Contact isolation in surgical patients: A barrier to care?

Contact isolation in surgical patients: A barrier to care? Heather L. Evans, MD, Mary M. Shaffer, BS, Michael G. Hughes, MD, Robert L. Smith, MD, Tae W. Chong, MD, Daniel P. Raymond, MD, Shawn J. Pelletier, MD, Timothy L. Pruett, MD, and Robert G. Sawyer, MD, Charlottesville, Va

Background. Contact isolation is commonly used to prevent transmission of resistant organisms. We hypothesized that contact isolation negatively impacts the amount of direct patient care. Methods. For 2 hours per day over a 5-week period, a single observer recorded provider/patient contact in adjacent isolated and nonisolated patient rooms on both the surgical intensive care unit (ICU) and surgical wards of a university hospital. Number of visits, contact time, and compliance with isolation were recorded, as was illness severity as assessed by APACHE II score. Results. Isolated patients were visited fewer times than nonisolated patients (5.3 vs 10.9 visits/h, P < .0001) and had less contact time overall (29 ± 5 vs 37 ± 3 min/h, P = .008), in the ICU (41 ± 10 vs 47 ± 5 min/h, P = .03), and on the floor (17 ± 3 vs 28 ± 4 min/h, P = .039), in spite of higher mean APACHE II scores in the isolated (10.1 ± 1.0 vs 7.6 ± 0.8, P = .05). Among floor patients with APACHE II scores greater than 10, patients in the isolated group had nearly 40% less contact time per hour than patients in the nonisolated group (19 ± 4 vs 34 ± 7 min/h, P = .05). Conclusion. Because of the significantly lower contact time observed, particularly among the most severely ill of floor patients, we propose a reexamination of the risk-benefit ratio of this infection control method. (Surgery 2003;134:180-8.) From the Departments of Surgery and Internal Medicine, UVA Health System, Charlottesville, Va

THE PROBLEM OF ANTIBIOTIC RESISTANCE has reached near epidemic proportions in U.S. hospitals, necessitating aggressive, mandatory infection control programs. Antibiotic restriction, increased education, and new products to promote hand hygiene, as well as strict use of barrier and isolation methods, now comprise the front line in the war against transmission of resistant organisms. Although long criticized for lack of clinical evidence to support their use,1-3 particularly in units endemic with vancomycin-resistant enterococci (VRE),4-6 several recent studies report efficacy and cost-effectiveness in the use of isolation protocols to control resistance both in epidemic7,8 and endemic9-12 settings. Controversy surrounding the use of barrier precautions and patient source isolation emanates at least in part from the change in clinical workflow imposed by donning gloves, a gown, and somePresented at the 64th Annual Meeting of the Society of University Surgeons, Houston, Texas, February 12-15, 2003. Reprint requests: Heather L. Evans, MD, Surgical Infectious Disease Research Laboratory, Building MR-4, Room 3150, Lane Rd, Charlottesville, VA 22908-1380. © 2003 Mosby, Inc. All rights reserved. 0039-6060/2003/$30.00 + 0 doi:10.1067/msy.2003.222

180 SURGERY

times a mask, before initiating patient contact. There is evidence that these mandatory practices have impacted daily behavior, even to the point of erecting disincentives to enter a patient’s room,13 suggesting that patients most at risk for nosocomial illness14 may receive less-than-optimal care15 and that severity of illness has no bearing on the compliance with isolation precautions.16 However, a direct comparison of patient care time between patients in the isolated and nonisolated groups has not been performed. We elected to prospectively examine the clinical practice of isolation precautions on the surgical service in our hospital with the hypothesis that patients in the contact isolation group would be seen less often than their nonisolated counterparts. METHODS This investigation was composed of 3 parts: a prospective observation of patient/provider encounters, a questionnaire-based survey of a sample of patients in the isolated and nonisolated groups, and a retrospective review of the incidence of infection caused by a single resistant organism. The prospective observations were conducted over a 5-week period during June and July 2001 at the University of Virginia Health System, a 700-bed tertiary care hospital, on both the surgical wards and in the surgical

