional information regarding the infection rates of patients treated in the prehospital setting by paramedics Importance Identification of patients at an increased risk for nosocomial infections is important for infection control and proper admission location. There are currently no guidelines to predict which patients are of greatest risk, and any strategy that can be deployed would be useful, starting in the ED. Reducing nosocomial infections in a known high-risk population not only would be beneficial to the patients' health but also would result in savings to hospitals Goals of this investigation This study determines the prevalence of nosocomial and community-acquired infections in admitted patients treated by advanced life support (ALS), in comparison to patients not treated by ALS. Our hypothesis is that patients treated by ALS have a higher rate of nosocomial and communityacquired infections. 2. Methods 2.1. Study design The study is a retrospective cohort. It has been approved by our university's institutional review board, which has a subcontract with our hospital Setting This study was conducted in an urban setting at a level I trauma center. The county population of approximately residents is made up of 68.4% white, 13.9% Asian, 13.6% Hispanic, and 9.1% African American residents. The county occupies 323 square miles with a combination of urban cities with high-crime regions and suburban communities with multiple nursing homes. In the county, there are 5 acute care hospitals and 3 ALS-providing agencies. The EMS system is 2-tiered, comprised a combination of paid and volunteer basic life support (BLS) units and paid hospital-based ALS units that contain 2 paramedics each. In our hospital, there are 8 BLS units and 6 ALS units that respond to approximately dispatches per year, 6500 of these being patients treated by ALS Selection of participants All patients admitted to our hospital between January 1, 2003, and December 31, 2007, as determined by queries of our hospital's patient information system, Sunrise Clinical Manager (SCM; Eclipsys Corporation, Atlanta, Ga), were selected for inclusion in this study Method of measurement Every day, the hospital laboratory's culture results are sent to the infection control department, which maintains a Microsoft Access database of infections. All cases are examined by the infection control team, which determines the infection type (nosocomial or community-acquired) and source (invasive procedure site, lower respiratory tract, primary bloodstream, pneumonia, surgery-related, surgical site, urinary tract, or other), according to Centers for Disease Control and Prevention criteria [3]. Additional demographic information contained in the database includes the patient's name, birth date, and admission date. Because each infection is entered as a separate record, the database was consolidated by patient admission, and new fields were created for the number of infections by type and source. Using this new, condensed database, patients having nosocomial or community-acquired infections were identified. A separate Access database, maintained by our hospital's EMS department, was used to identify patients who had been treated by our hospital's ALS units, transported to our

3 Nosocomial infection and patients with prehospital ALS hospital, and subsequently admitted. To identify patients with infections who had been treated by our hospital's ALS units, the infection data were cross-matched with the EMS database, based on first name, last name, date of birth, and date of call/date of admission. Records were also matched when the date of admission was 1 day ahead of the date of call because patients arriving to the hospital at night may not be admitted until the next day. For records that only exactly matched on 1 or 2 of first name, last name, and date of birth, information was more closely examined for possible discrepancies in spelling or date entry. In the rare instance of a nonexact match without a clear typo, SCM was consulted to determine whether there were 2 patients admitted with similar identifying information or if the slightly different records were, in fact, regarding the same person. Because patients' ED charts often lack data regarding method of arrival, we were unable to identify patients transported to our hospital by other hospitals' ALS units. Data from our hospital's EMS department only identify patients treated by our hospital's ALS units. Therefore, we needed to account for patients that arrive at our hospital after treatment by other hospital's ALS units. To do this, we obtained data from the regional communications center, which dispatches all of the ALS units for each of the hospitals in the county. From a database of county-wide ALS unit dispatches, a query was run to retrieve a breakdown of transports to our hospital by transporting ALS agency. This provided the fraction of all ALS patients transported to our hospital who were treated by other hospitals' paramedic units Outcome measures Nosocomial and community-acquired infection rates were calculated by dividing the number of patients with infections by the total number of patients treated by ALS and patients not treated by ALS Primary data analysis Fig. 1 shows the study design as an Euler diagram (not to scale). The total number of hospital admissions was determined from an SCM query. Patients with infections were identified from the infection control department's Access database. Patients transported by our hospital's ALS came from the