Bhama PK, Davis GE, Bhrany AD, Lam DJ, Futran ND. The Effects of Intensive Care Unit Staffing on Patient Outcomes Following Microvascular Free Flap Reconstruction of the Head and NeckA Pilot Study. JAMA Otolaryngol Head Neck Surg. 2013;139(1):37-42. doi:10.1001/jamaoto.2013.1132
Author Affiliations: Division of Facial Plastic Surgery, Department of Otolaryngology, Massachusetts Eye and Ear Infirmary, Harvard Medical School, Boston (Dr Bhama); Department of Otolaryngology–Head and Neck Surgery, University of Washington School of Medicine, Seattle (Drs Davis, Bhrany, and Futran); and Department of Otolaryngology–Head and Neck Surgery, University of Cincinnati, Cincinnati, Ohio (Dr Lam).
Objective To determine if the implementation of the closed intensive care unit (ICU) at our institution altered clinical outcomes in patients who had undergone microvascular free flap reconstruction of the head and neck by the Otolaryngology–Head and Neck Surgery Service.
Design Retrospective medical chart review.
Setting A single tertiary medical center.
Patients The open ICU cohort had 52 flaps performed on 50 patients, and the closed ICU cohort had 52 flaps performed on 52 patients.
Main Outcome Measures Fifty-two free flap reconstructions of head and neck defects were performed on 50 patients who were admitted to an open ICU. The length of stay (LOS) in the ICU and hospital and incidence of complications were compared with those of 52 patients who underwent 52 free flap reconstructions and were admitted to a closed ICU over a separate period.
Results The mean length of stay in the ICU was 44 and 45 hours in the open and closed ICU cohorts, respectively (P = .90). The incidence of surgical and medical complications was similar in the open and closed ICU cohorts (P > .05).
Conclusions There does not appear to be a significant difference in patient outcome between open and closed ICU care in our study.
In the past 2 decades, intensive care unit (ICU) staffing has become a topic of investigation. Intensive care units can be divided into 2 main groups based on the use of intensivists. High-intensity staffing models include ICUs in which intensivists manage or comanage care for all patients, whereas ICUs with low-intensity staffing have intensivists who care for some or none of the patients. Closed ICUs embody high-intensity staffing because critical care teams led by intensivists provide care for all patients. Open ICUs, however, represent a model of low-intensity staffing because intensivist consultation is not mandatory.1,2
Several studies have demonstrated the benefits of high-intensity staffing in the care of critically ill patients. In particular, one large systematic review of the literature by Pronovost et al2 demonstrated that high-intensity staffing is associated with a lower hospital mortality, lower ICU mortality, reduced hospital length of stay (LOS), and reduced ICU LOS compared with low-intensity staffing. A number of other studies have demonstrated positive outcomes with high-intensity staffing vs low-intensity staffing.3- 8 Because of these findings, The Leapfrog Group, a voluntary program that works with employers to study health care safety and outcomes, has established an ICU Physician Staffing (IPS) standard encouraging the use of intensivists in high-intensity staffing patterns.1 In their 2008 Hospital Survey Results, The Leapfrog Group reported that 31% of hospitals fully meet the Leapfrog standard for IPS.9
Head and neck microvascular free flap reconstruction patients frequently require postoperative care in the ICU for airway management and to ensure stability of the anastomosis. We are not aware of any studies addressing the impact of closed vs open ICU care on outcomes after head and neck free flap reconstruction. Our institution recently initiated the use of a closed ICU system. The goal of our study was to determine if this change altered clinical outcomes in patients who had undergone microvascular free flap reconstruction of the head and neck by the Otolaryngology–Head and Neck Surgery Service.
This study is a retrospective medical chart review. All subjects received their surgery and postoperative care from a single tertiary medical center. All surgical procedures were performed by otolaryngologists trained in head and neck microvascular reconstructive surgery. Prior to beginning this study, institutional review board approval was obtained from the University of Washington Human Subjects Division.
