HSC Section 8_April 2017

ranges from 6% to 9%. 7–10 The incidence of residual cho- lesteatoma is considerably higher and is frequently attrib- uted to insufficient disease resection or inadequate exposure. 11–14 As a result, second-look procedures are often employed to evaluate for residual or recurrent dis- ease and to perform ossiculoplasty, if indicated. Whereas the canal wall-up tympanoplasty is associ- ated with higher recurrence of cholesteatoma, the rou- tine execution of a second-look procedure is associated with significant cost and risks of a second procedure. There also exists significant difference in opinion regard- ing the optimal surgical strategy. 15 Costs of a second procedure include cost of the preoperative clinic evalua- tion, preoperative imaging, procedure, operative time, perioperative staff and perioperative consumables, post- operative prescription medications, and postoperative follow-up clinic visits. Other costs could include patient loss of productivity during the recovery period and the opportunity cost of the surgeon’s time. Risks of the pro- cedure include wound infection, hemorrhage, facial nerve paralysis, hearing loss, vestibulopathy, and risks of anesthesia. To date, no study has investigated outcomes of the second-look procedure from a cost perspective. Our aim is to investigate the outcomes and cost of both single- stage and second-look approaches to cholesteatoma management. This study was reviewed and approved by the Duke Uni- versity Institutional Review Board. We identified 420 adult and pediatric patients who underwent a tympanoplasty- mastoidectomy for cholesteatoma with a single-stage or second- look operative strategy from 2009 to 2014. During initial screen- ing, patients were included if they had a canal wall-up (CWU) or canal wall-down (CWD) primary tympanoplasty-mastoidectomy for cholesteatoma with or without a second-look procedure. Patients were excluded if they had history of a prior tympanoplasty-mastoidectomy, insufficient or incomplete medical records, or procedure for a diagnosis other than cholesteatoma. There are two parameters guiding the decision to perform a sec- ond look. One parameter for a planed second look was if the sur- geon believed not all cholesteatoma was removed. Another parameter for a second look was if there was evidence of exces- sive granulation and/or inflammation in the middle ear. In the event of excessive inflammation, our surgeons were concerned that a prosthesis would extrude. In these cases, reconstruction was planned to be completed at a second-look procedure. Outcomes Outcome variables tabulated included procedure approach, residual or recurrent cholesteatoma, ossicular chain reconstruc- tion (OCR) frequency, and operative complications. Recurrent cholesteatoma was defined as having found cholesteatoma in a new retraction pocket. 3 Residual cholesteatoma was defined as cholesteatoma found in the middle ear space and secondary to incomplete disease resection. 3 Audiologic outcomes included pre-/postoperative and latest visit air bone gap (ABG) and word recognition scores (WRS). Air conduction and bone conduction thresholds were tabulated at .5, 1, 2, and 3 kHz. MATERIALS AND METHODS Patient Population

Cost Analysis The cost data included charges, direct costs, and indirect costs for consultation and follow-up visits, surgical procedures, computed tomography (CT) scans of the temporal bones, and audiology visits. Our financial services office defined charges as the amount billed to the patient or insurance payor for hospital services. Specific cost and charge definitions for each component of care are used internally within our hospital finance depart- ment. Total cost includes both direct costs and indirect costs. Direct costs are a product of a variable direct cost and fixed direct cost. These cost types are directly related to patient care, but variable direct costs, such as supplies and direct nursing care, fluctuate with patient volume. Fixed direct costs (e.g., nurse managers, depreciation of medical equipment) do not fluctuate to the same degree. Indirect costs include costs not directly related to patient care but relevant to the upkeep of the hospital and facilities (e.g., administrative salary, utilities, grounds landscap- ing, building depreciation). When performing the cost analysis comparing patients who have had a second-look procedure versus those who did not, we used charges to set the perspective from that of a patient or insurance payor. We did not account for the probability of complications or the cost of managing these compli- cations in our cost analysis. Statistics Statistical analyses were completed using the JMP Pro 11 software suite (Cary, NC). Two-by-two contingency tables were created, and analysis of variance and Fisher’s exact tests were performed as appropriate. P values were reported with statisti- cal significance fixed at P 5 0.05. Statistical analysis was reviewed and approved by a statistician. RESULTS We reviewed 420 patients who presented for a tympanoplasty-mastoidectomy at our institution. Of these patients, 314 were excluded and 106 were included in subsequent analyses. Of the included cases, 65 (61.3%) were male and 76 (71.7%) were of adult age, with an average age 36.5 years and range of 2 to 90 years old. The most common initial surgical approach was CWU with 80 (75.5%), followed by 19 (17.9%) CWD, six (5.6%) tympanoplasty only, and one (0.94%) transcanal. For subsequent analyses, we focused on the 80 patients who underwent a CWU approach. Of these patients, 46 (57.5%) had a planned second-look proce- dure and 34 (42.5%) did not. When comparing CWU patients who had a planned second look versus no planned second look, there was no significant difference in any demographic variable, including age, % male gen- der, and ear sidedness. In CWU patients who had a planned second-look strategy, 22 (47.8%) had cholesteatoma identified at their second-look procedure. At the second-look proce- dure, two (4.35% of second-look patients; 2.5% of all CWU cases) had recurrent cholesteatoma, and 20 (43.5% of second-look patients; 25% of all CWU cases) had resid- ual cholesteatoma. In CWU patients who had no planned second look, four (7.5%) patients had an unanti- cipated second-look procedure frequency for clinical find- ings, suggestive of recurrent disease on a follow-up examination. All four of these patients had recurrent

Crowson et al.: Second-Look Tympanoplasty-Mastoidectomy

100