I. Patients' Perceptions of HCP Attire
We identified 26 studies (published from 1990 onward) that examined patients' perceptions of HCP attire[1–26] (Table 1). Most (23/26) studies surveyed patient preference for different types of HCP attire[1–6,8–18,20–25] using either pictures of models in various dress styles[3,4,7–9,15–18,20,22–24] or descriptions of attire.[1,5,11,14,21,25] Four studies[6,10,12,13] asked patients to assess the attire of their actual physicians. Attire descriptions and terminology varied among studies (eg, "formal," "business," "smart," "suit and tie," and "dress") and will be referred to hereafter as "formal attire." We use "casual attire" to refer to anything other than formal attire.
A. Formal attire and white coats: Most of the studies using pictures and models of HCP attire indicated patient preference for formal attire, which was favored over both scrubs[1,3,7,9,18,22] and casual attire.[7,9,15,16,19,22] However, several other studies revealed that physician attire was unlikely to influence patients' levels of comfort,[4,20] satisfaction, trust, or confidence in physicians' abilities,[2,4,9,19,20,25] even if patients previously had expressed a preference for one type of attire.[4,9,20,25]
Fifteen studies addressed white coats.[1,4,7–9,11–17,20–22] In 10 of these studies, patients preferred that physicians wear white coats,[1,7–10,12,15–17] and in 1 study patients reported feeling more confident in those physicians. Similarly, 2 studies showed a significant association between the presence of a white coat, especially on a female physician, and patients' trust and willingness to share sensitive information. Patients also indicated less comfort in dealing with an informally dressed physician, describing a shirt and a tie as the most professional and desirable attire for physicians[23–25] in addition to an overall well-groomed appearance.[5,15] Moreover, the following items were deemed as inappropriate or undesirable: jeans,[5,14] shorts, clogs,[14,15] and open-toed sandals. In the remaining 5 studies, patients showed no clear predilection for one dress style over another or did not consider a white coat either necessary or expected.[4,11,13,20,21]
Five studies assessed patient satisfaction, confidence, or trust on the basis of their treating physicians' dress,[2,6,10,12,13] showing little response variations regardless of apparel. A survey of patients seen by obstetricians/gynecologists who were randomly assigned formal attire, casual attire, or scrubs found high satisfaction with physicians regardless of the group allocation. Similarly, in a before-and-after trial, emergency department (ED) physicians were asked to wear formal attire with a white coat one week followed by scrubs the subsequent week. Using a visual analog scale, patients rated their physician's appearance, professionalism, and satisfaction equally regardless of the week of observation. Another ED study found no difference in patients' satisfaction with the care provided when their physicians wore white coats combined with either scrubs or formal attire. Similarly, 2 groups of patients who received preoperative care by the same anesthesiologist wearing either formal attire for one group of patients or casual attire for the other found no differences in patient satisfaction between the groups. In contrast, one crossover trial involving physicians dressed in "respectable" or formal versus "retro" or casual attire found that patient confidence and trust were higher with the respectable-dress protocol. Another study evaluating the attire of patients' treating physicians indicated preference for polished shoes and short hair for men, with jeans, clogs, trainers, and earrings on men being rated as undesirable. A survey among Japanese outpatients indicated a preference for white coats but no significant difference in satisfaction levels based on attire when presented with physicians wearing white coats or "noninstitutional clothes."
B. BBE: Preference for BBE was assessed in 6 studies originating in the United Kingdom following implementation of the nationwide BBE policy[1,3,23–25] and in 1 US study. In these 7 reports, patients did not prefer short sleeves. After informing patients of the BBE policy, older patients were more likely to prefer short-sleeved shirts without ties, while younger patients favored scrubs. After providing information about the potential for cross-contamination from shirt sleeve cuffs and neckties, responses changed from a preference for formal or long-sleeved attire to a preference for short sleeves or scrubs.[11,18,23] In addition, Shelton et al also found an association between physician gender and BBE attire: after a statement informing the participants of the potential cross-transmission of microorganisms by attire, patients preferred scrubs for female physicians but did not differentiate between scrubs and short-sleeved shirts for male physicians.
