Although they are medically necessary and beneficial for the patient, surgical procedures can result in surgical site infections (SSIs) at a rate of ~0.6%–8.8%, depending on the surgical specialty and wound classification. ^{Reference Birgand, Azevedo and Toupet1} SSIs are caused by microbes, such as bacteria and fungi, entering a patient’s incision site through airborne particles in the operating room (OR). SSIs lead to extensive physical, emotional, and financial consequences for individual patients due to elongated hospital stays, further surgeries, and possibly sepsis and death, all of which increase annual healthcare expenditures substantially. ^{Reference de Lissovoy, Fraeman, Hutchins, Murphy, Song and Vaughn2–Reference Graves, Nicholls and Morris5}

Environmental risk factors are unavoidably present in an OR. These include patient comorbidities, surgical procedure complexity, air circulation, equipment sterilization, and certain staff behaviors. ^{Reference Wanta, Glasgow and Habermann4} Operating rooms can reduce SSI risk by following sterilization guidelines and protocols, ^{Reference Durando, Bassetti and Orengo6–Reference Munoz-Price, Bowdle and Johnston11} and practicing disciplined behaviors (eg, restricted OR door openings) known to affect the SSI rate. ^{Reference Crolla, Van Der Laan, Veen, Hendriks, Van Schendel and Kluytmans12,Reference Van Der Slegt, Van Der Laan, Veen, Hendriks, Romme and Kluytmans13}

Operating rooms are “isolated, positive-pressure environments designed to recirculate air through filtered ventilation ducts.” ^{Reference Lynch, Englesbe and Sturm14} Laminar, or nonturbulent, airflow produced by the OR ventilation system helps maintain a sterile field around the wound. ^{Reference Hamilton, Balkam, Purcell, Parks and Holdsworth15} Door openings may reduce air pressure and possibly disrupt the laminar airflow, creating a path for new contaminants to enter the OR and making it difficult to clear potentially dangerous particles. ^{Reference Hamilton, Balkam, Purcell, Parks and Holdsworth15} Measuring the microbial activity in the OR can provide valuable insights about how door openings and OR traffic affect microbial activity in the OR. ^{Reference Anthony, Peterson, Polgreen, Sewell and Polgreen16,Reference Taaffe, Lee and Ferrand17} Although studies have reported mixed results, there appears to be evidence that door openings and OR traffic can increase microbial activity. ^{Reference Perez, Holloway and Ehrenfeld18–Reference Panahi, Stroh, Casper, Parvizi and Austin20}

Using air sampling during orthopedic and implant procedures, studies have demonstrated positive correlation between total traffic flow and total microbial colony forming units (CFU), and between CFU and infection. ^{Reference Andersson, Bergh, Karlsson, Eriksson and Nilsson21,Reference Darouiche, Green and Harrington22} However, the correlations among the number of people in the OR, CFU count, and SSI are not clear. ^{Reference Anthony, Peterson, Polgreen, Sewell and Polgreen16,Reference Wan, Chung and Tang23,Reference Babkin, Raveh and Lifschitz24} A weak positive correlation has been detected between the total number of people present and total CFU, ^{Reference Andersson, Bergh, Karlsson, Eriksson and Nilsson21} while a retrospective study of patients who developed a superficial SSI did not reveal an association between the number of personnel in the OR and infection rate. ^{Reference Wanta, Glasgow and Habermann4}

In a previous study, the research team collected data from 27 procedures using settle plates and air samplers to determine how movement and door openings within the OR affect microbial loads at various locations. ^{Reference Taaffe, Lee and Ferrand17} They also videotaped and observed these procedures to gain further insight. However, the settle plates and air samplers were not in the same locations for each procedure due to the OR team requesting that the devices be placed further away from the operating field or high-traffic areas. Another limitation was that the frequency and path of staff movements varied between surgeries resulting from unique procedure characteristics (eg, patient, OR staff, procedure length, etc).

In the current study, the research team addressed these limitations by creating a script to simulate movements and door openings during surgical procedures, allowing the OR environment to be controlled. The team reviewed 27 procedures then counted and measured all staff movements and door openings during 4 orthopedic cases that were videotaped. These 4 cases were between 3 and 4.5 hours in length (15 total hours). The script was enacted at 2 hospitals in upstate South Carolina while bacteria and fungi samples were collected. Enacting a simulation allowed data collection devices to be placed in identical locations across all tests and ensured that the exact level of staff movements and door openings were performed in each test.

