Unplanned admissions following outpatient otologic surgery: The University of Arkansas experience.

Abstract

We analyzed the outpatient otologic surgery experience at our institution to identify those factors that are associated with a high risk of postoperative complications that require an unplanned hospital admission. We found that among a group of 662 patients who underwent group II otologic procedures (i.e., tympanoplasty with or without mastoidectomy, stapedotomy, and middle ear exploration), the overall admission rate was 4.7%, of which 3.9% were unplanned. A significantly larger percentage of children were admitted than adults (5.7 vs. 2.3%), primarily for nausea and vomiting. Three factors were significantly associated with unplanned admissions: the type of surgery (tympanomastoidectomy with ossicular reconstruction), the duration of general anesthesia ([greater than]2 hr), and asthma as a coexisting condition. The choice of anti-emetic administered (ondansetron or droperidol) and the specific agents used for general anesthesia did not appear to have any significant impact on unplanned admissions. We recomm end that the three predisposing factors be taken into consideration when formulating the treatment plan. Scheduling an inpatient procedure for patients who have risk factors for complications requiring hospitalization would avoid the extra costs and stress associated with unplanned admissions.

Introduction

The past decade has seen the widespread growth and acceptance of outpatient otologic surgery in the United States. The advent of effective antiemetics and short-acting anesthetic agents has significantly reduced the incidence of postoperative nausea and vomiting. Moreover, improvements in pre- and perioperative instruction have made it possible for patients to better manage their own care after surgery. Finally, postoperative results with regard to hearing and graft-take have been shown to be similar to outcomes following hospital inpatient procedures. [1]

From a financial standpoint, outpatient otologic surgery, of course, is cost-effective, and most third-party payers provide incentives to encourage the use of ambulatory surgery. This, in turn, has led to a proliferation of freestanding outpatient surgery centers that have peripheral associations with inpatient facilities. However, not all freestanding surgery centers have inpatient facilities available, so some patients who require an unplanned admission as a result of postoperative complications of anesthesia or surgery must be transferred to another facility. This only adds to the inconvenience and stress for both patient and family. Thus, two questions arise:

1. Is it possible to identify which patient characteristics or surgical interventions are associated with a high risk of admission so that these patients can be treated in a hospital setting initially?

2. Can we identify which antiemetic and anesthetic agents are associated with the lowest risk of postoperative complications?

We analyzed the outpatient otologic surgery experience at the University of Arkansas for Medical Sciences (UAMS) and its affiliate, the Arkansas Children's Hospital, a large pediatric tertiary care center, both in Little Rock. Our aim was to identify any factors that are associated with a high risk of postoperative complications that require admission. By identifying pre- and perioperative parameters associated with unplanned admissions, we can better facilitate patient care in an outpatient setting.

Materials and methods

We performed a retrospective review of charts in a computerized otologic database to identify all patients who had undergone outpatient group II otologic surgery between July 1994 and July 1997. As defined by the American Academy of Otolaryngology--Head and Neck Surgery, group II procedures comprise tympanoplasty with or without mastoidectomy, stapedotomy, and middle ear exploration. [2] All surgeries were performed by the senior author and residents at UAMS or Arkansas Children's Hospital. We identified all patients who had either planned or unplanned admissions, and we compared the latter with an age-matched control group consisting of patients who were not admitted. The following information was extracted from each patient's chart: age at the time of surgery, sex, type of surgery, anesthesia technique, duration of surgery, medical comorbidities, and reason for admission.

The two types of anesthesia were local standby and general. Local standby anesthesia was used for all adult stapedotomies as well as for selected adult tympanoplasties, depending on the wishes of the patient. It consisted of propofol at a dose of 1 to 2 mg/kg, titrated to effect during the canal injections, with approximately 6 ml of 1% lidocaine with 1:100,000 epinephrine. Small doses of fentanyl (25 [micro]g) were used as needed to relax the patient.

General anesthesia was administered to all other adults and to all children. Canal injections identical to those used in the local cases were administered to facilitate hemostasis. Induction was performed with either propofol (2 mg/kg) or thiopental (4 to 5 mg/kg) in combination with fentanyl (50 to 100 [micro]g) and succinylcholine (1 to 2 mg/kg) to facilitate intubation. Halothane was used during induction in younger children prior to establishing intravenous access. Anesthesia was maintained with either isoflurane, halothane, or desflurane in combination with a mixture of air, [O.sub.2], and nitrous oxide at concentrations sufficient to maintain hemodynamic parameters. The choice of inhalational anesthesia was at the discretion of the anesthesiologist. Halothane and isoflurane were used at Arkansas Children's Hospital, and isoflurane and desflurane were used for adults at UAMS.

