HIGH RATE OF HAND CONTAMINATION AND LOW RATE OF HAND WASHING BEFORE INFANT CONTACT IN A NEONATAL INTENSIVE CARE UNIT

Abstract

Nosocomial infections are a monumental problem in the hospital setting. They result in 5.7 infections for every 100 hospital admissions, totaling ≈2 million infections per year, and are the direct cause of death in at least 80 000 persons a year in the US.1-3 Infections acquired in an intensive care unit (ICU) account for >20% of all the nosocomial infections acquired in the US even though ICUs make up only 5% of hospital beds.4 Nosocomial infection rates in a neonatal intensive care unit (NICU) are as high as 25%, and this is frequently attributed to the immunocompromised nature of the usual NICU patient.5 The treatment of hospital-associated nosocomial infections in the US adds billions of dollars to health care costs annually.6 Guidelines from the CDC clearly state that hand washing should occur before and after patient contact, particularly for compromised patients in an intensive care setting.7-9 In this study we use a new technique to evaluate hand washing compliance. In the process of documenting compliance with this technique we observed the potentially serious problem of hand contamination, defined as touching of self, another patient, another person or a high risk object, by staff and parents before contact with the patients. Hand contamination has been observed previously10 but has not described in the detail which this study does. Methods. The study site was a Level III NICU which serves as a referral center for a multistate region. Institutional Review Board approval for the study was obtained before commencing data collection. To record hand washing compliance data, a video camera was hung from the ceiling in one corner of the unit with the ostensible purpose of recording traffic patterns for optimal design of a new NICU. The camera was connected to a video recorder two floors above so that there was no need for any of the study staff to enter the NICU to change tapes. The camera recorded staff and visitor contact with many different beds and sinks in the NICU but provided unobstructed views of three patient beds. Only these three beds were observed for data collection. The patients in these beds changed during the data collection period as did their diagnoses, level of intervention, need for ventilatory support, etc. Hand washing compliance data were collected on random nursing shifts (all times of day) for a 2-month period. The videotapes were analyzed by a nurse research assistant who had previously worked in the NICU and was aware of the professional status of the various staff members and parents who were recorded. The recorded data were viewed looking at the patient contact episodes first and then observing the tape in reverse for hand washing compliance and/or hand contamination episodes. In this manner we were able to record precisely which staff member or visitor had contact with a patient and what activities occurred immediately before the contact. Data were analyzed using observations of four groups: nurses, physicians, respiratory therapists (RT), and parents. The activity that immediately preceded patient contacts was categorized using a predefined set of definitions: (1) Washed hands: hands were washed at one of the sinks in the nursery using water and a towel dry. Soap did not have to be used to be considered an adequate hand washing. (2) Contaminated hands (occurring with or without prior wash): (a) touched another baby (hand contact with a different patient directly preceded contact with the patient under study); (b) touched self (The staff member or visitor in question touched him/herself above the neck (nose, mouth, hair, etc.) or below the waist with his or her hands before contact with the patient. These areas of the body were believed to be high risk for contamination with organisms that could potentially cause nosocomial infections.); (c) touched another person: hand contact with another person in the nursery preceded contact with the patient; (d) touched high risk object (hand contact with an object considered to be at high risk of contamination with nosocomial pathogens, e.g. a suction unit, waste receptacle, dirty diaper, etc.). Episodes not associated with one of the above events (e.g. the caretaker entered from an unmonitored room) were excluded. Events preceding the defining event were not further evaluated. Results. Two hundred sixty-seven contacts with patients were reviewed. These included 186 nurse contacts, 22 physician contacts, 44 RT contacts and 15 parent contacts. The hand washing compliance (percent of patient contacts when hand washing directly preceded contact) for the various groups (see Table 1) was observed for: nurses 24.7%; physicians 31.8%; RTs 20.5%; and parents 6.6%. Hand washing percentages were not significantly different by chi square analysis (P = 0.31, 3 degrees of freedom). Hand contamination occurred before 76.4% of patient contacts (see Table 1) including 75.3% of nurse contacts, 68.2% of physician contacts, 79.5% of RT contacts and 93.4% of parent contacts. In all four groups the most common type (91.7%) of hand contamination event was touching of self. For nurses and RTs touching another baby was the second most common form of self-contamination. Discussion. Albert and Condie11 in one of the first studies of its type used an on-site observer to document hand washing compliance in an adult ICU setting. They found that physicians in a university hospital complied with hand washing recommendations 28% of the time compared with nurses (43%), respiratory therapists (76%) and radiology technicians (44%). The direct observation technique has become the standard method to record hand washing compliance data. Using an observer Crow and Greene12 found that 66.7% of anesthesiologists failed to wash their hands before a case. Donowitz10 showed in a pediatric ICU setting that physicians had “breaks” (failed to wash their hands after contact with a patient or support equipment and before contact with another patient or departure from the unit) in their hand washing technique 79% of the time compared with 63% for nurses, 62% for occupational therapists/physical therapists, 78% for RTs and 78% for radiology technicians. In a Brazilian NICU DeCarvalho et al.13 demonstrated that 15% of physicians and 32% of nurses did not wash before patient contact and 56% of physicians and 49% of nurses did not wash after patient contact. Raju and Kobler14 assessed hand washing compliance in an NICU and found that 37.5% of doctors and 53.9% of nurses demonstrated appropriate hand washing before handling an infant. In a medical-surgical ICU Preston et al.15 recorded hand washing rates of 16% in an open unit and 30% in isolation rooms. The video camera technique we used provided several advantages in recording hand washing compliance compared with the direct observation method. The direct observation method makes it difficult to watch the hands of a staff member or visitor without having the observer be an integral part of the clinical environment. Alternatively with the video camera technique data can be recorded unobtrusively, and hand washing activity can be reviewed multiple times, as necessary, for accurate determination of the events that occurred. By recording the data on video tape, both hand washing compliance and subsequent events before patient contact can be examined. A report by Larson16 reviews the literature on hand washing practices and points out the need for accurate methods of measuring hand washing compliance. In previous studies that attempted to increase the amount of hand washing as a way to decrease the rate of nosocomial infections, either no measure of the hand washing was done,17-19 a surrogate measure other than the amount of hand washing was used20, 21 or in several recent studies, the direct observation method was used to determine compliance.22-24 As an alternative the video recorded data gathering and review of the video tape in reverse data analysis technique used in this study is an accurate method by which to measure hand washing compliance and other related activities. The American Medical Association recently adopted a resolution that calls for reminding physicians to wash their hands before and after examining patients. Additionally the CDC's Guidelines for Handwashing and Hospital Environmental Control clearly states, “Personnel should always wash their hands before taking care of particularly susceptible patients, such as those who are severely immunocompromised and newborns.”8 The poor hand washing compliance recorded in this study is similar to that seen in previous studies.10-16, 24-28 Given the high cost, morbidity and mortality and emotional and physical suffering related to nosocomial infections, it is clear that there continues to be a need for aggressive educational and infection control programs to promote optimal hand washing compliance. The hand contamination described in this study may negate the cleansing effect of hand washing or increase the nosocomial pathogen load on the hands of non-hand washers. The high rate of hand contamination episodes documented in our study (76% of the time preceding patient contact) may increase the risk of nosocomial infection as the flora of the caretaker or parent's skin, nares, perineum, etc., other patients and the surface of high risk objects may harbor organisms frequently associated with nosocomial contagion (e.g. Staphylococcus aureus, coagulase-negative Staphylococcus, Enterobacteriaceae, Candida albicans and others). Infection control training in the future, particularly in intensive care settings, should be directed not only to increasing hand washing compliance but also to decreasing hand contamination before patient contact. Acknowledgment. Funding for this project was provided by a grant from The Children's Hospital Research Institute, The Children's Hospital, Denver, CO. Jeffrey Brown, M.D., M.P.H.; Ann Froese-Fretz, R.N., M.S.; Dennis Luckey, Ph.D.; James K. Todd, M.D. Department of Epidemiology The Children's Hospital of Denver (JB, AF, DL, JKT) Departments of Pediatrics, Microbiology, and Preventive Medicine (JB, JKT) University of Colorado School of Medicine The Children's Hospital Research Institute (JB, JKT) Denver, CO