Data were gathered in 2 phases of simulation experiments. Phase 1 was conducted in the spring of 2019 to test the hypotheses that increasing the number of door openings and staff movements increased the overall microbial load in the OR. Phase 2 was conducted in the fall of 2019 to gather more data and to adjust the study factors in response to an OR management request for more information. Phase 2 included the width and duration of door openings and excluded the number of door openings as experimental factors.

Methods

Simulation script development

The simulation script was developed by first identifying the minimum and maximum number of door openings and zone-to-zone staff transitions (Table 1) for each role (ie, surgeon, anesthesiologist, circulating nurse, and scrub nurse) during both the OR preparation and intraoperative phases for 4 orthopedic procedures. More than 15 hours of zone-to-zone transitions were summarized, with the preparation phase defined as “case cart emptied” to “incision made” and the intraoperative phase defined as “incision made” to “incision closed.” Each staff member’s movements were converted into rates appropriate for a 30-minute simulation. The data counted all zone-to-zone transitions (Fig. 1) as a movement, so the movement count was higher than actual origin-to-destination trip counts. Because our goal was to study the effect of movement, counting zone transition movements increased the differences between the low and high settings.

Table 1. Number of Door Openings and Staff Movements During the Operation Room Preparation and Intraoperative Phases for 30 Minutes

Fig. 1. Settle plate locations for phase 1 and phase 2.

Data collection

Five 30-minute trials were conducted each day. The first trial was a control that had no movement or door openings. Samples were collected from the same 8 locations for each of the 5 trials. As shown in Figure 1, the locations were (1) nonsterile door, (2) circulating nurse workstation, (3) instrument table, (4) intake vent 1, (5) sterile door, (6) anesthesia area, (7) intake 2, and (8) surgical table. At each location, settle plates were placed at 2 different heights labeled “H” for high and “L” for low. The “H” plates were placed about waist high, while the “L” plates were placed at floor level. Each location and each height had both an Sabouraud dextrose agar (SDA) plate to collect fungi and a blood agar (BA) plate to collect bacteria, resulting in a total of 32 settle plates per trial. All settle plates, which were labeled with their location, height, and trial number, were open for the entire 30-minute trial with the lid right side up. They were collected at the end of the trial, secured with parafilm, and placed into Ziploc bags for transport. The BA plates were placed in an incubator for 48 hours at 37°C and then refrigerated for 2 days. The SDA plates were kept at room temperature (23°C) for 4 days. After their incubation period, the number of CFUs on each plate were counted.

Phase 1 experiment

In phase 1, a total of 10 trials over 2 days had the experimental factor settings given in Table 1: staff movement (low vs high), number of door openings (low vs high), and surgical phase (preparation vs intraoperative). Each combination of factors was tested (using a preparation or intraoperative sequence), along with one control trial at the beginning of each day. The specific hypotheses tested in phase 1 are listed in Table 2. This table provides conclusions on tests that compare whether differences in microbial load were detected based on sample location height in the OR, staff movement, door openings, and surgical phase.

Table 2. Phase 1 Statistical Results on Hypotheses (P values recorded)

^a Reject at P = .05.

^b Reject at P = .01.

Phase 2 experiment

The phase 2 factor settings were amount of staff movement (low vs high), duration of door openings (6 seconds vs 12 seconds), and width of door opening (45° vs 90°). Phase 2 trials were conducted across 2 days in a different hospital than phase 1, but the locations in the OR and the methodology used to manage the settle plates were the same as in phase 1. We tested 5 hypotheses (Table 3); 4 examined the effects of the experimental factors. The fifth hypothesis compared all trials with movement and door openings to the control trial. In all trials, the number of door openings was constant at what had been the high setting in phase 1 by opening the door at 2-minute intervals. ^{Reference Birgand, Saliou and Lucet7} Instead, door openings varied based on duration and width as described above. To ensure consistency between trials, tape was placed on the floor to mark each door opening angle and a watch was used to confirm the door opening time.

Table 3. Phase 2 Statistical Results on Hypotheses (P values recorded)

^a Reject at P = .05.

^b Reject at P = .01.