EMS department's Access database. Using these 2 databases, the patients treated by our hospital's ALS and had infections were identified. The total number of patients transported to our hospital by other hospitals' ALS was determined based on dispatch records for all of the ALS agencies in the county. Because we only had access to our own hospital's ALS records, we were unable to identify individual patients transported to our hospital by other hospitals' paramedic units. Therefore, we were unable to determine the exact fraction of patients treated by other hospitals' ALS who were admitted to our hospital and those who had infections, Fig. 1 Euler diagram of patients transported to our hospital (figure not to scale; shading represents estimated intersection). represented by the shaded intersection in Fig. 1. To estimate these figures, we assumed that the number of patients treated by other hospitals' ALS, admitted to our hospital, and had infections was proportional to the number of patients treated by our hospital's ALS, admitted to our hospital, and had infections. Using our hospital's ALS and admissions records, we first determined the percentage of patients admitted to our hospital who were treated by our own ALS units. Then, the percentage of all ALS patients brought to our hospital by our hospital's ALS units was determined from dispatch queries. With this information, the total number of patients admitted to our hospital after being treated by any ALS service was estimated. This was obtained by dividing the number of patients treated by our own ALS service by the fraction of patients our ALS service transported to our hospital out of those who were treated by any hospital's ALS service. Using the same fraction of patients treated by our hospital's ALS units out of patients treated by all ALS units, the total number of ALS-treated patients with infections (nosocomial and community-acquired separately) was calculated. Then, the percentage of admitted ALS-treated patients with nosocomial and community-acquired infections was determined. Next, the rate of infections among admitted patients who were not treated by ALS was computed. The differences between the total number of patients admitted to our hospital and ALS-treated patients admitted to our hospital, as well as the total number of patients with infections and ALS-treated patients with infections were calculated. These numbers were then divided to arrive at the percentage of admitted patients not treated by ALS who had nosocomial and communityacquired infections. Using the fractions of patients with infections treated by ALS and not treated by ALS, the odds 59

4 60 S.M. Alter, M.A. Merlin Table 1 Baseline characteristics of admitted patients by transporting ALS agency ratio was calculated for both nosocomial and communityacquired infections Sensitivity analyses Because we were unable to identify patients transported to our hospital by other ALS agencies, we used dispatch data to approximate the rates of infections. In the primary data analysis, we estimated that ALS patients were equally likely to have an infection whether transported by our paramedics or another ALS agency. The fraction of total ALS patients transported by our ALS units was used to calculate the odds ratios in the main analysis. To account for possible bias, a sensitivity analysis was performed to determine alternative odds ratios for nosocomial and community-acquired infections. The percentage of ALS patients with infections who were treated by our own agency out of all ALS-treated patients was varied in increments of 10% from 20% to 100%, and the odds ratios were calculated. An additional sensitivity analysis was performed, keeping the percentage of patients transported by our ALS units constant and varying the infection rates of patients transported by other ALS services in increments of 0.05% from 0% to 4%. 3. Results All patients Our hospital's ALS 3.1. Characterization of study participants Not our hospital's ALS N Age, y, (26.92) (20.04) (26.98) mean (SD) Sex, % male 49.28% 53.88% 48.87% Age and sex baseline characteristics do not include data from 2003, due to the unavailability of admissions statistics. From 2003 to 2007, patients were admitted to our hospital. Of these admissions, patients were treated by our hospital's ALS units (Table 1). In the same time, there were patients treated by all the regional ALS providers combined and transported to our hospital. Of these, (81%) were treated by our hospital's ALS. Estimating that the admission rate for other hospitals' ALS units is the same as that of our own hospitals' ALS units, the approximate number of admissions of ALS patients from all providers would be Thus, the number of admissions not treated by ALS would be During the same 5-year period, of all admitted patients, 3653 had nosocomial-acquired infections and 1765 had community-acquired infections. During the matching process of the EMS database to the infection database, there were several records that only matched some, but not all fields due to typos and errors such as reversal of first and last names. Only 2 records of patients with similar names required reconciliation via SCM to determine whether the patients were the same. Table 2 shows the baseline characteristics of these admitted patients with infections, comparing those transported by our hospital's ALS with patients not transported by our ALS, which includes both patients brought in by other hospitals' ALS units and patients not treated by ALS at