Is same-day discharge suitable following rigid esophagoscopy? Findings in a series of 655 cases. (Original Article).

Abstract

It has been suggested that more otolaryngologic procedures should be performed on an outpatient basis, and that rigid upper aerodigestive tract endoscopy might be a particularly suitable procedure in this regard. To determine if this is indeed the case, we retrospectively reviewed the records of 563 patients who had undergone 655 rigid esophagoscopies in our unit between Jan 1, 1991, and July 31, 1998. We ascertained the rate of complications (primarily esophageal perforation) following such procedures and, when they did occur, we determined the length of time between surgery and the onset of the complications' signs and symptoms. Our aims were to establish the minimum duration of postoperative observation that is required following esophagoscopy and to propose criteria for safe same-day discharge. We found that perforation rates were 4.5% following therapeutic procedures (dilation, biopsy, and foreign-body removal) and 1.2% following diagnostic procedures. In 40% of the patients who experienced perforations, no such signs or symptoms were noted within the first 8 hours following surgery. This finding has important implications for surgeons who wish to perform rigid esophagoscopy on an outpatient basis.

Introduction

Same-day-discharge surgery (also called outpatient, ambulatory, and, in the United Kingdom, day-case surgery) allows surgeons to treat patients more quickly and for less money than is possible with inpatient surgery. As a result, same-day discharge has become more popular in recent years. It has been suggested that approximately 50% of all otolaryngologic operations in the U.K. can be performed on an outpatient basis; currently, fewer than 25% are done in this way. (1)

In 1992, the Royal College of Surgeons of England published guidelines in which it suggested that same-day discharge is appropriate for patients who undergo rigid upper aerodigestive tract endoscopy. (2) Since then, two retrospective studies were undertaken to investigate the safety of same-day discharge following rigid upper aerodigestive tract endoscopy. (3,4) Between them, the two studies included fewer than 100 esophagoscopies, and no complications occurred. These reports are interesting, but as yet there is still insufficient evidence to support the Royal College's proposal that same-day discharge following rigid esophagoscopy is safe or to indicate which patients should be considered unsuitable for same-day discharge.

Rigid esophagoscopy carries a low complication rate; the most common and serious problem is esophageal perforation. Perforation rates have been reported to be as low as 0.1 to 1.1 % (5) and, in patients who undergo esophageal dilation or biopsy, as high as 12%. (6-8)

It is possible that the complications of esophagoscopy do not become apparent within the short postoperative observation period during which patients remain in the day-surgery unit. There are few data in the literature regarding the time of onset of the signs or symptoms of postesophagoscopy perforation, so the minimum duration of postoperative observation that is necessary remains unknown. In some series, many perforations of the esophagus were not diagnosed until 6 hours postoperatively and occasionally as late as 72 hours. (9,10) Such delays might well be attributable to the absence of signs or symptoms in the early postoperative period.

In order to determine both the complication rate following rigid esophagoscopy and to establish the minimum amount of postoperative observation that is required to identify complications, we retrospectively reviewed the records of all rigid esophagoscopies that had been performed in our unit over a period of 7-plus years. Our findings allowed us to propose criteria for the safe same-day discharge of these patients.

Patients and methods

We reviewed the records of all patients who had undergone rigid esophagoscopy in the Department of Otolaryngology at Crosshouse Hospital in Kilmarnock, Scotland, between Jan. 1, 1991, and July 31, 1998. Patients were identified by a search of our hospital's computerized coding database. We began by isolating any procedure that could conceivably relate to rigid esophagoscopy, and then we checked the resulting list on a case-by-case basis to exclude incorrectly coded procedures. This laborious process was necessary to detect any errors in the coded data. The recorded data included age, sex, and the type of procedure performed in each case.

All patients had been routinely assessed by the operating surgeon at the end of the day's surgeries and again during ward rounds the next morning. Postoperative investigations, such as chest x-rays, were not performed routinely; they were done only when a perforation was suspected clinically.