Surgery Volume 134, Number 2

intensive care unit (ICU). Because of the observational nature of this part of the study and the anonymous method of data collection, this portion of the investigation was granted exemption from informed consent by the Human Investigation Committee at the University of Virginia. For 2 hours daily an observer was positioned in the work area of either the surgical ICU or the surgical ward to view a pair of adjacent, private patient rooms, with and without isolation precautions. A single investigator (M.M.S.), who was not involved in direct patient care or decision-making regarding treatment, recorded the duration and nature of all health care provider entries into the designated patient rooms. Observations took place during Monday through Friday between 8:00 AM and 1:00 PM. Subjects were not informed of the nature of the observations, and the presence of the observer, a member of the health care system, was accepted on the hospital units without question. All observed patient rooms housed patients admitted to the general, transplantation, or trauma surgery services. Because there were a limited number of hospitalized isolated patients, on different days, some patient rooms and therefore some of the same patients were observed multiple times. Patient demographic data, comorbidities, and severity of illness as calculated by APACHE II score17 were recorded at the time of observation, as were the reason for and type of isolation. Compliance with designated isolation precautions was recorded, including whether the 5-foot radius surrounding the patient was penetrated, and the donning of required isolation clothing as appropriate. Hand washing was evaluated after the provider left the patient’s room. In the second part of the study, patients in both the isolation and nonisolation groups were interviewed to examine their subjective impressions regarding the care they had received during their hospitalization. After informed consent, patient interviews were conducted with a 16-item questionnaire created for this study and approved for use by the Human Investigation Committee. Patients were asked to recall the number of times per day they were visited by their physician, the average length of interaction, and to rate on a modified Likert scale the level of contact, and level of comfort communicating with their physician. Patients indicated whether physician interactions occurred while the physician stood in the doorway, stood at the bedside, or performed a physical examination, scoring 1 to 3, respectively. The recollections of nursing interactions were similarly recorded. In addition, patients who were isolated were asked to report their level of

Evans et al 181

understanding of the reason for isolation and the degree of change in the amount and quality of interaction after isolation, as well as their perception of change in their condition after isolation. Finally, as a measure of the efficacy of contact precautions, the incidence of gentamicin-resistant Enterococcus sp (GRE) among surgical patients was calculated for the 1-year periods before and after the discontinuation of contact precaution for GRE with hospital administrative databases. Definitions. At the University of Virginia Health System, hospital isolation precautions for antibioticresistant organisms are divided into contact precautions and contact-droplet precautions.18,19 Contact precautions require a private room, or cohorting with another patient on similar precautions. These precautions are commonly used for patients with documented infection or colonization with either VRE or Clostridium difficile. Gowns are donned before room entry if provider’s clothing will touch the patient or anything in the patient’s room. Providers are required to wear gloves on entering the room and wash their hands after gloves are removed before leaving the immediate environment. Acceptable hand washing includes use of alcohol-based hand wash solution (available outside all isolation rooms), or soap and water. Patients in the isolation group are given designated equipment (eg, thermometers, stethoscopes, and blood pressure cuffs) that does not leave their room. Transport of the patient is limited. Contact-droplet precautions differ only in that everyone within 5 feet of the patient must wear a mask, in addition to following contact precautions. Because of the risk of nasal carriage,8 methicillin-resistan Staphylococcus aureus requires contact-droplet precautions in our hospital, whereas all other antibiotic-resistant organisms require only contact precautions. Direct contact is defined as either physical contact with the patient or object in the patient’s room. In the case of contact-droplet precautions, entering the 5-foot radius around the patient was considered direct contact. Contact time was defined as the total amount of time a provider spent in a patient’s room, not including the time taken to don or remove appropriate isolation gear. Isolation compliance required gloves and hand washing for all encounters, gowns if direct contact occurred, and mask if within 5 feet of a patient in the contactdroplet-isolated group. Statistical analysis. Observations of patients in the isolation group were compared against those in the nonisolation group. The mean values of continuous variables with normal distributions were compared by use of Student’s t test; otherwise, the