C. Ties: Neckties were specifically addressed in several studies from the United Kingdom.[5,21,24] In one study, patients reported that attire was important but that neckties were not expected. Similarly, in a survey among individuals in the public concourse of a hospital, 93% had no objection to male physicians not wearing ties. None of these studies evaluated neckties in the context of patients' perceptions of infection prevention.
D. Laundering of clothes: In one study, patients identified "daily laundered clothing" as the single most important aspect of physicians' appearance.
E. Other factors: Several additional variables may influence patient preference for physician attire, including age of either the patient or the managing physician, gender of the practitioner, time of day, setting, and the attire patients are accustomed to seeing. In Japan, older patients were more likely to prefer white coats. Similarly, older patients in England found scrubs less appealing than did younger patients. Pediatric dental patients were more likely than their parents to favor casual attire. Patients preferred formal attire for senior consultants but thought that junior physicians should be less formal. Patients identified female physicians' attire as more important than the attire worn by male physicians. Formal attire was less desirable by patients seen during the night shift. Parents of children being seen in the ED favored surgical scrubs. Additionally, 2 trials evaluated attire preference on the basis of what patients often see their HCP wearing. In one trial, patients accustomed to seeing their anesthesiologist in a suit were more likely to find suits and ties desirable. Similarly, the practice to which a patient belonged was found to be an independent factor in the patient's choice of preferred attire; however, another study found poor agreement between patient preferences and their physicians' typical attire.
In summary, patients express preferences for certain types of attire, with most studies indicating a predilection for formal attire, including a white coat, but these partialities had a limited overall impact on patient satisfaction and confidence in practitioners. This is particularly true in trials that evaluated the effect of attire on patient satisfaction in real-world settings. Patients generally do not perceive white coats, formal attire, or neckties as posing infection risks; however, when informed of potential risks associated with certain types of attire, patients appear willing to change their preferences for physician attire.[11,18]
II. HCP Perceptions Regarding Attire
Few studies evaluated HCP preferences with regard to attire.[5,6,14,26] While most studies addressed specific elements of HCP attire, one looked at the overall importance of attire and found that 93% of physicians and nurses versus 83% of patients thought that physician appearance was important for patient care (P < .001).
White coats: In a survey exploring perceptions of surgeons' apparel performed among surgeons themselves, inpatients, and the nonhospitalized public, all 3 groups were equally likely to consider a white coat necessary and blue jeans inappropriate. Surgeons were more prone to consider scrubs and clogs appropriate. In another survey of 15 obstetricians/gynecologists, 8 preferred casual attire, while 7 preferred formal attire. Three studies assessed HCP alongside patient perception of infection risk or lack of hygiene associated with white coats, formal attire, or neckties,[3,24,26] with one finding that HCP were more likely than patients to consider white coats unhygienic.
Ties: In a survey performed in a public concourse of a UK hospital, HCP were more likely than non-HCP to prefer physicians' wearing of neckties for reasons of professionalism.
Laundering of clothes: A recent survey showed that nonsurgical providers preferentially (and without prompting) laundered their scrubs every 1.7 ± 0.1 days (mean ± standard error) compared with white coats, which were laundered every 12.4 ± 1.1 days (P < .001); however, the reasons for this divergent behavior remain unclear.
III. Studies of Microbial Contamination of Apparel in Clinical and Laboratory Settings
No clinical studies have demonstrated cross-transmission of healthcare-associated pathogens from a HCP to a patient via apparel; however, a number of small prospective trials have demonstrated the contamination of HCP apparel with a variety of pathogens ( Table 2 ).[5,28–37]
White coats/uniforms: The 5 studies we evaluated indicate that physician white coats and nursing uniforms may serve as potential sources of colonization and cross-transmission. Several studies described contamination of apparel with Staphylococcus aureus in the range of 5% to 29%.[30,33–35,38] Although gram-negative bacilli have also been identified, these were for the most part of low pathogenicity;[30,35] however, actual pathogens, such as Acinetobacter species, Enterobacteriaceae, and Pseudomonas species, have been reported.