All patients were given antiemetic drugs prior to induction. In patients who were seen before 1995, droperidol (0.05 to 0.1 mg/kg) was given 10 minutes prior to induction and repeated as needed for nausea and vomiting in the postanesthesia care unit. This antiemetic regimen was altered in 1995, when patients began receiving ondansetron (0.05 mg/kg) 30 minutes prior to induction and as needed postoperatively, up to a total dose of 4 mg.

Results

During the 3-year study period, 662 group II oto-logic procedures were performed on 317 children (age: [less than]18 yr) and 345 adults (table 1). A total of 31 patients (4.7%) were admitted to the hospital, 26 of whom (3.9%) had been scheduled only for outpatient surgery (table 2). The other five admissions had been planned.

Children vs. adults. Of the 26 patients who had unplanned admissions, 18 were children (5.7% of all pediatric cases) and eight were adults (2.3% of all adult cases). The difference between the percentage of unplanned adult and pediatric admissions was statistically significant (p[less than]0.05).

General vs. local anesthesia. Local anesthesia was used in 112 cases, and general anesthesia in the remaining 550 cases. Among the 550 patients who received general anesthesia, 25 (3.8% of the total population) had unplanned admissions. The reasons for admission in the pediatric group included intractable nausea and vomiting (14 patients) and pulmonary complications, defined as a failure to maintain pulse-oximeter scores above 95% on room air (four patients). In the adult group who received general anesthesia, the reasons for unplanned admission were intractable nausea and vomiting (six patients) and difficulty in treating brittle diabetes (one patient).

There was only one unplanned admission for a patient who received local anesthesia. An adult who had undergone revision stapes surgery was admitted for vertigo, nausea, and vomiting.

Control-group comparison. We compared the perioperative parameters of the 26 patients who had unplanned admissions with those of 26 age-matched controls who were not admitted (table 3). Statistically significant differences between the two groups were seen with respect to the type of surgery, the duration of anesthesia, and the presence of asthma as a medical comorbidity (p[less than]0.05 for each). Patients who underwent a tympanomastoidectomy with ossicular reconstruction and those whose duration of anesthesia exceeded 2 hours had a significantly higher incidence of postoperative nausea and vomiting and a significantly higher admission rate (p[less than]0.05). (These two parameters are obviously related because tympanomastoidectomy is a more extensive group II procedure that typically requires at least 2 hours of anesthesia time.) The particular antiemetic or anesthetic agent administered did not have any demonstrable impact on the rate of unplanned admissions.

Asthma was also associated with a significant risk for unplanned admission (p[less than]0.05). Ten of the 12 patients with asthma had to be admitted following outpatient surgery--six for nausea and vomiting and four for pulmonary complications.

Other medical problems--such as diabetes, cardiovascular disease, and renal disease--were not associated with unplanned admissions. Of the five planned admissions, two were for cardiovascular monitoring because of a history of heart disease, two were for obstructive pulmonary disease, and one was for asthma.

Discussion

The results of outpatient group II otologic procedures with regard to hearing result, graft-take, and complications such as facial nerve injury and cerebral spinal fluid leaks have been shown to be no different from those of inpatient procedures. [1]

Children vs. adults. Our overall rate of unplanned admissions (3.9%) was somewhat higher than that reported in the literature (0.4%). [1] One reason for the difference might be that almost half of the procedures we performed were on children, all of whom received general anesthesia; their unplanned admission rate (5.7%) was more than twice that of adults (2.3%), one-third of whom were given local anesthesia. The unplanned admission rate for all the general anesthesia cases (adult and pediatric) was 3.8% of the entire population.

Asthma. Another reason for the different admission rates between adult and pediatric patients was the presence of asthma. Of all the medical comorbidities we analyzed, only asthma was associated with a statistically significant difference in the rate of unplanned admissions (p[less than]0.05). Although the incidence of asthma among the adults and children in this study was not significantly different, there was a greater prevalence in the pediatric group (nine children versus two adults). Likewise, four children were admitted for pulmonary complications, a complication not reported in the adult group.

The reason that medically controlled asthma is associated with a significantly higher rate of unplanned admission than other medical comorbidities is difficult to ascertain. It is interesting to note that of the 10 patients with a history of asthma who were admitted unexpectedly, only four were admitted for pulmonary complications; nausea and vomiting accounted for the other six unplanned admissions. It is possible that the history of asthma lowered the threshold for admission in this group and biased the personnel in the postanesthesia care unit toward conservative treatment. Unfortunately, this is difficult to document from the postoperative record. We have now initiated steps to closely monitor patients with asthma to document the actual reason for admission. This documentation will help us determine whether a planned admission should be considered for this subset of patients.

Nausea and vomiting. The primary complications of outpatient otologic surgery seem to be associated not so much with the surgical procedure itself, but with the anesthetic techniques used. In our study, as in others, nausea and vomiting were the most common reasons for unplanned admission. Nausea and vomiting might be more common after otologic surgical procedures (the incidence has been reported to be as high as [50%.sup.3]) because otologic procedures might stimulate the vestibular labyrinth. Labyrinth stimulation activates the chemoreceptor trigger zone and can lead to emesis.