all. Infections were also categorized by source (invasive procedure site, lower respiratory tract, primary bloodstream, pneumonia, surgery related, surgical site, urinary tract, or other). The percentage of infections by source for nosocomial and community-acquired infections is shown in Figs. 2 and 3, respectively Main results Considering that the number of infected patients treated by ALS would be composed of and proportional to the percentage of patients transported to our hospital by ALS agency, there would be 459 patients with nosocomial infections and 175 patients with community-acquired infections who were transported by all ALS agencies combined. This leaves 3194 patients with nosocomial infections and 1590 patients with community-acquired infections who were not treated by any ALS providers. Table 2 Characteristics of admitted patients with infections by transporting ALS agency All patients Our hospital's ALS Not our hospital's ALS Nosocomial N Age, y, mean (SD) (24.02) (18.34) (24.37) Infections, mean no. (SD) 1.37 (0.92) 1.47 (0.98) 1.36 (0.91) Community N Age, years, mean (SD) (25.64) (17.56) (25.21) Infections, mean no. (SD) 1.01 (0.10) 1.01 (0.12) 1.01 (0.10) Not our hospital's ALS includes both patients brought in by other hospitals' ALS units and patients not treated by ALS at all.

5 Nosocomial infection and patients with prehospital ALS 61 Fig. 2 Classification of nosocomial infection sources by ALS treatment. Using the number of infected patients and the total number of hospital admissions, infection rates were calculated (Table 3). The rate of patients with nosocomial infections was 3.20% for ALS-treated patients and 2.28% for non ALS-treated patients. The odds ratio was 1.42 (95% confidence interval [CI], ). The rate of community-acquired infections was 1.22% and 1.14%, for ALS and non-als patients, respectively, with an odds ratio of 1.08 (95% CI, ) Sensitivity analyses When varying the fraction of infected patients transported to our hospital by our own ALS agency compared with ALS patients treated by all hospitals' units, the odds ratio for nosocomial infections always remains significant (Fig. 4). In the range immediately around the estimated 81% of ALS patients transported by our own agency, the odds ratio for nosocomial infections ranges from 1.43 (95% CI, ) Fig. 3 Classification of community-acquired infection sources by ALS treatment.

6 62 S.M. Alter, M.A. Merlin Table 3 Infection rates of admitted patients by ALS treatment ALS Non-ALS Odds ratio (95% CI) Nosocomial 3.20% 2.28% 1.42 ( ) Community 1.22% 1.14% 1.08 ( ) at 70% to 1.41 (95% CI, ) at 90%. The odds ratio for community-acquired infections remains not significant for all percentages above 20%; however, percentages less than 70% are extremely unlikely to reflect reality. Another analysis varying the infection rates of patients treated by other ALS agencies, assuming 81% of ALS patients transported by our own agency, is shown in Fig. 5. For all percentages of patients with nosocomial infections above 0.05%, the odds ratio is significant. It is unlikely that the nosocomial infection rate of patients transported by other ALS agencies would be this low, considering the main results that 3.20% of ALS patients acquire nosocomial infections and 2.28% of non-als patients acquire nosocomial infections. Odds ratios of patients with communityacquired infections remain insignificant for all percentages less than 1.69%. The actual rate of community-acquired infections for patients treated by other agencies' ALS is unlikely to be this high, as the rates of infections found in the main results were only 1.22% and 1.14% for ALS and non ALS-treated patients, respectively. 4. Limitations Several limitations exist in our study. Patients transported to our hospital via other hospitals' ALS agencies could not be directly linked to our infection database. Therefore, we could only estimate the admission data based on ALS dispatches. To account for potential bias, a sensitivity analysis was performed, which revealed little variation in odds ratios. However, had these patients been linked to and extracted from the non-als patient groups, more accurate baseline characteristics and main results could have been calculated. Another limitation of the study was the accuracy of entered data. The information obtained from the Access databases was copied from other sources and was not directly linked to the patients' hospital charts. Some discrepancies were identified when matching patients between databases, including errors in spelling and dates. The data were thoroughly reviewed to identify these errors, but the possibility that not all were detected exists. Any potential missed cross-references would have resulted in higher percentages of infections in the non-als groups and lower rates in the ALS groups. Therefore, the presented nosocomial odds ratio may be slightly lower than the actual ratio. Finally, because this study was conducted retrospectively, no cause and effect relationships could be established. Although an association was found between patient treatment by ALS and nosocomial infections, this may be due to confounding variables. It is possible that these findings are the result of age-related effects. 5. Discussion All patients admitted to the hospital had the same rate of community-acquired infections, despite treatment by ALS. Although there is no intuitive reason for this finding, Fig. 4 Sensitivity analysis, varying the percentage of admitted ALS patients transported by our ALS agency.