182 Evans et al

Surgery August 2003

Table I. Demographics and comorbidities of observed patients Observations (n) Age Male Race=White APACHE II score* Transplantation Cardiac Hepatic Pulmonary Malignancy Diabetes Renal insufficiency Hemodialysis Mechanical ventilation Cellular transfusion

Isolated

Nonisolated

P value

48 47.8 ± 2.0 41 (85%) 45 (96%) 12.0 ± 1.0 23 (48%) 6 (13%) 20 (42%) 14 (29%) 15 (31%) 13 (27%) 26 (54%) 18 (38%) 12 (25%) 8 (17%)

48 58.3 ± 2.4 36 (75%) 44 (92%) 14.3 ± 1.0 13 (27%) 14 (29%) 11 (23%) 7 (15%) 17 (35%) 17 (35%) 6 (13%) 2 (4%) 5 (10%) 11 (23%)

— .0013 .20 .40 .12 .035 .04 .05 .08 .665 .38 <.0001 <.0001 .06 .44

*APACHE II Score calculated at the time of observation

Wilcoxon rank sum test was used. Categorical variables and proportions were compared by use of the χ-square test or Fisher’s exact test when individual cell counts were 5 or less. Contact time was tallied per observation period, as well as per encounter. It was possible to achieve more than 60 minutes of contact time per hour of observation, because multiple providers were sometimes present in patient rooms simultaneously. Subgroup analysis of contact time was also performed on the basis of location and APACHE II score. Provider compliance with isolation precautions and hand washing was evaluated for all isolation encounters, and contact time was compared between compliant and noncompliant encounters. RESULTS Twenty-four observations of paired isolation and nonisolation rooms were completed on the surgical floor, as well as 24 similar observations in the surgical ICU. During 5 observation sessions of nonisolated patient rooms and 5 sessions of isolated patient rooms, the full 2-hour observation period was truncated because of patient transport. Patients on isolation precautions were younger and as a group had more comorbid illnesses than the nonisolated patient sample, with the notable exception of more heart disease in the nonisolated patients (Table I). There was a trend toward higher severity of illness in the isolated group, as measured by APACHE II score at the time of observation (14 ± 1 vs 12 ± 1; P = .12). Subgroup analysis by location revealed no significant difference between the severity of illness of patients in the isolated ICU and nonisolated ICU groups (18.5 vs 16.5; P = .30), but patients in the isolation group

were found to have significantly higher APACHE II scores than patients in the nonisolation group (10.1 vs 7.6; P = .05). Among the observed isolation rooms, 65% housed patients colonized with a resistant organism; 58% were associated with resistant organism infection. Only 17% required isolation at the time of admission. Six percent were colonized with more than one organism, and 4% were infected with 2 resistant organisms. Although 15 (31%) patients simultaneously harbored more than 1 resistant organism, the most common organism prompting isolation was methicillin-resistan Staphylococcus aureus (69%), followed by VRE (40%), C difficile (19%), multidrug-resistant Acinetobacter spp (4%), and multidrug-resistant S maltophilia (4%). Consequently, most of the isolation rooms observed required contact-droplet precautions. A total of 485 patient/provider encounters were observed in 48 isolated patient rooms over 91.6 hours (5.3 encounters/h). There were 1002 encounters in 48 nonisolated patient rooms observed over 91.2 hours (10.9 encounters/h), significantly more per hour than in the isolation rooms (P < .0001); this was observed on both the surgical floor and in the ICU (Table II). Nurses were the most commonly observed provider in both isolation and nonisolation rooms, followed by patient care assistants (PCAs), physicians, and medical students. Although nonisolated encounters outnumbered isolated encounters 2:1, the proportion of encounters by provider type was not statistically different (Table III). Direct contact in isolation rooms could be assessed in 476 of 485 encounters, and occurred in 314 encounters (66%). Contact time was calcu-

Evans et al 183

Surgery Volume 134, Number 2

Table II. Encounters and contact time per hour of observation Encounters/h (no.) ICU encounters/h (no.) Floor encounters/h (no.) Contact time/h ICU contact time/h Floor contact time/h