A number of factors were found to influence the magnitude of contamination of white coats and uniforms. First, the degree of contamination was correlated with more frequent usage of the coat, recent work in the inpatient setting, and sampling certain parts of the uniform. Higher bacterial loads were found on areas of clothing that were more likely to come into contact with the patient, such as the sleeve. Additionally, the burden of resistant pathogens on apparel was inversely correlated with the frequency of lab coat change. Apparel contamination with pathogenic microorganisms increased over the course of a single patient care shift. Burden et al demonstrated that clean uniforms become contaminated within only a few hours of donning them. Similarly, a study testing nurses' uniforms at both the beginning and the end of their shifts described an increase in the number of uniforms contaminated with one or more microorganisms from 39% to 54%, respectively. The proportion of uniforms contaminated with vancomycin-resistant enterococci (VRE), methicillin-resistant S. aureus (MRSA), and Clostridium difficile was also noted to increase with shift work.
In the first report of a positive correlation between contamination of hands and contamination of white coats, Munoz-Price et al cultured the hands, scrubs, and white coats of intensive care unit staff. The majority of bacteria isolated from hands were skin commensals, but HCP were also found to have contamination of hands, scrubs, and white coats with potentially pathogenic bacteria, including S. aureus, Enterococcus species, and Acinetobacter baumannii. Among dominant hands, 17% of 119 hands were contaminated with one of these species, and staff members with contaminated hands were more likely to wear a white coat contaminated with the same pathogen. This association was not observed with scrubs.
BBE: Two observational trials evaluated the bacterial contamination of HCP's hands on the basis of BBE attire versus controls, finding no difference in total bacterial counts or in the number of clinically significant pathogens.[40,41] In contrast, Farrington et al, using a fluorescent method, examined the efficacy of an alcohol hand wash among BBE providers versus controls. The authors found decreased efficacy of hand hygiene at the wrist level in the non-BBE group, suggesting that the BBE approach may improve wrist disinfection during hand washing.
The United Kingdom has adopted a BBE approach, on the basis of the theory that it will limit patient contact with contaminated HCP apparel and to promote better hand and wrist hygiene. However, a randomized trial comparing bacterial contamination of white coats against BBE found no difference in total bacterial or MRSA counts (on either the apparel itself or from the volar surface of the wrist) at the end of an 8-hour workday.
Scrubs: The use of antimicrobial-impregnated scrubs has been evaluated as a possible solution to uniform contamination. In a prospective, randomized crossover trial of 30 HCP in the intensive care unit setting, when compared with standard scrubs, antimicrobial-impregnated scrubs were associated with a 4–7 mean log reduction in surface MRSA burden, although there was no difference in MRSA load on HCP hands or in the number of VRE or gram-negative bacilli cultured from the scrubs. The study did not assess the HAI impact of the antimicrobial scrubs.
Ties: Several studies indicated that neckties may be colonized with pathogenic bacteria, including S. aureus. Lopez et al reported a significantly higher bacterial burden on neckties than on the front shirt pocket of the same subject. In 3 studies, up to 32% of physician neckties grew S. aureus.[5,31,37] Steinlechner et al identified additional potential pathogens and commensals from necktie cultures, including Bacillus species and gram-negative bacilli. Two reports found that up to 70% of physicians admitted having never cleaned their ties.[5,31]
Laundering of clothes: Numerous articles published during the past 25 years describe the efficacy of laundering hospital linens and HCP clothing, but most investigations of the laundering of HCP attire have employed in vitro experimental designs that may or may not reflect real-life conditions. A 2006 study demonstrated that while clothes lost their burden of S. aureus, they concomitantly acquired oxidase-positive gram-negative bacilli in the home washing machine. These bacteria were nearly eliminated by tumble drying or ironing. Similarly, investigators found that recently laundered clothing material acquired gram-negative bacteria from the washing machine, which were subsequently eliminated by ironing. Another in vitro study in the United Kingdom compared the reduction of microorganisms on artificially inoculated nurses' uniform material after washing at various temperatures as well as with and without detergents. Washing uniforms contaminated with MRSA and Acinetobacter species at a temperature of 60°C, with or without detergent, achieved at least a 7-log reduction in the bacterial burden of both microorganisms. There is no robust evidence that centralized industrial laundering decontaminates clothing more effectively than home laundering.