The types of complications of esophagoscopy were identified by the hospital's coding data and by a detailed review of the case notes of all patients (1) who had stayed in the ward for more than the usual 24 hours after surgery, (2) who had been readmitted within 14 days of surgery, and (3) who had died within 14 days of surgery. Particular attention was paid to the nursing records and observation charts. Complications were categorized as either definite perforation, probable perforation, or other (table 1).

Results

During the study period, 563 patients had undergone 655 rigid esophagoscopies. Our study population was made up of 296 males (52.6%) and 267 females (47.4%), aged 2 to 90 years (median: 60).

Type of procedure. Of the 655 procedures, 221 (33.7%) were therapeutic and 434(66.3%) were diagnostic. Of the therapeutic procedures, 96 (14.7% of the 655 procedures) had involved dilation of a stricture, 84(12.8%) had involved an esophageal biopsy, and 68 (10.4%) had involved the removal of a foreign body; these figures include those cases when patients underwent the procedure for more than one indication (table 2). These procedures had been performed by staff members of all grades, from registrar to consultant.

Complications. Complications of esophagoscopy were identified in 21 cases (3.2% of the 655); four (0.6%) were definite perforations, 11 (1.7%) were probable perforations, and six (0.9%) were categorized as other (table 3). Of the 96 patients who had undergone stricture dilation, three (3.1% of the 96) developed either a definite or probable perforation. Of the 84 patients who had undergone esophageal biopsy, six (7.1%) developed a definite or probable perforation. Of the 68 patients who had undergone removal of a foreign body, one (1.5%) had a probable perforation.

Perforations occurred during 10 of the 221 (4.5%) therapeutic procedures and during five of the 434 (1.2%) diagnostic procedures. The difference in the two rates was statistically significant according to the chi-squared ([chi square]) test with Yates' correction ([chi square] = 6.01; p<0.02). Two perforations (one definite and one probable) resulted from esophagoscopy in the presence of a previously unsuspected pharyngeal pouch. Of the 15 cases of definite or probable perforation, the medical records of only five (33.3%) contained any mention of a mucosal laceration.

Time to onset of signs or symptoms of complications. The length of time between surgery and the onset of signs or symptoms of perforation in the 15 patients ranged from 0 to 24 hours (median: 8). The first signs or symptoms of perforation had been recorded within 4 hours in four patients (26.7%). By 6 hours, that number had increased to six patients (40.0%), and by 8 hours it had risen to nine patients (60.0%). The remaining six patients (40.0%) did not exhibit any signs or symptoms until between 8 and 24 hours postoperatively.

Readmission. Only one patient had been readmitted within 14 days of discharge, and she had already been identified as having a perforation that had been treated conservatively.

Mortality. Three patients died (overall operative mortality: 0.5%), two as a result of an esophageal perforation. One of these patients had become confused and pyrexial 24 hours postoperatively, and a perforation of her esophagus was assumed. This assumption was confirmed when she developed surgical emphysema of the neck the next day. She was managed conservatively to no avail. No contrast studies had been performed to localize the site of her perforation.

Another patient who died had developed chest pain, back pain, and pyrexia on postoperative day 1. She was immediately transferred to the thoracic surgical unit, but she died 4 weeks later despite treatment. The third death occurred in a patient who had become stridulous immediately following surgery. She was thought to have aspirated.

Discussion

Rigid esophagoscopy is safe in most patients. However, if we are to introduce a shorter hospital stay following this procedure, we must be aware of the possibility of complications and we must tailor our postoperative observation accordingly. We undertook our study to determine the complication rates of the various procedures (stricture dilation, biopsy, foreign-body removal, and diagnostic esophagoscopy) and to record the time of the onset of the signs and symptoms of perforation.

Complications of esophagoscopy are sufficiently uncommon to make a prospective study impractical, hence the retrospective study design. Certainly, we must acknowledge the limitations of a retrospective study as well as the possibility of inaccuracies in the coded data. We tried to make our review as comprehensive as possible, and we searched the coded data case by case to address these issues. There always remains, however, the likelihood that some cases will be missed. For example, we cannot account for any patients who might have sought treatment elsewhere for complications of surgery done at our hospital. However, since Crosshouse Hospital has the only otolaryngology unit in Ayrshire (County), Scotland, there is little overlap between our service and that of neighboring units.