Isolated

Nonisolated

P value

5.3 (485) 6.1 (319) 4.2 (166) 29.2 ± 5.3 41.5 ± 9.7 16.9 ± 2.9

10.9 (1002) 13.8 (658) 7.9 (344) 37.4 ± 3.4 47.0 ± 4.7 27.9 ± 4.3

<.0001 <.0001 <.0001 .0078 .039 .0379

Table III. Encounters by provider type (P = .92) Provider type Nurse PCA Physician Medical student

Isolated

Nonisolated

247 (51%) 184 (38%) 50 (10%) 4 (1%)

501 (50%) 394 (39%) 97 (10%) 10 (1%)

lated without regard to direct contact by summing the total time spent by each provider per room during each 2-hour observation. A difference in mean contact time was observed between isolated and non-isolated rooms overall (29 ± 5 vs 37 ± 3 min/h; P = .008). This was again demonstrated in subgroup analyses comparing isolated with nonisolated rooms in the ICU (42 ± 10 vs 47 ± 5 min/h; P = 0.03), as well as on the surgical wards (17 ± 3 vs 28 ± 4 min/h; P = .039), in spite of the fact that the patients in the isolated ward group had higher mean APACHE II scores at the time of observation (10.1 ± 1.1 vs 7.6 ± 0.8; P = .05). Simultaneous conditioning on location and severity of illness revealed that patients in isolation rooms on the surgical floor with APACHE II scores greater than ten had approximately 40% less contact time per hour (19 ± 4 vs 34 ± 7; P = .05) than their nonisolated counterparts (Fig 1). Overall, the contact time per encounter was on average 2 minutes longer in the isolation rooms. In subgroup analysis by provider type, PCAs spent significantly more time in isolation rooms per encounter, and there was a trend toward this in physician and medical student encounters (Table IV). Conditioning on location of encounter, it is apparent that the difference in mean contact time is almost entirely accounted for in the ICU encounters (Fig 2). In fact, on the floor, both physician and nurse contact time per encounter was shorter for patients in the isolation group than in the nonisolation group. Compliance with isolation precautions was evaluable in 441 encounters (91%). Two hundred fiftyone encounters (57%) exhibited noncompliance. The most common reason for noncompliance was

failure to cleanse the hands after exiting the patient room (215 encounters, 86% of noncompliant encounters), but a surprising number of providers did not use gloves appropriately (172 encounters, 69% of noncompliant encounters). Overall handwashing rates were 51%. The 2 medical students observed were the most compliant (100%), followed by physicians (57%), nurses (52%), and PCAs (47%). In comparing contact-droplet isolation with contact isolation, there was more noncompliance associated with contact-droplet isolation (182 [60%] vs 69 [50%]; P = .05) and less handwashing (137 [46%] vs 88 [63%]; P = .0007). Among all noncompliant encounters, mean contact time per encounter was significantly less than the compliant encounters (1.7 ± 0.2 vs 8.9 ± 0.8 min; P < .0001). There was a trend toward higher severity of illness in the compliant encounters (Table V). A total of 26 patients, 9 of which were (or had previously been) in the isolation group, were interviewed by use of the survey instrument created for this study; results are reported in Table VI. In comparing patients in the isolated versus nonisolated groups, there was no significant difference observed in the perception of care delivery, but there was a trend toward greater comfort interacting with nursing staff among patients in the nonisolated group. There was no significant difference between recollections of patients in the isolated and nonisolated groups of the number of physician visits per day (3.3 ± 0.4 vs 4.4 ± 0.7 visits; P = .46) or average length of visit (6.7 ± 1.5 vs 8.5 ± 1.4 minutes; P = .43). Patients in the nonisolation group recalled a substantially longer average physician visit time than was actually observed. Among the 9 patients in the isolation group interviewed, only amount of interaction was perceived as less during isolation, whereas quality of care was reported to be better (data not shown). Five (56%) patients reported no change in their condition after isolation; the remaining patients were evenly divided between improved (2 [22%]) and worsened (2 [22%]). GRE was removed from the list of organisms requiring contact isolation precautions by hospi-

184 Evans et al

Surgery August 2003

Fig 1. Comparison of mean contact time per hour of observation in isolated and nonisolated rooms, conditioned simultaneously on severity of illness and hospital unit. Bars indicate 95% confidence intervals.