Footwear: Although restrictions on HCP footwear are influenced by a desire to meet patients' preferences for appropriate attire,[10,14,15] most are driven by concerns for HCP safety.[47–50] Studies have found that wearing of shoes with closed toes, low heels, and nonskid soles can decrease the risk of exposure to blood or other potentially infectious material,[47,48,50,51] sharps injuries,[48,50,52] slipping, and musculoskeletal disorders.
Casual, open footwear, such as sandals, clogs, and foam clogs, potentially expose feet to injury from dropped contaminated sharps and exposure to chemicals in healthcare facilities. A comparison of needlestick injury surveillance data from the standardized Exposure Prevention Information Network program revealed a higher proportion of hollow-bore needle injuries to the feet of Japanese HCP, with 1.5% of 16,154 total injuries compared with 0.6% of 9,457 total injuries for US HCP (2.5 times higher; P < .001). Although multiple factors were linked to these injuries, one included the common practice in Japan to remove outdoor shoes and replace them with open-toed slippers on hospital entry.
Footwear is an area of increased concern in the OR. The Association of periOperative Registered Nurses (AORN) recommends that OR footwear have closed toes as well as backs, low heels, and nonskid soles to prevent slipping. The US Occupational Safety and Health Administration (OSHA) requires the use of protective shoes in areas where there is a danger of foot injuries from falling objects or objects piercing the soles. One study that measured the resistance of shoes to penetration by scalpels showed that of the 15 pairs of shoes studied, only 6 were made of material that was sharp resistant, including sneaker suede, suede with inner mesh lining, leather with inner canvas lining, nonpliable leather, rubber with inner leather lining, and thicker rubber. The OSHA bloodborne pathogens standard mandates that employers determine the workplace settings in which gross contamination with blood or body fluids is expected, such as the OR, and to provide protective shoe coverings in those settings.[47,48,50,51] Shoe covers are not meant to prevent transmission of bacteria from the OR floor; in fact, preliminary data show that the OR floor may play a dynamic role in the horizontal transmission of bacteria due to frequent floor contact of objects that then directly touch the patient's body (eg, intravenous tubing, electrocardiogram leads).
When HCP safety concerns or patient preference conflict with a HCP's desire for fashion, a facility's dress code can be the arbiter of footwear. OSHA allows employers to make such dress code determinations without regard to a worker's potential exposure to blood, other potentially infectious materials, or other recognized hazards.
IV. Outbreaks Linked to HCP Apparel
Wright et al reported an outbreak of Gordonia potentially linked to HCP apparel. In this report, postoperative sternal wound infections with Gordonia bronchialis in 3 patients were linked to a nurse anesthetist. Gordonia was isolated from the HCP's scrubs, axillae, hands, and purse and from multiple sites on the HCP's roommate.
V. Studies From Developing Countries
In Nigeria, factors identified increasing the likelihood of bacterial contamination of white coats included daily laundering and use limited to patient care rather than nonclinical duties. In India, medical students' white coats were assessed for bacterial contamination, paired with surveys about laundering habits and attitudes toward white coats. Coats were contaminated most frequently with S. aureus, followed by Pseudomonas species and coagulase-negative staphylococci. A similar trial of white coats used by staff in a rural dental clinic also revealed predominantly gram-positive contamination.