Subjects were selected based on Current Procedural Terminology (CPT) codes for microvascular free flap reconstruction that were performed between November 10, 2008, and March 15, 2010. Subjects were excluded from the study if their care was managed primarily by another surgical team following transfer out of the ICU. Subjects were also excluded who transferred to another facility following surgery.
At the University of Washington Medical Center (UWMC), all patients who undergo free flap reconstruction by the Otolaryngology–Head and Neck Surgery (OTOHNS) Service are admitted postoperatively to the surgical ICU for overnight mechanical ventilation and frequent monitoring of the microvascular free flap. Prior to June 6, 2009, all free flap reconstruction patients admitted to the ICU following surgery by the OTOHNS Service were cared for primarily by the OTOHNS Service with optional medical consultation. After June 6, 2009, the UWMC ICU changed to a closed system so that the same patients were care for primarily by a surgical critical care team with input provided by the OTOHNS team. Once the patient was transferred out of the ICU, the patient was cared for by the OTOHNS team with optional medical consultation, unless readmitted to the ICU.
The subjects were divided into 2 cohorts: “open” and “closed.” The open cohort consisted of subjects who were admitted to the open ICU, and the closed cohort encompassed subjects who were admitted to the closed ICU.
Both the electronic and paper medical charts were reviewed for demographics, comorbidities, and outcomes of interest from progress notes, nurse charting, and administrative record keeping.
A variety of potential confounders were included in our analysis. These included preoperative or demographic risk factors such as age (years), sex, body mass index, and alcohol or tobacco use, as well as the presence of medical comorbidities. The American Society of Anesthesiologists (ASA) Physician Status Classification is a grading system used to assess the fitness of patients prior to surgery.10 The Charlson Comorbidity Index (CCI) assigns a score to each patient based on selected comorbid conditions and has been used to predict 1-year mortality in patients with cancer.11 Both indices were used in our study to stratify patients based on comorbid disease states. We also noted the presence of the following medical conditions: coronary artery disease, chronic obstructive pulmonary disease, diabetes mellitus, stroke, and hematologic malignancy. In patients with a history of neoplasm, tumor site and stage were assessed based on the American Joint Committee on Cancer Staging guidelines,12 as well as any previous surgical or medical treatment for this condition.
In addition, the following perioperative risk factors were also reviewed: presence of tracheostomy (yes/no), presence of neck dissection (bilateral vs unilateral), length of anesthesia (minutes) volume infused during the procedure (liters), estimated blood loss (liters), and type of free flap used, and postoperative ventilation time (minutes).
The primary outcome measure was LOS in the ICU measured in hours. The “ICU admit time” was defined as the end of the anesthesia noted on the anesthesia record for the free flap procedure. The “ICU discharge time” was determined by reviewing progress notes, nursing notes, and vital sign charting. A transition from hourly vital sign monitoring to intervals of 4 to 6 hours was also used to estimate the time of transfer out of the ICU. This transition occurs when the intensivist team places an order to transfer the patient to acute care status. The ICU LOS was defined as ICU discharge time minus ICU admit time.
A secondary outcome was the total hospital LOS, which comprised the duration of the entire postoperative hospital course including both ICU LOS and LOS after transfer out of the ICU. Total hospital LOS was defined as hospital discharge time minus ICU admit time. We also addressed additional secondary outcomes measures based on postoperative adverse events occurring within 30 days of the initial procedure such as death, ICU readmission, and other surgical or medical complications. Surgical complications included wound dehiscence, fistula, hematoma, surgical site infection, anastomotic revision, and airway distress. Medical complications were categorized as pneumonia/pulmonary issues, delirium, acute myocardial infarction, and pulmonary embolism.
Testing of differences among covariates and outcomes between the closed and open ICU cohorts was done using a 2-sample t test for continuous variables and a χ2 test for categorical or dichotomous variables. Univariate linear regression analysis was used to test for unadjusted associations between the primary outcome of interest, ICU LOS, and potential risk factors. Multivariate linear regression analysis was used to adjust for potential confounding by a variety of covariates. All statistical analyses were performed using Stata SE version 10.1 (StataCorp).