Table IV. Mean length of encounter in minutes by provider type Provider type All Nurse PCA Physician Medical student

N

Isolated

Nonisolated

P value

147 748 578 147 14

5.5 ± 0.5 4.9 ± 0.7 5.9 ± 0.7 5.9 ± 1.6 17.8 ± 10.0

3.4 ±0.2 3.7 ± 0.3 3.0 ± 0.2 3.4 ± 0.7 5.2 ± 2.1

.012 .6625 .0002 .1535 .1367

tal infection control in March 2000. From March 1999 through February 2000, there were 1703 admissions to the surgical ICU and floors, and 18 episodes of GRE infection were observed in surgical patients. In the year after GRE contact precautions were no longer required, there were 1753 admissions and 12 episodes of GRE. Average length of stay was 6 days for both periods. There was a decrease in the incidence of GRE infection, but this did not reach statistical significance (1.1 vs 0.7 episodes/100 admissions, P = .32; 1.4 vs 0.8 episodes/1000 patient days, P = .24).

DISCUSSION In this prospective observational study, patients in the isolated group were seen less often by health care providers than patients in the nonisolated group, both in number of contacts and total contact time per hour, whether on the surgical floor or in the ICU. In a sense, these results are not unexpected, because isolation precautions are intentional barriers created to limit the spread of resistant organisms by limiting the contact to patients who harbor them. Indeed, the benefit of source isolation is not intended for the isolated

Evans et al 185

Surgery Volume 134, Number 2

Fig 2. Comparison of isolated with nonisolated mean contact time per encounter, stratified by location and provider type. Bars depict 95% confidence intervals. No encounters with medical students were observed on floor, so all medical student encounters are excluded from this analysis.

Table V. Compliance with isolation precautions Encounters (n = 485)* Mean contact time/encounter APACHE II score

Compliant

Noncompliant

P value

190 (39%) 8.9 ± 0.8 16.0 ± 0.5

251 (52%) 1.7 ± 0.2 14.9 ± 0.4

— <.0001 .11

*Forty-four encounters could not be assessed for compliance.

patient, but rather other patients in the environment, who may be at risk for hospital-acquired infections. But at what cost? Examination of our data reveals that the patients in the isolation group were more immunosuppressed, with more comorbidities, and needed more invasive critical care support, as evidenced by the higher incidence of mechanical ventilation and hemodialysis. These factors may in fact reflect a need for more contact time relative to their nonisolated neighbors. Although this study was not designed to assess the outcomes of individual patients, the observation that providers visit patients at potentially higher risk less often sug-

gests that the practice of contact isolation may selectively harm the patients most at risk for infectious complicationsCby erecting barriers to the provision of care. On the other hand, although patients perceived less amount of time spent with their providers after isolation, they reported overall better care during isolation, but this difference was not statistically significant. In another study of isolation and hand washing compliance, Kirkland and Weinstein13 found that health care providers were only half as likely to enter the ICU room of a patient in isolation, prompting them to conclude that isolation could be harmful through limiting access to care, partic-

186 Evans et al

Surgery August 2003

Table VI. Patient perception of isolation questionnaire results Total interviews (n) Amount of contact with MD* Amount of contact with RN* Comfort talking with MD* Comfort talking with RN* Level of contact with MD† Level of contact with RN†

Isolated

Nonisolated

P value

9 4.3 ± 0.2 4.2 ± 0.2 4.3 ± 0.3 4.3 ± 0.2 2.4 ± 0.2 2.9 ± 0.1

17 4.4 ± 0.2 4.5 ± 0.2 4.7 ± 0.1 4.8 ± 0.1 2.8 ± 0.1 2.9 ± 0.1

— .95 .40 .16 .07 .13 .96

*Rated by patients on a 1-5 scale: 1 = “Very unsatisfactory” to 5 = “Very satisfactory.” †Rated by patients on a 1-3 scale: 1 = “Provider stood in doorway”; 2 = “Provider stood at bedside”; 3 = “Provider performed examination or obtained vital signs.”