Certainly, the choice of an antiemetic and its effect on the unplanned admission rate was of paramount interest in this study. From a theoretical standpoint, one would anticipate that ondansetron would be advantageous in the treatment of nausea and vomiting associated with otologic procedures because of its method of action. [4] Ondansetron is thought to function centrally, possibly as a selective antagonist of the [5-HT.sub.3] receptors in the chemoreceptor zone, directly reducing vestibular labyrinth-induced nausea and vomiting. [5] Droperidol, another effective antiemetic, inhibits dopaminergic receptors in the chemoreceptor trigger zone, although it does not always influence labyrinth-induced vomiting. [6]

It is interesting to note that while the total number of admissions was small in both groups that received antiemetics, there was no significant difference in the rate of unplanned admissions between those who received droperidol and those who received ondansetron. This finding was seen previously in a prospective study comparing these two antiemetics and placebo. [4] While both antiemetics lowered the incidence of nausea and vomiting to a greater degree than did placebo following routine otologic procedures, there was no difference between the two drugs themselves. The ramification of this finding has to do with cost-effectiveness. At our institution, ondansetron is 18 times more expensive than droperidol, a fact that must be considered when this and other studies find no significant advantage of ondansetron over droperidol.

These comparisons also substantiate the fact that the nausea and vomiting associated with otologic procedures might not be related to stimulation of the labyrinth by surgical manipulation, but to factors related to general anesthesia. This theory is supported by our data on stapes surgery. Although the labyrinth is directly violated by this surgery, there were no unplanned admissions following primary surgery, all of which were performed with local anesthesia. In fact, the type of anesthesia (local vs. general) used had a significant impact on the rate of unexpected admissions. Only one of the 112 patients who received local anesthesia was admitted, and this was because the patient experienced vertigo after revision stapes surgery, not because of any complication of the anesthetic. Therefore, the use of general anesthesia should be considered to be a relative risk for unplanned admission.

The specific type of induction or inhalational agent used for general anesthesia appeared to have no significant bearing on nausea and vomiting. There is some evidence in the literature suggesting that the use of propofol as a maintenance agent might reduce the incidence of nausea after ear surgery, [7] but we were not able to make such an assessment because propofol was used only as an induction agent.

Duration of anesthesia. The duration of anesthesia was significantly associated with the rate of unplanned admissions. The mean duration of anesthesia for those who had unplanned admissions was 2 hours and 17 minutes, compared with only 1 hour and 26 minutes for the controls. Earlier data reveal that general anesthesia that lasts longer than 2 hours is associated with a significant risk for nausea and vomiting, [8] which correlates well with the data from our study.

Type of procedure. The duration of anesthesia is, of course, dictated by the type of surgical procedure performed. In our study, tympanomastoidectomy with ossicular reconstruction was significantly associated with unplanned admission. General anesthesia for this type of surgery typically lasts 2 to 2.5 hours, which is longer than that needed for any of the other procedures. The significance of this finding certainly has implications for resident training, where the duration of all surgeries is typically longer. Therefore, time management should be emphasized, especially for tympanomastoid surgery, to reduce unnecessary admissions and costs.

From the Department of Otolaryngology--Head and Neck Surgery, University of Arkansas for Medical Sciences, Little Rock.

References

(1.) Dickins JR, Graham SS. Otologic surgery in the outpatient versus the hospital setting. Am J Otol 1989;10:252-5.

(2.) Dickins JR. Comparative study of otologic surgery in outpatient and hospital settings. Laryngoscope 1986;96:774-83.

(3.) McGoldrick KE. Anesthesia for elective ear, nose and throat surgery. In: McGoldrick KE, ed. Anesthesia for Ophthalmic and Otolaryngologic Surgery. Philadelphia: W.B. Saunders, 1992: 101-5.

(4.) Jellish WS, Leonetti JP, Fluder E, Thalji Z. Ondansetron versus droperidol or placebo to prevent nausea and vomiting after otologic surgery. Otolaryngol Head Neck Surg 1998;118:785-9.

(5.) Bunce KT, Tyers MB. The role of 5-HT in postoperative nausea and vomiting. Br J Anaesth 1992;69(Suppl 1):60S-62S.

(6.) Stoelting RK. Pharmacology and Physiology in Anesthetic Practice. 2nd ed. Philadelphia: J.B. Lippincott, 1991.

(7.) Jellish WS, Leonetti JP, Fahey K, Fury P. Comparison of 3 different anesthetic techniques on 24-hour recovery after otologic surgical procedures. Otolaryngol Head Neck Surg 1999;120:406-11.

(8.) Bonica JJ, Crepps W, Monk B, et al. Postoperative nausea, retching and vomiting. Anesthesiology 1958;19:532-40.