A total of 104 consecutive free flaps were performed during the study period. Of these, 50 were admitted during the open ICU period, and 52 were admitted after conversion to the closed ICU. Both groups were similar with a few exceptions. The open ICU cohort had more male patients (P = .06). The open ICU cohort also had a higher degree of preoperative comorbidity, as demonstrated by higher CCI scores (P = .002) (Table 1). Of the 50 patients (52 free flaps) who were admitted to the open ICU, 17 (34%) had an internist assisting in the management of the medical comorbidities during their hospital stay.
The open ICU cohort had significantly more patients who underwent tracheostomy (P = .02) and bilateral neck dissection (P = .006) (Table 2). There were no other significant differences in the operative characteristics of the 2 cohorts. The mean length of anesthesia during the operation was approximately 12 to 13 hours. Most patients had disease confined to the oral cavity, but oropharyngeal and cutaneous neoplasms were also common in both cohorts. Fibular and radial forearm free flaps were most commonly used for reconstruction in both cohorts. The most common pathologic subtype was squamous cell carcinoma, followed by benign disease such as osteoradionecrosis. Those with malignancy typically had stage IV disease (Table 2).
Intensive care unit and hospital readmission rates were the same in both cohorts. The overall rate of surgical and medical complications was similar between both cohorts (Table 3).
Hospital LOS was 221.5 and 217.3 hours in the open and closed ICU cohorts, respectively. The LOS in the ICU was also similar in both cohorts (44 hours in the open ICU cohort and 44.9 hours in the closed ICU cohort) (Table 3). There were no significant differences in ICU, acute care unit, and total hospital LOS between the open and closed ICU cohorts. When we placed various risk factors into a univariate linear regression model, we found that increasing age was associated with increased ICU LOS (P < .05) (Table 4).
A multivariate model was then used to control multiple possible confounders including age, sex, CCI score, pathologic condition, stage, type of neck dissection, tracheostomy, and type of free flap. Patients who underwent radial forearm free flap stayed in the ICU for 28 hours less than their counterparts. Albeit nonsignificant, tracheostomy appeared to decrease ICU LOS by approximately 17 hours. Increasing age by 1 year resulted in a 1 hour corresponding increase in ICU LOS. Neck dissection did not significantly impact ICU LOS. Increasing TNM stage by 1 category resulted in a corresponding 22-hour increase in ICU LOS (Table 5).
Patients who undergo microvascular free flap reconstruction of the head and neck require special postoperative care for which head and neck surgeons are specifically trained. The most important of these are monitoring of the airway and free flap. Although high-intensity ICUs have several benefits,2,3 closed ICUs may result in distancing the primary head and neck surgical team from the patient in the acute postoperative period. Consequently, specific elements of head and neck surgical care such as adequate wound care and intraoral suctioning may not be delivered to the patient. Moreover, if specific events occurred during the patient's operation requiring specific assessments on the part of the surgical team, distancing the primary surgical team from the patient could result in missing important information regarding the patient's postoperative course. However, a critical-care perspective on patient care in the ICU by intensivists may result in improved outcomes in these patients. Regardless of the primary team managing the patient care, early detection of both detrimental and beneficial effects of major changes in hospital care permits early intervention to prevent poor outcomes or encourages interventions that result in positive outcomes.
To our knowledge, no study in the literature has focused exclusively on the effect of an open vs closed ICU on patients undergoing head and neck microvascular reconstruction. For this reason, we investigated if the implementation of the closed ICU at our institution changed the outcomes of our free flap reconstruction patients. We recognize that the need for postoperative ICU care varies from institution to institution based on factors such as availability of a step-down unit, a consistent group of staff caring for the free flap reconstruction patients, free flap volume, and in-house otolaryngology residents. In fact, recent studies have demonstrated that immediate extubation in the operating room is feasible and may even be beneficial with regard to LOS.13 Nonetheless, our practice provides what we believe to be the safest way to care for our patients based on the resources we have available at our institution, especially with regard to ensuring safe airway management. In our institution, the presence of a free flap by itself is not the critical aspect of why the patients are in the ICU. The purpose of this study was not to suggest a particular disposition for postoperative care of head and neck free flap patients. Rather, our goal was to determine if changes taking place on a national basis would have an overall effect on the care of the complex head and neck surgical patient.