ularly for patients requiring intensive care. Koss et al15 reported a higher rate of nosocomial pneumonia among patients in the ICU and in isolation when compared with their nonisolated neighbors, in spite of greater than 91% compliance with contact precautions during the study period. Although contact time was not measured in their study, isolation precautions may have sufficiently reduced the frequency of oropharyngeal care, or even discouraged more aggressive ventilatory weaning, to encourage the development of pneumonia. Furthermore, we observed that although physician encounters with patients in the isolated ICU group are on average longer than with patients in the nonisolated ICU group, on the floor they are shorter. Perhaps of most concern is the determination that the sickest patients on the floor under isolation precautions are seen 15 minutes less per hour than patients in the nonisolated group with the same severity of illness. The average admitting APACHE II score to our surgical ICU is now 18. In an era where bed space is at a premium, patients with greater severity of illness than ever before are now cared for on the general wards of the hospital. Isolation imposes fewer visits of shorter length, and less contact time with patients overall, with the most severely ill isolated patients at a potentially high risk-to-no benefit ratio. In this investigation, hand washing rates were similar to those reported in previous observational studies,16,20 but in contrast to the study by Pettinger et al,16 we found that the more intense form of isolation, contact-droplet, was associated with less hand washing compliance. Other studies have linked increased hand washing rates with increasing levels of isolation precautions. In a study to evaluate the efficacy of gown use in isolation, Slaughter et al6 demonstrated that in encounters in rooms assigned to both gown and glove use, hand washing rates were higher in comparison to encounters in rooms requiring glove use alone. Puzniak et al11 saw increased compliance with isolation precautions

during a period of universal gowning and gloving for all the beds in an ICU, and an associated decrease in the incidence of new VRE isolates when the colonization pressure of the ICU was less than 15. In all of these studies, the authors concluded that the greater awareness of infection control policies through use of more visible isolation practices (donning gowns) may have impacted the subjects’ increased rates of compliance. Although the rate of isolation protocol compliance ranged from 60% to 91% in these studies, our own observed compliance rate was only 43%. The higher rate of compliance during the periods of gown use in the studies by Puzniak et al11 (78%) and Slaughter et al6 (79%) studies could be attributed to their subjects’ awareness of the research protocols and of their actions being monitored (Hawthorne effect). Because it has been reported that long-term changes in the behavior of health care providers are uncommon after educational interventions and that most increases in compliance with isolation precautions are transient,2,21 these high rates of compliance are probably not representative of everyday clinical practice, and the impact of these interventions should be viewed with caution. On the contrary, our observational study was specifically designed and conducted to avoid bias; subjects were unaware, and intentionally not informed, of the observations performed by a medical student whose presence was inconspicuous on the surgical units. We believe therefore that our data are an accurate representation of the current state of compliance with isolation precautions in our hospital. This study has several important limitations. The data examined are only a representative sample of the isolated patients in our hospital. Although all of the observations in this study were conducted at the same time each morning to ensure consistency, observations conducted during a different time of day, particularly early morning rounds, might reveal different patterns and rates of contact, length of visits, and compliance. A more comprehensive evalua-

