HSC Section 3 - Trauma, Critical Care and Sleep Medicine

Otolaryngology–Head and Neck Surgery 156(4)

Data regarding overall and item-specific protocol compli- ance were then determined in an all-or-none fashion from these assessments. After the initial 3-month audit period, the full responsibility for patient assessments was transitioned to the unit nursing staff. TRAPU data collected retrospectively for all adult patients receiving tracheostomies between July 1, 2013, and June 30, 2014, prior to the implementation of the quality improvement protocol, were utilized for comparison (n = 183). All patients who were at least 18 years old and who underwent open sur- gical placement of a tracheostomy tube in the operating room from July 1, 2014, through June 30, 2015, at a single large urban tertiary care center were included in the study analysis (n = 155). Patients who were \ 18 years old, who underwent percutaneous tracheostomy, or who underwent tracheostomy tube placement at the bedside were excluded. Patients admit- ted to or transferred to the study institution with an existing tracheostomy tube in place were also excluded. Data at the research phase of the project were extracted with the quality improvement project database. Additional data regarding population characteristics and demographics were extracted from the electronic medical records/hospital billing data, merged with the primary database, and reconciled for valid- ity. Forty-eight patients were excluded from the final analy- sis: 21 who underwent percutaneous tracheostomy, 5 who underwent bedside tracheostomy, 1 who underwent cricothyr- otomy, and 21 who were \ 18 years old. A chi-square test of statistical independence with 1 degree of freedom was used to determine if a significant difference existed between the incidence of hospital-acquired TRAPUs prior to the establishment of the quality improvement initia- tive and after initiation of the protocol. Baseline characteris- tics were analyzed to assess the relative equivalence between the pre- and postintervention groups. A similar analysis was performed to compare patients with and without TRAPUs. Continuous variables (mean age, mean length of stay, mean intensive care unit days) were analyzed with an independent- sample Student t test. All other categorical variables were analyzed through chi-square analysis or Fisher exact test. A P value \ .05 determined statistical significance. To analyze the degree of comorbidity in each group, we used the Deyo-adapted Charlson Index as a measure of comorbid illness. 12,13 This index is based on 17 diagnostic categories identified from International Classification of Diseases, Ninth Revision, Clinical Modification discharge diagnoses. The Deyo-adapted Charlson Index was then con- sidered an ordinal variable with categories of 0, 1, 2, 3, 4, and 5 comorbid illnesses. Patients with 5 comorbid ill- nesses were categorized as 1 group. Data were stored and analyzed with SPSS 23. The Institutional Review Board of the Rush University Medical Center approved this study and waived the requirement for individual informed consent. Results Table 1 demonstrates the patient characteristics in the pre- and postintervention groups. All demographic variables and

comorbidities were similar in the 2 groups, with the most notable being the Deyo-Charlson score. There was a clini- cally small difference in renal disease in the postinterven- tion group (28% vs 17%, P = .023), but on analysis of the overall TRAPU cases, very few (n = 3) had comorbid renal disease and there was no statistically significant association ( P = .277). Table 2 illustrates a comparison between patients with and without TRAPUs overall. Mean length of stay in the hospital was longer in patients with TRAPUs (28.59 days) versus those without TRAPUs (22.35; P = .033). The distribution of the staging classifications of the TRAPUs is summarized in Table 3 . There were 20 TRAPUs identified out of 183 tracheos- tomies in the preintervention group (10.93%). After adop- tion of the standardized perioperative tracheostomy care bundle, the incidence of TRAPUs decreased, with only 2 incident ulcers out of 155 tracheostomies (1.29%). The chi- square analysis showed a significant difference between the pre- and postintervention groups, with a P value of .0003 ( Figure 4 ). The 2 patients who had a TRAPU in the postin- tervention period were reviewed by the WOCN-certified nurse, as with all ulcer cases. It was determined that the bundle was not fully implemented in both cases. In case 1, the patient did not have the sutures removed or tracheost- omy tube changed at postoperative day 7. In case 2, the patient’s dressing was not adequately positioned under the tracheostomy flange, allowing direct contact with the skin surface and subsequent ulceration. Protocol compliance data for each of the 4 elements in the study protocol are shown in Table 4 . Among the 19 audited patients during the first 3 months of the study, care of 3 patients was not in compliance with the study protocol. Two patients had their tracheostomy sutures removed after postoperative day 7. One patient’s sutures were removed on postoperative day 8 and the other patient’s on postoperative day 11. Another patient was observed to have 2 violations in compliance with nonneutral positioning of the head and neck on 2 of 9 observations and with no hydrocolloid dres- sing placed in the immediate postoperative period. Overall, the care of 84% of the randomly selected eligible patients during the 3-month audit period was found to be in com- plete compliance with the study protocol. Discussion The Centers for Medicare & Medicaid Services has included stage III and IV pressure ulcers on its list of preventable hospital-acquired conditions. Although stage III or IV TRAPUs may be rare, TRAPUs are probably an underreported complication of tracheostomy overall. In this study, our prein- tervention group had an incidence of hospital-acquired TRAPUs of approximately 10.92 ulcers per 100 tracheos- tomies performed. Audits from the prior year demonstrated an incidence of 12.5 per 100 cases. It was the high incidence of TRAPUs observed at our institution that provided the motiva- tion for this quality improvement intervention. In 2001, the Institute for Healthcare Improvement devel- oped the concept of a bundle: a group of 3 to 5 evidence-

112