In our study, there was no significant difference in ICU or hospital LOS for postoperative head and neck free flap patients with the implementation of the closed ICU. This finding is likely a result of the fact that the majority of patients undergoing microvascular reconstruction of the head and neck are likely placed in the ICU postoperatively, not for management of critical medical conditions, but for airway monitoring and optimization of flap care in the immediate postoperative period. Thus, a change to management by an intensivist for these less medically complex patients is unlikely to result in a significant change in length of ICU stay. Furthermore, once patients leave the ICU and transfer to the floor, overall hospital LOS is less likely to be dependent on the type of treatment received in the ICU and more likely related to total time spent in the ICU and to medical and surgical complications that arise during the stay. Furthermore, in the open ICU cohort, for patients who were more medically complex, an internal medicine physician was consulted. The internal medicine physician assisted in managing medical issues in the ICU, essentially acting as an “intensivist” for the primary surgical team.
Surgical and medical complication rates were similar for both groups. With regard to surgical complications, it is unlikely that any of these would be influenced significantly by changes in ICU staffing patterns. Nevertheless, there are certain situations where surgical outcome could be more affected by intensive care philosophy. For instance, liberal use of vasopressors in the ICU for hypotension may compromise perfusion of the free flap, possibly contributing to flap failure. Moreover, use of excessive anticoagulation early following surgery may result in bleeding or hematoma formation requiring surgical treatment. Placing the patient in positions that compromise geometry of the flap vessels may lead to hypoperfusion or congestion of the flap. Nonjudicious use of antibiotics in the ICU may predispose the patient to developing surgical site infection in the future. Decreased inclination to use of anti-inflammatory drugs (such as systemic steroids) may contribute to airway obstruction following extubation. However, it is more likely that surgical complications are due to events occurring during the operative procedure itself.
It is reasonable to assume that the incidence of medical complications would be more dependent on ICU care compared with the rate of surgical complications. For instance, poor pulmonary toilet in the ICU may lead to the development of pneumonia during the acute care unit stay. The failure to provide prophylactic treatment to patients and encourage early ambulation may lead to pulmonary embolism. Failure to address myocardial oxygen demands may result in myocardial infarction. Nonetheless, these issues are universal to care in an ICU or acute care ward managed by the primary head and neck surgical team. Thus, the finding that there is no difference in medical complications between the 2 groups is not surprising, given that head and neck surgical patients are again typically less medically complex and any such complex patients in the open ICU cohort had a medical consultant assisting in managing these issues in the ICU. Regardless, these examples illustrate the importance of good communication between the surgical and critical care teams to address the specific needs of each patient based on their anatomy and the events that occurred in the operating room, particularly since free flap reconstruction patients have very specific and complex necessities. Over the study period, we did not encounter any miscommunications that resulted in changes in patient quality of care. However, with time and repeated communication errors, it is inevitable that a lapse in patient care may occur. For this reason, we believe that good two-way communication between care health care providers is paramount to providing good quality of care to our patients. In addition, our institution has partially dealt with the issue by educating our intensivist colleagues and staff about specific head and neck free flap care in the postoperative period. Some centers may elect to establish specific protocols to standardize postoperative care of the head and neck surgical patient.