We also acknowledge that written records are often an inadequate source of information. Nevertheless, we found that the nursing notes we reviewed were invariably comprehensive, and we used the onset of tachycardia and pyrexia, as recorded on the observation charts, as two of our diagnostic signs. The use of contrast swallow imaging to detect the presence and position of a suspected perforation was not common in our series, so the diagnosis of a perforation was presumptive in many cases. This necessitated our decision to divide perforations into the two categories (definite or probable) based on the findings at the time. These limitations notwithstanding, we feel that certain conclusions can be drawn from our data.

Taking into account the case mix at our unit (including patients with benign and malignant disease and post-radiotherapy patients), our overall morbidity and mortality rates from rigid esophagoscopy are low and comparable to those reported in other published series. (5-8) The perforation rate for therapeutic procedures in our series was significantly higher than that for diagnostic procedures (4.5 vs 1.2%; p<0.02), so it would seem prudent to recommend overnight admission and observation for all patients who undergo biopsy or dilation procedures and possibly for those who undergo foreign-body removal (table 4).

It should be borne in mind that complications might not become apparent within the short observation period (4 hr) that is typical in a busy day-surgery unit. We believe that an observation period of at least 8 hours is required to identify most perforations; even so, in our series, 40% of perforations did not manifest until at least 8 hours had passed. It is possible that symptoms appear only after the patient has been allowed to drink fluids. The late diagnosis of a perforation might therefore be attributable to the surgeon's instructions regarding the length of time for which the patient must be kept "nil by mouth" postoperatively. It is our experience that surgeons' individual customs vary widely in this regard, but we were not able to quantify these practices in our study because they were not adequately addressed in the written records.

We were interested to know how many perforations had been recognized at the time of surgery. The presence of a mucosal tear was recorded in the surgical notes in only one-third of the 15 cases, but the true proportion cannot be judged with any certainty. It might be prudent to recommend overnight admission for any patient with a recognized mucosal tear.

Regardless of the move toward same-day discharge, we believe that our study has highlighted the need for a more aggressive approach to the management of suspected esophageal perforations in our unit. We will continue to institute conservative measures as soon as a perforation is suspected. These steps include the passage of a nasogastric tube, the prevention of oral fluid or solid intake, and the commencement of an intravenous broad-spectrum antibiotic. We will also now stress the importance of radiologic investigation to diagnose and localize a perforation and the need for early referral to the thoracic surgical unit of all patients who have a perforation of the thoracic esophagus. (11,12)

Same-day discharge will be unsuitable for many patients because of anesthetic or social reasons. We did not have sufficient data, however, to determine how many patients would have been poor candidates for same-day discharge in our series. Assuming that all patients in our series had satisfied the anesthetic and social criteria for same-day discharge, a minimum of 8 hours of postoperative observation for patients who had undergone a diagnostic procedure and overnight admission for those who had undergone a therapeutic procedure would have resulted in a failure to identify one definite and three probable perforations (26.7% of all perforations). This rate amounts to an overall risk of 0.9% (4/434) for a missed diagnosis for those discharged the same day.

An esophageal perforation is, of course, potentially fatal, and the prognosis is poorer when diagnosis is delayed. The risk of a missed perforation must be weighed against the benefits of a shorter hospital stay. Although patients might prefer same-day discharge, its perceived financial benefits might not be as great as was once anticipated in the U.K., (13) especially if we take into account the need for postdischarge telephone follow-up and the possible need for patients to consult their family physicians to a greater extent during the early postoperative period. Surgeons who wish to discharge their patients on the same day as a rigid esophagoscopy must be aware of this balance between risk and benefit. They should also be very aware that not all esophageal perforations manifest themselves clinically during the immediate postoperative period.