VI. Hospital Policies Addressing HCP Attire
We reviewed and compared policies related to HCP attire from 7 large teaching hospitals or health systems. In general, policies could be categorized into 2 groups:
General appearance and dress of all employees
Standards for HCP working in sterile or procedure-based environments (OR, central processing, procedure areas, etc)
Policies were evaluated for the following elements:
Recommended clothing (eg, requirement for white coats, designated uniforms) or other options (eg, BBE)
Guidance regarding scrubs
Use of name tags
Wearing of ties
Requirements for laundering or change of clothing
Footwear and nonapparel items worn or carried by HCP
Personal protective equipment
All institutions' human resources policies outlined general appearance or dress code requirements for professional standards of business attire; however, institutions varied in job-specific policies and for the most part did not address more specific attire requirements except for OR-related activities. Few institutional policies included enforcement provisions. The institutions that required accountability varied from detailing the supervisor's administrative responsibilities to more specific consequences for employee noncompliance.
Three institutions recommended clothing (such as color-coded attire) for specific types of caregivers (eg, nurses, nurses' assistants, etc). Policies specific to clinical personnel were most frequently related to surgical attire, including scrubs, use of masks, head covers, and footwear in restricted and semirestricted areas and surgical suites, and to central processing, as consistent with AORN standards. Scrubs were universally provided by the hospital in these settings. Laundering policies clearly indicated that laundering of hospital-provided scrubs was to be performed by the hospital or at a hospital-accredited facility. Use of masks, head covers, footwear, and jewelry were generally consistent with AORN standards.
Excluding surgical attire, only one institution provided guidance specific to physicians, outlining a recommendation for BBE attire during patient care. This policy specified not to use white coats, neckties, long sleeves, wristwatches, or bracelets. Institutional policies also varied in recommendations for laundering and change of clothing other than for surgical attire. No specific guidance was issued for other uniforms, other than cleanliness and absence of visible soiling; however, one institution referred to infection control specifications for maintenance of clothing. Guidance regarding frequency of clothing change was variable for scrubs, from nonspecific requirements (eg, wearing freshly laundered surgical attire on entry to restricted/semirestricted areas) to specific requirements (clean scrubs once per shift to once daily and if visibly soiled). In addition, most policies included instructions for HCP to remove scrubs and change into street clothes either at the end of the shift or when leaving the hospital or connected buildings.
VII. Survey Results
A total of 337 SHEA members and members of the SHEA Research Network (21.7% response of 1,550 members) responded to the survey regarding their institutions' policies for HCP attire. The majority of respondents worked at hospitals (91%); additional facilities included freestanding children's hospitals (4%), freestanding clinics (1%), and other facility types (5%), such as long-term acute care hospitals, multihospital systems, short-term nursing facilities, and rehabilitation hospitals (rounding of numbers accounts for the sum of percentages being greater than 100). The majority of responses were from either university/teaching hospitals (39%) or university/teaching-affiliated hospitals (28%). We received additional responses from nonteaching hospitals (24%), Veterans Affairs hospitals (3%), specialty hospitals (2%), and miscellaneous facilities (4%).
Enforcement of HCP attire policies was low at 11%. A majority of respondents (65%) felt that the role of HCP attire in the transmission of pathogens within the healthcare setting was very important or somewhat important.
Only 12% of facilities encouraged short sleeves, and 7% enforced or monitored this policy. Pertaining to white coats, only 5% discouraged their use and, of those that did, 13% enforced or monitored this policy. For watches and jewelry, 20% of facilities had a policy encouraging their removal. A majority of respondents (61%) stated that their facility did not have policies regarding scrubs, scrub-like uniforms, or white coats in nonclinical areas. Thirty-one percent responded that their hospital policy stated that scrubs must be removed before leaving the hospital, while 13% stated that scrubs should not be worn in nonclinical areas. Neckties were discouraged in 8% of facilities, but none monitored or enforced this policy.
Although 43% of respondents stated that their hospitals issued scrubs or uniforms, only 36% of facilities actually laundered scrubs or uniforms. A small number of hospitals provided any type of guidance on home laundering: 13% provided specific policies regarding home laundering, while 38% did not.
In contrast to other items of HCP attire, half of facilities required specific types of footwear, and 63% enforced and/or monitored this policy.
Infect Control Hosp Epidemiol. 2014;35(2):107-121. © 2014 The Society for Healthcare Epidemiology of America