Interestingly, more patients in the open ICU cohort underwent tracheostomy. When placed in a linear regression model, tracheostomy was associated with shorter ICU LOS (albeit not statistically significant). This difference could be because tracheostomy patients can more readily be weaned from mechanical ventilation and require less intensive airway monitoring than patients without a tracheostomy. More patients in the open ICU cohort had oropharyngeal tumors, which could account for the difference in tracheostomy incidence. However, the closed ICU cohort had more patients with oral cavity tumors, and fewer of these patients underwent tracheostomy. This can be attributed to a recent trend in our facility of using nasotracheal intubation for patients with tumors of the oral cavity as opposed to tracheostomy.14 There were also significantly more patients in the open ICU cohort with greater comorbid disease, as illustrated by their CCI scores in Table 1. It would be reasonable to hypothesize that patients with a greater degree of comorbidity would stay longer in the ICU, However, this was not the case in our study, since CCI score was not associated with increased LOS. This is of limited clinical significance, however, because comorbid conditions listed in the CCI may not be important factors in determining ICU LOS.
Increasing age was associated with a longer ICU LOS. This may be secondary to health care provider beliefs that elderly people require more careful monitoring, thereby prolonging their ICU LOS. Furthermore, older patients may actually require more time to recover postoperatively compared with their younger cohorts. Older patients may also have more comorbid disease, contributing to an increased ICU LOS.
Increasing TNM stage was also associated with a longer ICU LOS, possibly a result of the need for more extensive surgery, which can often result in longer and more intensive postoperative care. Not surprisingly, patients who underwent radial forearm free flap had significantly shorter ICU LOS. This is likely due to the lack of impact of the donor site on ambulation. In addition, radial forearm defects are more commonly used in our practice for anterior defects of the oral cavity, which result in less of a concern for airway monitoring. From a clinical standpoint, these findings are important in assessing how much time a patient may require in the ICU prior to ICU admission. This knowledge can also be helpful in planning the patient flow in the ICU for the intensivist team.
Perhaps the greatest limitation of this study was its retrospective nature. Because we relied on data that had already been collected, estimations had to be made. Specifically, the LOS outcome was estimated as described in the “Methods” section. In a prospective study, we would design a more accurate and precise method to measure the exact transition time from the ICU to the floor. Nonetheless, the metrics used in our study were likely sufficient to obtain a general sense of trends in data. Moreover, because the methods to calculate the LOS for each patient were performed in a consistent manner by the same author (P.K.B.), random error was minimized. It is for this reason that larger prospective studies are required in the future.
In conclusion, high-intensity staffing in ICUs is becoming more common and may have effects on outcomes of patients undergoing microvascular reconstruction of the head and neck. We found that patients admitted to a closed ICU have similar ICU LOS, hospital LOS, and incidence of complications compared with patients admitted to an open ICU following microvascular free flap reconstruction in the head and neck.
Correspondence: Prabhat K. Bhama, MD, Division of Facial Plastic Surgery, Department of Otolaryngology, Massachusetts Eye and Ear Infirmary, Harvard Medical School, Boston, MA 02108 (email@example.com).
Submitted for Publication: June 15, 2012; final revision received September 24, 2012; accepted October 17, 2012.
Author Contributions: Drs Bhama and Davis had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Bhama, Bhrany, and Futran. Acquisition of data: Bhama and Bhrany. Analysis and interpretation of data: Bhama, Davis, Bhrany, Lam, and Futran. Drafting of the manuscript: Bhama and Davis. Critical revision of the manuscript for important intellectual content: Bhama, Davis, Bhrany, Lam, and Futran. Statistical analysis: Davis and Lam. Administrative, technical, and material support: Bhama and Futran. Study supervision: Davis, Bhrany, and Futran.
Conflict of Interest Disclosures: Dr Futran is a consultant for Stryker Corporation and received payment for technical advice on medical device design and presentation regarding Stryker medical devices. There is no direct conflict with the subject of this article, and the information within as it does not pertain to any of Dr Futran's consultant activities with Stryker Corporation.
Funding/Support: This publication was made possible by grant KL2 RR025015 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH).
Disclaimer: The contents of this article are solely the responsibility of the authors and do not necessarily represent the official view of NCRR or NIH.
Previous Presentation: This research was presented at the American Academy of Facial Plastic and Reconstructive Surgery section of the Combined Otolaryngology Spring Meetings; April 19, 2012; San Diego, California.