Table 1

Conditions that were considered to be complications in this study


Definite perforation  Perforation on contrast study
                      Endoscope noted to be outside the
                      esophageal lumen during surgery
                      Surgical emphysema of the neck
                      Pneumothorax or mediastinal
                      emphysema

Probable perforation  Chest or back pain
                      Tachycardia >100 bpm
                      Pyrexia >37.5[degrees]C

Other                 Vomiting
                      Confusion
                      General complications of surgery

Table 2

Number and typeof procedures


 identified in this study
Therapeutic procedures                 n (%)
    Dilation of stricture            71 (32.1)
    Biopsy of the esophageal mucosa  64 (29.0)
    Removal of a foreign body from
     the esophagus                   60 (27.1)
    Dilation and biopsy              18 (8.1)
    Dilation and foreign-body
     removal                         6 (2.7)
    Dilation, biopsy, and foreign-
     body removal                    1 (0.5)
    Foreign-body removal and biopsy  1 (0.5)

Total therapeutic procedures         221 (100.0)  221 (33.7)

Total diagnostic procedures                       434 (66.3)

Total overall procedures                          655 (100.0)

Table 3

Number and type of complications Identified in this series


Complication                    n

Definite perforation       4 (2 deaths)

Probable perforation      11

Other
  Chest infection          1
  Vomiting                 2
  Aspiration and stridor   2 (1 death)
  Melena and hematemesis   1

Total                     21 (3 deaths)

Table 4

Suggested criteria for identifying patients who are not suitable for
same-day discharge following an 8-hour postoperative observation period


Patient does not meet anesthetic or social criteria

Patient underwent dilation of a stricture

Patient underwent biopsy of the esophageal mucosa

Patient underwent foreign-body removal

Mucosal tear was noted during surgery

Signs or symptoms of esophageal perforation are present

Acknowledgment

The authors thank Mr. John Dempster for reviewing the manuscript.

References

(1.) Brown PM, Fowler S, Ryan R, Rivron R. ENT day surgery in England and Wales--an audit by the Royal College of Surgeons (Eng.) Comparative Audit Service. J Laryngol Otol 1998;112:161-5.

(2.) Royal College of Surgeons of England Guidelines for day ease surgery (revised). London: RCSEng, 1992.

(3.) Lee CM, Hinrichs BA, Terris DJ. Routine hospital admission for patients undergoing upper aerodigestive tract endoscopy is unwarranted. Ann Otol Rhinol Laryngol 1998;107:247-53.

(4.) Whinney D, Vowles R, Harries M. Appropriate use of the day care unit for rigid endoscopy of the upper aerodigestive tract. Ann R Coil Surg Engl 1998;80:111-4.

(5.) Prinsley PR, Murrant NJ. Cervical oesophageal perforation caused by diagnostic flexible oesophagoscopy. J Otolaryngol 1989;18:314-6.

(6.) Ritchie AJ, McManus K, McGuigan J, et al. The role of rigid esophagoscopy in esophageal carcinoma. Postgrad Med J 1992;68:892-5.

(7.) Hatzitheofilou C, Kakoyiannis S, Charalambides D, et al. Iatrogenie oesophageal perforations in patients with cancer of the oesophagus. S Afr J Surg 1993;31:90-3.

(8.) Swaroop VS, Desai DC, Mohandas KM, et al. Dilation of esophageal strictures induced by radiation therapy for cancer of the esophagus. Gastrointest Endosc 1994;40:311-5.

(9.) Mizutani K, Makuuchi H, Tajima T, Mitomi T. The diagnosis and treatment of esophageal perforations resulting from nonmalignant causes. Surg Today 1997;27:793-800.

(10.) Nashef SA, Pagliero KM. Instrumental perforation of the esophagus in benign disease. Ann Thorac Surg 1987;44:360-2.

(11.) Moghissi K, Pender D. Instrumental perforations of the oesophagus and their management. Thorax 1988;43:642-6.

(12.) Goldstein LA, Thompson WR. Esophageal perforations: A 15 year experience. Am J Surg 1982;143:495-503.

(13.) Drake-Lee A, Harris S. Day case tonsillectomy: What is the risk and where is the economic benefit? Clin Otolaryngol 1999;24:247-51.

From the Department of Otolaryngology, Crosshouse Hospital, Kilmarnock, Scotland.

Reprint requests: Haytham Kubba, Clinical Fellow in Paediatric Otolaryngology, Great Ormond Street Hospital for Children, London WC1N 3JH, U.K. Phone: [phone]; fax: [phone]; e-mail: [e-mail]