7 Nosocomial infection and patients with prehospital ALS 63 Fig. 5 Sensitivity analysis, varying the infection rate of patients transported by other ALS agencies. perhaps some infected patients wait until they are very sick and require the assistance of EMS for transport to the hospital. Other patients may seek medical care earlier in the infection process and drive themselves to the ED. Rarely, patients may visit their primary care physicians first and, subsequently, be directly admitted to the hospital. These types of patients are all grouped together when considering community-acquired infections. Variables including time since exposure to the pathogen and progression of the disease are difficult to quantify because they all occur before medical attention is sought. When examining patients with nosocomial infections, ALS-treated patients admitted to the hospital had a significantly greater infection rate than patients admitted by other means. This may be due to several possible deficiencies, including the sterility of the prehospital environment, infection control practices of paramedics during procedures, and immune system of patients. It is possible that locations outside the hospital are less sterile and may result in an increased rate of infection during invasive procedures. Conversely, there may be a decreased risk of infection during procedures because paramedics only treat 1 patient at a time. Since the environment is usually void of other sick individuals, the risk of spreading organisms directly between patients is nonexistent. Another possibility is that patients treated by paramedics are sicker than patients not treated by paramedics. These individuals may be immunocompromised and thus more susceptible to inhospital organisms, likely to be of increased virulence. In reality, a combination of these factors probably all play a role in infection rates. When comparing baseline characteristics of patients by whether or not they were treated by our hospital's ALS, patients receiving ALS treatment tended to be older. There may be several potential reasons for this including older people relying EMS for transportation and having more chronic medical problems than younger people do. Age alone may be a confounding variable as to why ALS patients have a higher rate of nosocomial infections. However, if this were the case, one would expect ALS patients to have a greater rate of community infections, which they did not have. When examining the sources of nosocomial and community-acquired infections, there was not much difference whether or not patients were treated by our hospital's ALS. However, the largest differences were in pneumonias patients who had either nosocomial or community-acquired pneumonia were more likely to have been treated by our hospital's ALS. This may be related to the typical types of illnesses ALS treats. Most ALS calls are cardiac or respiratory in nature. Paramedics may be less likely to treat patients with other infection sources. Patients transported only by BLS were not included in this study due to unavailability of data and lack of generalizability of the potential results. Emergency department dispositions of BLS patients are not recorded in the EMS records, so it would be impossible to know which patients were admitted. Even if this information were obtained by other hospital information systems, the results would not be externally valid for several reasons. Although the primary coverage area of our ALS service is most of the county, our BLS service only primarily covers one city. In addition, the ALS service of our hospital is similar to all other ALS services in the state hospital based and staffed by 2 paramedics; however, BLS services vary from town to town. Our hospital-based BLS service is staffed by 2 paid emergency medical technician basics, whereas other