Evans et al 187

Surgery Volume 134, Number 2

tion of contact time and compliance with isolation precautions would require much more intensive surveillance not feasible by a single observer. Other investigators have uesd simple electronic sensors on sink faucets to determine compliance with handwashing,22 but a more sophisticated monitoring system, perhaps such as continuous video surveillance,20 would be required to adequately evaluate compliance with isolation protocols, as well as provider type, patient contact and length of encounter. Because the study focused on a sample of provider/patient encounters, it was not designed to examine outcomes of individual patients, and it was not possible to formally assess the independent risk of isolation. However, the significant decrease in the amount of care received by patients in isolation, particularly in light of the lack of benefit evidenced by the fact that contact precautions had no impact on the rate of GRE infection in our hospital, prompts us to question the risk/benefit ratio of this method of infection control. Even if this lack of benefit is dismissed because of the relatively low incidence of GRE infection, it appears that with the observed rates of contact, coupled with the current level of compliance, our practice of contact isolation may indeed work not through preventing contact transmission, but by preventing contact with isolated patients altogether. Further studies should analyze the impact of isolation on patient outcomes to quantify the true cost to the isolated patient, that is, to determine whether the inequity of care observed may contribute to the increased morbidity and mortality rates of resistant organisms infections. Infection control policies may need to be reconsidered, instead targeting interventions that are less likely to reduce contact time, while still controlling the transmission of resistant organisms. For example, we have previously published that physicians were more than 85% compliant with a scheduled empiric antibiotic rotation in the surgical ICU of our hospital, an intervention associated with a 68% decrease in the rate of resistant gram negative infections, and a 24% decrease in infections overall.23 In a recent report, Puzniak et al11 quantify for the first time the colonization pressure threshold under which gowning is ineffective in reducing the transmission of resistant organisms. On the basis of this evidence and in an effort to increase compliance, a better strategy for barrier-based infection control might require glove use and handwashing for resistant organism-carrying patients, adding gowns only when the isolated organisms are highly endemic to the hospital unit. Recent infection control initiatives reemphasize the importance of handwashing,24 and with the increasing availability of

alcohol-based handwashing solutions, it would seem that clinical workflow may be less impeded by their use, thus increasing compliance over traditional cleansing with soap and water.25,26 Resistant organisms pose a grave threat to hospitalized patients and demand a multifaceted approach to infection control. At the same time we must minimize the transmission of these difficult-totreat microbes, we are obligated to encourage provider behaviors that do not prevent the regular, timely provision of appropriate patient care. Barrierbased isolation precautions are associated with reduced contact time, less frequent patient visits, and a high degree of noncompliance—it is time to consider the impact of these environmental policies on the patients who are subjected to them. REFERENCES 1. Donowitz LG. Failure of the overgown to prevent nosocomial infection in a pediatric intensive care unit. Pediatrics 1986;77:35-8. 2. Lynch P, Jackson MM, Cummings MJ, Stamm WE. Rethinking the role of isolation practices in the prevention of nosocomial infections. Ann Intern Med 1987;107:243-6. 3. Pelke S, Ching D, Easa D, Melish ME. Gowning does not affect colonization or infection rates in a neonatal intensive care unit. Arch Pediatr Adolesc Med 1994;148:1016-20. 4. Lai KK, Kelley AL, Melvin ZS, Belliveau PP, Fontecchio SA. Failure to eradicate vancomycin-resistant enterococci in a university hospital and the cost of barrier precautions. Infect Control Hosp Epidemiol 1998;19:647-52. 5. Saint S, Atherton S, Lipsky BA, McDonald L, Strausbaugh L. Controlling the spread of vancomycin-resistant enterococci with contact precautions: time for a randomized trial. Int J Infect Dis 1999;3:179-80. 6. Slaughter S, Hayden MK, Nathan C, Hu TC, Rice T, Van Voorhis J, et al. A comparison of the effect of universal use of gloves and gowns with that of glove use alone on acquisition of vancomycin-resistant enterococci in a medical intensive care unit. Ann Intern Med 1996;125:448-56. 7. Chaix C, Durand-Zaleski I, Alberti C, Brun-Buisson C. Control of endemic methicillin-resistant Staphylococcus aureus: a cost-benefit analysis in an intensive care unit. Jama 1999;282:1745-51. 8. Lacey S, Flaxman D, Scales J, Wilson A. The usefulness of masks in preventing transient carriage of epidemic methicillin-resistant Staphylococcus aureus by healthcare workers. J Hosp Infect 2001;48:308-11. 9. Montecalvo MA, Jarvis WR, Uman J, Shay DK, Petrullo C, Horowitz HW, et al. Costs and savings associated with infection control measures that reduced transmission of vancomycin-resistant enterococci in an endemic setting. Infect Control Hosp Epidemiol 2001;22:437-42. 10. Montecalvo MA, Jarvis WR, Uman J, Shay DK, Petrullo C, Rodney K, et al. Infection-control measures reduce transmission of vancomycin-resistant enterococci in an endemic setting. Ann Intern Med 1999;131:269-72. 11. Puzniak LA, Leet T, Mayfield J, Kollef M, Mundy LM. To gown or not to gown: the effect on acquisition of vancomycin-resistant enterococci. Clin Infect Dis 2002;35:18-25. 12. Karchmer TB, Durbin LJ, Simonton BM, Farr BM. Cost-effectiveness of active surveillance cultures and contact/droplet