8 64 S.M. Alter, M.A. Merlin services in our catchment area are extremely heterogeneous. These services may be paid, volunteer, or a combination, or may be hospital based, municipal based, fire based, or private transport, and staffing may consist of 0 to 5 emergency medical technicians. All of these factors would make any analyses of patients treated by our BLS service not representative of the patients admitted to our hospital. Because this study does not establish causality, it remains unknown whether the increased nosocomial infection rate of ALS-treated patients is due to the environment of treatment, actions of paramedics, or patient population. Future research should include prospective studies examining rates of infections caused by specific interventions that are performed by paramedics. When the source of a nosocomial infection is identified, any procedure that may have caused the infection should be identified and logged in a standard manner. If procedures are recorded, in addition to the infection source, then the cause of the infection can be isolated. This information could then be used to determine whether ALS patients have a greater rate of nosocomial infections due to actions of the paramedics. Patients treated by ALS are probably more susceptible to infections; however, this is not a modifiable factor when a patient presents to the hospital for medical treatment. Therefore, care should be taken by prehospital providers to ensure their work environment is as clean as possible and to maintain sterile technique when performing invasive procedures. Once patients have been admitted to the hospital, treatment by ALS should be used as a potential identifier of patients at an increased risk of acquiring nosocomial infections. References [1] Scott II RD. The direct medical costs of healthcare-associated infections in U.S. hospitals and the benefits of prevention. Atlanta (Ga): Centers for Disease Control and Prevention; [2] Rosenthal MD. Nonpayment for performance? Medicare's new reimbursement rule. NEJM 2007;357: [3] Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health care associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008;36: [4] Rutala WA, Weber DJ. Healthcare Infection Control Practices Advisory Committee. Guideline for disinfection and sterilization in healthcare facilities, Atlanta (Ga): Centers for Disease Control and Prevention; [5] Siegel JD, Rhinehart E, Jackson M, Chiarello L. Healthcare Infection Control Practices Advisory Committee. Guideline for isolation precautions: preventing transmission of infectious agents in healthcare settings Atlanta (Ga): Centers for Disease Control and Prevention; [6] Siegel JD, Rhinehart E, Jackson M, Chiarello L. Healthcare Infection Control Practices Advisory Committee. Management of multidrugresistant organisms in health care settings, Am J Infect Control 2007;35:S [7] Tablan OC, Anderson LJ, Besser R, Bridges C, Hajjeh R, CDC, Healthcare Infection Control Practices Advisory Committee. Guidelines for preventing health-care associated pneumonia, 2003: recommendations of CDC and the Healthcare Infection Control Practices Advisory Committee. MMWR Recomm Rep 2004;53(RR-3):1-36. [8] Sehulster L, Chinn RY, CDC, HICPAC. Guidelines for environmental infection control in health-care facilities. Recommendations of CDC and the Healthcare Infection Control Practices Advisory Committee (HICPAC). MMWR Recomm Rep 2003;52(RR-10):1-42. [9] Boyce JM, Pittet D, Healthcare Infection Control Practices Advisory Committee, HICPAC/SHEA/APIC/IDSA Hand HygieneTask Force. Guideline for hand hygiene in health-care settings. Recommendations of the Healthcare Infection Control Practices Advisory Committee and the HICPAC/SHEA/APIC/IDSA Hand Hygiene Task Force. MMWR Recomm Rep 2002;51(RR-16):1-45. [10] O'Grady NP, Alexander M, Dellinger EP, Gerberding JL, Heard SO, Maki DG, et al. Guidelines for the prevention of intravascular catheter-related infections. MMWR Recomm Rep 2002;51(RR-10): [11] Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline for prevention of surgical site infection, Infect Control Hosp Epidemiol 1999;20: [12] Wong ES. Guideline for prevention of catheter-associated urinary tract infections. Am J Infect Control 1983;11: [13] Alves DW, Bissell RA. Bacterial pathogens in ambulances: results of unannounced sample collection. Prehosp Emerg Care 2008;12: [14] Lawrence DW, Lauro AJ. Complications from i.v. therapy: results from field-started and emergency department-started i.v.'s compared. Ann Emerg Med 1988;17: [15] Levine R, Spaite DW, Valenzuela TD, Criss EA, Wright AL, Meislin HW. Comparison of clinically significant infection rates among prehospital-versus in-hospital-initiated i.v. lines. Ann Emerg Med 1995;25:502-6.