188 Evans et al

13. 14.

15.

16. 17.

18.

19.

precautions for control of methicillin-resistant Staphylococcus aureus. J Hosp Infect 2002;51:126-32. Kirkland KB, Weinstein JM. Adverse effects of contact isolation. Lancet 1999;354:1177-8. Donowitz LG, Wenzel RP, Hoyt JW. High risk of hospitalacquired infection in the ICU patient. Crit Care Med 1982;10:355-7. Koss WG, Khalili TM, Lemus JF, Chelly MM, Margulies DR, Shabot MM. Nosocomial pneumonia is not prevented by protective contact isolation in the surgical intensive care unit. Am Surg 2001;67:1140-4. Pettinger A, Nettleman MD. Epidemiology of isolation precautions. Infect Control Hosp Epidemiol 1991;12:303-7. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818-29. VRE Information for Healthcare Workers. In: UVA Infection Control Manual. Charlottesville, Va: UVA Health System; 2000. Available at: http://www.healthsystem.virginia .edu/internet/infection-control/ICManual/icmvrehcw .cfm. MRSA Information for Healthcare Workers. In: UVA Infection Control Manual. Charlottesville, Va: UVA Health System; 2000. Available at: http://www.healthsystem.virginia .edu/internet/infection-control/ICManual/icmmrsainfo .cfm.

Surgery August 2003

20. Nishimura S, Kagehira M, Kono F, Nishimura M, Taenaka N. Handwashing before entering the intensive care unit: what we learned from continuous video-camera surveillance. Am J Infect Control 1999;27:367-9. 21. Larson E, Kretzer EK. Compliance with handwashing and barrier precautions. J Hosp Infect 1995;30(Suppl):88-106. 22. Swoboda S, Earsing K, Lipsett P, Lane S, Strauss K. Isolation status and voice prompts improve hand hygiene. Crit Care Med 2003;30:A17. 23. Raymond DP, Pelletier SJ, Crabtree TD, Gleason TG, Hamm LL, Pruett TL, et al. Impact of a rotating empiric antibiotic schedule on infectious mortality in an intensive care unit. Crit Care Med 2001;29:1101-8. 24. Boyce JM, Pittet D. Guideline for hand hygiene in healthcare settings: recommendations of the healthcare infection control practices advisory committee and the HICPAC/SHEA/APIC/IDSA Hand Hygiene Task Force. Am J Infect Control 2002;30:1-46. 25. Maury E, Alzieu M, Baudel JL, Haram N, Barbut F, Guidet B, et al. Availability of an alcohol solution can improve hand disinfection compliance in an intensive care unit. Am J Respir Crit Care Med 2000;162:324-7. 26. Pittet D, Hugonnet S, Harbarth S, Mourouga P, Sauvan V, Touveneau S, et al. Effectiveness of a hospital-wide programme to improve compliance with hand hygiene. Infection Control Programme. Lancet 2000;356:1307-12.

Availability of Journal back issues As a service to our subscribers, copies of back issues of Surgery for the preceding 5 years are maintained and are available for purchase from Mosby until inventory is depleted. Please write to Mosby, Subscription Customer Service, 6277 Sea Harbor Dr, Orlando, FL 32887, or call (800)654-2452 or (407)345-4000, for information on availability of particular issues.